PHOTOSYNTHESIS AND RELATED PROCESSES VOLUME II Part 2 « '«;■- py as- b^ Section through Chlorella pyrenoidosa show- ing the cell wall, the cupshaped chromato- phore with its pyrenoid (py) and the cell nu- cleus (n), X 57,000. Inset: The pyrenoid, X 75,000 (both after Albertson and Leyon, Expii Cell Research, 7, 288, 1954). PHOTOSYNTHESIS nnd Related Processes By EUGENE I. RABINOWITCH Research Professor, Photosynthesis Research Labora- tory, Department of Botany, University of Illinois. Formerly Research Associate, Solar Energy Research Project, Massachusetts Institute of Technology. VOLUME II • Pari 2, Kinetics of Photosyutiiesis (continued)^ Addenda to Volume I and Volume II, Part 1 19 5 6 INTERSCIENCE PUBLISHERS, INC., NEW YORK Interscience Publishers Ltd., London © 1956, by Interscience Publishers, Inc. ALL RIGHTS RESERVED. This book or any part thereof must not be reproduced in any form without permission of the publisher in writing. This applies specifically to photostat and microfilm reproductions. Library of Congress Catalog Card Number 45-7293 INTERSCIENCE PUBLISHERS, Inc., 250 Fifth Avenue, New York 1, N. Y. For Great Britain and Northern Ireland: INTERSCIENCE PUBLISHERS Ltd., 88/90 Chancery Lane, London W. C. 2 PRINTED IN THE UNITED STATES OF AMERICA BY MACK PRINTING COMPANY, EASTON, PA. PREFACE to Volume II, Part 2 This volume brings to completion the review of Photosynthesis and Re- lated Phenomena which the author rashly undertook to prepare, for his own orientation, in the summer of 1938. Two months at Woods Hole, with its splendid hbrary, seemed, at that time, an adequate period to complete the undertaking; and Interscience Publishers temptingly offered to publish the result in book form. Seventeen years and 2000 printed pages later, it is time to stop — even if this closure has to come in the midst of rapid and promising developments in several areas covered by the narrative. A mistake, which proved almost fatal to the completion of the task, was to postpone the publication of the whole manuscript in 1943, when it Avas first finished in draft form, at a time when the war had imposed a hia- tus upon the progress of non-applied research. Instead, only one half of the manuscript was prepared for publication at that time (and actually published, as Volume I, in 1945) ; while the second half was set aside for final revision until after the war. When, in 1946, the author returned from the Manhattan District to the Solar Energy Research Project at the Massa- chusetts Institute of Technology (to move on, soon afterward, to the Photo- synthesis Laboratory at the University of Illinois), the rapidity with which new research data began to accumulate made it difficult to digest and fit them into the framework of the monograph. As a result, the "second volume" in turn burst its confines, and became divided in two. The first half (Volume II, Part 1) was published in 1951; completion of the second has taken five more years. Chapters 31 to 34 of the present volume (II, Part 2) follow the original outline, bringing to a close the discussion of Kinetics of Photosynthesis begun in Chapter 25. Chapters 35 and 36 deal with two areas of knowledge that have been vastly enlarged since 1945 — the Photochemistry of Chloro- phyll (in solution and in chloroplast preparations), and Chemical Path of Carbon Dioxide Reduction. Chapter 37 was originally intended as a catch-all for new information in the various areas covered by Volumes I and 11,1 ; but in the course of preparation, division into four parts became advisable: 37A (Structure and Composition of Chloroplasts) , 37B (Chem- istry of Pigments), 37C (Spectroscopy and Fluorescence of Pigments), and 37D (Kinetics of Photosynthesis), the latter bringing up-to-date, VI - PREFACE but not to a close, among other things, the controversial subject of maximum efficiency. This arrangement left out many not unimportant "miscellaneous" topics, such as the more recent work on photosynthesizing and chemo- synthesizing bacteria, photochemical nitrate metabolism, etc. The mono- graph as a whole therefore does not quite live up to the ambitious stand- ard of coverage set in the first volume. No attempt has been made to treat the problems of large-scale culturing of microscopic algae for food (and other practical purposes), to which so much public attention has been drawn lately. Reference must be made in this connection to the symposium "Algal Culture, from Laboratory to Pilot Plant," edited by J. S. Burlew, and published by the Carnegie Insti- tution of Washington in 1953 (Publication No. 600), which contains con- tributions by leading investigators in this field. ^ Each of the three volumes — 1, 11,1 and 11,2 — contains an Author Index of the Main Investigations. A Subject Index was provided in Volume I, but not in 11,1 ; the present Volume 11,2 contains a comprehensive Subject Index for all three volumes. If the writing of this monograph were to be started now, a somewhat different plan would have been adopted, with a different distribution of emphasis. Some chapters now resemble a scaffolding, erected years ago, behind which no building has appeared, and which one would now be in- clined— perhaps too hastily — to dismantle. There is, however, relatively little material, even in the earliest chapters, which does not contribute to the establishing of a proper perspective in the whole field. The poUcy — adopted at the beginning of the work — of discussing all alternatives and suggestions, has paid off in leaving the author free of commitment to any by now flagrantly obsolete theory. In fact, systematic reading— which is more than can be expected for a monograph of this bulk — would re- veal that many of the current "new" ideas in photosynthesis have been proposed or discussed somewhere in it. At the end of the book (p. 1994) the author expresses hope that it will not rapidly become obsolete, even if the rate at which new developments follow in the field makes it inevitable that it will be incomplete already at the date of its publication. Some of the developments of the years 1954-55 could at least be mentioned in the "Epilogue" (Chapter 38); others were published (or came to the author's attention) too late for this purpose. The electron microscopic study of the chloroplast structure has entered a promising new stage with the improvement of techniques for the preserva- 1 More recent summaries, presented at the Phoenix, Arizona, Conference on Applied Solar Energy (Oct. 30-Nov. 4, 1955) will be published in the proceedings of this con- ference. PREFACE Vll tion of the intimate structure of the specimens (some striking pictures obtained with this technique are reproduced in Chapter 38). Leyon,2 ^nd Vatter/^ have reported interesting new data concerning the ontogenetic development of the chloroplast structure. Important new observations were made concerning the photochemical activity of chloroplast preparations. Thomas and Haans"* thought that the loss of much or all of the chloroplast stroma in the preparation of granular chloroplast fragments from leaves may be responsible for their incapacity to utilize carbon dioxide as photochemical oxidant. They accordingly pre- pared macerates from stroma-free, laminar chloroplasts of Spirogyra, and found that these preparations could take up manometrically measurable amounts of carbon dioxide and liberate roughly equivalent amounts of oxygen in light. Arnon and co-workers^ expanded, in a series of papers, the interesting findings to which brief reference only could be made on pp. 1537 and 1982. They found that whole chloroplasts, separated from protoplasmic particles (mitochondria), as well as fragments of such protoplasma-free chloroplasts, have very little aerobic metabolism — respiration and correlated ATP- formation from morganic phosphate. On the other hand, they show evi- dence of two photosynthetic processes — the uptake of CO2 (revealed by tracer measurements), with C(14) incorporation in sugar phosphates and carbohydrates (starch) and liberation of oxygen (measured manometrically, after qualitative gas identification by luminous bacteria) ; and of anaerobic formation of high-energy phosphate (''photosynthetic phosphorylation"). The photochemical processes can be directed preponderantly towards photosynthesis, or towards ATP accumulation, by varying external condi- tions or adding appropriate inhibitors. In chloroplast fragments (as contrasted to whole chloroplasts) the "photosynthetic" CO2 fixation could be observed only upon the addition of an aqueous extract from chloroplast maceration. The anaerobic forma- tion of ATP in light could be stimulated, in such fragments, by "co-factors," which included Mg++-ions, vitamin-K type compounds {e.g., menandione), riboflavin, and ascorbic acid, until it was much stronger than in intact chloroplasts. The photochemical C02-incorporation was not stimulated (or even mhibited) by these additions, with the exception of ascorbic acid. The apparent necessity of the latter (in amounts much larger than that of 2 Leyon, H., Exptl. Cell Reserch, 7, 609 (1954). 3 Vatter, A. E., Thesis, Univ. of Illinois, 1955. * Thomas, J. B., and Haans, A. M. J., Biochim. et Biophys. Acta, 18, 287 (1955). 6 Arnon, D. I., Whatley, F. R., and Allen, M. B., J. Am. Chem. Soc, 76, 6324 (1954); Allen, M. B., Arnon, D. I., Capindale, F. R., Whatley, F. R., and Durham, L. J., ibid., 77, 4149 (1955); Arnon, D. I., Science, 122, 3157 (1955); Arnon, D. I., et al, Gatlinburg Conference on Photosynthesis of the NAS, Oct. 1955 (in press). Viii PREFACE carbonate) appears as a complication in the interpretation of these observa- tions, particularly if one recalls the complex action of ascorbic acid on the Hill reaction (c/., p. 1568) and its role in the Krasnovsky reaction (cf., pp. 1514-1522) . The difference between the C(14)-tagged intermediates identi- fied in Arnon's photosynthesizing chloroplast preparations and those found by Calvin and co-workers in intact photosynthesizing cells also needs clari- fication. In any case, the observations of Arnon, and of Thomas, bring nearer to closure the gap between the photochemical activities of chloroplast prepara- tions and whole cells, and between the Hill reaction and true photosynthesis (first narrowed by the coupling of Hill reaction to the malic enzyme system via pyridine nucleotides, described on pp. 1578-1585). Vishniac^ reported that the colorless "acetone powder," prepared from chloroplasts, can be restored to photochemical activity (reduction of TPN in light) by addition of chlorophyll solution. Similar activation of macer- ates from white leaves by added chlorophyll was described by Rodrigo^; the material obtained in this way was weakly fluorescent and had an ab- sorption maximum at 678 m/i. Kok^ has made new flash light experiments, and suggested a simple two- step enzymatic mechanism, permitting the observations made with instan- taneous flashes (Emerson and Arnold) to be reconciled with those ob- tained by means of flashes lasting several milliseconds (Tamiya). This mechanism, if correct, means that of the two constants previously derived from flash light experiments, one, F™^^ (- 5 X 10 -^^ Chlo), retains its significance as titer (multiplied by a simple fraction, 1/n) of a "stabilizing" enzyme; but that the other, designated in Chapter 34 as fc^A (50 sec."^) loses its significance as rate constant of the same enzyme, and becomes a function of the rate constants of the two enzymes, catalyzing the two reac- tion steps; under certain conditions, it may be practically equal to the rate constant of the second of them. The course of the reaction decay in the dark period is, in this case, more complex than that expected for a simple first order reaction— in agreement with the flash light findings of Gilmour, and Kok, and with the chemiluminescence data of Arnold and Strehler. New contribution to the minimum quantum requirement -problem were published by Yuan, Evans, and Daniels,'^ and by Bassham, Shibata, and Calvin'"; the first-named study supported the view that this requirement 8 Vishniac, W., Gatlinburg Conference on Photosynthesis of the NAS, Oct. 1955 (in press). 7 Rodrigo, T. A., Thesis, Univ. of Utrecht, 1955. « Kok, B., Gatlinburg Conference on Photosynthesis of the NAS, Oct. 1955 (in press). 9 Yuan, E. L., Evans, R. W., and Daniels, F., Biochim. et Biophys. Acta, 17, 185 (1955). "> Bassham, J. A., Shibata, K., and Calvin, M., ibid., 17, 332 (1955). PREFACE IX is close to 8, the second one suggested that it may be 6.5 or 7.0 in strong light, and as low as about 4 in weak light (as suggested earlier by Kok). Significant new observations of the reversible changes in absorption svec- trum and fluorescence of chlorophi/Il in vivo have been described by Witt,^^ as well as by Chance and co-workers, '^ by Strehler and Lynch,!^ ^^^ by Coleman, Holt, and Rabinowitch.'" The latter suggested that in Chlorella, chlorophyll undergoes, in light, a reversible change (probably reduction), leading to a bleaching in the red and the appearance of a band at 525 mn. However, discrepancies between the observations of Coleman et al, on the one hand, and of Duysens (and Strehler and Lynch), on the other, call for further study, especially in the red region of the spectrum. The chemical reduction -path of carbon dioxide, elaborated by Calvin, Benson, and co-workers on the basis of C(14) studies (c/., Chapter 36) has received a suggestive confirmation in the demonstration by Racker'^ that reduced pyridine nucleotide, adenosine triphosphate, and bicarbonate will in fact produce sugar if provided with the eleven enzymes postulated in Calvin and Benson's cycle. Franck'6 made the interesting suggestion that the transfer of hydrogen from hydrated chlorophyll to an acceptor, such as the carboxyl group in PGA, is achieved by the intervention of an excited triplet state of the pig- ment, formed by cooperation of two photons— one producing the meta- stable triplet state, and the other (supplied by resonance transfer from near-by chlorophyll molecules) transferring the metastable molecule into the excited state. Obviously, this mechanism requires a minimum of 8 photons per reduced COo molecule. The existence of a triplet metastable state of chlorophyll b has been confirmed, and its energy found equal to about 33 Kcal./mole, by phosphorescence studies of Becker and Kasha." This is only a short (and arbitrary) selection of important new facts and speculations added to the field under review within a few months after this monograph was "completed!" The author's thanks are due to Dr. Robert Emerson, the Photosynthesis "Witt, M. T., Natunvissenschaften, 42, 72 (195,5); Z. physik. Chem., 4, 1920 (1955); Z. Elektrochem., 59, 981 (1955); Gatlinhurg Conference on Photosynthesis of the NAS, Oct. 1955 (in press). 12 Chance, B., Gatlinburg Conference on Photosynthesis of the NAS, Oct. 1955 (in press). '3 Strehler, B. L., and Lynch, V. M., Science, 123, 462 (1956). " Coleman, J., Holt, A. S., and Rabinowitch, E., Second GatHnburg Conference on Photosynthesis of the NRC, Oct. 1955 (to be pubUshed); Science (in press). 15 Racker, E., Nature, 175, 249 (1955). 16 Franck, J., Daedalus, 86, 17 (1955) (Rumford Medal Lecture, American Academy of Arts and Sciences in Boston). " Becker, R. S., and Kasha, M., in The Luminescence of Biological Systems, Prince- ton Univ. Press, Washington, 1955, p. 25. X PREFACE Laboratory, and the Botany Department of the University of IlHnois for the patience with which they have allowed him to use their facilities for the completion of this work. The author owes much to discussion with his co-workers, Dr. A. Stanley Holt and Dr. E. E. Jacobs. Mr. Paul Latimer, Mr. John Coleman, and Dr. Sylvia Frank have helped in the preparation of the index. Miss Natalie Davis kindl}^ supplied one of the drawings. Miss Carolyn Prouty has helped to minimize, in this volume, the vexa- tious errors in bibliography, much too many of which have been permitted to go undiscovered in Volume I : she and Mrs. Ruth Adams have also helped with proofreading. Interscience Publishers have shown infinite patience and forbearance with the author's unreliability, procrastination, and tendency to enlarge and change the text up to the very last moment. The author's warm thanks are due Dr. Eric S. Proskauer and his staff. The author early acquired a prejudice against dedicating scientific treat- ises to parents, teachers, or wives; it somehow seemed to him that such homage should be reserved to more personal works of art, and those who inspired them. This prejudice has prevented him from expressing his thanks to the one man to whom he owes both his interest in photochemistry and photobiology, and whatever special qualifications he may have to deal with these subjects, in the most natural and adequate way — by dedicating this monograph to him. The author has had the privilege of studying or working with several great scientists of our time; but Dr. James Franck is the one of whom he likes to consider himself a pupil — not only in the nar- rower field of common scientific specialization, but in the whole approach to the world of atoms and molecules. While the author has not been able to match the persistence, concentration, and clarity of thinking that have made James Franck one of the great pathfinders in this enchanted world (not to speak of acquiring his humility and deep understanding of the world of men), he can plead that these have been among the strongest influences he has experienced, and guiding lights he has tried to follow. I hope Dr. Franck will accept these words of gratitude at the end of the long work, in lieu of a dedication which was due to him on its first page. Eugene I. Rabinowitch Urhana April 1956 CONTENTS Page Preface . V PART FOUR— KINETICS OF PHOTOSYNTHESIS (continued) Chap. 31. The Temperature Factor 121 1 A. Internal Temperature of Plants 1211 B. Temperature Range of Photosynthesis 1217 1. Lower Temperature Limit and Chill Injury 1218 2. Optimum and Upper Temperature Limits; Heat Injury 1220 3. Thermal Adaptation 1225 C. Temperature Coefficient and Heat of Activation of Photosynthesis . 1231 1. Absence of Temperature Influence in Weak Light 1232 2. Temperature Effect in Strong Light 1235 3. Temperature Curves in the Carbon Dioxide-Limited State. . . . 1242 4. Temperature Curves with Several Optima 1243 5. Temperature Effect on Fluorescence 1245 6. Theoretical Remarks 1247 Bibliography 1254 Chap. 32. The Pigment Factor 1258 1. Relation between Light Absorption and Pigment Content 1258 2. Influence of Natural Variations of Chlorophyll Content on Photosynthesis 1261 3. Influence of Artificial Changes of Chlorophyll Content on Pho- tosynthesis 1267 4. Chlorophyll Concentration and Yield of Photosynthesis in Flashing Light 1272 5. Energy Migration and the Hypothesis of the Photosynthetic Unit 1280 6. Energy Transfer between Different Pigment Molecules 1299 Bibliography I'^l" Chap. 33. Time Effects. I. Induction Phenomena 1313 A. Gas Exchange during the Induction Period 1314 1. "Long" and "Short" Induction 1314 2. Oxygen Exchange during the Short Induction Period 1317 3. Carbon Dioxide Exchange during the Short Induction Period. . 1334 4. Induction after Change to Lower Light Intensity 1355 5. Gas Exchange during the Long Induction Period 1361 6. Influence of Anaerobiosis on Induction 1364 7. Induction after Photoxidation 1371 XI 7113 xii CONTENTS Chap. 33, contd. PagE 8. Induction in the Photoreduction by Algae 1374 9. Induction in Hill Reaction 1375 B. Fluorescence and Absorption Changes during the Induction Period 1375 1. Fluorescence Induction Phenomena in Leaves, Algae, Chloro- plasts and Chlorophyll Solutions 1375 2. Influence of Different Factors on Fluorescence-Time Curves .. . 1382 (a) Duration of Incubation 1382 (b) Light Intensity 1386 (c) Temperature 1390 (d) Carbon Dioxide Concentration 1391 (e) Poisons 1395 (f ) Oxygen; Effects of Anaerobiosis 1399 (g) Oxidants and Neutralizers 1404 3. Changes in Absorption Spectrum during Induction 1406 C. Interpretation of Induction Phenomena 1407 1. Diffusion and Buffer Effects 1407 2. Building Up of Intermediates 1408 (a) Photochemical Intermediates 1409 (b) Thermal Intermediates 1412 3. Role of the Carbon Dioxide Acceptor in Induction 1413 4. Inactivation of the Catalytic System as the Primary Cause of Induction. Gaffron-Franck Theory of Induction 1419 5. Role of Respiration in Induction Phenomena 1425 6. Other Theories of Induction 1428 Bibliography 1429 Chap. 34. Time Effects. II. Photosynthesis in Intermittent Light 1433 A. Alternating Light 1435 1. Yield of Photosynthesis in Relation to Frequency of Alterna- tions 1435 2. Theoretical Discussion of the Effect of Alternating Light 1441 B. Flashing Light 1447 1. Intermittency Factor in Flashing Light 1447 2. Maximum Flash Yield 1452 3. Effect of Cyanide on Photosynthesis in Flashing Light 1455 4. Influence of Temperature, Carbon Dioxide Concentration, Nar- cotics and Ultraviolet Light on Flash Yield 1460 5. Flash Yield in Heavy Water 1465 6. Flashing Light Experiments Calling for Revision or Supple- mentation of the Emerson-Arnold-Franck Mechanism 1467 7. Hill Reaction in Flashing Light 1478 8. Flashing-Light Experiments with Bacteria and Hydrogen- Adapted Algae 1481 Bibliography 1483 PART FIVE— ADDENDA TO VOLUME I AND VOLUME II, PART 1 Chap. 35. Photochemistry of Chlorophyll in vitro and in vivo 1487 A. Photochemistry of Chlorophyll in Solution 1487 1. Bleaching of Chlorophyll in Methanol 1487 CONTENTS Xl^l Page Chap. 35, contd. 2. Photoxidation of Chlorophyll 1499 3 Reversible Photoxidation of Bacteriochlorophyll 1501 4'. Reversible Photoreduction of Chlorophyll and Its Derivatives. 1501 5. Chlorophyll-Sensitized Oxidation-Reductions 1507 6. Chlorophyll-Sensitized Autoxidations 1525 B Photochemistry of Chloroplast Preparations 1528 1. Can Carbon Dioxide Serve as Oxidant in Hill Reaction? 1530 2 Chloroplast Preparations from Different Plants • • • 153/ 3. Preparation, Preservation and Activation of Chloroplast Ma- , 1 IOt:^ (a) Preparation of Chloroplast Suspensions 1542 (b) Loss of Activity. Stabilization J543 (c) Activation by Anions ^ (d) Fractionation of Chloroplast Material 1554 4. Different Oxidants (a) Ferric Salts (b) Other Inorganic Oxidants J^^^ (c) Oxygen as Hill Oxidant J^°^ (d) Quinones ^^^2 (e) Organic Dyestuffs ; • (f) Respiration Intermediates and Other Cellular Materials. . . 157b (g) Miscellaneous Organic Compounds J 589 5. Kinetics __ Methods for Measuring Reduction i^»^ (b) Rate and Yield under Different Conditions 1592 (c) Dark Reactions (d) Inhibition and Stimulation J^ (e) Survival of Photochemical Reductant in the Dark Ibl5 , , . ... 1616 6. Mutations C. Photochemistry of Live Cells „., ,. , lozo Bibliography Chap. 36. Chemical Path of Carbon Dioxide Reduction 1630 1 AQn A. Isotopic Carbon Tracer Studies ■■■^■■- 1 Photosynthetic and Respiratory Fixation of C O, lo-i" 2. C*02 Fixation in Darkness with and without Preillumination. A Surviving Reductant? _ 3. C*02 Fixation in Light: Phosphoglyceric Acid as First Inter- ^^^^ mediate - „ , „ 4 The Experiments of the Chicago Group -■ J^^o 5 Early Intermediates Other than PGA; Paper Chromatography . 1655 6. The Role of Malic Acid: One or Two Carboxylations in Photo- ^^^^ synthesis? , ' '. .,^„^ 7 The C5 and C7 Sugars as Intermediates in Photosynthesis ib/u ^ Ti i. 1675 8. The C2 Fragment ^^^^^ 9. Kinetic Studies 10. The Sequence of Sugars 11. Effect of Poisons and pH on CO2 Fixation J^so 12. Evolution of the CO2 Reduction Mechanism l^»« xiv CONTENTS Chap. S6, contd. PagE 13. The Lipoic Acid Hypothesis 1698 14. Tracer Studies of Special Forms of Photosynthesis 1701 B. Photosynthesis and Phosphate Metabolism 1702 C. Nitrate Metabolism and Photosynthesis 1709 Bibliography 1710 Chap. 37. Miscellaneous Additions to Volumes I and II. 1 1714 A. Structure and Composition of Chloroplasts, Chromoplasts and the Chromatoplasm 1714 1. Light Microscopy 1714 2. Electron Microscopy 1718 (a) Structure 1718 (b) Shape, Dimensions and Composition 1733 (c) Location of Chlorophyll 1736 3. Ultracentrifuge Study 1740 4. Optical Evidence of Chloroplast Structure 1741 5. Composition of Chloroplasts 1742 (a) Proteins, Lipids and Nucleotides 1742 (b) Enzymes 1744 (c) Heavy Metals 1749 (d) Ascorbic Acid 1749 6. State of Pigments in Chloroplasts 1750 (a) Chloroplastin 1750 (b) Crystallized Liproprotein-Chlorophyll 1751 (c) Is All Chlorophyll in the Cell in the Same State? 1752 (d) Location of Accessory Pigments 1754 Bibliography 1755 B. Chemistry of Chloroplast Pigments (Excluding Photochemistry). 1759 1. Biosynthesis of Chlorophyll; The Protochlorophyll 1759 2. Biogenesis of Chlorophyll; Earlier Precursors 1768 3. Isomers and Solvates of Chlorophyll 1771 4. Oxidation, Allomerization and Reduction of Chlorophyll 1773 5. Crystallization and Stability of Chlorophyll and Bacteriochloro- phyll 1782 6. Nature of Chlorophyll c 1786 7. Chemistry of Bacteriochlorophyll and Bacterioviridin 1786 8. Molecular Structure and Properties of Phycobilins 1788 Bibliography 1790 C. Spectroscopy and Fluorescence of Pigments 1793 1. Theory of Chlorophyll Spectrum 1793 2. New Measurements of Absorption Spectra of Pigments in Solu- tion 1799 (a) Chlorophyll, Chlorophyllides and Pheophorbides 1799 (b) Bacteriochlorophyll and Derivatives 1806 (c) Protochlorophyll and Other Porphin Derivatives 1808 (d) Spectra of Accessory Pigments 1809 (e) Infrared Spectra of Chlorophyll and Its Derivatives 1811 3. Spectra of Crystalline and Colloidal Chlorophyll Derivatives. . 1815 4. Fluorescence of Chlorophyll in vitro 1827 CONTENTS XV Chap. 37, conld. PaGE (a) Quantum Yield 1828 (b) Polarization 1830 (c) Quenching 1831 (d) Activation I835 (e) Sensitization I835 (f) Fluorescence of Colloidal Chlorophyll Solutions 1837 (g) Fluorescence of Bacteriochlorophyll and Protochlorophyll . 1837 (h) Fluorescence of Phycobilins 1838 5. Chemiluminescence of Chlorophyll in vitro and in vivo 1838 6. Light Absorption bj^ Pigments in vivo 1841 (a) Absorption Spectra of Leaves and Algae 1841 (b) Two Forms of Chlorophyll in vivol 1847 (c) Absorption Spectra of Purple and Green Bacteria 1849 (d) Phycobilin Spectra in vivo 1856 (e) Changes in Absorption Spectrum during Photosynthesis .. . 1856 (f) Calculation of True Absorption Spectra from Transmission and Fluorescence Spectra of Suspensions 1863 6A. Scattering by Pigment Bodies in vivo 1866 7. Fluorescence of Pigments in vivo 1867 (a) Absolute Yield 1867 (b) Sensitized Fluorescence in vivo 1868 (c) Re-absorption of Fluorescence 1879 (d) Fluorescence of Protochlorophyll in vivo 1881 (e) Fluorescence of Chlorophyll c in Diatoms 1882 (f) Changes of Fluorescence Intensity Related to Photosyn- thesis 1882 Bibliography 1882 D. Kinetics of Photosynthesis 1886 1. The Carbon Dioxide Factor 1886 (a) Utilization of Bicarbonate 1886 (b) Carbon Dioxide Curves 1892 (c) Carbon Dioxide Compensation Point 1898 (d) Carbon Dioxide Supply through Roots 1901 (e) Isotopic Discrimination 1901 2. Other Chemical and Physical Factors 1902 (a) Catalyst Poisons 1902 (b) Oxygen 1912 (c) Water Factor 1918 (d) Inorganic Ions 1919 (e) Various Organic Compounds 1921 (f) Ultraviolet Light 1921 3. Photosynthesis and Respiration 1925 (a) C(14) Studies 1926 (b) 0(18) Studies 1929 (c) Polarography 1936 4. Maximum Quantum Yield 1940 (a) Experiments by Warburg and Co-workers since 1950 1941 (b) Other New Manometric Measurements 1956 (c) New Nonmanometric Measurements of Quantum Yield. . . 1956 XVI CONTENTS Chap. S7, contd. PaGE (d) Franck's Interpretation of Warburg's 1950-1952 Experi- ments 1962 (e) Quantum Yield of Hill Reaction 1967 (f ) Quantum Requirement of Bacteria 1968 5. Thermochemical Considerations 1969 Bibliography 1975 Chap. 38. Epilogue 1979 Author Index of the Main Investigations in Vol. II, Part 2 1995 Subject Index to Vols. I, II-l, and II-2 2022 PART FOUR KINETICS OF PHOTOSYNTHESIS (continued) Chapter 31 THE TEMPERATURE FACTOR Photosynthesis is a sequence of photic and catalytic chemical reactions, combined with physical processes of diffusion and convection. The pri- mary photochemical reaction probably is independent of temperature; but all other partial processes of photosynthesis, physical as well as chemi- cal, must be influenced by it. The adsorption and hydration equilibria, which affect the colloidal state of the protoplasm and of the chloroplasts and thus, indirectly, alter the efficiency of the photosynthetic apparatus, also are sensitive to heat or cold. Therefore, the influence of temperature on photosynthesis is a complex phenomenon. The rate of photosynthesis can be expected to be insensitive to temperature changes — at least within a certain range — only in the "light-hmited" state, in which the velocity of the over-all process is equal to that of the primary photochemical reaction. Under all other conditions, the rate of photosynthesis will change with temperature, and the character of this change will depend on what factor exercises the strongest influence on the rate under the specific condi- tions of the experiment. It was mentioned on page 1137 that the maximum experimentally realizable quan- tum yield of photosynthesis may be smaller than the theoretical photochemical quantum yield because a certain fraction of the products may be lost by back reactions, independ- ently of light intensity. This proportion, and thus the maximum quantum yield, could depend on temperature. In other words, photosynthesis could be somewhat dependent on temperature even in the light-limited state. Before considering the "temperature curves" of photosynthesis {i. e., curves in which rate is plotted against temperature, at constant light in- tensity and carbon dioxide concentration) , we will first discuss the differ- ence between the internal temperature of plants and the temperature of the surrounding medium. Obviously, this internal temperature, rather than the temperature of the medium, should be used as the independent vari- able in the analysis of the temperature curves. A. Internal Temperature of Plants* The internal temperature of plants can be higher or lower than that of the ambient medium. It may be different in different parts of the * Bibliography, page 1254. 1211 1212 THE TEMPERATURE FACTOR CHAP. 31 same plant, e. g., the chloroplasts, which absorb Ught, may be warmer than the epidermis of the leaf, which is cooled by evaporation. Fifty years ago, Timiriazev (1903) thought that local heating of chloroplasts by light absorption may be sufficient for thermodynamic reversal of combustion processes, and that ttiis may explain photosynthesis. However, it has since become clear that the mechanism of activation of chemical reactions by light is different from that of activa- tion by heat. The former is based on the formation of electronically excited molecules, or of free radicals, wliile the latter depends on the production of molecules with liigh kinetic energies. Unlike warm-blooded animals, plants have no mechanism for automatic regulation of oxidation and transpiration processes, capal)le of maintaining the organism at a constant temperature. However, they, too, continuously produce heat by exothermal metabolic processes, and a plant enclosed in a dark vessel full of saturated water vapor (to prevent both transpiration and photosynthesis) warms itself up until the internal evolution of heat is com- pensated by increased conduction losses to the gas and by thermal radia- tion losses to the walls. If plants are allowed to absorb hght, and to transpire, the above three items of energy exchange are augmented by two additional terms: energy supply by light absorption and energy losses by transpiration. A certain fraction of the former is lost, from the point of view of heat balance, by conversion into chemical energy (photosynthesis). Brown and Escombe (1905) and Brown and Wilson (1905) made the first attempt to calculate the stationary temperature of plants by esti- mating all these energy terms. They concluded that the stationary tem- perature of ordinary leaves, enclosed in a dark space saturated with water vapor, can rise only a few hundredths of a degree above the temperature of the medium. The difference may be somewhat larger in plant organs in which the volume/surface ratio is small, such as fruits, stalks and succulent leaves. In strongly illuminated leaves, the conversion of light into heat is by far the largest item on the credit side of the heat balance, this physical process outweighing by far the chemical heat production by respiration and other exothermal metabolic processes. On the debit side, too, the two physical energy-consuming processes, transpiration and heat transfer (the latter we consider to include thermal conduction, convection and radiation losses), account for a much larger amount of energy than the chemical process of photosynthesis. Photosynthesis is most important in weak light (of the order of 1000 lux), where it may consume 30% or more of absorbed Ught en- ergy; but in direct sunlight this proportion does not exceed 5% (c/. chapter 28). Consequently, the internal temperature of leaves exposed to the sun can be attributed (in the first approximation) to the balance of three physical processes: light absorption, transpiration and heat transfer. INTERNAL TEMPERATURE OF PLANTS 1213 A controversy has arisen as to the relative importance of the two last- named processes. Brown and Escombe (1905) considered transpiration I)y far the most important of the heat-dissipating processes; and this view was supported by the experiments of Shull(1919), Eaton and Belden (1929), Arthur and Stewart (1933) and Clements (1934). Eaton and Beldon (1929) pointed out that only transpiration can account for the fact that sun-exposed leaves often are cooler than the air. Smith (1909) and Clum (1926), on the other hand, came to the con- clusion that transpiration has a relatively small effect in preventing an ex- cessive rise of temperature in strongly illuminated leaves. According to Clum, the leaves in which transpiration is prevented by a layer of vaseline are, in direct sunlight, only 2 or 3° C. warmer than similar, freely trans- piring leaves. He therefore suggested that heat transfer is the main factor responsible for keeping the temperature of sun-exposed leaves within narrow, comfortable limits. Watson (1933, 1934) pointed out that even if it were true that heat transfer normally dissipates less energy than transpiration, its relative importance must increase with increasing dilTer- ence in temperature between leaf and air. The low estimate of the energy loss by heat transfer made by Brown and Escombe might have been due, at least in part, to the neglect of infrared reradiation; an item, the impor- tance of which was pointed out in particular by Curtis (1936 ^■2). The concept of transpiration as the all-important heat-dissipating and temperature-regulating process in green leaves, which was for a time gener- ally accepted in plant physiology, needs correction. The combined heat- dissipating effect of convection currents and thermal radiation may some time equal or even exceed that of transpiration. The relative importance of the several heat-dissipating processes depends on the structure of the leaf, and on atmospheric conditions, such as wind, exposure to cold surfaces and the humidity of the air. Because of these special conditions (and also because of the use of insufficiently reliable experimental methods), the data given in the literature for the temperature of illuminated leaves vary widely. Table 31.1 contains the most important results. A discussion of the experimental methods can be found, e. g., in Miller's Plant Physiology (1931) and in an article by Seybold and Brambring (1933). The earliest investigations were carried out by means of mercury thermometers with bulbs pressed against, or wrapped into, leaves. Later, small thermoelements were substituted for thermometers; some authors used them to measure the surface tempera- ture of the leaves, whereas others attempted to introduce them into the leaves to deter- mine the internal temperature of the latter. Table 31.1 shows that Shreve (1919), Miller and Saunders (1923), Eaton and Belden (1929), and Seybold and Brambring (1933) found only 1214 THE TEMPERATURE FACTOR CHAP. 31 a; OS 42 > o 00 P T3 O J3 o 0(N + + o o + + -^eO rH ' 00 + + +0 + O o O lO o -t-' -IJ -u . -1^ + D, a, S S 3 ID c3 CO CO O c |||l.S+J.S|| 03 o s3 P ^ b. t; > o 0) a) ""^ 3 3 eo S w e CO ^ -tj'-:j w 3 ^ O r— I ^H *■ — ■' ^— ^ >> bC to c o3 3 C ^ O a _tij to to Lh <^ 3 3 e 3 O -fe a cc ■^ -c e 3 o3 IE -t-3 s.. 3 -2 O 3 ^ o o s 3 W •73 t*-. tJ-H rf o3 a m "a 'a, a 3 3 3 0 0 0 0 0 0 0 0 0 a a a ;-( u Sh D 0 m J3 -3 ^ H H H ^_^ »o 0 03 1-H • i-f lU o3 rd -tJ -!-:> 03 § 03 >o a" 0 o> t— 1 03 1—1 1 1 -« * • ^— ^ ^ 0 -1-3 ^ 03 a 1— < s M w O -1-2 o c3 I a a 3 p. >a as 3 £ -3 P o ° ^ O a".2^ a> gi ^ gi £^ 3, ■" '& °3 3'+- 3 <*- O w O 03 o « o o o— ' o~ gx: 3-3 ^ H CO (N 03 l-H ro I-, a 3 o INTERNAL TEMPERATURE OF PLANTS 1215 CO I -M o o o & 3 O -r< CO + + 00 t> + + CO o coco ++ Q0O5 + + COC^' ++ d -50 S 00 > m -5 o ^ 0) o o o :2:2-^:2:2 » 4J 03 „ o 3tr:,a^[2 ^ CO ^- 03 -i OJ 03 s p. o += += 03 00 c fl ■3 '3 hC bC o3 ^ 03 03 ^ 03 'a 0 ao 3 0 o3 -t-H §^ 0 0 3 OQ " 3 0 ffi a ■+H a^ t-. 03 t-. o3 03 03 03 03 JU ""* si-^ H CO CO 05 •—1 ,. V V, ' bC ^H 'G ^— • ^ 1=1 a 03 o3 s t^ "oS pq M 2" fl "o 0 Xi -1-= CO a 03 03 03 -D C O 'So 03 03 G 03 a 03 'S a O +3 C O • <-( O a o a 03 O c 03 03 03 fl o3 43 a 03 03 c a w c c3 O . 03 .i3 bca go Sh 03 O O 1216 THE TEMPERATURE FACTOR CHAP. 31 comparatively small temperature differences between leaf and air. Ac- cording to Miller and Saunders, the average surface temperature of leaves in the field is practically equal to that of the air. (In direct sunlight, the surface temperature of the leaf often was, in their experiments, 1 or 2° above that of the air; but, when the sun was covered by a passing cloud, transpiration caused the leaf temperature to drop immediately to 1 or 2° below that of the surrounding atmosphere. Averaging over all atmospheric conditions gave a mean value of temperature difference close to zero.) Eaton and Belden (1929) found that turgid leaves of cotton plants had a temperature lower than that of the air during the greater part of the hot and dry summer day; only on mornings and evenings was this relation reversed. Wilted cotton leaves, on the other hand, showed positive tem- perature differences during the whole day. (The authors considered these results evidence of the strong influence of transpiration.) In contrast to these examples of negligible or even negative tempera- ture differences between illuminated leaves and air, others, from Asken- asy (1875) to Clum (1926), have observed in direct sunlight, internal tem- peratures 10°, 15° or even 20° C. above the temperature of the ambient air. The fact that Miller and Saunders conducted their experiments with much thinner leaves than Blackman and Matthaei, and Clum may explain some of the discrepancies. Askenasy (1875) first suggested that thick suc- culent leaves are particularly hable to overheating in the sun. Table 31.1 shows that this conclusion was confirmed by Ursprung (1903), but not by Smith (1909). The figures of Seybold and Brambring in this table (c/. also fig. 31.1) also show no significant differences among the temperatures of xeromorphic, hygromorphic and succulent leaves. This is not as strange as it may appear; because of the lower ratio of surface to volume, succulents undoubtedly are handicapped in the dissipation of heat; but, since Hght absorption is proportional to the area rather than to the volume (while heat capacity is proportional to the volume) of the leaf, the heating by light absorption also is much slower in the case of succulents than in that of thin leaves (c/. Bnjophylhmi curve in fig. 31.1). (Conditions are dif- ferent in the dark, since heat production by respiration is approximately proportional to the volume of the tissue.) The difference between leaf temperature and air temperature may be of paramount importance for the photosynthetic activity of evergreens in winter. According to Ehlers' data in Table 31.1, the needles of conifers may have, during a winter day, an internal temperature 7 or 8° C. above that of the air. The temperature of alpine plants may rise even higher, because of the intense radiation to which they are exposed, even while the temperature of the air may be quite low. The danger of overheating is particularly great in experiments with TEMPERATURE RANGE OF PHOTOSYNTHESIS 1217 leaves exposed to artificial light of high intensity, because this light may contain ten times more infrared rays than sunlight of equal visual bright- ness. Although leaves are comparatively transparent above 800 m/x {cf. chapter 22), they absorb enough in this region to affect their heat balance seriously. In the experiments of Seybold and Brambring, illuminated white leaves, which absorb only infrared light, acquired temperatures not much lower than those reached by green leaves (which absorbed both infra- red and visible hght). For example, the temperature difference was -1-2.7° C. for the upper surface of a white Pelargonium leaf, and -}-3.0° for the upper surface of a green leaf of the same species (cf. fig. 31.1) ; the corre- sponding figures for Abutilon were +2.3° and ^-3.7° C, respectively. I 2 3 4 5 6 TIME, min. Fig. 31.1. Changes in leaf temperature upon illumination and darkening (after Seybold and Brambring 1933): R, Rhododendron hybridum; P«, Perlargonium zonule, green leaf; Pio, same, white leaf; B, Bryophyllum, succulent leaf. Karmanov (1951) found that leaves exposed to a flux of 1000 kerg/ (cm. 2 sec.) from a 750 watt incandescent lamp, acquired a temperature up to 20° C. above that of the ambient air. Franck and French (1941) found the temperature of Hydrangea leaves, illuminated by collimated light from a 1000 watt lamp (about 80,000 lux) to be 1.3° C. above that of the medium; these leaves were cooled by rapid circulation of gas in a thermostated vessel. Because of the effective cooling of aquatic plants by the surrounding water, such plants have no opportunity to acquire internal temperatures markedly different from those of the medium, even if subjected to very intense illumination. They thus provide the most appropriate material for the quantitative study of the influence of temperature on the rate of photosynthesis. B. Temperature Range of Photosynthesis* All life processes are restricted to a certain "biokinetic" range of tem- peratures. Above and below this range, the organisms suffer a more or * Bibliography, page 1254. 1218 THE TEMPERATURE FACTOR CHAP. 31 less rapid, and more or less irreversible "chill injury" or "heat injury." If the rate of a specific biochemical process in vivo is measured as a function of temperature, one often finds a region of temperatures in which the effect of heating is reversible, and similar to that observed in the study of simple reactions in vitro. However, at the two ends of this range, the reversible influence of temperature becomes obscured by irreversible, destructive processes, such as changes in the colloidal structure of the protoplasm, which affect indirectly all chemical reactions in the living organism. In the case of photosynthesis, the region of reversible changes extends, for plants adapted to moderate climates, only from about 0° to about 30° C. Below +5°, and above +25° C, a slow chill injury or a slow heat injury may set in, so that the observed rate of photosynthesis depends not only on the momentarily prevailing temperature, but also on how long the plant has been exposed to it. The exact limits of the biokinetic range of photosynthesis depend on the individual (ontogenetic) adaptation of the plants, as well as on the phylogenetic adaptation of the species. Analogous to the existence of heliophilic and umbrophilic species and individuals (described in chapter 28), certain plants exhibit thennophilic, others cryophilic properties. This "thermal adaptation" will be discussed in section 3. 1. Lower Temperature Limit and Chill Injury Conifers, mosses and lichens retain their chlorophyll in winter, even though the temperature of the air may drop to — 50° C. It has been observed (c/. Ehlers 1915) that the starch deposits in conifers sometimes are replenished in winter (while those of evergreen deciduous plants are used up during the same period) ; this points to continued photosynthetic ac- tivity of conifers in cold weather. How low the temperature must be to inhibit all photosynthesis has not yet been determined by reliable experi- ments under laboratory conditions ; while observations made in the open gave contradictory results. Boussaingault (1874) found no photosynthesis below the freezing point, not even in conifers. Kreusler (1887, 1888), on the other hand, noticed photosynthesis in Ruhus at -2.4° C, and Jumelle (1891) claimed to have observed oxygen evolution by conifers {Picea ex- celsa, Juniper communis), and lichens (Evernia prunastri) even at -30° and -40° C. These extreme results found little acceptance; Ewart (1896, 1897), among others, rejected them as erroneous, asserting that tropical plants cease to reduce carbon dioxide at temperatures as high as +4° or +8° C, that subtropical and aquatic plants stop reduction at 0° or +2° C, and that land plants from temperate, arctic and alpine zones do so slightly below the freezing point. Ewart noted that oxygen production ceases LOWER TEMPERATURE LIMIT AND CHILL INJURY 1219 upon cooling gradually rather than suddenly, e. g., in some tropical plants, after 1 hour exposure to 5°, and after 15 minutes exposure to 1° C. Mat- thaei (1904) found that the photosynthesis of cherry laurel leaves declines rapidly below 0° and ceases practically immediately at —6°. An exten- sion of the temperature range of active photosynthesis, down to —16° for certain alpine phanerogams and to -20° for alpine lichens, was again claimed by Henrici (1921). In the interpretation of her results, the pos- sibility of wide difference in temperature between the interior of the leaves and the surrounding air should be taken into account. In the mountains, where strong irradiation combines with low air temperature, the heating of leaves by light absorption may be particularly strong. Ivanov and Orlova (1931) found photosynthesis in pine trees down to -7° C, Printz (1933), only down to —2° or —3°. Freeland (1944) noted that apparent photo- synthesis still was positive, in three Pinus species, at an external tempera- ture of —6°. It was mentioned above that the inhibition of photosynthesis by cold requires time. Conversely, once photosynthesis has been stopped by cold, plants may require a certain time for the recovery of their photosynthetic ability. Ewart (1896) found that, when Elodea was chilled to 0° for 6 hours and then transferred to warm water, photosynthesis did not reappear before 10 or 15 minutes; after chilling for 1 or 2 days, the recovery required 3 hours, and after 5 days, from 5 to 24 hours. The suspension of photosynthesis in chilled leaves and its recovery upon thawing is one aspect of the broad problem of frost resistance of plants, which is of vital importance in agriculture. Largely because of this practical importance, extensive investigations have been carried out on the way in which plant cells are injured by cold. Ice formation — inside as well as outside the cell — certainly is a factor of importance, and frost re- sistance has often been associated with the capacity of the protoplasm for undercooling. Some plants can be cooled to —20° C. or lower without visible formation of ice. This capacity for undercooling has been as- sociated with the living state of the protoplasm; thus, Lewis and Tuttle (1920) found that ice was first formed in living cells of Pyrola rotundifoUa at —32° C, while in leaves of the same species, killed by immersion into solid carbon dioxide, water was observed to freeze at —3.5° C. Ice formed on the outer walls drains water from inside the cells and thus causes injuries similar to those induced by drought. Ice formed in- side the cell can cause injury by mechanical pressure. Warburg (1919) found that Chlorella cells can resist immersion into liquid air for several hours without loss of their capacity for photo- synthesis. He attributed this striking property to the fact that the single chloroplast contained in a Chlorella cell has the form of a bell, spread over. ■i k a «vi^ 1220 THE TEMPERATURE FACTOR CHAP. 31 the inside of the sturdy cell walls; this position may enable the chloro- plast to sustain without injury a pressure that would destroy the chloro- plasts freely suspended in the protoplasm. Whatever the macroscopic and microscopic cause of frost injury to plant cells, the submicroscopic phenomenon is probably always a change in the colloidal structure of the cytoplasm or of the chloroplasts. It is a fundamental fact of life that the protoplasm colloids become granulated or coagulated under the influence of mechanical forces. Mechanical stresses, rather than direct temperature effects, probably explain the destruction of cells by freezing. This is why dried cells, in which no ice formation is possible, can sustain much lower temperatures than the same cells while they still contain water. (These questions are discussed in the book of Lepesclikin 1924.) One immediate effect of cooling below about 10° C. is a rapid increase in viscosity and decrease in permeability of protoplasm. This change must impede the diffusion of carbon dioxide to the chloroplasts, as well as the translocation of the intermediates and products of photosynthesis. It can thus contribute to the rapid decline of the photosynthetic efficiency, which most plants experience in this temperature region. 2. Optimum and Upper Temperature Limits; Heat Injury Above the lower temperature limit, the rate of photosynthesis increases with temperature, first rapidly, then more slowly, until an optimum is reached, followed by a rapid decrease to zero. For land plants in moder- ate climates, the optimum is situated at 30-35° C. However, it lies much lower for land plants and algae adapted to low temperatures, and much higher for "thermophihc" algae that live in hot springs, and for tropical desert plants. Some authors, e. g., Henrici (1921), Lundegardh (1924). Ehrke (1929, 1931) and Stocker (1927, 1935), have observed temperature curves with several (two or three) maxima. Even if these results (to be dis- cussed in more detail in section C) are correct (which is doubtful), the oc- currence of several temperature optima certainly is an exception and not the rule. The decline of the net gas exchange (P — R) at high temperatures (which eventually leads to a change in sign) is partly caused by a continued rapid rise in respiration. (The latter increases exponentially up to 45 to 50° C.) However, even if correction is made for enhanced respiration, the true photosynthesis (P) also is found to possess an optimum, even though it is less sharp, and is situated at a higher temperature, than the maximum of net oxygen liberation (c/. fig. 31.2). OPTIMUM AND UPPER TEMPERATURE LIMITS; HEAT INJURY 1221 Thomas (1950) found for the purple bacterium, Rhodos-pir ilium ruhrum, a peak of efficiency at 40° C, followed by a sharp drop. As previously mentioned, P becomes time-dependent when the limit of the "biokinetic" range is approached, usually even before the optimum has been reached. The position of the optimum thus becomes a function of the duration of the experiment. Matthaei (1904) found slow heat injury to cherry laurel leaves at temperatures above 25° C; figure 31.3 was 10 20 TEMPERATURE, "C. Fig. 31.2. Temperature curves of respiration (R) and photosynthesis (P) in the lichen Ramalina farinacea (after St§,lfelt 1939). constructed from her data by Jost (1906). By extrapolating them to zero time, Blackman (1905) extended the exponentially ascending tem- perature curve of the photosynthesis of cherry laurel leaves up to 37.5° C. Decker (1944) observed a decline of 45% in the (apparent) photosynthesis of two species of pine between 30° and 40° C, in light of about 40,000 lux. Green algae behave similarly: Wurmser and Jacquot (1923) found com- plete stoppage of photosynthesis in Ulva after 2 minutes exposure to 45° C. Stalfelt (1939) found in Usnm dasypoga (a lichen) time dependence above 21° C, independently of light intensity (2-32 klux), but subject to adaptation to the temperature of the ambient medium. Noddack and Kopp (1940) found that the photosynthesis of Chlorella became time-dependent above 22° C. By using experiments of not over 30 miimtes duration, they could extend the exponentially ascend- ing curve up to 30° C. ; but a 150 minute exposure to the latter temperature gave an average rate 25% less than that found in 30 minute experiments 1222 THE TEMPERATURE FACTOR CHAP. 31 (c/. fig. 31.4). Craig and Trelease (1937) found no marked time effect in Chlorella up to 46° C, if the exposure was restricted to 20 minutes. 150 min 10 15 20 25 30 35 40 45 TEMPERATURE, "C. Fig. 31.3. Temperature curves of photo- synthesis of cherry laurel (Jost 1906, after Blackman and Matthaei 1905). *-• 5 10 15 20 25 30 35 TEMPERATURE, "C. Fig. 31.4. Temperature curve of photosynthesis in Chlorella (after Noddack and Kopp 1940). 25? 20 u n . E £ 15 o 10 lO . e E o o o o o 8 *-" 10 20 30 40 HEAT TREATMENT, min. Fig. 31.5. Effect of heat treatment (at 45° C.) on photo- synthesis of Chlorella (after Kennedy 1940). The use of rapid recording devices (cf. chapter 25) may permit follow- ing the ascending temperature curve of photosynthesis beyond the range attained by Blackman and Noddack. OPTIMUM AND UPPER TEMPERATURE LIMITS; HEAT INJURY 1223 If a plant is subjected to superoptimal temperatures for more than a few minutes, reversible thermal inhibition may be replaced by an irrevers- ible (or only slowly reversible) thermal injury. Van Amstel (1916) observed thermal injury to the photosynthetic apparatus of Elodea at 40° C. Wurm- ser and Jacquot (1923) observed that the inhibition of photosynthesis in 200 10 14 18 22 26 30 10 14 18 22 26 30 TEMPERATURE, "C. Fig. 31.6. Temperature curves of photosynthesis and res- piration of four algae (after van der Paauw 1934). green, red or brown algae, caused by a 2 minute exposure to temperatures of 36-45° C, was not entirely reversible, but left a permanent decrease in photosynthetic efficiency after return to 16°. Kennedy (1940) studied quantitatively the effects of preheating to 40 and 45° C. on the photosynthesis of Chlorella at 25°. One hour ex- posure to 40° had no effect on chlorophyll concentration and respiration, but caused a decline in photosynthesis by about 30%. Exposure to 45° caused a very rapid decline in the capacity for photosynthesis at 25° (cf. 1224 THE TEMPERATURE FACTOR CHAP. 31 fig. 31.5). Kennedy argued that if this heat injury were caused by the inactivation of an enzyme, as suggested by Blackman, its influence could be balanced, in flashing light experiments (c/. chapter 34) by longer dark intervals between flashes (as is actually possible in the case of cyanide poisoning; cf. chapter 12, page 307). However, he found that the reduc- tion in oxygen yield per flash, caused by 15 minutes preheating to 45°, was the same, whether the dark intervals lasted 0.0175 or 0.37 second. (The cyanide effect would disappear in the latter case; cf. fig. 34.12.) Progressive heat injury is not a specific property of the photosynthetic ap- paratus, but is common to all biochemical functions, as well as to many en- zymatic reactions in vitro. However, photosynthesis is more sensitive to heat than most other life processes. Respiration of yeast, for example, shows the first signs of inhibition at 46° C. and is rapidly destroyed only by temperatures in excess of 50° {cf. van Amstel and van Iterson 1911). A comparison between the temperature curves of photosynthesis and respiration of the lichen Rantalina farinacea was given in figure 31.2; figure 31.6 gives a similar comparison for four unicellular algae. Sooner or later, all vital functions of the cell become totally inhibited by heat. What ensues is known as "heat coma." In its first stages, it is still reversible, but finally it leads to "thermal death." With most leaves and algae, this happens at 55-60° C, although for organisms adapted to extreme cold (or heat) the lethal temperatures may be considerably lower (or higher) . The comparatively early onset of the thermal injury of photosynthesis seems to indicate that its origin lies in an impairment of the photosjmthetic apparatus itself, rather than in the general decline in the "vitality" of the protoplasm. The fact that, with short exposures, the ascending tempera- ture curve can be followed for some distance above the optimum indicates that the thermal inhibition is caused by a destructive process (e. g., slow deactivation of an enzyme), and is not associated with an intrinsic property of the kinetic mechanism of photosynthesis (such as thermal dissociation of the ACO2 complex, as was suggested by Willstatter and StoU). Attempts have been made to calculate the activation energy Ea of the process responsible for heat injury, by measuring its rate at difterent tem- peratures {cf. Belehradek 1935, page 174). Values between 51 and 95 kcal have been calculated for the heat inhibition of photosynthesis between 20 and 65° C, with the smaller values derived from measurements at the lower temperatures. Several hypotheses have been suggested as to the nature of the process to which this remarkably high activation energy may correspond. Denaturation and coagulation of proteins was the first explana- tion; and it finds much credence despite the fact that the "heat inhibition" of photosynthesis occurs at temperatures considerably below those usually THERMAL ADAPTATION 1225 associated with the denaturation of proteins. The destruction of enzymes, which also was suggested as a possible cause of the decline of photosynthe- sis at high temperatures (Blackman 1905) may be but a consequence of changes in structure of their proteinaceous components. Lipides (the role of which in the composition of the chloroplasts was discussed in chapter 14) also may be responsible for heat sensitivity. (It was mentioned in chapter 14, page 361, and in chapter 24, page 817, that the "melting" of grana in chloroplasts, observed under the fluorescence microscope, was at- tributed by Metzner to a liquefaction of lipides.) It may be asked whether the rapidly reversible thermal inhibition (by comparatively low temperatures and short exposures) is fundamentally dif- ferent from the irreversible, or only slowly reversible, injury caused by higher temperatures and longer exposures. In addition to reversible shifts of chemical equilibria, such as the one responsible for the formation of the ACO2 complex, various other reversible changes could affect reversibly the rate of photosynthesis at high temperatures. One of them is the increased viscosity of the protoplasm. It wa smentioned on page 1220 that, at low temperatures, the viscosity of the protoplasm, hke that of most other mate- rials, decreases upon heating. Characteristically, however, it passes through a minimum, usually in the neighborhood of 15° C, and then in- creases again. This increase can slow down the diffusion of carbon di- oxide and thus bring photosynthesis from the carbon dioxide-saturated into the carbon dioxide-limited state. Like the effect of the dissociation of the ACO2 complex, this kind of thermal inhibition should disappear upon an increase in the concentration of carbon dioxide. Thus, in order to clarify the role of viscosity in the decline of photosynthesis at low and high temperatures, it would be useful to measure these effects at different con- centrations of carbon dioxide, and to compare the results with the change in viscosity. Much quantitative work remains to be done in this field. Thermal deactivation of enzymes by the denaturation of proteins also may be more or less rapidly reversible, depending on how far the denatura- tion has been allowed to proceed. Reversible denaturation as an explana- tion of the temperature effect on enzymatic processes has been discussed, e. g., by Johnson, Brown and Marsland (1942) on the basis of experiments with lucif erase. 3. Thermal Adaptation Ewart's figures, given on page 1219, indicated already that the lower limit of photosynthesis differs in plants from different climatic zones, and thus reveals a considerable degree of "thermal adaptation." This adapta- tion also affects the positions of the optimum and of the upper limit of photo- 1226 THE TEMPERATURE FACTOR CHAP. 31 synthesis. While higher plants and algae adapted to moderate climates attain the maximum rate of photosynthesis at 30-35° C. and suffer rapid injury above 40°, cold-adapted (cryophilic) plants often show a maximum of true photosynthesis below 10°, and a maximum of net oxygen evolution near the freezing point. Some of them have been found to suffer thermal injury at temperatures as low as 12° {Phaeocystes poucheti). Arctic land plants, polar sea algae and the flora of snow fields and glaciers ("red snow") are examples of such cold-adapted species. Plants 3- 2- I 1 1 ' 1 1 i 1 J^ — ' ^^° ; ; ^^ — [ 10° / 1 1 1 / 1 ] 1 1 ] 1 1 1 1 1 1 i_ o o I 0 ce I 3 5 7 9 II 2 6 10 14 18 22 LIGHT INTENSITY, klux LIGHT INTENSITY, klux Fig. 31.7. Light curves of two arctic plants — Salix glauca (left) and Chamaenerium latifolium (right) at different temperatures (after Miiller 1928). adapted to high temperatures do not reach the maximum rate of net photo- synthesis before 40° C; they are capable of organic synthesis even at 50° and survive without permanent injury at 80° or even 90°. Algae that live in hot springs and tropical desert plants are extreme examples of such "thermophilic" plants. Among studies of the temperature dependence of photosynthesis of cold-adapted species, we can mention that of Miiller (1928) with Salix glauca and Chamaenerium latifolium at Disco, Greenland. Figure 31.7 shows the relative positions of light curves of net oxygen evolution by the two plants at 0°, 10° and 20°. In weak light (ordinate a) the largest oxygen production is obtained at 0° ; at 2000 lux (ordinate h), the optimum of net synthesis is shifted to 10°, while in stronger light (ordinate c), it lies at (or above) 20° C. Similar behavior is shown by cryophilic marine algae. Thus, Harder (1915) found that Fucus serratus liberates, in weak light, twenty times as much oxygen at 0 as at 17° C. It must be pointed out, however, that within the linear range of the light curves, where true THERMAL ADAPTATION 1227 photosynthesis is independent of temperature, the net oxygen evolution of all plants — cryophilic as well as thermophilic — must decrease with tem- perature (because of accelerated respiration). In cryophilic plants, this behavior is accentuated by the early "thermal saturation" of true photo- synthesis (which is not paralleled by an equally early saturation of respira- tion). Furthermore, these plants live in regions where the average inten- sity of sunlight is low, and this shifts their maximum efficiency under natural conditions toward lower temperatures. In July, the average illumination at Disco is only 450 lux, a light intensity at which the opti- mum of net oxygen production by Salix lies close to 0° C. ro . E E to to LlI X >- o h- o X Q- 10 15 20 25 30 35 40 TEMPERATURE, "C. Fig. 31.8. Variation in rate of photosynthesis with tem- perature at high light intensities for Nitzschia closterium (O) and N. palea (A) (after Barker 1935). The temperature optima of true photosynthesis probably never lie as low as one may think from the consideration of the net gas exchange of cryophihc plants. Barker (1935) (c/. fig. 31.8) found that the optimum of true photosynthesis of the marine diatom Nitzschia closterium lies at 27° C, although this organism usually lives at 8-12°. The fresh water diatom Nitzschia palea had an optimum at 33°. (This high optimum temperature probably enables the species to withstand the high temperatures that shal- low waters may reach on warm summer days.) At the opposite extreme from the arctic plants, adapted, as far as their net organic synthesis is concerned, to temperatures near the freezing point, we find some algae (e. g., Phormidium) as well as certain sulfur bacteria thriving in hot springs, with temperatures up to 80 or 90° C. (c/., for ex- ample, Harvey 1924). Their cells must be capable of sustaining such high 1228 THE TEMPERATURE FACTOR CHAP. 31 temperatures without injury. Furthermore, unless the water in which they Hve cools down perceptibly during certain periods of the day or sea- son, these algae must be able to carry out photosynthesis at a tempera- ture that would bring immediate and complete thermal inhibition — in fact, thermal death — to most other plants. Photosynthesis of these "thermo- philic algae" certainly is worth closer investigation, under natural as well as under laboratory conditions. Inman (1940), who studied some algae from Yellowstone Park geysers, was concerned primarily with the proof of the spectroscopic identity of their chlorophyll with that of ordinary plants; the only observation he made concerning their photosynthesis was that they liberate oxygen when irradiated at room temperature. Desert plants, exposed to direct sunhght, sometimes are heated to temperatures approaching those of hot springs. Their photosynthetic ap- paratus, too, must remain uninjured by heat. MacDougal and Working (1921) found that Opuntia actually continues to grow, and thus presumably to carry out photosynthesis, at 58° C. ; growth is stopped, and shrinkage ensues, when the temperature reaches 62°. Wood (1932) found that the optimum of photosynthesis of some Australian desert plants lies between 40° and 50°, and that their net oxygen production does not become zero until 55°. Extended heat resistance of thermophihc algae, bacteria and desert succulents must be due to a different structure of the protoplasmic con- stituents responsible for heat injury. Lipides might, perhaps, bear the main responsibihty for this difference, because it is known that the ther- mal stability of fats and lipides depends on the temperature at which they have been formed in the organism. One chemical peculiarity of thermophilic algae was noted by Harvey (1924): they contain no catalase, a unique occurrence in the whole plant world. Harvey suggested that thermal algae have no need for catalase because, at the high temperatures at which they live, the decomposition of hydrogen peroxide proceeds rapidly enough by itself. This caused him to speculate generally on the possibility of a primeval "life without enzymes" in a medium hot enough for the organisms to dispense with catalysts. However, thermophilic algae are more likely to represent adaptations of normal species to high temperatures, than remnants of such prehistoric, enzyme-free flora. (To prevent misunderstanding, it must be pointed out that thermophihc algae contain many enzymes — e. g., oxidases — and are only deficient in catalase.) A very interesting result was obtained by Sorokin and Myers (1953) in the course of experiments on mass cultivation of Chlorella. By starting with inocula from warm local surface water, and incubating the cultures at 32° C, several strains were isolated which showed decided "thermo- THERMAL ADAPTATION 1229 philic" behavior. One of them grew fastest at 39° C; fig. 31. 8A shows its growth rate as a function of temperature, compared to that of an ordinary strain of CJdorella pyrenoidosa. The rate of growth of the two strains ap- pears the same below 25° C, but the thermophihc strain grows much faster at the higher temperatures; a much higher light intensity is needed to reach growth "saturation" in this case. PreHminary manometric studies of the thermophilic strain at 39° C. indicated a rate of glucose respiration equiv- alent to 18 volumes of oxygen, and a rate of photosynthesis equivalent to 186 volumes of oxygen per volume of cells per hour (Sorokin 1954) — by far the highest rate ever observed with any organism ! 30 TEMPERATURE Fig. 3 1.8 A. Growth rates of an ordinary and a thermophilic strain of Chlorella (Sorolcin and Myers 1953): (A) Chlorella pyrenoidosa (Emerson's strain) at 1600 foot-candles; (X) Tx 71105 at 1600 foot-candles; (O) Tx 71105 at 500 foot-candles; (0) Tx 71105 at 2800 foot-candles. We have spoken so far of the adaptation of whole species (or strains) to heat or cold. As in the case of heliophilic and umbrophilic plants, this "phylogenetic" thermal adaptation is paralleled, although on a reduced scale, by the adaptation of individual organisms. Harder (1924) studied, as an example, the behavior of the aquatic plants Elodea canadensis, Fontinalis antipyretica, Hypnum, CJiara and Chladophora, grown at 4.6° and Table 31.11 Photosynthesis of Fontinalis Plants Grown at 4.6° and 20° C. (after Harder 1924) I, lux Temp., " c. P, relative units Cold-adapte d Warm-adapted 1,017 29,545 8 18 8 18 224 151 1564 1894 162 182 944 2639 1230 THE TEMPERATURE FACTOR CHAP. 31 20° C. The specimens showed no difference in appearance, but their responses to temperature were quite different. An example is given in Table 31.11. In weak light, the photosynthesis of cold-adapted Fontinalis plants declined between 8° and 18° C. while that of the warm-adapted plants increased; in strong light, it increased in the same range, but propor- tionately much less than the photosynthesis of the warmth-adapted in- dividuals. Since the figures in Table 31.11 appear to represent true photosynthesis, and not net gas exchange, the negative sign of the temperature effect in the case of cold-adapted plants in weak light is noteworthy. The analogy between light and temperature adaptations caused Harder to ask whether cold-adapted plants have an enhanced capacity for photosynthesis at low temperatures (analogous to the higher efficiency of shade-adapted plants in weak light). He carried out no experiments with plants of equal dry weight (or equal leaf surface), but attempted in- stead to answer this question by analyzing the figures given in Table 31.11 for plants of unknown weight and leaf area. He concluded that at low temperatures cold-adapted individuals actually are more efficient than warmth-adapted ones. Theoretically, the situation in cryophilic plants is somewhat different from that in the umbrophiles; in the latter case, an improved efficiency (rate per unit area) in weak light could be easily ex- plained by a higher chlorophyll content and consequent enhanced absorp- tion of fight (c/. page 422). For cryophilic plants to have an enhanced ef- ficiency at low temperatures (in strong light), they should contain a higher concentration of rate-limiting enzyme ; and if this is the case, one can ask : why have they not a superior efficiency at the higher temperatures as well? One conceivable explanation is to assume that two different enzymatic processes limit the rate at the two temperatures. In contrast to the low respiration of umbrophilic plants, the respira- tion of cryophilic plants is, according to Harder, not weaker, but even stronger than that of the thermophilic individuals; consequently, their compensation points are higher. Plants that do not interrupt photosynthesis in winter experience a re- versible adaptation to cold during this season as shown, e. g., by meas- urements of Staifelt (1937, 1939) on hchens. The temperature curve of respiration is practically unaffected by the season, while the optimum of photosynthesis, and consequently also that of net organic synthesis, is shifted in winter toward lower temperatures. For the eight species inves- tigated by Stalfelt the average shift of the optimum of net photosynthesis was from 18.5° in summer to 14.1° C. in winter. The fact, observed by Beljakov (1930), that barley plants suddenly cooled to 3-10° interrupt TEMPERATURE COEFFICIENT AND HEAT OF ACTIVATION 1231 photosynthesis for a while, but resume it afterward, also can be inter- preted as an example of individual thermal adaptation. (This phenomenon contrasts with the gradual suspension of photosynthesis in chilled plants, observed by Ewart, and mentioned on page 1219.) Similar questions have been much discussed in connection with the increase in the abundance of algae observed as one proceeds northward in European waters. Kniep (1914) and Harder (1915) suggested that this is due to the greater excess of photosynthesis over respiration at the lower temperatures. This theory received support from experiments by Ehrke (1931) ; but Lampe (1935), who conducted experiments on numerous green, red and brown algae, under laboratory conditions and in natural habitats, concluded that the average light intensity under natural conditions often is not low enough to permit application of the Kniep-Harder theory. Furthermore, he observed that some algae can adjust themselves wdthin a few days to higher or lower temperatures; the position of the optimum of net organic synthesis is shifted, by this adaptation, nearer to the tempera- ture of the medium. These algae can act as cryophiles in winter and as thermophiles in summer. Only some deep water red algae, and the umbro- philic green algae, have rigid cryophilic characteristics, and actually in- crease in weight more rapidly in cold water (and dim hght). C. Temperature Coefficient and Heat of Activation OF Photosynthesis* As stated before, if we leave aside the two extreme ends of the "bio- kinetic range" of photosynthesis, and restrict ourselves to its middle part, where no time-dependent inhibition effects occur, we can obtain a short segment of a temperature curve that apparently reflects the direct effect of temperature on the kinetic mechanism of photosynthesis. The shape of this temperature curve depends on several parameters, such as light in- tensity and carbon dioxide supply. This was not sufficiently realized by earlier plant physiologists, who gave figures for the "temperature coef- ficient" of photosynthesis without identifying the specific conditions under which this coefficient was determined. In weak light and in the presence of an adequate supply of carbon dioxide, the rate of the over-all process of photosynthesis is practically equal to that of the primary photo- chemical reaction, and the temperature curve reflects the influence (or, more likely, lack of influence) of temperature on this primary reaction. When the supply of carbon dioxide is low, the temperature coefficient may be essentially that of the supply process, e. g., of the diffusion of carbon diox-ide through an aqueous phase. In this case, Qio (the proportional increase in rate caused by an increase in temperature by 10° C.) will be of * Bibliogi-aphy, page 1255. 1232 THE TEMPERATURE FACTOR CHAP. 31 the order of 1.2 or 1.3. If the supply of all reactants, as well as of Hght, is abundant, the over-all rate will be determined by the efficiency of a non- photochemical "bottleneck" reaction, and the temperature coefficient may reach or exceed 2 (which is the common value for enzyme reactions in vitro and in vivo). Obviously, by working in intermediary regions, one can obtain values of Qio ranging all the way from 1 to 2 or more (as illus- trated by figs. 28.6 and 28.7). Most of the earlier investigators worked under "natural" conditions, which means partial or complete hght saturation, but usually incom- plete saturation with carbon dioxide. No wonder a controversy arose as to whether the temperature coefficient of photosynthesis is about 1.3 or about 2. Prjanishnikov (1876), Kreusler (1887, 1888, 1890) and Lubimenko (1906), among others, found Qio values of less than 1.5, whereas Matthaei (1904), Blackman and Matthaei (1905) and Blackman and Smith (1911) obtained values close to 2. Blackman was the first to realize that high temperature coefficients were associated with high light intensities. He took this as an indication that photosynthesis involves, in addition to the photochemical reaction proper, also an enzymatic process, which becomes rate-determining when light and carbon dioxide are supplied in abun- dance. This conclusion was criticized by Brown and Heise (1917), who referred to the above- mentioned papers of Prjanishnikov and others for proof that the temperature coefficient of photosynthesis is lower than that of typical enzymatic reactions; this criticism was rejected by Smith (1919) as showing a lack of understanding of the dependence of the temperature coefficient on external conditions. The designation of the nonphotochemical stages of photosynthesis as the "Blackman reaction," first used by Willstatter and Stoll in 1918 and Warburg in 1919, was based on this argument of Blackman. We know now that photosynthesis involves not one, but many nonphotochemical catalytic reactions; therefore, references to the Blackman reaction should be avoided. Whenever possible, one should specify which of the several known or suspected catalytic reactions one has in mind. 1. Absence of Temperature Influence in Weak Light Primary photochemical processes ordinarily are independent of tem- perature. This is so because light usually provides more than the required minimum activation energy, so that no additional thermal activation is needed. Exceptional situations occur, however, in which light energy alone is insufficient to accomplish a certain transformation that becomes possible if a certain amount of thermal energy is supplied during the ex- citation period. ABSENCE OF TEMPERATURE INFLUENCE IN WEAK LIGHT 1233 When it was generally assumed that 4 quanta are sufficient to reduce one molecule of carbon dioxide, the scarcity of energy available in 4 quanta of red light caused Franck and Herzfeld (1937) to examine the possibility that light energy might be supplemented by thermal energy, available in the many degrees of freedom of the chlorophyll molecule. However, recourse to such hypotheses becomes unnecessary if the lati- tude of at least 6 quanta per molecule of carbon dioxide is allowed to the theorists. Another potential source of temperature effects in the light- limited state (?". e., of a temperature dependence of the maximum quantum e " 60 High intensity, high [CO2] High intensity, low [003] -o- Low intensity, high [CO2] 8 12 16 TEMPERATURE, "C. Fig. 31.9. Rate of Gigartina photosynthesis plotted against temperature (after Emerson and Green 1934). 60 w. incandescent lamps 8 cm. below vessel; 10% transmission filter used for low light curve. Artificial sea water; 35 X 10~^ M CO2/I. for "high CO2," and 2.5 X 10-^ M CO2/I. for "low CO2." yield of photosynthesis) also was discussed before (page 1138) — the possible influence of temperature on the relative probability of the forward and the back reaction of the primary photochemical products. This effect, if it exists, must be determined by the difference of two activation energies, and therefore could be comparatively small. Experimental results speak against any theory requiring a strong tem- perature dependence of the rate in low light. Wlierever the illumination was weak and the supply of carbon dioxide adequate, the rate of photo- synthesis was found to be more or less exactly constant over a considerable range of temperatures. Thus, Matthaei (1904) found that the rate of production of oxygen by cherry laurel leaves in weak light is practically constant between 0° and 20° C. Warburg (1919) observed that the temperature coefficient of photosynthesis of Chlorella, Q ^ P (20° C.)/P (10° C), roughly equal to 2 in strong light (/ = 16-45 relative units), de- 1234 THE TEMPERATURE FACTOR CHAP. 31 clined to unity at / = 1. Figure 31.9 shows similar results obtained more recently by Emerson and Green (1934) with the red alga Gigartina; here the rate obtained at the lower light intensities (and with an ample supply of carbon dioxide) is practically independent of temperature between 4° and 16° C. Emerson and Lewis (1940, 1941) in their work on the maximum quantum yield, 70, of photosynthesis in Chlorella (cf. chapter 29) found no significant differences between the 70 values at 0°, 10° and 20° C. Similar results were obtained by Wassink, Vermeulen, Reman and Katz (1938) and Noddack and Kopp (1940), (c/. figs. 28.6 and 28.7). Figure 28.7A refers to experiments in white light, and figure 28.7B, to those in monochromatic (red) light. It will be remembered (c/. page 1160) that Noddack and Eichhoff found that the "light curves" bend early in white light, but remain linear until close to saturation in red light. The difference between figures 28.7A and B corresponds to these findings. In the first case, the rate is independent of temperature only in ex- tremely weak light, whereas, in the second one, the light curves corresponding to 10° and 20° C. (corrected for respiration) coincide over a considerable range of light inten- sities. We said (pp. 1098 and 1162) that the origin of differences between the shapes of light curves in white and red light, claimed by Noddack and co-workers, is obscure; here we must be satisfied with the fact that, whether the light curves bend early or late, the rate is independent of temperature so long (and only so long) as they are linear. In other words, temperature dependence becomes evident as soon as some dark process interferes to reduce the over-all rate below the maximum value allowed by the primary photochemical process. Table 31.III Compensation Point and Temperatube Observer Organism le (in lux) 0° c. 10° c. 16° C. — 299 457 — 270 408 — 247 299 300 750 — 175 500 — Ehrke (1931) Entheromorpha compressa Fucus serratus Plocamium coccineum Muller (1938) Salix glauca Chamaenerium latifolium Of course, the rate can be independent of temperature only within the "biokinetic" range. The drop in efficiency at the two ends of this range cannot be avoided, however weak the illumination. Blackman and Mat- thaei (1905) found that the rate of photosynthesis of cherry laurel leaves declines sharply below 0° C, in weak as well as in strong light. There appear to be no data available on the rate of photosynthesis in weak light at temperatures above 30° ; but one can expect that, sooner or later, super- optimal temperatures will affect photosynthesis also in the hght-limited state. However, if the exceptional sensitivity of photosynthesis to heat is due mainly, or partly, to the enzyme that limits the rate in strong fight, the quantum yield in weak light may remain unimpaired, at least for some TEMPERATURE EFFECT IN STRONG LIGHT 1235 time, even at temperatures at which the saturation rate is strongly de- pressed. Speaking of the temperature independence of photosynthesis in weak Hght, one thinks, of course, of true photosynthesis and not of net gas ex- change. Since respiration is accelerated by heating, while true photo- synthesis in weak light is independent of temperature, the net oxygen hberation in weak light must decline with increasing temperature; the compensation point is thus shifted toward the higher light intensities. This fact already was mentioned in chapter 28 (page 984); Table 31. Ill gives some additional examples. 2. Temperature Effect in Strong Light We now consider another extreme case— that of strong illumination and ample supply of carbon dioxide. The transition from temperature- independent photosynthesis in weak light to temperature-dependent photo- synthesis in strong light is best illustrated by light curve families with tem- perature as parameter (figs. 28.6 and 28.7), and temperature curve families with light intensity as parameter (fig. 31.9). In Table 31. IV are collected the results of a number of investigations dealing with the effect of tem- perature on photosynthesis in light- and carbon dioxide-saturated state. We have omitted measurements in which light and carbon dioxide satura- tion probably was incomplete, for example those of van Amstel (1916), carried out in light of only 2500 lux, and of Lundegardh (1924), Yoshii (1928), Beljakov (1930) and Stalfelt (1939), in which ordinary air was used as the source of carbon dioxide. (We saw in chapter 27 that most plants require considerably more than the 0.03% CO2 in the atmosphere for the saturation of their photosynthetic mechanism.) Table 31. IV contains values of the temperature coefficient, Qio, and of the heat of activation, Ea,* as characteristics of the temperature effect. The theoretical significance of these constants will be discussed in section 6. The constancy of Ea over a certain temperature range indicates that the rate in this range follows the Arrhenius function : (31.7) logF = A + {B/T) where A and B are constants and T is the temperature on the absolute scale. Figures 31.10 to 31.14 serve as illustrations to Table 31.IV. They clearly show the difference between the results of van der Paauw on CJilamijdomonas, Craig and Trelease on Chlorella and Noddack and Kopp * The similarity of this symbol and of the symbol Ea used throughout this book for the carbon dioxide-fixing catalyst should cause no confusion. 1236 CO < a D, o c3 sn 03 -1-3 & IB ~:^ 13 ^^ a 03 c3 .2 '^ a 'S >i 01 Ql ' — ^ i 1- b 03 a^ a THE TEMPERATURE FACTOR .a r3 a S^ 0} en • — .5 03 o3 CO IN O '-' t^ 0(N O -H -* -H i^ 00 CO (N (N t-H T^ (M lO O O O t^ O <-< iM CO CO CO I I I I 1 OO OOO i-H (M CO (M C3 o aj _d J3 .S _o *t-i "C c c I. M O O O o fcj . O !- X tJO O o o o CO o6~ IM CO S bc^ e „ a s 2 > ■« > e 2 2 S rt g 03 ■< aj S a; ?i. 00 . &5 1::d b cii -t :aJ ^-, o3 -^ O -^^ r^ _:2 r-i ::: 05 "g <^ "^ c3 -rt CO C "0 .ti -1-i —5 fcj -s 05 e3 -If o3 0) o ,-1 1^ ^ m oco 1 1 (M rH O lO CO CO I 1 O >o IN (N "3 ^ c CO ""• M o O o O -u CO ^ o (N 3 P 03 O ^ O o3 o 5=1 fej o m 03 ^- ^ § 5 d -A i>^v G >-liO 03 ^'§P^ — • 1— ( bC^^ • — C ^ !5 g^ .13 ^ OD CO o IN CO 7 o o > o a in CO c3v_. a -jj -i-i ooooo ^CN ^(N d <- < o o fc- '-3 t- VS S^ 3 F 3 >-5 >-:3 o o ^ CO M CO IN CO O a O a O o U o c 1— « ^W ^'^ '^ lO o -. (N 4-:> X. aS _bD si to "-Ij ^^ 03 o a 3 O o 03 lO CO !^ *^ e e C35 ::» 1-^ e "s ■^ ^ S. Si Ss e e •■^3 >-** •^ V lij o o-^ O c3 P o » a> a; CD J:: o< (M-T^ IM •^^ > > o -C -* oj -u to o y 0) •^ t: a< x: <» b, s< H 'x O o p. r» O -4-J a ^ 3 r' -^ « > CC 3 r'i C s- 13 < rr (n C C O o -i-J -tJ o o c:! OJ ;-< tH 073 (N CO oi^ 05 o o lO iC COiM iCt^ O CO(M ^ (N CD 00 lO lO C-l t^Tj* (M r-l CO Ot^ -—I t-~(M<— I— iiOCNOJ'— iO>iOCOOGO._ (N 00 fO(M O (M O CD iC 00 "-I ^ (M (M I I I I I ■*0OiM O IC CO CD I -O T-iCOCO CO I -IM I OO O lOOioooioooioooiooaD '-'C^C^CO(N(MCO(MClCO(N(MCO(N I I I I 1 I 1 I I I I I I I lOOiOOOiOOO'OOO'OO'-i CDO o 00 CO I I III o o o o o o o ;? ^ ^ ti t-, ^ 01 OJ HJ sc to to 3 3 3 « M « B8Sd c3 CO o o o »*.i— to 03 -2 -I T ' V 0) bO c o e Si ^ X S - o o s o S 1— I c 03 D. en o3 . CD Or-I -2 » — i '—J ^- e ^ e O 4c O CO o o S5, fc « S 5 CD T 1—1 o S to O o to e s o S o e CD *-H I O CJ e to o o • CD I ^ aj .— I CJ ^§ S -►^ ►; CO ^ :s ^S o o CO o O 3 o ceo o . on bD c^ 5-co ^^ o3^ii C ^ " c ^^^^^ ^^ e3t^-jC 5gO CiO t-(Mf^ fO.sScocj-to^co . M05 CO (TiCs'OOJ'Cc^ bfl •3^ '^JaJ^Tj-h _— (Crt bC ! CO. W CC 1 2 2 o3 u ;2; c oj > o3 00 c3 > C Si "■■ 0) CO 3 ^ & C >> o; ^ 03 o ^ 1— t a X O CD o CO ^ UJ ft (D a c 03 o3 -»-> a) ^ 73 O CD a CJ X 00 bC bi) W K bD bC ■^ K K Si 1 CO l» ^— , i- ^-^ 3 » 03 «i tr. C -C 03 ■^ 3 o ^ 5^ TJ 03 T3 IS i: e ^ S e !;d 5=^ O) . !-. bC to CO 5i ^ ^ ^ W CO C3 03 W 1238 THE TEMPERATURE FACTOR CHAP. 3 also on Chlorella, on the one hand, and those of Emerson, and Emerson and Green on Chlorella and Grigartina, on the other. The first-named investi- gators (as well as Blackman and Matthaei) found the Arrhenius equation o 2.400 2.300 2.200- 2.100 2.000 1.900- 1.800- 1.700 1.600 1.500 Chlamydomonas species J_ J_ _L 0.003300 0.003400 0.003500 2.400 0.003300 0.003400 0.003500 1/7" 1.500 0.003300 0.003400 0.003500 0.003600 \/T Fig. 31.10. Heat of activation of photosynthesis (after van der Paauw 1934): log TEMPERATURE EFFECT IN STRONG LIGHT 1239 to be valid, at least approximately, over a considerable range of tempera- tures; whereas the log P = fO-/T) curves of the second group of investi- gators (which also includes Warburg and Yabusoe) show practically no straight section at all. The observations of van der Paauw on Hormidium and Stichococcus also indicate continuous curvature of the logarithmic temperature curve, even if this curvature is less strong than in Emerson's curves. -0.8 - 20,300 21,000 18,700 18,400 33 34 35 36 \/T X 10^ Fig. 31.11. Rate of photosynthesis as a function of temperature in the range 10-30° C. (after Craig and Trelease 1937). log P = f(l/T) for Chlorella. Figures on curves are Ea values; curves 1-3 for HoO; curves 4—6 for D2O. Ordinate scale given is for curve 3; curve 1 moved up 0.6; curve 2, up 0.2; curve 4, up 0.05; curve 5, down 0.1; curve 6, down 0.25. In the region around 15° C, the activation energies found by most observers are in approximate agreement (Ea = 10-12 kcal), although Barker gave Ea = 30 kcal for diatoms, and Craig and Trelease, and Wassink and co-workers found for Chlorella Ea = 20 and 18 kcal, respectively. Larger discrepancies are encountered below 10° where the activation energies measured by some investigators were about the same as at higher temperatures, whereas those found by others were up to 30 or even 40 kcal/ mole. At 20-30° the Ea values sometimes were found to remain constant, and some- times to decrease to as little as 5 or 8 kcal. Tamiya, Huzisige and Mii (1948) found that the temperature curve of the photo- synthesis of Chlorella ellipsoidea can be interpreted, between 1° and 28° C, in the carbon dioxide saturated state, by two consecutive reactions with activation energies of 6 and 27 kcal, respectively. This applied to measurements at pH 4.6 as well as at pH 7.0; however, the first reaction was slowed down by a decrease in pH, while the second one was unaffected by this change. 1240 THE TEMPERATURE FACTOR CHAP. 31 The decline in Ea above 20° undoubtedly is associated with an incipient heat inhibition. Since the temperature at which this inhibition first oc- curs, and its rapidity, are quite different for different species or strains, it may well explain the variability of the apparent Ea values in the upper tem- perature range. Some of these discrepancies could perhaps be removed if all investigators would take the precaution of reducing the time of exposure at the higher temperatures to a few minutes, or of extrapolating the results to zero time, as was done by Blackman and by Noddack and Kopp. More 1.900 Q 1.200 QQ036 0.003300 0.003400 0.003500 0.0034 0.0035 \/T l/TCK.) Fig. 31.12. log P = f{l/T) for Fig. 31.13. Continuous variation of heat of Chlorella pyrenoidosa (after Nod- activation for Chlorella at two different chloro- dack and Kopp 1940). phyll concentrations (after Emerson 1929): log P = Kl/T). difficult to interpret are the high Ea values observed by some authors (par- ticularly by Emerson and Green 1934, and Emerson 1929) in the low temperature region (<10'^ C), a region in which Blackman, and Noddack and Kopp, for example, noticed no increase in temperature coefficient above the values observed at 10-20° C. (The theory of nonlinear "Arrhenius curves" will be discused in section 5.) The low temperature coefficients obtained by Willstatter and StoU (cf. Table 31. IV) also are noteworthy. In their experiments the light was strong (over 50,000 lux) and the carbon dioxide supply ample (5% CO2); nevertheless, the observed Qw values, at 15-25° C, were as low as 1.5 for green leaves and 1.3 for yellow leaves. As mentioned before, Willstatter suggested that the increase in the velocity of the limiting enzym- atic process with temperature is counteracted by the thermal dissociation of the ACO2 complex, and that this reduces Qio below the "normal" value of 2. However, if this were so, the situation could be amended by an increase in the concentration of carbon dioxide^which is not the case. Results obtained with Chlorella suspensions also make TEMPERATURE EFFECT IN STRONG LIGHT 1241 Willstatter and Stoll's hypothesis improbable. It seems as if in their experiments the rate was not free of carbon dioxide supply limitations, despite the high external value of ICO2]. The marked difference between the temperature coefficients of green and chlorophyll-deficient leaves, found by Willstatter and Stoll, also offers a problem for interpretation. Perhaps the yellow leaves were incompletely light-saturated under the conditions of the experiment. Emerson (1929), working with chlorotic ChloreUa cells, found that the temperature curves in the region 4-6° C. had the same slope for chlorophyll concentrations C vulgaris 6 harveyana 0.0034 0.0035 O0036 \/T CK.) Fig. 31.14. Comparison of log P — /(l/T) for Gigartina harveyana, ChloreUa vulgaris, ChloreUa pyrenoidosa and Horvii- dium (after Emerson and Green 1934). varying in the ratio of 1 to 6 (c/. fig. 31.13); in the case, apparently, hght intensity was sufficiently high to produce light saturation at all tempera- tures and chlorophyll concentrations. It should be recalled (c/. page 968) that Emerson found no dependence of the saturating light intensity on chlorophyll concentration in ChloreUa, whereas such a difference was noted by Willstatter and Stoll in green and aurea leaves (c/. fig. 32.2). Warburg (1919) and Yabusoe (1924) suggested that the rate of photosynthesis is a linear, rather than exponential, function of temperature, and considered this an essen- tial characteristic of this process. (For the respiration of the same algae, they found the normal exponential curve.) As pointed out by Emerson (1929), the number of points determined (only three in Yabusoe's paper) was insufficient to warrant the unusual conclusion. Emeison and Arnold (1932*) found that, in flashing light, intervals of about 0.02 second between flashes are necessary at room temperature (25° C.) to obtain the maximum oxj^gen yield per flash — i. e., to let the rate- limiting dark reaction run to completion before the next flash. Similar 1242 THE TEMPERATURE FACTOR CHAP. 31 experiments at 5° C. (c/. Emerson and Arnold 1932') showed that at this temperature dark intervals required for "flash saturation" are longer — of the order of 0.08 second. This difference corresponds to a Qio of about 2, or an activation energy of about 12 kcal, in agreement with the steady light results of Noddack and Kopp and many others, listed in Table 31. IV. This supports the view that in Chlorella, at least, the same "finishing" dark reaction that limits the maximum yield per flash also limits the rate in strong continuous light. 3. Temperature Curves in the Carbon Dioxide-Limited State No systematic measurements of the temperature dependence of photo- synthesis in the "carbon dioxide-limited" state {i. e., in strong light and with a reduced concentration of carbon dioxide) are available in the litera- ture. The case nearest to this limiting one, which was much studied, was that of free atmosphere (0.03% CO2). In this case, the carbon dioxide con- centration, although not strictly "limiting," exercises a strong influence on the rate, and the effect of heating may be largely due to the accelerated supply of carbon dioxide. It was mentioned on page 1232 that Prjanishni- kov (1876), Kreusler (1887, 1888, 1890) and Lubimenko (1906) found, for photosynthesis in the open air, Qw values between 1 and 1.5. Similar or smafler values were found by Stalfelt for mosses and Uchens (Table 31.V). Table 31.V Qio Values for the Photosynthesis of Mosses and Lichens (after Stalfelt 1937, 1939) (0.03% CO2) Species 0-10" 10-20° 20-30° C. MOSSES Hylocomium proliferuni 1.56 1.13 0.74 Hylocomium triquetrum 1 . 43 1 . 00 0 . 60 Hylocomium squarrosum — 1.13 0 . 65 Hylocomium parietinum 1 .52 1 . 13 0.60 Ptilium crista castrensis 1 . 43 1.15 0 . 69 Sphagnum Girgensohrii 2 . 67 (5-15 ° ) 1 . 35 0 . 66 LICHENS Usnea dasypoga 2.65 0.97 0 . 43 Cladonia silveslris 2.42 1 .04 0.63 Ramalina fraxinea 1 . 55 1.15 0 . 62 Ramalina farinacea 1 . 95 1 . 03 0 . 68 Umbilicaria pustulata 1 . 40 0 . 86 0 . 48 Evernia prunastri 2 . 07 1.14 0 . 75 Cetraria islandica 1 .08 0.81 0.46 Peltigera aphtosa 1.06 1.14 0.73 Cetraria glauca 1.0 0. 84 0. 62 TEMPERATURE CURVES WITH SEVERAL OPTIMA 1243 In figure 31.9, the middle curve referred to the temperature dependence of photosjTithesis in a red alga (Gigartina) in high light and comparatively- low carbon dioxide concentration (initial concentration, [C02]o = 2.5 X 10~^ mole/1., about twice that in air). The rate is about doubled between 4 and 14° C, corresponding to Qio = 2; but the influence of temperature still is very much weaker than in the presence of abundant carbon dioxide (upper curve), where the rate is increased, in the same temperature inter- val, by more than a factor of six. Tamiya, Huzisige and Mii (1948) found that, at 1 X 10 -« mole/1. CO2, the temperature curves of Chlorella ellipsoidea could be interpreted by a sequence of three reactions, with Ea = 0, 6, and 27 kcal, respectively (c/. above their results obtained at CO2 saturation). Only the first reac- tion depends on CO2 concentration. The factor that may be important for the temperature dependence of photosynthesis at low temperatures, in the carbon dioxide-limited state, is the viscosity of the protoplasm (with the stomata fully open, the main diffusion resistance between air and chloroplasts may be in the aqueous phase within the cell; cf. chapter 27, page 916). According to Weber (1916) the temperature coefficient of the viscosity of the protoplasm is about 1.4 (between 10° and 20° C). Miller and Burr (1935) noted (in experiments described on page 898) that, at ap- proximately 20,000 lux, potted plants of Pelargonium, Tolmica, Coleus, Bryophyllum, Eichhornia, Primula, Saxifraga, Zebrina and Begonia, as well as an unidentified Cras- sulacea, reached a balance between respiration and photosynthesis at approximately 0.01% CO2 in the air, at all temperatures between 4° and 37° C. (This balanced state was maintained for many hours at temperatures <30°, but was soon disturbed, ap- parently by a progressive heat inhibition of photosynthesis, at 35-37° C.) Since, at 0.01% CO2 and 20,000 lux, carbon dioxide supply probably is the main rate-limiting factor in photosynthesis, one could consider these results as proof of a practical equality of the temperature coefficients of respiration and carbon dioxide supply. This would be remarkable, since the temperature coefficient of respiration is high (Qw ^^ 2). However, it must be taken into account that, in the state of compensation, a large part of the carbon dioxide used for photosynthesis is supplied internally, by respiration, and there- fore has only a very short diffusion path (or is even used directly in situ). The tempera- ture coefficient of photosynthesis may thus be that of the carbon dioxide production by respiration. 4. Temperature Curves with Several Optima It was mentioned on page 1220 that temperature curves with two or more optima have been observed with some plants. The first such observation was made by Henrici (1921) on alpine lichens. Lunde- gardh (1924) found, in the study of tomato, potato and cucumber leaves, a main maxi- mum at 30-35° C, which was most prominent in strong light and with an abundant 1244 THE TEMPERATURE FACTOR CHAP. 31 supply of carbon dioxide, and two subsidiary maxima at 20° and 10°; in weak light (e. g., 1/25 of full sunlight) and relatively low carbon dioxide concentration (e. g., 0.03%), the 20° maximum became more prominent than the one at 30°. Subsequent studies of bean leaves by Yoshii (1928) and of barley leaves by Beljakov (1930) — in Lundegardh's laboratory — also gave temperature curves with several maxima. An example is shown in figure 31.15. Beljakov found that the relative promi- nence of the two main maxima — in the region of 20° and 30° C. — was different for two barley strains, and he considered this a sign of phylogenetic adaptation to dilTerent cli- matic conditions. Stocker (1927) found two maxima in the temperature curves of northern lichens (thus confirming the observation of Henrici with alpine lichens) and later (1935) also 22 21 < 18 o o d. 15 E o < • (A) 003% CDs , '/4 full sunlight ■ (B) 003% COj , '/lefull sunlight " (C) OI2%C02, 'A full sunlight 1 - (D) 0.12% COz '/i6 full sunlight /l - l\j\ - 1 r-^ \ _ rf- ^^J"^^ \ \ ■ 1/ '^ A. ^k/\ \ /^J^ B / ^- — ^ A \ / X /-^/^ ^ \ \ y - //'^ '"■^ ^ 1,1,1 ^ 12 18 24 30 36 TEMPERATURE, °C. 42 Fig. 31.15. Temperature curves of bean leaves showing several maxima (after Yoshii 1925). in the temperature curves of tropical trees. Stalfelt (1939), however, found only one optimum for northern lichens (c/. fig. 31.1). Ehrke (1929, 1931) found very irregular temperatures curves, with several "waves," for marine algae (e. g., the brown Fucus and the red Plocamium). Similar curves were obtained by him also for the respiration of these algae — a process for which a smooth ascent to an optimum is generally accepted. We strongly doubt whether the multiple maxima of Henrici, Lunde- gardh, Stocker and Ehrke are at all real, and not caused by experimental errors. Theoretically, no sequence of reactions with different temperature coefficients can give rise to curves of this type ; it is, however, undoubtedly' TEMPEIIATUKE EFFECT ON FLUORESCENCE 1245 possible for systems of competing reactions to exhiliit such complex relation to temperature. Henrici (1921) associated the decline in the P = f{T) curves between 20° and 30° C. with the formation of starch in the chloroplasts, and its inhibiting effect on photosynthe- sis (c/. Vol. I, page 331 ), and attributed the renewed rise of the rate between 25° and 30° to the disappearance of starch by accelerated conversion into sugars and dislocation of the latter from the chloroplasts. 5. Temperature Effect on Fluorescence Knowledge of the influence of temperature on the yield of chlorophyll fluorescence in the living cell may contribute to the understanding of the nature of the temperature effect in photosynthesis. For a given fluorescent molecule and given composition of the complex in which this molecule is imbedded, the yield of fluorescence should be (at least, in the first approxi- mation) independent of temperature, because the probabilities of absorp- tion, of reradiation, of "quenching" by photochemical reaction within the complex and of energy dissipation by "internal conversion" can be ex- pected not to be strongly affected by minor changes in temperature. There- fore, whenever a strong dependence of fluorescence intensity on temperature is observed, the most likely explanation is a change in the composition of the fluorescent complex. The effect of temperature changes on the light curves of fluorescence in hving plants was described in chapter 28 (page 1055). These effects were observed by Kautsky and Spohr (1934) and Franck, French and Puck (1941) in leaves, by Wassink, Vermeulen, Reman and Katz (1938) in ChloreUa, Wassink and Kersten (1945) in diatoms and by Katz, Wassink and Dorrestein (1941) and Wassink, Katz and Dorrestein (1942) in purple bacteria. The characteristic results are represented in figures 28.36-28.40. In most cases, the yield of fluorescence was found to be higher at lower temperatures. Franck and co-workers found, more specifically, that lowering of temperature caused in Hydrangea leaves a downward shift of the light intensity, Ic, at which the quantum yield of fluorescence changed from the "low light value," vi) (c/. page 1049 and fig. 28.26). The observations of Wassink and co-workers with purple bacteria (fig. 28.40) can be interpreted in the same way. On the other hand, Wassink, Vermeulen, Reman and Katz (1938) found no effect of temperature on

^, while at lower temperatures the two phenomena occur simultaneously, then indi- cates that the activation energy of the carbon dioxide fixation process is higher than that of the enzymatic process (presumably, according to Franck, the stabilization of the primary photoproducts by the catalyst Eb), which is responsible for light saturation at high temperatures. Franck was thus led to the hypothesis that the (higher) temperature coefficient of photo- synthesis observed in measurements in the lower temperature range (such as 5-15° C.) is characteristic of the carbon dioxide-fixation process, while the (lower) temperature coefficient observed around room temperature (15- 25°) is determined by the "finishing" reaction, catalyzed by Eb- In confirmation of this hypothesis, one may quote the observation (noted on page 1057) that the effect on fluorescence of lowering the tem- perature is quite similar to that of carbon dioxide deprivation. The results of Wassink and Kersten with diatoms do not fit well into this picture, because, apart from the reversal of the relationship between e, p A »a = [8]ae-ea/RT where S stands for substrate, and P for product. The two reactions velocities, V and v, are equal when : (31.13) In {A/a) = (Ea - ea)/RTa The transition temperature is thus: (31.14) Tc = .-5° ~/;, , ^ ^ 4.57 log {A /a) Above Te, the temperature dependence of the over-all rate of conversion THEORETICAL REMARKS 1251 of S into P will approach that of the noncatalytic reaction; below Tc, it will approach that of the catalytic reaction. Thus, the absolute value of the slope of the log v = f{l/T) plot will change, with increasing tempera- ture, from a smaller to a larger absolute value. As shown by Burton (1936), among others, the transition will be gradual, rather than sudden, i. e., the log v = J{l/T) plot (where v = v -\- V means the total reaction velocity) will be curvilinear over a range of 20° or more, depending on the difference between the two activation energies. Changes in the slope of the temperature curves also can be caused by a sequence of consecutive reaction steps — a so-called "catenary reaction series" {cj. chapter 26), which resembles a radioactive decay series (except that the velocity of each step depends on temperature). This concept has been much used in the explanation of the temperature dependence of biological and biochemical processes. It stems from Blackman's idea of "limiting factors," already discussed in chapter 26, and is thus historically connected with the theory of photosynthesis. As described in that chap- ter, Blackman first spoke vaguely of the "slowest factor" in a process deter- mined by several external factors; this notion was later replaced (c/., for example, Romell 1926, 1927) by the more precise concept of the slowest ste'p in a sequence ("catenary series") of chemical reactions. This step was often called the "master process" or "master reaction," not because of its intrinsic importance for the chemical result of the transformation, but for its "limiting" effect on the rate. Putter (1914) applied the master reaction concept to the interpretation of temperature coefficients. Crozier, in a series of papers (1924 and later), analyzed the temperature curves of numerous biological processes by dividing them into straight sections corresponding to different "master processes," connected by short curvilin- ear segments at the transition temperatures, Tc. This procedure was criti- cized, particularly by Burton (1936), who showed — for the case of a se- quence of monomolecular reactions — ^that the transition regions can never be as narrow as suggested by Crozier. In photosynthesis, the supply of carbon dioxide (by diffusion and car- boxylation), and several enzymatic dark reactions of the primary photo- chemical products are steps in a "catenary series." It can thus happen that, starting with the state of "enzymatic hmitation" {i. e., saturation with respect to both light and carbon dioxide), we may pass, by a mere in- crease in temperature, into the state of "carbon dioxide limitation." In contrast to the first-considered case of two competing reactions (in which the reaction having the lower activation energy predominates at the lower temperatures) the case of two consecutive reactions is characterized by the fact that the one with the higher activation energy is rate-determining at lower temperatures. 1252 THE TEMPERATURE FACTOR CHAP. 31 In photosynthesis, the temperature coefficient dechnes with tempera- ture, thus making an interpretation in terms of a series of consecutive reactions possible at least formally. An example of this kind of hypothesis is the above-mentioned suggestion of Franck, that the higher activation energy, lia, which determines the temperature coefficient in the low tem- perature range, is characteristic of enzymatic carbon dioxide fixation (catalyst E^), w4iile the lower activation energy prevailing at room tem- perature is characteristic of a finishing enzj^matic action. The analysis of Burton (1937) indicates that the gradual change in Ea with temperature, as observed in photosynthesis, does not preclude such an explanation. However, it must be pointed out that in the kinetics of photosynthesis (as in the treatment of many biological processes) we usu- ally deal with a steady state, rather than with the transformation of a limited quantity of a substrate. Burton (1937) thought that the concept of a master process cannot be applied to the steady state at all, since to reach this state, the rates of all processes in the catenary series must have adjusted themselves to that of the first irreversible step. This undoubtedly is true; but it is also true that, under different conditions, different steps in a catenary series can assume the role of the "first irreversible step." For example, if an enzymatic reaction consists of two stages : (31.15) S+E;;=L=iSE — ^^^ P + S (the first being the reversible formation of the substrate-enzyme complex SE, and the second its transformation into the reaction product P and the free enzyme S), the over-all rate equation (£"0 = total amount of enzyme) is; (31.16) V = d[F]/dt = kMS]Eo/ik[ + k. + k^ [S]) and this is reduced to: (31.17) V = d[P]/dt = A-2^0, if A"i[S] »A-; + k2 and to (31.18) V = ki'[S]Eo, if A-2»A-:; + AdS] The transition from the case (31.17), in which the over-all rate is deter- mined by the "first irreversible reaction," SE -> P -|- E, to the case (31.18), in which the reaction S -j- E -^ SE becomes "irreversible," can be brought about by a change in external conditions, e. g., an increase in temperature. Changes of this kind may well occur in photos3aithesis. Wohl (1937, 1940) thought that the high apparent activation energy of photosynthe- sis at low tempei'atures may be attributed to liberation of a comj)lete glucose molecule from a "reduction center" to which it was attached by six links. This process obviously requires high energy, but can perhaps occur at comparatively low temperatures if it also THEORETICAL REMARKS 1253 has a high value of A in equation (31.4). According to the theory of the activated com- plex (c/. equation 31.7) a high value of A ( = ce^'^"/'^^) means a large entropy, Sa of the activated state; and a reaction in which six bonds are disrupted simultaneously must increase the molecular disorder, i.e., lead to an increase in entropy. (This hypothesis is similar to the theory of denaturation of proteins of Stearn and Eyring.) Wohl analyzed the P'"'""' = f{T) curves given ty Warburg and by Emer- son for Chlorella, and by Emerson and Green for Gigartina, by assuming two consecutive dark reactions, both occurring at the same "reduction site" {i. e., at the same enzyme molecule), one requiring the time tai and the other the time ta2- Under these conditions, the dark reaction time, ta, derived from flashing light experiments (chapter 34) is simply the sum of the two consecutive reaction periods, ta = t,i + 1^2- Wohl found it possible to reproduce the experimental data closely by assuming that the first reac- tion has an A value in the Arrhenius formula characteristic of a bimolecular reaction (with the reaction partner present in a concentration of 10 "^'^ to 10 ~^^ mole/1.) , and a low activation energy (0-9 kcal/mole), while the second one is a monomolecular reaction with a very high activation energy (23-58 kcal/mole), and a high activation entropy. This second reaction was the one he interpreted as the liberation of a complex (Ce) molecule, by simultaneous dissociation of several (six?) bonds attaching it to the en- zyme. The bimolecular reaction accounts for 73-85% of the total dark reaction time at 15-25° C, and for only 12-24% at 5° C. Wohl pointed out himself that with the four arbitrary constants avail- able (two A values and two Ea values), the possibility of representing the experimental curves by the theoretical equation is not in itself significant ; but he considered it significant that the two calculated reactions are so dif- ferent in character, and, in particular, that the reaction that seems most important at low temperatures has the character of a monomolecular reac- tion with an extraordinarily high activation energy. The attempt of Tamiya, Huzisige and Mii (1948) to analyze the temper- ature curves of photosynthesis, in terms of two or three consecutive dark reactions, each obeying the Arrhenius law, was mentioned in sections 2 and 3. The success of mathematical analyses such as those of Wohl and Tamiya does not prove that the assumption on which they are based is necessarily correct. It may be doubted whether the rapid decrease in the rate of photo- synthesis at low temperatures (as well as the similar phenomenon occur- ring above 30-35° C.) is at all due to a reaction that constitutes a step in the "catenary series" of photosynthesis. It seems more probable that these changes are due to alterations in the colloidal structure of the proto- plasm, which affect, although to a different degree, all physiological proc- esses taking place in the cell (as the freezing or evaporation of a solvent 1254 THE TEMPERATURE FACTOR CHAP. 31 affects the kinetics of reactions between all the solutes which may be con- tained in it). Bibliography to Chapter 31 The Temperature Factor A. Internal Temperature of Plants 1875 Askenasy, E., Botan. Z., 33, 441. 1903 Ursprung, A., Bibliotheca hotanica, 12, 68. Timiriazev, C, Proc. Roy. Soc. London, 72, 424. 1905 Blackman, F. F., and Matthaei, G. L. C, ibid., B76, 402. Brown, H. T., and Escombe, F., ibid., B76, 69. Brown, H. T., and Wilson, W. E., ibid., B76, 122. 1909 Smith, A. M., Ann. Roy. Botan. Gardens Peradnyia, 4, 229. 1915 Ehlers, J. H., Am. J. Botany, 2, 32. 1919 Shreve, E. B., Plant World, 22, 100, 172. Shall, C. A., School Sci. and Math., 19, 1. 1923 Miller, E. C, and Saunders, A. R., /. Agr. Research, 26, 15. 1926 Clum, H. H., Am. J. Botany, 13, 194, 217. 1929 Eaton, F. M., and Belden, G. 0., U. S. Dept. Agr. Tech. Bull. No. 91. 1931 Miller, E. C., Plant Physiology. McGraw-Hill, New York, p. 362. 1933 Arthur, J. M., and Stewart, W. D., Contrib. Boyce Thompson Inst., 5, 483. Seybold, A., and Brambring, A., Planta, 20, 201. Watson, A. N., Ohio J. Sd., 33, 435. 1934 Clements, H. I., Plant Physiol, 9, 165. Watson, A. N., Am. J. Botany, 21, 605. 1936 Curtis, 0. F., ibid., 23, 7. Curtis, 0. F., Plant Physiol, 11, 343. 1941 Franck, J., and French, C. S., J. Gen. Physiol, 25, 309. 1951 Karmanov, V. G., Compt. rend. (Doklady) acad. sci. USSR, 77, 913. B. Temperature Range of Photosynthesis 1874 Boussaingault, J. B., Agronomic, chimie agricole et physiologic. Vol.V,Mallet- Bachelier, Paris. 1887 Kreusler, U., Landw. Jahrb., 16, 711. 1888 Kreusler, U., ibid., 17, 161. 1890 Kreusler, U., ibid., 19, 649. 1891 Jumelle, H., Co7npL rend., 112, 1462. 1896 Ewart, A. J., J. Linnean Soc. London, Botany, 31, 364. 1897 Ewart, A. J., ibid., 31, 554. 1904 Matthaei, G. L. C, Trans. Roy. See. London, B197, 47. 1905 Blackman, F. F., Ann. Botany, 19, 281. 1906 Jost, L., Biol. Zentr., 26, 225. 1911 van Amstel, J. E., and van Iterson, G., Proc. Acad. Sci. Amsterdam, 19, 106, 534, 1914 Kniep, H., Intern. Rev. ges. Hydrobiol. Hydrog., 7, 1. BIBLIOGRAPHY TO CHAPTER 31 1255 1915 Harder, R., Jahrh. wiss. Botan., 56, 282. 1916 van Amstel, J. E., Rec. trav. botan. neerland., 13, 1. 1919 Warburg, 0., Biochem. Z., 1., 100, 230. 1920 Lewis, F. J., and Tuttle, G. M., Ann. Botany, 34, 405. 1921 Henrici, M., Verhandl. Naturforsch. Ges. Basel, 32, 107. MacDougal, D. T., and Working, E. B., Carnegie Inst. Yearbook, 20, 47. 1923 Wurmser, R., and Jacquot, R., Bull. soc. chim. biol, 5, 305. 1924 Harder, R., Jahrb. wiss. Botan., 64, 169. Harvey, R. B., Science, 60, 481. Lepeschkin, W., Kolloidchemie des Protoplasmas. Springer, Berlin. Lundegardh, H., Biochem. Z., 154, 195. 1927 Stocker, 0., Flora, 121, 334. 1928 Miiller, D., Planta, 6, 22. Yoshii, J., ibid., 5, 681. 1929 Ehrke, G., ibid., 9, 631. 1930 Beljakov, E., ibid., 11, 727. 1931 Ehrke, G., ibid., 13, 221. Ivanov, L. A., and Orlova, I. M., Zhur. Russ. Botan. Obshchestva, 16, 139. 1932 Wood, J. G., Australian J. Exptl. Biol. Med. Sci., 10, 89. 1933 Printz, H., Nytt Magazin Naturvidensk., 73, 167. 1934 van der Paauw, F., Planta, 22, 396. 1935 B§lehrd,dek, J., Temperature and Living Matter. Borntraeger, Berlin. Barker, H. A., Arch. Mikrobiol, 6, 141. Lampe, H., Protoplasma, 23, 534. Stocker, 0., Planta, 23, 402. 1937 Craig, F. N., and Trelease, S. F., Ain. J. Botany, 24, 232. 1939 StMfelt, M. G., Planta, 29, 11. Stalfelt, M. G., Svensk Botan. Tid., 33, 383. 1940 Inman, 0., J. Gen. Physiol, 23, 661. Kennedy, S. R., Am. J. Botany, 27, 68. Noddack, W., and Kopp, C, Z. physik. Chem., A187, 79. 1944 Freeland, R. 0., Plant Physiol, 19, 179. Decker, J. P., iUd., 19, 679. 1950 Thomas, J. B., Enzymologia, 5, 186. 1953 Sorokin, C., and Myers, J., Science, 117, 330. 1954 Sorokin, C., Thesis, Univ. of Texas. C. Temperature Coefficient and Heat of Activation of Photosynthesis 1876 Prjanishnikov, J., V. Congress of Russian Naturalists, Warsaw, 1876. 1887 Kreusler, U., Landw. Jahrb., 16, 711. 1888 Kreusler, U., ibid., 17, 161. 1890 Kreusler, U., ibid., 19, 649. 1904 Matthaei, G. L. C., Trans. Roy. Soc. London, B197, 47. 1905 Blackman, F. F., and Matthaei, G. L. C., Proc. Roy. Soc. London, B76, 402. Blackman, F. F., Ann. Botany, 19, 281. 1906 Lubimenko, V. N., CompL rend., 143, 609. 1911 Blacliman, F. F., and Smith, A. M., Proc. Roy. Soc. London, B83, 389. 1256 THE TEMPERATURE FACTOR CHAP. 31 1914 Putter, A., Z. allgem. Physiol, 16, 574. 1915 Kanitz, A., Temperatur und Lebensvorgdnge. Borntraeger, Berlin. 1916 van Amstel, J. E., Rec. trav. botan. neerland., 13, 1. 1917 Brown, W. H., and Heise, G. W., Philippine J. Sci., C12, 1. 1918 Willstiitter, R., and Stoll, A., Untcrsuchungen ilber die Assimilation der Kohlensciure. Springer, Berlin. 1919 Osterhout, W. J. V., and Haas, A. R. C, J. Gen. Physiol, 1, 295. Smith, A. M., Ann. Botany, 33, 517. Warburg, 0., Biocliem. Z., 100, 258. 1921 Henrici, M., Verhandel Naturforsch. Ges. Basel, 32, 107. 1924 Crozier, W. J., J. Gen. Physiol, 7, 189. Lundegc^rdh, H., Biochem. Z., 154, 195. Yabusoe, M., ibid., 152, 498. 1926 Romell, L. G., Jalirh. wiss. Botan., 65, 739. 1927 Romell, L. G., Flora, 121, 125. Stocker, 0., ibid., 121, 334. 1928 Yoshii, Y., Planta, 5, 681. 1929 Emerson, R., J. Gen. Physiol, 12, G23. Ehrke, G., Planta, 9, G31. 1930 Beljakov, E., ibid., 11, 727. 1931 Ehrke, G., ibid., 13, 221. 1932 Emerson, R., and Arnold, W., /. Gen. Physiol. 15, 391. Emerson, R., and Arnold, W., ibid., 16, 191. van der Paauw, F., Rec. trav. botan. neerland., 29, 497. 1934 Emerson, R., and Green, L., J. Gen. Physiol, 17, 817. van der Paauw, F., Planta, 22, 396. Kautsky, H., and Spohn, H., Biocliem. Z., 274, 435. 1935 Barker, H. A., Arch. Mikrobiol, 6, 141. B^lehrddek, J., Temperature and Living Matter. Borntraeger, Berlin. Miller, E. S., and Burr, G. 0., Plant Physiol, 10, 93. Singh, B. N., and Kumar, K., Proc. Indian Acad. Sci., Bl, 736, Stocker, 0., Planta, 24, 402. 1937 Burton, A. C., J. Cellular Comp. Physiol, 9, 1. Craig, F. N., and Trelease, S. F., Am. J. Botany, 24, 232. Franck, J., and Herzfeld, K. F., J. Phys. Chem., 41, 97. Wohl, K., Z. physik. Chem., B37, 169. Stalfelt, M. G., Planta, 27, 30. 1938 Miiller, D., Planta, 29, 215. Wassink, E. C., Vermeulen, D., Reman, G. H., and Katz, E., Enzymologia, 5, 100. 1939 StMfelt, M. G., Planta, 29, 11. 1940 Emerson, R., and Lewis, C. M., Carnegie Inst. Yearbook, 39, 154. Shri Ranjan, J. Indian Botan. Soc, 19, 91. Noddack, W., and Kopp, C., Z. physik. Chem., A187, 79. Wohl, K., New Phytologist, 39, 33. 1941 Franck, J., French, C. S., and Puck, T. T., J. Phys. Chem., 45, 1268. BIBLIOGRAPHY TO CHAPTER 31 1257 Emerson, R., and Lewis, CM., Am. J. Botany, 28, 789. Katz, E., Wassink, E. C, and Donestein, R., Photosynthesis Symposium, Chicago (unpublislied). 1942 Johnson, F. H., Brown, D., and Marsland, D., Science, 95, 200. Wassink, E. C, Katz, E., and Dorrestein, R., Enzymologia, 10, 285. 1945 Wassink, E. C, and Kersten, J. A. H., ihid., 11, 282. 1948 Tamiya, H., Huzisige, H., and Mii, S., Bot. Mag. Tokyo, 61, 39. Chapter 32 THE PIGMENT FACTOR* 1. Relation between Light Absorption and Pigment Content The Hght energy used in photosynthesis is taken up by the pigments. In the reaction kinetics of photosynthesis, the "pigment factor" is there- fore closely related to the "light intensity factor." The rate of absorption (number of quanta absorbed per unit time in unit volume) is affected by changes in the concentration of the absorbing pigments, c, as well as by alterations in the intensity of illumination, I. In a homogeneous system, the relation between these two factors is determined by Beer's law: (32.1) A = 7(1 - 10-"'^'^) where a is the absorption coefficient of the material and d the thickness of the absorbing layer. We note that, whereas / is a proportionality factor, c stands in the exponent. Therefore, the rate of absorption increases proportionately with light intensity, but slower than proportionately with the concentration of the absorbent. Only in the limiting case, when acd 'a) = 156/i^a (seconds) (44 is the molecular weight of carbon dioxide, and 900 the approximate average molecular weight of chlorophyll) . For "normal" green leaves, the ^a values of Willstatter and Stoll were between 20 and 25 seconds. Values of this order of magnitude have been found for leaves the chlorophyll content of which differed by a factor of three or even more. The absorption of light by these leaves probably did not vary by more than 10 or 20% (even though some of them were "dark green" and others "light green"). In the saturation region, such a varia- tion of absorption could not in itself cause a marked change in the rate of photosynthesis. Therefore, the fact that the saturation rate did change approximately proportionately with [Chi ] is an indication that the velocity of the dark, rate-limiting process was in these plants proportional to the content of chlorophyll. Since the maximum rate of an enzymatic reaction usually is proportional to the available amount of the enzyme, we conclude that in this case the "limiting" catalyst was either chlorophyll itself, or a compound the quantity of which in "normal" plants is (approximately) pro- portional to that of chlorophyll. The possibility of identifying the rate- limiting catalyst with chlorophyll was discussed in chapter 28 (p. 1030) . The alternatives, presented there, were either to attribute saturation (in continuous light) to a catalyst the concentration of which is roughly equiva- lent to that of chlorophyll (one such possibility being chlorophyll itself), and the working period of which is of the order of 10 seconds, or to attribute it to a catalyst whose concentration is a thousand times .smaller and working time a thousand times shorter, i. e., of the order of 0.01 sec. Flashing light experiments, to be described in chapter 34, were quoted there as demon- strating directly the existence of a catalyst of the second type, with respect to both concentration and working period. Chlorophyll could be a "com- peting" Hmiting factor, imposing a "ceihng" only shghtly higher than that imposed by the catalyst revealed by flashing light experiments; but the deviations from constancy of va, shown by the "abnormal" objects of Will- statter and Stoll (Table 28. V) , do not support this hypothesis. These devia- tions can be as wide as 100% (va = 15) in green leaves and very much larger in aurea leaves. In Table 28. V we find, for three aurea varieties, the values Pa = 78, 82 and 117, respectively. If the photosynthesis of these plants were limited, in strong light, by the requirement that, after having partici- pated in the primary photochemical act, each chlorophyll molecule has to 1264 THE PIGMENT FACTOR CHAP. 32 spend 10 or 20 seconds in a photochemically inactive state, assimilation times of less than 10 seconds would be impossible; in fact, however, the above-quoted va values of yellow leaves correspond to assimilation times of 1.5 to 2 seconds. Apparently, nature had deprived these leaves of nine tenths of the normal chlorophyll content, without a corresponding reduc- tion in the concentration of the rate-limiting enzyme. One can conclude, from the behavior of aurea leaves, that, intrinsically chlorophyll molecules do not require periods of the order of 10 sec. to com- plete the part of photosynthesis in which they are directly involved, but are available for a new photochemical act within a much shorter time. Related to the same subject are the experimental results of Shcheglova (1940). She found that the light-green parts of leaves of Polygonum sacchalinense have a higher natural rate of photosynthesis in the first half of the summer than the dark-green ones, while in the second half of the summer (when the average light intensity is lower) the relation is reversed. In laboratory experiments with the same plant at different tem- peratures, she found that at 20-30° C, the light-green sections of the leaves (1.2-1.8 mg. chlorophyll per 100 sq. cm.) give a higher yield in strong light than the dark-green ones (2.2-3.1 mg. chlorophyll per 100 sq. cm.), and thus have a much higher assimilation number (6.6-5.5 vs. 3.7-2.5); while at 13-17° C, the relation is reversed, with the result that the assimilation numbers are about equal (2.2 and 2.4). In interpreting differences in assimilation numbers as indicative of varia- tions in the content of a rate-limiting enzyme, one has to keep in mind that the maximum rate of an enzymatic reaction may be affected, in addition to the concentration of the enzyme, also by the properties of the medium in which the enzyme has to act. Not only the reaction (the pH) of the medium, but also its colloidal properties may influence this rate. This is particularly true if the rate-limiting step is bimolecular (e. g., if it involves the encounter of a substrate molecule with an enzyme molecule) ; but even if the rate-limiting step is monomolecular (transformation of the reaction complex {enzyme -f- substrate! ), its rate nevertheless may be affected by association with colloidal particles. Dastur (1924, 1925) and Dastur and Buhari walla (1928) noted that in aging leaves the water content declined more rapidly than the chlorophyll content, and suggested that variations in water content could provide an explanation for the drop in the as- similation number, reported by Willstatter and StoU for aging leaves. The state of hydration of the chloroplasts could perhaps affect the rate constant of the limiting en- zyme, even if the concentration of the latter remains unchanged. Conditions similar to those in aurea Avere found in some autumn leaves. According to Willstatter and Stoll, the autumnal decrease in photo- synthesis usually parallels the disappearance of chlorophyll, so that the assimilation numbers remain approximately constant. Sometimes, how- ever, the decrease of P'^^''- lags behind that of [Chi], so that a transient increase in va to 12-20 occurs in early autumn, before the final dechne has EFFECT OF NATURAL VARIATIONS OF CHLOROPHYLL CONTENT 1265 set in. On the other hand, leaves may remain green in the autumn and nevertheless show declining v^ values; sometime these green but inactive leaves recover their efficiency after having been returned to higher temper- ature for several hours. To sum up, Willstatter and Stoll's determinations of va. fmd Ia suggest that the concentration of the limiting catalyst has a tendency to change proportionately with [Chi], but that this rule admits of many, and sometimes^ striking, exceptions. o 00 o o E CO CO z >- CO o H o X a. 120 100 Green leaves 0.25 0.50 075 1.00 LIGHT INTENSITY, fractions of full sunlight Fig. 32.2. Light curves of green and aurea leaves of Sambucus (after Willstatter and Stoll 1918). In addition to numerous experiments on the assimilation numbers of various leaves in strong light, Willstatter and Stoll also determined the complete light curves of the green and aurea leaves of Sambucus. The results are represented in figure 32.2. The yellow leaves contained only 5-10% of the normal chlorophyll content, and probably absorbed not more than one third of the amount of white light absorbed by the green forms (c/. page 685). This difference may explain the slower rise in the rate of photosynthesis of yellow leaves in weak light, and their failure to reach complete saturation even in the strongest light used. The extrapolated saturation value of the yellow variety appears, however, to be equal to, or only slightly below, that of the green one. In other words, figure 32.2 agrees approximately with the prototype of the two solid curves in figure 32.1B (and differs from that of fig. 28.20(1)). 1266 THE PIGMENT FACTOR CHAP. 32 In some aurea leaves, characterized by extreme chlorophyll deficiency, the rate of photosynthesis in light of 45,000 lux was found to be much lower than in the moderately chlorophyll-deficient yellow leaves, to which figure 32.2 refers. For example, a leaf of Sanibucus containing less than 0.1 mg. chlorophyll/ 10 g. fresh weight assimilated, in strong light, only 12 mg. CO2/hr./10 g. (as against 90 mg. taken up by a leaf containing 0.75 mg. chlorophyll, and 150 mg. taken up by a fully green leaf with 23.5 mg. chloro- phyll in 10 g.). These figures show that the assimilation number of the leaf with 0.5% of the normal quantity of chlorophyll was about the same (~120) as that of the leaf with 3% of the normal chlorophyll content, both being about eighteen times larger than the assimilation number of a normal leaf. However, by analogy with the curves in figure 32.2, it appears possible that the leaf with extreme chlorophyll deficiency was far from Hght-saturated at 45,000 lux. If the light curve of this leaf, followed to much higher intensities, would also approach the normal saturation value, this would mean an assimilation number of over 1000 {i. e., an assimilation time of less than 0.15 second)! It would be interesting to find out whether such extremely high assimilation numbers actually occur. It was mentioned above that Gabrielsen (1948) found the rate of photosynthesis of aurea leaves in weak light to be even lower than expected from their light absorption, and attributed this deficiency to "inactive" light absorption in cell walls, plasma and vacuoles. No such deficiency is apparent in Willstatter and StoU's curves (fig. 32.2); these curves, however, cannot be evaluated quantitatively because of the absence of correlated absorption data. Gabrielsen's absorption estimates were based not on his own measurements, but on Seybold and Weissweiler's data (chapter 22); consequently, they, too, are not very reliable. Correct estimation of scattering losses is very important in the determination of true absorption by leaves, particularly when the latter are poor in pigments; and figures given in the hterature for "inactive absorption" of visible light by "colorless" leaf constituents (p. 684) may be much too high because of unsatisfactory methods of integrating scattering losses. Another experiment which also indicated the essential independence from chlorophyll of the catalyst that limits the rate of photosynthesis in strong light was described by Emerson (1935). He found that, when a suspension of Chlorella was illuminated by strong light for 16 hours, the volume of the cells was almost trebled, without appreciable change in the total quantity of chlorophyll. Thus, the average concentration of chloro- phyll in the suspension was unchanged, but its concentration within the cells was smaller than before, by a factor of three. Photosynthesis was found to be about twice as large as before {v^ about 4 before the treatment, and about 8 afterward) . Only a small part of this improvement can be at- tributed to a more efficient absorption of light by the preilluminated sus- pension (caused by more uniform distribution of chlorophyll and conse- quent dechne in the "sieve effect"); the result can therefore be taken as indication that, after multiplication, the cells contained about twice as much of the rate-limiting catalyst as was present in the original suspension, while their total content of chlorophyll was unchanged. A striking demonstration of the existence of an enzymatic component that can limit the rate of photosynthesis, and is independent of chlorophyll INFLUENCE OF ARTIFICIAL CHANGES OF CHLOROPHYLL CONTENT 1267 (since it can be inactivated without noticeable damage to chlorophyll), is provided by the experiments of Davis (1948, 1952), who obtained Chlorella mutants containing normal chlorophyll, but incapable of photosynthesis. 3. Influence of Artificial Changes of Chlorophyll Content on Photosynthesis The experiment of Emerson, described at the conclusion of the preced- ing section, forms a transition from the study of natural to that of artificial changes in chlorophyll content. Since we cannot extract or destroy a part of chlorophyll present in the cells without killing the cells, the desired artificial changes in chlorophyll content have to be obtained indirectly, by varying the conditions of culturing, so as to induce the plants to produce less (or more) than their normal complement of pigments. As the first experiments of this kind, we may consider those in which Willstatter and Stoll (1918) found abnormally high va values in etiolated plants, i. e., seedlings grown in the dark, just in process of becoming green. For example, the assimilation number of yellowish-green etiolated Phaseo- lus vulgaris plantules was as high as 133 (as compared with 9.4 in a green control plant). The chlorophyll concentration was 0.7 mg. in 10 g. fresh leaves of the etiolated plant, and 18.6 mg. in the control specimen. These results of Willstatter and Stoll disagreed with the earlier conclusions of Irving (1910) (who worked in Blackman's laboratory); the latter had found that etiolated seedlings do not photosynthesize at all until they have acquired a considerable amount of chlorophyll. Inman (1935) found that the photosynthesis of etiolated plantules begins simultaneously with the appearance of green color. Beber and Burr (1937) reported that, in etiolated oat seedlings, photosynthesis did not begin until some chlorophyll h was formed (not confirmed bj' Smith, cf. p. 1766). Smith (1949) plotted the rate of photosynthesis (at 45 klux) of etiolated bean and corn seedlings in the process of greening, as observed by Willstatter and Stoll, as function of (1 — e~ <'°'^s*- [Chi] ^^ g^j^jj obtained a straight line. He interpreted this as an indication that the increase of P with [Chi] was a trivial consequence of increasing absorption. However, this kind of relationship is to be expected only in the light-limited state, and even if etiolated seedlings might not have been completely light-saturated at 45 klux, they are not likely to have been in the light-limited state in such intense illumination — particularly after the chlorophyll content had increased to 10-15 mg. in 10 g. (fresh weight). Furthermore, Smith's interpretation disregards the \).\^\ absolute assimilation numbers obtained by Willstatter and Stoll for etiolated leaves. Smith (1949) also made observations of the rate of dyestuff reduction (Hill reaction) by chloroplasts extracted from etiolated barley seedlings at different stages of greening, and found a systematic increase in activity, similar to that observed for photosynthesis. More recent studies of greening, indicating transient formation of a photosyntheti- cally inactive chlorophyll form, will be described in chapter 37B (p. 1767). 1268 THE PIGMENT FACTOR CHAP. 32 Chlorotic plants, i. e., plants in which the clilorophyll development has been arrested by nutritional deficiencies, also have been investigated by Willstatter and Stoll. The assimilation numbers of plants of Helianthus annuus and Zea mats, made chlorotic by iron deficiency, were found to be not very different from those of normal plants (sometimes smaller, some- times larger). It thus seems that the development of the enzymatic ap- paratus has been held back together with that of chlorophyll. 100 to > (/) o I- o I Q. li. O UJ I- < 0.025 0.050 0.075 0.100 RELATIVE CHLOROPHYLL CONCENTRATION Fig. 32.3. Relation between chlorophjdl concentration and photosynthesis of Chlorella in strong Ught (after Emerson 1929). Two series of experiments. These results of Willstatter and Stoll were confirmed and amplified by Emerson (19290, who investigated cultures of Chlorella vulgaris grown in nutrient solutions with a variable concentration of iron. Glucose was added to support growth in iron-deficient solutions. In this way, varia- tions of [Chi] in the ratio 1:5 could be obtained. The maximum photo- synthetic rate (in Warburg's buffer No. 9, and in light of about 10^ lux) was found to increase regularly — but more slowly than proportionately — with the chlorophyll content {cf. fig. 32.3). The high light intensity used by Emerson made it improbable that the increase in P with [Chi] could be in- terpreted as a conseciuence of incomplete light saturation of the chlorophyll- deficient cells; this conclusion was confirmed by a second investigation (1929^), in which complete light curves were determined for two Chlorella INFLUENCE OF ARTIFICIAL CHANGES OF CHLOROPHYLL CONTENT 12G9 cultures with chlorophyll contents differing in the ratio of 4:1. As shown by figure 32.4, the saturation yields of these two forms in fact differed widely (by a factor of 3.4, which is only 15% smaller than the ratio of the chlorophyll contents). Emerson's light curves conform closely to the prototype of figure 28.20(/). In this case the reduction of the chlorophyll content by three c]uarters, l)rought about by iron deficiency, had the same effect on photo- synthesis as would have been caused by removal of three fourths of all cells from a nonchlorotic suspension. In other words, chlorotic Chlorella suspen- sions, hke Willstatter and Stoll's chlorotic leaves, behaved as "normal" leaves. However, since we have described above several cases in which the Chlorotic 200 RELATIVE LIGHT IMTENSITY Fig. 32.4. Light curves of two Chlorella suspensions with different content of chlorophyll (after Emerson 1929). Double arrow indicates the "linear range." relation was different {aurea leaves, autumn leaves, shade leaves and etio- lated leaves), we are bound to conclude (as we did once before) that the two components of the photosynthetic apparatus — chlorophyll and the rate- limiting catalyst — are not identical ; and that even if the ratio of their quan- tities often has a tendency to remain constant, this gives us no right to con- sider the tw^o components as associated in a proportion that remains con- stant under all conditions. The three different relationships between [Chi] and P""*^- are again compared in figure 32.5, where light curves A are those of Emerson's normal and chlorotic Chlorella cells (P'"'^*- proportional to [Chi]), while light curves B correspond to Willstatter and Stoll's green and yellow leaves (•pmax. roughly independent of [Chi]), and light curves C to the umbro- philic and heliophilic plants of Lubimenko (P'"^"- antiparallel with [Chi]). The dependence of P"''''' on temperature was found by Emerson to be unaffected by changes in the chlorophyll content of Chlorella cells. This result is different from Willstatter and Stoll's observations with aurea 1270 THE PIGMENT FACTOR CHAP. 32 leaves, shown in Table 31.IV; but the reason for this difference may be simply that the aurea leaves were studied in a state of incomplete light saturation (c/. fig. 32.2), while Emerson used Hght intensity sufficient to saturate both the normal and the chlorophyll-deficient Chlorella cells (c/. fig. 32.4). Theoretically, there seems to be no reason why the temperature coefficient of P™^'"' should change with [Chi], at least as long as the same dark reaction is rate-limiting in all the cell systems under comparison. The photosynthesis of chlorotic Chlorella cells was found by Emerson to to be more strongly affected by cyanide than that of the normal ones. This may indicate that chlorotic cells contained less of the cyanide-sensi- tive carboxylating enzyme, Ea {cf. Volume I, page 306). I I Fig. 32.5. Light curves of optically dense (— ) and optically thin (— ) systems: (A) normal and chlorotic Chlorella cells (cf. fig. 32.4); (B) normal and yellow Sarnbucus leaves {cf. fig. 32.2); and (C) shade and sun leaves {cf. fig. 28.10-18). Fleischer (1935) repeated Emerson's experiments, varying the chloro- phyll concentration of Chlorella by three types of changes in the nutrient medium: (a) changing the iron concentration, (6) changing the nitrogen concentration and (c) changing the concentrations of magnesium {cf. chapters 13 and 15). Results obtained by the methods a and h, (using Warburg's buffer No. 9, and fight of 75,000 lux) weresimilar; likeEmerson's observations, they showed an approximately linear increase in yield with the chlorophyll concentration. Method c, on the other hand, gave more complex results, indicating that magnesium deficiency had a specific effect on photosynthesis, beside the indirect influence caused by a lowering of chlorophyll concentration. Van Hille (1937), was unable to confirm Fleischer's results; he found a systematic decrease in both [Chi] and P'"*''' in plants deprived of magnesium. However, the existence of a direct ef- fect of magnesium concentration on the rate of photosynthesis was con- firmed by Kennedy (1940) {cf. chapter 13, page 337). In a second paper, van Hille (1938) reported that the photosynthetic capacity of Chlorella pyrenoidosa cultures decreased steadily with age, even if the chlorophyll content continued to increase. Experiments with varied nutrient solutions showed that this decline was connected with a develop- INFLUENCE OF ARTIFICIAL CHANGES OF CHLOROPHYLL CONTENT 1271 ing deficiency of nitrogen ; it could be remedied by repeated addition of nitrate to the suspension medium. (This influence of nitrogen concentra- tion, too, was discussed in chapter 13; cj. page 339). To sum up, experiments with nitrogen-deficient and magnesium-de- ficient cell cultures gave less consistent results as to the correlation be- tween chlorophyll concentration and the capacity for photosynthesis than did experiments in iron-deficient media. However, in the latter case, too, we probably are dealing not with direct effects of changes in chlorophyll concentration, but with variations in the general development of the photo- catalytic apparatus, which includes, in addition to chlorophyll, also the several catalysts participating in the "dark" stages of photosynthesis. Myers (1946) noted that, when Chlorella pijrenoidosa was grown in a continuous growth apparatus, the chlorophyll content per unit cell volume increased by a factor of 5 when illumination was decreased from 360 to 6 foot-candles. At the same time, however, the number of cells in one cubic centimeter of packed cell material increased from 6 X 10^ to 29 X 10^ so that the average number of chlorophyll molecules in a single cell was ahnost unchanged, the small "shade cells" containing about as much chlorophyll as the large "light cells." Among these different cultures, the highest saturation rates of photosynthesis per unit cell volume were shown —in alkahne buffer as well as in acid Knop's medium— by those grown in fight of 25-50 foot-candles; the saturation rate per unit cell volume de- cfined sharply for cultures grown < 25 foot-candles and gradually for cul- tures grown >50 foot-candles. The saturation rate per cell (and thus, be- cause of the above-mentioned constancy of the chlorophyll amount per cell, also the saturation rate per chlorophyll molecule, i. e., the assimilation number), increased continuously with increasing intensity of the culture light. (It should be noted that the "saturation rate" was measured at 600 foot-candles— a light intensity that may be insufficient to completely saturate fight-adapted Chlorella cells.) Myers noted that the maximum rate of growth of Chlorella was reached at about 100 foot-candles, and suggested that in stronger light the cells are incapable of utilizing all the immediate products of photosynthesis, and develop a special mechanism of dissimila- tion ("hght respiration," or photoxidation?) which disposes of surplus photosynthates. In other words, below 100 foot-candles growth is limited by the rate of photosynthesis, while above 100 foot-candles it is limited by a dark process (such as nitrogen assimila- tion), which the primary products of photosynthesis (carbohydrates?) must undergo to be converted into balanced cell material. In Chapter 31, (p. 1228) we described Sorokin and Myers' discovery of a "thermophilic" Chlorella pyrenoidosa strain, with a peak capacity for growth at 39° C. (instead of the usual 26° C). At its optimum tempera- ture, this strain required a higher fight intensity for saturation, and pro- 1272 THE PIGMENT FACTOR CHAP. 32 duced considerably more oxygen per unit cell volume per unit time than the ordinary strains of Chlorella pyrenoidosa do at their optimum tempera- ture. The assimilation number of such thermophilic algae at 39° C. may be considerably higher, and their assimilation time correspondingly shorter, than those that other strains can reach under the most favorable conditions. 4. Chlorophyll Concentration and Yield of Photosynthesis in Flashing Light In chapter 34, we will deal with experiments on photosynthesis in periodically interrupted light; we must here anticipate some of the results, which concern the relation between the chlorophyll content and the maxi- mum amount of oxygen that can be liberated by a single short flash of hght. This quantity can be interpreted in different ways. The interpreta- tion that seemed most natural and was therefore the first to be discussed was based on the identification of the maximum amount of substrate that can be reduced by a single flash, with the quantity of this substrate available in the cells, after a period of darkness, in a form suitable for immediate photochemical reduction (e. g., as an acceptor-carbon dioxide compound or complex, [CO2] or ACO2). In a dark-rested chloroplast, in presence of sufficient carbon dioxide, each acceptor molecule must be associated with a molecule of carbon dioxide. A sudden flash of light will send all these molecules through the photochemical reduction stage (or stages), and thus produce a quantity of primary photoproducts equivalent to the amount of the available acceptor. After this, a certain interval of time may be re- quired for the "recharging" of the photosynthetic apparatus, e. g., by the slow formation of a new quantity of the acceptor-carbon dioxide complex ("preparatory" dark reaction). The yield of oxygen, produced by a single strong flash, will then be equivalent to the quantity of the carbon dioxide acceptor present in the cells. In discussing the nature of the carbon dioxide acceptor in chapter 8, we found that it is a compound the content of which in green cells is about equivalent to that of chlorophyll. In chapter 34, we will find evidence that this acceptor may itself be a transient product of photosynthesis, and therefore decrease in concentration in the dark and increase again as photo- synthesis gets under way — perhaps approaching approximate equivalency with chlorophyll in the stationary state in strong light. If the maximum oxygen yield per flash were limited by the available quantity of this ac- ceptor, we would expect this yield, too, to be approximately equivalent to the chlorophyll content of the cells; in other words, it should be of the order of 0.01 mole per fiter of cells, or as much as 0.2 volume of oxygen per unit volume of cells per flash. CHLOROPHYLL CONTENT AND YIELD IN FLASHING LIGHT 1273 Emerson and Arnold (1932) found, however, in their first flashing Ught experiments with suspensions of Chlorella pijrenoidosa, that the maximum oxygen production per flash was less than 0.0001 cubic centimeter per flash per cubic centimeter of cells, corresponding to only one molecule of oxygen for each 2660 molecules of chlorophyll! In a second paper (1932) the same observers investigated whether the maximum production per flash had any relation to the chlorophyll concentration at all. For this purpose, they used Chlorella cells grown in light of different intensity or spectral c o a 71 O X < o o UJ _) o .5 0.5 1.0 [Chi], (moles/mm.') x 10^ Fig. 32.6. Flash yield and chlorophyll content (after Emerson and Arnold 1932). composition (neon tubes, mercury lamps and incandescent lamps). The [Chi] values of these cultures ranged from 4 X 10-^ mole/1, to 16 X lO^' mole/1, (referred to the volume occupied by the cells). The oxygen yield per flash was found to be approximately proportional to the chlorophyll con- tent {cf. fig. 32.6) ; but the proportionality factor (r) was again found to be of the order of 5 X lO-\ and not of the order of unity. In Table 32.1 we list the values of 1/r taken from the fundamental investigations of Emerson and Arnold (1931, 1932), as well as from the subsequent determinations of Arnold and Kohn (1934) and Emerson, Green and Webb (1940). The last set of figures in Table 32.1 shows that r was found to decline steadily in aging cultures. (The yield per flash diminished with age al- though the chlorophyll content remained constant or even increased.) This result reminds one of the observations of van Hille on the effect of age on rate of photosynthesis of Chlorella in continuous light (Vol. I, p. 239). Perhaps the striking decline of t also could be checked, at least to a certain extent, by avoiding nutritional deficiencies. With the exception of aged 1274 THE PIGMENT FACTOR CHAP. 32 cultures, the r values in Table 32.1 obtained with plants belonging to five different phyla all he in the range between 2000 and 5000. The value of l/r indicates the participation in the photosynthetic proc- ess of a catalytic component, x, the total available amount of which is from 2000 to 5000 times smaller than that of chlorophyll: [x]o = 2 to 5 X lO"* [Chl]o. We may consider [x]o = 4 X lO"" [Chljo as the "normal" value. This estimate may need a correction : If the catalytic component in ques- tion has to operate n times in order that one molecule carbon dioxide can be reduced to the carbohydrate level, the factor r must be multiplied by n to obtain the number of catalyst molecules required. For example, if we assume n = 8 (a plausible assumption in an "eight-quanta theory"), the concentration of the yield-limiting catalyst required to explain the "normal" value of t, will be [x]o = 3.2 X 10-^ [Chl]o. It is also possible that the limiting catalyst has to act only on one set of intermediates {e. g., on the reduction intermediates of carbon dioxide, and not on the oxidation intermediates of water, or, vice versa). In this case, t will have to be multi- phed by n/2 instead of by n and the "normal" value of [x]o will be 1.6 X 10~'[Chl]o. It is more difficult — although not impossible — to imagine Table 32.1 Maximum Oxygen Yield" in Flashing Light [Chi], Observers Species (moles/mm.') X 10» I/7 Emerson, Arnold Chlorella pyrenoidosa 0 . 428 2,660 ^10-^2^ 0-606 1,980 ^^^•^^^ 0.620 2;660 0 . 766 2,380 Emerson, Arnold Chlorella pyrenoidosa {25° C.) 0.756 2,010 (■1932^') with variable chlorophyll 1.07 2,380; 2,780 ^ ^ content 1.14 3,120 1.49 2,310 1.59 2,460 Bryophyllumcalycinuin {31.5°) — 2,500; 2,600 Chlorella vulgaris {25°) — 2,800 Arnold, Kohn (1934) Lemna sp. (25, 26, 29°) — 2,600; 3,200; Nicotiana lon.gsdorffii (30, 29 °) — 2,800 ; 2,500 Selaginella sp. (28°) — 4,200 Stichococcus bacillaris (25, 30°) — 3,700; 5,000 Emerson, Green, Chlorella pyrenoidosa, different age Webb (1940) 2 days^ — 3,750 4 days" — 6,180 8 days'- - 11.120 15 days<= — 6,650 21 days" — 11,000 29 days" — 14,500 Arnold {cf. van Niel Rhodospirillurn rubrum — 400 1941) <• T = oxygen yield, molecules per flash per molecule chlorophyll. " Four 40 w. lamps 10 cm. from cultures during growth. " One 40 w. lamp 15 cm. away. CHLOROPHYLL CONTENT AND YIELD IN FLASHING LIGHT 1275 mechanisms in which the limiting catalyst will have to act only once (or, more generally, less than four times) in the reduction of one molecule of carbon dioxide and the production of one molecule of oxygen. Even with [x]o = 8[Chl]oT, the concentration [x]o remains much too low for x to be identified with the carbon dioxide acceptor, A. It must thus be another component of the catalytic mechanism. According to the classification we have used often before, it may be either a preparatory or a finishing catalyst. The preparatory catalysts work on stable sub- strates ; therefore nothing seems to prevent them from accumulating more and more products as the dark intervals between flashes increase in length ; and this should permit the maximum yield per flash to increase indefinitely. The kind of preparatory catalysts to which this consideration does not apply are "acceptors" (similar to the repeatedly discussed carbon dioxide ac- ceptor, A) that have a limited capacity and, once "filled," can be emptied only by, or in consequence of, the light reaction. We can postulate, ad hoc, such an acceptor, e. g., on the "reduction side" as an additional intermediate between the first carbon dioxide compound, ACO2, and the chlorophyll complex, or, on the "oxidation side," where a water acceptor compound, A'H20 (or, more generally, an intermediate hydrogen donor, RH2), was postulated on previous occasions. Another, and more plausible possibility is, however, to relate the maximum yield per flash to the limited availability of a finishing catalyst, i. e., a catalyst acting on the products of the primary photochemical process. Of course, if these intermediary photoproducts were stable (meaning by "stability" that their life-time is much longer than the working period, t^, of the catalyst, x), then the catalyst could continue working, in the dark, until it has completely exhausted the products of the preceding flash; in this case, the yield per flash would again become capable of increasing indefinitely with the intensity of the flash, if the dark intervals between flashes are increased correspondingly. However, it is plausible to assume that the intermediary photoproducts are unstable; and, if their ex- istence is so fleeting that a "second batch" cannot wait until the "first batch" had been processed and stabilized by the catalyst, the maximum yield per flash will be equal to the size of this first batch, and therefore equiv- alent to the available amount of the stabilizing catalyst (more exactly, it win be equal to this amount, [x], divided by the number of times the cata- lyst has to operate in the reduction of one molecule of carbon dioxide to carbohydrate). No useful purpose will be served, in this case, by further increasing the energy of the flash, or prolonging the dark intervals. One could suggest that the figure obtained in this way (3.2 X 10~'[Chl]o) is not necessarily equal to the concentration of the catalyst x, but may represent a multiple of it— the product of [x]o and the number of batches the catalyst is allowed to process after a flash. We have assumed that this number is 1, i. e., that the Ufe-time of the 1276 THE PIGMENT FACTOR CHAP. 32 photoproduct is small compared with the working period of the catalyst. But what if this life-time is long enough to enable each molecule of the catalyst to work an average of two or three times before it has to stop because the substrate has disintegrated? If this were so, the necessary number of catalyst molecules would be only one half or one third of the above-calculated figures. However, the duration of the dark interval neces- sary to obtain the maximum yield per flash, would then change with increasing intensity of the flash. After a weak flash, a single working period would be sufficient to process all photoproducts; after a stronger flash, two such periods would be required, to permit a second "round" of catalytic activity, and so on, until the full life-time of the photo- products is utilized. No such dependence of the length of dark intervals on flash in- tensity required to obtain the maximum yield per flash has been noted by Emerson and co-workers. The attribution of the maximum flash yield to a back reaction destroy- ing all photoproducts that cannot be immediately stabilized by a catalyst was suggested by Franck, and the concept was developed in detail by Franck and Herzfeld (1941) in conjunction with the reaction mechanism illustrated by scheme 7.VA, in which the "stabilizing" catalyst, Eb, is as- sumed to provide the limiting influence. However, the same hypothesis can be combined also with the other reaction schemes discussed in chapters 7, 9, 24 and 28, insofar as these schemes, too, contain "finishing" catalytic reactions that follow the primary photochemical process (or processes). In chapter 28 we concluded, from an analysis of the phenomena of light saturation, and of changes in chlorophyll fluorescence in strong light, that the hght saturation of photosynthesis is caused (usually, but probably not always) by the bottleneck of a finishing catalytic reaction. "Flash satura- tion" may thus be due to the same limiting agent as saturation in continu- ous Hght; and this hypothesis is strengthened by quantitative comparison of the two saturation yields. In making this comparison, we assume that the maximum velocity of the rate-hmiting catalytic reaction can be represented by a product of a con- centration factor, such as [^b], and a velocity constant, /cb (i. e., that this reaction is a monomolecular transformation of a catalyst-substrate com- plex). If the maximum yield in continuous light is hmited by the same "bottleneck" as the maximum yield per flash, the concentration of the catalyst limiting the rate in continuous light must have the value derived above from the ratio r, namely (assuming n = 4), [x] = 0.0016 [Chl]o. Since PJSax: is of the order of 0.05 [Chl]o (^a = about 20 seconds), we can write 0.0016 [Chljo X k = P^H] = 0.05 [Chl]o, and thus obtain k = 30 sec. -\ When the yield per flash was measured as function of the length of the dark interval between flashes, a value of about 50 sec.-^ was in fact found for the monomolecular rate constant of the reaction by which photo- synthesis is completed after the flash. (For the description of these experi- ments, see chap. 34, section B2.) In other words, the product k\E] de- CHLOROPHYLL CONTENT AND YIELD IN FLASHING LIGHT 1277 rived from P^ll'., the saturation rate in continuous light, agrees with the product of the values [Eb] and [/cb] derived separately, from the saturation yield in flashing hght and the flash yield dependence on the length of the dark intervals, respectively. The agreement supports strongly the as- sumption that saturation in continuous light and the maximum flash yield are determined by the same "bottleneck" reaction, brought about by a finishing catalyst present in a concentration of the order of 10 ~^ mole/1. (~0.1% of fChlJo) and having an "action period" of the order of 1/50 second (at room temperature). No flashing Hght experiments have been carried out with objects for which abnormal assimilation numbers had been found in continuous hght, such as aurea leaves, etiolated seedlings or autumn leaves. If the hypothe- sis that the rate-limiting reaction is the same in continuous and in flashing light applies to all of them (which is not necessarily true, because under abnormal conditions the limiting role may be taken over by a different re- action), then the plant objects showing exceptionally high assimilation numbers, pa, should exhibit also exceptionally high values of the ratio r. In Myers' (1946) experiments with Chlorella cells grown in light of different intensity, the saturation rates per unit chlorophyll amount (and thus also the ratios va) varied by a factor of 4 as the culture light increased from 6 to 360 foot-candles, and the chlorophyll concentration within the cells dechned by a factor of 5. In table 32.1, Emerson and Arnold's figures show no significant change associated with chlorophyll concentration changes by a factor of 2, produced in a similar way {i. e., by changes in illumination during the growth of the algae). However, only experiments performed on the same algae in both steady and flashing light could pro- vide truly significant information as to the correlation between va and t. In chapter 34 we will describe experiments by Tamiya and co-workers in which flash yields far exceeding the Emerson-Arnold hmiting value of 5 X 10~^ [Chi] have been obtained by further increasing the energy of the flashes and by prolonging the dark intervals beyond the Emerson-Arnold limiting duration of about 0.02 sec. For reasons to be presented there, we are reluctant to consider these experiments as sufficient to cast aside all the conclusions based on the observations of Emerson and Arnold; but a renewed experimental study of flashing light saturation seems necessary. In identifying the hmiting catalyst with the catalyst Eb in schemes 28.1 and 28.11, we follow Franck and Herzfeld. In their scheme, (7.VA), Eb was supposed to act (with equal efficiency) on four (different) intermedi- ate products "on the reduction side" and four (identical) intermediate products "on the oxidation side." The assumption of several processes on the reduction side, all catalyzed by Eb, can be avoided by postulating only one photochemical reduction process, followed by catalytic dismuta- 1278 THE PIGMENT FACTOR CHAP. 32 tions (in fact, Eb can be a "mutase" bringing about the first of these dismu- tations). The assumption of a stabilizing reaction on the oxidation side, also catalyzed by Eb, seems unnecessary, since the limitation of the jdeld of finished reduction products will automatically bring about also a limita- tion of the yield of finished oxidation products. (Whenever the production of AHCO2 exceeds the capacity of Eb, the intermediates AHCO2 will accu- mulate, until their back reaction with A'HO will be able to compete with the oxygen-liberating reaction, catalyzed by Eel the photosynthetic quotient will thus be rapidly reduced, after an initial, unbalanced "induction" pe- riod to the normal value of unity. Alternatively, if the stabilizing catalyst is assumed to operate on an instable oxidation product (|0H} or A'OH), there would seem to be no need to assume also a stabilizing reaction on the reduction side. The postulate of a single photochemical oxidation-reduction step, fol- lowed by dismutation, is found, in chemically more concrete form, in re- cent speculations on the mechanism of carbon dioxide reduction in photo- synthesis, derived from C(14) experiments (c/. chapter 36, section 12). We will now consider, from the point of view of the theory of the "finishing bottleneck," the high quantum yield of photosynthesis and the absence of an extended induction period in weak light — these being the two kinetic observations that have played a decisive role in the development of the theory of the "photosynthetic unit" (to be described in section 5). In the early discussions of this subject, Warburg's original value of the quantum yield (7 = 0.25) was used; but even if we substitute the smaller value (7 = 0.12-0.15), which now appears more plausible (c/. chapters 29 and 37D), it still remains true that probably all, or almost all, of the light quanta absorbed by chlorophyll in weak light actually can be utilized in the photo- synthetic process. The high quantum yield in weak fight (the explanation of which offers some difficulty for a theory which would assume that only a small fraction of chlorophyll molecules are associated with appropriate reduction sub- strates) obviously does not embarrass a theory such as Franck's, which postulates that, except for cases of carbon dioxide deficiency or specific poisoning of the carboxylase, Ea, all chlorophyll molecules are associated with photosensitive complexes, even when light saturation sets in. The absence of an extended induction period in weak fight {cf. chapter 33) re- quires, however, somewhat closer consideration. This fact indicates that the collection of the 4 or 8 quanta required for the reduction of one molecule of carbon dioxide does not require that these quanta be absorbed by the same molecule of chlorophyll. In chapter 25 (page 838) we derived a relationship between light intensity and the fre- quency of absorption acts by a single molecule : (32.4) n = 4 X 10-2iaAr^^sec.-i CHLOROPHYLL CONTENT AND YIELD IN FLASHING LIGHT 1279 where A^^, is the rate of incidence of Hght quanta per second per square centimeter. The highest quantum yields of photosynthesis have been ob- served in light of the order of 100-1000 lux, corresponding to A^^, = 10 1'^ 10^^; with a ^ 10\ this means n = 4 X 10-» to 4 X 10"^, i. e., each chloro- phyll molecule absorbs a quantum every 25-250 seconds, jf a carbon di- oxide molecule were to remain anchored, during the whole reduction process, at one chlorophyll molecule, waiting for 4 (or 8) quanta to be absorbed by the latter, this should take, in the light of the above-mentioned intensity (100-1000 lux), from 100 to 1000 sec. for 4, and from 200 to 2000 sec. (i. e., up to one half hour) for 8 quanta. In completely absorbing, dense cell suspensions, such as were used by Warburg and Negelein, Emerson and Lewis, and others, in quantum yield determinations, the average light inten- sity was lower than 100 lux, and the frequency of absorption acts by a single chlorophyll molecule must have been only one every 10 or 20 minutes, cor- responding to 1.5 to 3 hours for 8 quanta! If one would start, after a dark period, with chlorophyll deprived of all intermediates, and all chlorophyll molecules associated with ACO2 mole- cules, over 1 hr. of illumination would thus be required, under the assumed conditions, for the uptake of new carbon dioxide to reach the steady rate. The Hberation of oxygen, on the other hand, could become steady after a quarter or an eighth of this period, because, in the most plausible reaction mechanisms, the final oxidized photoproduct (ROH in scheme 7.VA and A'OH in scheme 28.1, etc.) is obtained in consequence of a single photo- chemical step, and is converted to molecular oxygen entirely by dark reac- tions (e. g., combination and dismutation of radicals, such as 2 [OH] -^ [H2O2]— [H20] + 02). Experiments have shown, however, that not only the liberation of oxy- gen, but also consumption of carbon dioxide begins very quickly upon il- lumination after a dark period, even in very concentrated suspensions. This fact can be explained, from the point of view adopted in this section (z. e., without recourse to the hypothesis of a "photosynthetic unit") in two ways, depending on whether one adopts Franck and Herzf eld's scheme, (7.VA), in which the reduction of carbon dioxide is achieved by four consecutive photochemical steps, or chooses one of the reaction schemes (e. g., scheme 28. lA) in which a single photochemical reduction step is combined wdth repeated catalytic dismutations. In the first case, one has to assume either that the intermediate reduc- tion products, AHCO2, AH2CO2 . . . , are so stable that they can survive a very prolonged period of darkness, or that the stock of these intermedi- ates is continuously replenished in darkness in consequence of slow reversal of photosynthesis by an autoxidation process, different from normal respira- tions, occurring within the chloroplasts. In either of these two ways, the 1280 THE PIGMENT FACTOR CHAP. 32 photosynthetic apparatus can be kept stocked, in darkness, with a full as- sortment of intermediates, so that, when illumination starts, carbon di- oxide absorption can begin immediately, even in very weak light. If we assume only one photochemical step, followed by catalytic dis- mutations, the co-operation of 4 or 8 light quanta in the reduction of one molecule of carbon dioxide can be explained, without the assumption of stable or continuously regenerated intermediates, by catalytic, nonphoto- chemical reactions among several identical, primary photochemical prod- ucts. The velocity of these reactions is in no way limited by the rate of absorption of light quanta by a single chlorophyll molecule, since the inter- mediate photoproducts, formed by several chlorophyll molecules, can dif- fuse and react with each other. In this case, even if after a dark period the cells were entirely devoid of reduction intermediates, oxygen evolution and carbon dioxide consumption could begin immediately upon illumina- tion. 5. Energy Migration and the Hypothesis of the Photosynthetic Unit We consider the assumption of a "finishing" dark reaction of limited maximum rate as the cause of flash saturation (as well as of saturation in steady light) to be preferable to the assumption that this saturation is due to the limited quantity of the reduction (or oxidation) substrate. The arguments that influence our choice are {1) the observations of fluorescence, which make it probable that usually all chlorophyll molecules are associated with the photosensitive substrate, even after the light saturation has set in (c/. page 1075), and (2) the observations which indicate that the primary carbon dioxide absorber is present in the cells in a quantity roughly equiva- lent to that of chlorophyll. As described in chapter 28, this theory, proposed by Franck, can explain also the possibility of utilizing for photosynthesis all the light quanta absorbed by chlorophyll in weak light and the absence, under these conditions, of an extended induction period for the uptake of carbo]! dioxide. Leaving aside for a while the arguments derived from fluorescence, and from the apparent abundance oi the primary carbon dioxide absorber, we ask whether the alternative theory of "substrate limitation" could explain the two last-named phenomena: the high quantum yield and the absence of induction losses in weak light. Offhand this appears difficult. If, at best, only 0.1% of the chlorophyll molecules present in the cell can be as- sociated with the reaction substrate (since this is the total amount of this substrate supposed to be available in the cells), how can the quanta ab- sorbed by all chlorophyll molecules be made useful for photosynthesis? And if each chlorophyll molecule has to absorb 4, or, more likely, 8 quanta ENERGY MIGRATION AND THE PHOTOSYNTHETIC UNIT 1281 in order to complete the reduction of the carbon dioxide molecule associated with it, how can the uptake of carbon dioxide start immediately upon illu- mination, even in light which is so weak that individual chlorophyll mole- cules absorb quanta at an average rate of only one everj^ 30 minutes? The second difficulty could be eliminated by the assumption of a single photochemical reduction step, followed by nonphotochemical dismutations, while the first one could be resolved by the assumption that the reaction substrate (or an intermediate catalyst) moves freely through the photo- synthetic apparatus, and can thus collect the required energy quanta from several excited pigment molecules. The quantity of the substrate, or catalyst, required for efficient energy collection may well be only 0.1 mole per cent of the amount of the pigment itself, particularly if the activa- tion of chlorophyll is prolonged by transfer into a long-lived active state (c/. Vol. I, chapter 18). An alternative, more exciting and more controversial interpretation has been suggested. It was submitted that the absence of an induction period in weak light, as well as the high quantum yield, can be explained by postulating the existence of a "photosynthetic unit" of 300-2400 chlorophyll molecules. This concept was developed by Gaffron and Wohl (1936) from initial suggestions by Emerson and Arnold (1932\ 1932^). In discussing, on the basis of substrate limitation, the figures in table 32.1, Emerson and Arnold (1932^, 1932^) and Arnold and Kohn (1934) sug- gested that a "chlorophyll unit" of 2500 molecules may be associated with one "reduction center" (for example, a molecule of the carbon dioxide- acceptor complex, ACO2) in such a way that all these pigment molecules can co-operate in bringing about the reduction of the molecule of carbon dioxide attached to the one "center." In discussing the possible attribution of flash yield limitation to a finishing cata- lyst, we suggested that to calculate the available amount of this catalyst the factor r should be multiplied by n (or n/2) (n being the number of quanta required for the reduc- tion of one molecule of carbon dioxide), because the catalyst might have to operate n (or n/2) times in the reduction of one carbon dioxide molecule to the carbohydrate level and the liberation of one oxygen molecule. (For example, in the Franck-Herzfeld scheme, 7.VA, eight unstable intermediates must be stabilized by catalyst Eb.) In the hypothesis of flash yield limitation by the available substrate now under discussion, the necessity of dividing r by n (or n/2) is less certain, because a reaction center can con- ceivably take up, during a single flash, all the n (or n/2) quanta that might be required to complete the reduction of the associated carbon dioxide molecule (and the oxidation of the associated water molecule). It is, however, equally possible that each reaction center can utilize only 1 (or 2) quanta per flash, and that the intermediate photochemical products must complete their conversion into the final products, carbohydrate and oxygen, by nonphotochemical dismutations (or coupled reactions, discussed in chapter 9 under the name "energy dis- mutations"). If one of these mechanisms is used in photosynthesis, the number of 1282 THE PIGMENT FACTOR CHAP. 32 needed reaction centers is n (or n/2) per oxygen molecule liberated in the flash. We therefore conclude that, in the "substrate Umitation" theory of flash saturation, the required number of "reaction centers" can be T[Chl]o, or 4T[Chl]o or 8r[Chl]o depending on whether the postulated reaction mechanism permits each center to utihze 1, 4 or 8 quanta in each flash. Gaffron and Wohl (1936) sought an explanation of how the quanta ab- sorbed anywhere 'in the "unit" can be utilized, without loss, for photochemi- cal action in a single "center." They suggested that the "unit" may be a closely packed system in which the individual pigment molecules are so intimately associated that a light quantum absorbed by one of them can be exchanged, from neighbor to neighbor, until it reaches the reduction center. In other words, instead of the energy quanta being collected (as sug- gested above) by chemical agents diffusing through the system, the quanta themselves were supposed to move around until they find the substrate for photochemical action. Weiss (1937) suggested that the "units" may be identical with the chloroplast grana (Vol. I, p. 357). Frey-Wyssling (1937) pointed out that each granum contains about 10^ chlorophyll molecules, while "units" were supposed to consist of only about 10^ molecules each. However, the unit may also be interpreted statistically as a structure containing about one molecule of an enzyme per about 10^ molecules of chlorophyll. It could be suggested, for example, that each protein disc in the grana (c/. chapter 37 A), carrying about 10^-10'^ chlorophyll molecules, is provided with about 10^-10* "reaction centers," e. g., molecules of an enzyme. Each of the latter can be conceived of as "servicing," preferentially an area, or being associated exclusively with an "island" of about a thousand chlorophyll molecules (c/. Rabinowitch 1951). The "servicing" may be accomplished either by the diffusion of material particles, or by energy migration be- tween pigment molecules, or by a combination of both mechanisms. Aside from any special assumption concerning the nature an ddistribu- tion of the "reduction centers" in the cell, we may ask: Is the postulated efficient exchange of excitation energy between a large number of chloro- phyll molecules physically possible? Can evidence be adduced for (or against) the occurrence of such an exchange in living chloroplasts? The study of energy transfer in liquids or solids is a rather new de- velopment. It was known for a long time that in simple gases a trans- fer of excitation energy from molecule to molecule does occur in colli- sions, and that its efficiency is a function of the "resonance" between the collision partners. The closer the resonance, the stronger the interac- tion, the higher the probability of energy transfer, and the greater the dis- tance over which it can occur. For example, when an excited helium atom approaches a normal atom of the same gas, the perfect resonance between the two states, (He* -}- He) and (He -1- He*), causes the energy early to ENERGY MIGRATION AND THE PHOTOSYNTHETIC UNIT 1283 begin fluctuating back and forth between the two atoms. The frequency, V, with which this fluctuation occurs, determines the interaction energy, E, according to the quantum-mechanical equation relating energy to fre- quency: (32.4) E = hv If 1/v is small compared with the duration of the collision (which, for average thermal velocities of the molecules, is of the order of 10~^^ second) a large number of energy exchanges will occur during a single coUision, and, when the atoms separate after the collision, the chance of finding the exci- tation energy associated with the formerly unexcited helium atom will be equal to that of finding it in the originally excited particle. When identical molecules are present in close mutual proximity in a concentrated solution, a pure liquid or a crystal, the probability of excita- tion energy exchanges can be so high that the energy quantum will change its location several times before it is re-emitted as fluorescence, dissipated as heat or utilized for a photochemical transformation. This phenomenon of energy migration includes two extreme cases (and all transitions between them). In the one extreme, the frequency of intermolecular exchanges is much higher than that of intramolecular vibrations. In this case, the ex- citation energy migrates from molecule to molecule without tarrying in any one of them long enough to warrant the application of the Franck-Condon principle; in other words, the electronic excitation energy is in and out of the molecule before the sluggish nuclei can adjust themselves to its pres- ence. The excited state then belongs to the system of many identical molecules as a whole, rather than to any individual molecule; the excita- tion energy "package" is in a condition reminiscent of that of an electron in the "conductivity band" of a metal (or a "pi electron" in the benzene ring). In the other extreme case, the excitation energy stays with each mole- cule long enough for intramolecular vibrations to be acquired (or lost) in accordance with the Franck-Condon principle. At each given moment, then, the excitation energy belongs to a definite molecule. It moves from one molecule to another by a diffusion mechanism reminiscent of the Brownian movement of material particles. There is some confusion in the terminology used to describe the different cases of resonance migration of energy. The term "exciton" was first proposed by the Russian theoretical physicist Frenkel in a discussion of the "fast" propagation mechanism, and it has been suggested that this term should be reserved for this case (although Frenkel him- self has used it later also in referring to "slow" migration). The first phenomenon could also be described as "non-localized" or "communal" absorption of light quanta by a system of coupled resonators; while the second is reminiscent of repeated "collisions of the second kind," such as are responsible for sensitized fluorescence in gases; it has been 1284 THE PIGMENT FACTOR CHAP. 32 therefore sometimes called, rather awkwardly, the "fluorescence mechanism." We will use the terms "fast" and "slow" migration, keeping in mind that the time yardstick by reference to which the two cases are distinguished is the period of intramolecular vibra- tion, about 10 ~" sec. Theoretical estimates indicate, and experimental results confirm, that the exchange of excitation energy between resonating molecules can occur, in a condensed system, not only in actual collisions or (to use a term more appropriate for such systems) "encounters," but also while these mole- cules are separated by a solvent layer of several molecular diameters. The occurrence of such "remote" transfers was first derived from observations of the "concentration depolarization" of fluorescence in dyestuff solutions. When fluorescence of a dyestuff is excited by polarized hght, the fluorescent light is found to be more or less strongly polarized. To explain this we have to assume that excitation by polarized light occurs preferentially in pigment molecules having a certain orientation, and that, in viscous media, this orientation is not, or not completely, lost by thermal agitation in the time between excitation and re-emission, thus resulting in an at least partly polarized fluorescence. The degree of polarization of the fluores- cent light proves to be a function of concentration, being highest in very dilute solutions; this is the phenomenon of "concentration depolari- zation." This depolarization occurs without change in the absorption or fluorescence spectrum, the yield of fluorescence, or its life-time. It there- fore cannot be attributed to dimerization (or generally, polymerization) of the dyestuff molecules or ions. The alternative is then between attributing depolarization to kinetic encounters, or to "remote" interactions between dyestuff molecules. Probably, both phenomena occur. Perrin (1932), Vavilov (1942-1950) and Forster (1946-1948, 1951) have been particularly concerned with the remote interaction. The following observation speaks in favor of this interaction as the main or only source of depolarization. The concentration at which depolarization reaches 50% is (in the case of fluores- cein solution in glycerol) of the order of 10 ~^ mole/Hter. Offhand, this seems sufficiently high for kinetic encounters to produce the observed ef- fect; however, if encounters were actually responsible, the concentration of the dyestuff required for a certain degree of depolarization would in- crease with increasing viscosity of the medium. No observations of such a dependence have been made; furthermore, it was found that concentra- tion quenching of the fluorescence of the same dye in sugar-glycerol solu- tions (which occurs in a much higher concentration range, 10~^ to 10~^ mole/1.) actually is independent of viscosity. The quenching thus appears to be a function of the average mutual distance of the pigment molecules (which is independent of viscosity) rather than of the frequency of their encounters (which decreases with increasing viscosity). A fortiori, this ENERGY MIGRATION AND THE PHOTOSYNTHETIC UNIT 1285 conclusion should also be applicable to depolarization, which occurs at con- centrations ten or a hundred times lower than that required for quenching. (It seems unlikely that complete dissipation of excitation energy, implied in quenching, should occur by remote interaction while the much "gentler" depolarizing interaction required direct encounters.) The concentration depolarization and concentration quenching of dyestuff solutions were again studied experimentally by Sevchenko (1944), and found to be in agreement with the predictions of the theory of "molecular induction" (which is another term for resonance exchange). Pekerman (1947) found that, when dyestuff solutions are taken up by sintered glass, whose pore diameters are such as to cause dye molecules to form one-dimensional trains, self-depolarization and self-quenching are decreased— probably because, under these conditions, resonance transfer is possible in one direction only, and is therefore less effective than in bulk solution. According to this concept, concentration depolarization indicates (a) that a considerable proportion of the fluorescent light is emitted, not by the primary excited pigment molecules, but by molecules to which the excita- tion energy has been transferred between excitation and emission, and (6) that this transfer takes place without actual encounters of the emitting molecules with the primarily excited molecules. (It may be useful to point out that the assumed mechanism of energy exchange is different from absorption and secondary re-emission of fluores- cence, a phenomenon that also is possible in concentrated solutions. Es- timates indicate that, because of the displacement of the fluorescence band toward the longer waves compared with the absorption band, the probability of "secondary fluorescence" of this type is much too small to account for the observed depolarization.) If the energy transfer does not require (or prefer) parallel orientation of the molecules (an admittedly extreme assumption), the "self -sensitized" fluorescence will be completely unpolarized. In this case, the decrease of polarization with increased concentration will provide a direct measure of the average number of energy transfers during the excitation period. (Equal distribution of the probability of re-emission over the primarily excited and one secondarily excited molecule will produce 50% relative de- polarization; distribution over three molecules, 67% depolarization, and so on.) In a IQ-^ M solution of fluorescein in glycerol, the polarization is only 20% of that in dilute solution. If remote energy transfer is the only mechanism of depolarization, this figure indicates the distribution of the excitation probability over five pigment molecules. At the same concentration, the fluorescence yield is reduced by about one half compared with dilute solutions. If this "self-quenching" effect also is ascribed to remote energy transfer, the two phenomena together indi- cate an energy exchange over an average of ten molecules. Perrin sug- 1286 THE PIGMENT FACTOR CHAP. 32 gested that concentration quenching of fluorescence is generally due to remote energy transfer. This is certainly not true as a general rule, since in many dyestuffs (such as methylene blue) self-quenching is due pre- dominantly to the formation of nonfluorescent dimers at the higher con- centrations. Furthermore, at concentrations as high as 10"^ mole/hter, quenching (and depolarization) by actual kinetic encounters is unlikely to be entirely negligible even in a medium as viscous as glycerol. In methylene blue and similar dyestuffs, the self-quenching occurs at concentra- tions at which dimer formation is revealed by changes in the absorption spectrum. In other dyestuffs, however, quenching occurs when the equilibrium concentration of di- meric molecules is too small to permit the assumption that only the Ught quanta directly absorbed by such molecules are lost for fluorescence. The concentration quenching can nevertheless be attributed to dimer formation, by means of either or both of the following two hypotheses : first, that short-lived, nonfluorescent dimers can be formed in encounters of excited and normal pigment molecules; and second, that the excitation can "seek out" the few dimeric molecules present in equilibrium, by making numerous jumps from molecule to molecule during the excitation period (Forster's hypothesis). According to Franck and Livingston (1949), migration of excitation energy could lead to quenching also by another mechanism, not requiring dimers: the perambulating quantum could be dissipated whenever in its travel it visits a molecule containing an abnormally high amount of vibrational energy, since, in such a molecule, electronic exci- tation may produce a configuration permitting conversion of the electronic energy into vibrations of the ground state (for an objection to this hypothesis, see Forster 1951, p. 252). A mathematical theory of quenching by energy migration was developed also by Vavilov (1942, 1944, 1950), who postulated that each transfer of electronic energy from molecule to molecule involves a certain probability of its loss, without specifying the mechanism of this dissipation. Perrin (1932) calculated, using a classical harmonic oscillator as a model of pigment molecules, that the probability of energy transfer from the orig- inally excited to a second, resonating oscillator, becomes equal to the prob- abihty of re-emission of energy by the primary absorber, when the distance between the two oscillators is of the order of X/27, corresponding to about 100 myu for visible light. Neither the concentration quenching nor the de- polarization occur at such extreme dilutions. Forster (1946) ascribed this to lack of exact resonance (assumed in Perrin's calculations) . The resonance is not exact for two reasons: the displacement of the fluorescence band relative to the absorption band, and the finite width of both bands. Forster made a rough calculation, taking into account two requirements : (a) that the frequencies of the two interacting oscillators must be within the region where the fluorescence band and the absorption band overlap, and (6) that these frequencies must differ by not more than E/h (where E is the inter- action energy). The result of this calculation is that the "critical distance" {i. e., the intermolecular distance at which the probabilities of re-emission ENERGY MIGRATION AND THE PHOTOSYNTHETIC UNIT 1287 and transfer are equal) is reduced, compared with Perrin's estimate, by a factor of [A'v/riAvy]'^': (32.5) do = i\/27)[A'v/t(Auy]'^' Here, A'j^ is the range of overlapping fluorescence and absorption frequen- cies, Av the band width and i the average duration of excitation. With A'j//Av = 0.1, ^ = 5 X 10-9 second and Av ^ 1.5 X 10l^ Forster calculated (Iq o:=i 7.5 m/x, corresponding to about 20 molecular diameters. An im- proved, quantum-mechanical calculation leads (Forster 1948, 1951) to the same order of magnitude for the distance over which excitation energy can be transmitted. The frequency of energy transfer is, according to this calculation, proportional to the inverse sixth power of the distance between molecules, and therefore directly proportional to the sq^iare of concentra- tion. The following equation was obtained by Forster (1948) as a substi- tute for equation (32.5) : (32.6) rfo -^ SicTl/STSim'^ul where t ^ average life-time of excitation, c = velocity of light, n = index of refraction of the medium, A'"' = number of molecules in a millimole; vo, the frequency of the band (average of the frequencies of the absorption and fluorescence peaks) ; and T^ the integral : (32.4) n = 2.302y^" a,«2. .a. which is a measure of the overlapping of the fluorescence and the absorption spectra (a^ being the decimal molar extinction coefficient at frequency i'). Applying this equation to chlorophyll a in ether (/i = 1.35) Forster ob- tained do = 8.0 mju as the distance at wliich energy transfer becomes equally probable with re-emission. This is the average intermolecular distance at 7.7 X 10—* m./l. Treating chlorophyll grana as a homogeneous system 0.1 ilf in chlorophyU (cf. Vol. I, page 411), i. e., neglecting the probable orderliness of the arrangement, Forster calculated that about 10'' transfers can occur during the excitation period of 3 X 10^^ sec. According to the equation AE ^ hv (where v is the frequency of the energy exchange) this rate of exchange should cause a change, AE, in the energy of the excited state, equivalent to about 10 cm.-^ [10V(3 X 10"* X 3 X 10 1°) = 10]. This corresponds to a shift of the absorption band, leading to this state, by about 0.4 /x. It will be noted that, with 10* trans- fers per second, 30 to 300 molecular vibrations (with periods between 10~^* and 10-^^ sec.) wifl be possible per visit; consequently, the band shape — determined by the couphng of electronic excitation and intermolecular vi- brations (according to the Franck-Condon mechanism) — wiU be preserved more or less unchanged. 1288 THE PIGMENT FACTOR CHAP. 32 These estimates seemed to fulfill the requirements of the theory of the photosynthetic unit (300-2400 transfers during the excitation period) and to be at least not inconsistent with the spectroscopic facts (the red band shift in vivo is »0.4 m/x, but it could be attributed largely to complexing with proteins; the shape of the red absorption band is approximately the same in vivo as in vitro). However, the numerical values used in Forster's calculations require correction. He used ^ = 3 X 10"^ sec. as the average hfe-time of the excited state. A larger value, 8 X 10~^ sec, was derived on page 633 from the integral of the absorption curve, but much more recent calculations (cf. chapter 37C, section 1) gave only 1.3 X 10"^ sec. A much more radical correction is required by the fact— neglected by Forster — that the yield of fluorescence of chlorophyll in the living cell is low, perhaps only 0.1% (as against 10% in organic solvents). This indicates that the "natural" life- time of excitation (as derived from band intensity) is shortened — -perhaps by a factor of 10^ — ^by energy dissipation. The effective t thus may be of the order of 10"", rather than 10-^ sec. Even a few hundred excitation jumps during this period will reduce the "visiting time" below 1 X lO^'^ sec, and thus destroy the coupling with molecular vibrations. (Inciden- tally, a "visiting time" of such brevity should also affect unfavorably the probability that excitation Avill actually cause the photochemical change while "visiting" the chlorophyll molecule associated ^^•ith the reaction center, since this energy transfer, too, depends on the conversion of elec- tronic energy into the kinetic energy of atomic nuclei.) From considerations of this type, Franck and Teller (1938) concluded that, in the hving cell, energy propagation by the "slow" transfer mech- anism cannot extend over the number of molecules required by the theory of the photosynthetic unit. The concept was nevertheless revived by Duysens (1952). He derived from Forster's equations the conclusion that at a concentration [k]o = 1000/NAdl(m./l.), where do is the "critical distance" given in equation (32.6), and TVa, Avogadro's number (6 X 10'^), an average of about 15 energy transfers will occur during the lifetime of the excited molecule — assuming an at random distribution of the acceptor. With increasing con- centration, [k] > [k]o, the average number of transfers will increase propor- tionally. It was estimated above that with t = 3 X 10~^ sec, [k]o for chloroph3dl a is 7.7 X lO^"* mole/1.; the pigment concentration in the grana is much higher, ^0. 1 mole/1. True, the lower actual yield of fluores- cence in vivo does not permit an estimation of the average extent of energy migration simply from this ratio of concentrations; nevertheless (using a natural life-time of 4 X 10~^ sec.) Duysens arrived at a relatively optimistic ENERGY MIGRATION AND THE PHOTOSYNTHETIC UNIT 1289 estimate by postulating an "intrinsic" fluorescence yield of 10% (the same in vivo as in vitro), and an actual yield of the order of 1%. According to Duysens, at an average concentration of 0.1 mole/1, (which is a plausible value for the grana), the average number of energy transfers, during an excitation period of 4 X 10 "^ sec, is about 750. If, before the end of the transfer chain, excitation hits a "reaction center" in which it is utihzed for a photochemical reaction, fluorescence is quenched. Assuming that one such center exists per 200 chlorophyll molecules (in a random, three-dimensional array), the chance of hitting it before the trans- fer chain of 750 links is terminated can be estimated as 0.8; the fluores- cence must then be reduced by 80%— from the "intrinsic" value of 10% to an "actual" value of about 2%. Duysens considered this figure as close to the actual true yield of fluorescence in vivo (about 1%, according to his estimates; c/. chap. 37C, section 7), and the whole calculation as proving that the concept of the photosynthetic unit, with energy exchange by the "slow" resonance transfer mechanism, can be reconciled with the spectro- scopic evidence, if the unit is reduced to about 200 chlorophyll molecules (and the yield of fluorescence in vivo is assumed to be of the order of 1%). About 750 transfers during 4 X 10-^ sec. means an exchange frequency of 2 X 10^^ sec. -^ corresponding to a band shift by only 7 cm. -1 (2 X lO^^V 3 X 10^°), which is small compared to the actual red shift in vivo (about 200 cm.-i). The vibrational structure of the absorption band wifl be unafi"ected by such slow transfer, and the band shift caused by it will be practically negligible. A simplified procedure — similar to that repeatedly used above — would be to set 4 X 10-'" sec. as the available migration time (indicated by a 1% fluorescence yield), and to conclude that 200 transfers during this time mean an exchange frequency of 5 X IQii sec.-i, and a visiting time of 2 X IQ-'^ sec— just enough to preserve the coupling with molecular vibrations with frequencies of the order of 10'' sec."'. The theoretical migration range is cut to Vs if t is only 1.3 X 10 -» sec. The case of energy transfer to a minor constituent of the pigment sys- tem, mixed at random with the main pigment, and having an absorption band in resonance with the fluorescence band of the latter, is clearly analogous to the above-discussed case of energy transfer to a "reaction center" in a photosynthetic unit. Duysens (1952) considered from this point of view the— apparently highly effective— loss of excitation energy of chlorophyll a in red algae by transfer to an unidentified minor pigment (probably, chlorophyll d). To compete effectively with the "reduction centers" as ultimate energy acceptors, the "chlorophyll d" molecules must be present in a similar concentration (one for a few hundred chlorophyll molecules, which is compatible with experimental evidence). This as- sumes eriual probability of transfer to both acceptors; the capacity of 1290 THE PIGMENT FACTOR CHAP. 32 "chlorophyll c?" to divert quanta from the "reduction centers" would be enhanced if the overlap integral for the transfer chlorophyll a -> "chloro- phyll d" were larger than for the transfer chlorophyll a -»• "reduction center" — as it may well be. (The "reduction centers" may be properly located or complexed molecules of chlorophyll a itself.) One may ask whether all these estimates could be affected by the probable existence of a long-lived, metastable state of the chlorophyll molecule (or of the pigments-protein- lipide complex in the living cell). If the low yield of fluorescence of chlorophyll in vivo is the result of the transfer of most of the excited chlorophyll molecules (or complexes) into a metastable state containing considerable electronic energy, and not of total con- version of this energy into heat (or chemical energy), why should it not be possible for the electronic energy of the metastable state to be exchanged "at leisure," as it were, between adjacent molecules? (The time available for the exchange could be, in this case, a million — -or more^ — -times longer than the natural life time of the fluorescent state.) If the metastable state is a tautomeric form of the molecule, no resonance trans- fer of energy is possible, because it would require a displacement of the nuclei; but if the long-lived state is an electronic mesomer (e. g., a triplet state, as envisaged by Terenin*) and by Lewis and Kasha, cf. pp. 486, 790), energy transfer is feasible. However, the rate of transfer by the "slow" mechanism is proportional to the fourth power of the transition probability between the ground state and the excited state; while the natural life-time of the excited state is inversely proportional to the square of the same probability. There- fore, if the natural life-time of the excited state is tf, and that of the metastable state, t„, the number of transfers during the full natural life-time of the latter will be smaller (by a factor of t^/t„) — and not larger than the number of transfers during the full life- time of the fluorescent state. If we assume that the low yield of chlorophyll fluorescence in vivo (0.1%, or 1%) is caused by the transfer of 99 (or 99.9%) of the excited chloro- phyll molecules from the fluorescent into the metastable state (rather than by the dissi- pation of total energy of the fluorescent state by internal conversion), the number of energy transfers during the actual excitation period will be increased in consequence of this transfer, only if t„ is <103 (or 10") X tf. (We now compare the number of jumps during the full natural life-time of the metastable state with that during one-thousandth or one-hundredth of the natural life-time of the fluorescent state.) With <^ = 4 X 10~* sec, this means that the existence of a metastable state could favor migration only if the natural life-time of this state were <4 X 10 ~* (or 4 X 10"^) sec. If the first figure is correct, a metastable state with a natural life-time of 4 X 10 ~« sec. would increase the frequency — and thus also the range — of energy migration by a factor of 10, and a meta- stable state with a life-time of 4 X 10 "^ sec, by a factor of 100 — presuming this state actually survives for its full natural life-time. However, if this were the case, the meta- stable state would produce a marked phosphorescence. Since this state is situated con- siderably below the initial excitation state (as judged by the nonoccurrence of delayed red fluorescence, which could be caused by return from metastable to the fluorescent state by thermal energy fluctuations), an emission originating in this state should be located in the infrared. Until evidence is presented that chlorophyll in vivo does emit infrared phosphorescence, with a quantum yield higher than that of the known red fluorescence, we have to assume that the metastable triplet state, if it occurs in vivo at all, survives for only a small fraction of its "natural" life — and this should make it im- * Terenin (1940, 1941) had suggested the interpretation of the metastable state of organic molecules as a triplet "biradical" state before this was proposed by Lewis and Kasha (1945), but his work did not become known abroad, because of wartime con- ditions, until considerably later. (For a review of the work by Terenin and co-workers, see Terenin's Photochemistry of Dyes, 1947.) ENERGY MIGRATION AND THE PHOTOSYNTHETIC UNIT 1291 possible for this state to contribute significantly to resonance energy migration. In other words, the number of energy transfers that occur while the molecule is in the meta- stable state is smaller than those occurring in the brief period (4 X 10 "i", or 4 X 10 ~" sec.) the molecule spends in the original excited state. In chapter 23 (p. 795) we mentioned the— inconclusive — attempts by Calvin and Dorough to identify a long-Uved, infrared fluorescence of chlorophyll in vitro; no similar experiments have been made with chlorophyll in vivo. Terenin and Ermolaev (1952) saw evidence of energy exchange in the metastable state in benzaldehyde-sensitized phosphorescence of naphthalene in a frozen mixture of these two compounds. The energy content of the excited singlet state of benzaldehyde is too small to lift naphthalene into its excited singlet state. The authors suggested that excited benzaldehyde is first converted into the metastable triplet state, and that the energy of the latter is then transferred to naphthalene. The triplet state of naphtha- lene, resulting from this transfer, slowly decays by phosphorescence. It was suggested long ago, from the study of sensitized fluorescence in gases, that, for the resonance transfer to be a "permitted" process, the total electronic spin of the system must remain constant — and that this condition can be fulfilled by the excitation of a "prohibited" singlet -^ triplet transition at the cost of another "prohibited" triplet -> singlet transition, as well as by the more trivial replacement of one "permitted" tran- sition by ajiother. There is a certain contradiction between this statement and the argu- ment used above in discussing the probability of energy migration in the metastable state, since in the latter the low oscillator strengths of the "prohibited" transitions were supposed to be unaffected by the mutual approach of the two partners in the exchange. Which of the two conclusions is correct must depend on the strength of the coupling be- tween the electron spins of the two molecules at the distances over which the energy ex- change occurs. We will now consider whether the "fast" exchange mechanism ("com- munal absorption") could operate in the chloroplasts, and whether the spectroscopic properties of chlorophyll in the Hving cell are consistent with such an exchange. Since chlorophjdl is not distributed uniformly in the granum (instead, it probably is arranged in monomolecular layers, cj. chapter 37 A), resonance exchange may be considerably faster than was calculated by Forster from the average concentration of chlorophyll in the granum. If we attribute the total red shift of the absorption band in vivo (about 10 m^ from its position in ethereal solution, and about 25 m/i from its extrapolated position in vacuum), to such an exchange, we calculate for the exchange process a wave number of from 250 to 625 cm.~^ and a frequency of from 7.5 to 20 X lO^^ sec.-^; this would permit 300-1200 exchanges, enough to satisfy the requirements of the "photosynthetic unit" during an excitation period as short as 4 X 10"^^ sec. (not to speak of the 4 X 10-'° sec. available if Duysens' revised estimate of the fluores- cence yield in vivo is correct). A difficulty arises, however, when we consider the sha'pe of the absorp- tion band. As pointed out before, this shape is determined, in solution, by the coupling of electronic excitation with intramolecular vibrations; and one can expect this coupling to be destroyed if the excitation does not stay with each visited molecule »10-i^ sec, to permit molecular vibra- tions to get excited according to the Franck-Condon mechanism. With a transfer occurring each 5-15 X lO"'* sec. this is impossible; excitation 1292 THE PIGMENT FACTOR CHAP. 32 shoots through the molecules without setting them in vibration, like a bullet passes through a glass pane without shattering it. One therefore expects this type of energy exchange to strip the absorption band of its vibrational structure and, by preventing electronic energy dissipation into vibrations, to produce a high yield of fluorescence of the "resonance" type {i. e., fluorescence with a wave length practically identical with that of the absorption band). Absorption and fluorescence phenomena of this type have been ob- served by Scheibe et at. (1936-41; see also Katheder 1940; Jelley 1936) in concentrated solutions of certain dyestuffs. These investigators found that the absorption spectra of isocyanine, pinacyanol and certain other dyestuffs change radically when their concentration in aqueous solution is increased above a certain value. The original absorption band loses its intensity and a new sharp and narrow band appears on the long-wave side 150,000 - I I vX I ) N N'' CI' c 1 ' CsHj C2H5 > / - C= 1 05-10"^ 2 - C= 52310"' 100,000 3 - C= 1 20- 10"" 0 , 50,000 40,000 30,000 20,000 10,000 - .'" \;. / 1 1 1 1 " 0 \ ^iSa, 22,000 20,000 18.000 cm Fig. 32.7. Absorption spectrum of pseudo-isocyanine at three different concen- trations, showing the formation of dimors at 5 X 10"' mole/1, and of polymers with a sharp absorption band at 1 X 10 "^ mole /I. (after Scheibe 1938). of the original band. Figure 32.7 shows this phenomenon for pseudo- isocyanine. This change is due to linear 'polymerization of the dyestuff ; it disappears sharply upon heating to a certain temperature. The extinction coefficient of the polymer per single link is of the same order of magnitude as that of the monomeric form, thus showing that the "super-molecule" acts as the sum of as many chromophores as it contains monometric mole- cules. Figure 32.7 shows that, in a typical case, the polymer band is shifted ENERGY MIGRATION AND THE PHOTOSYNTHETIC UNIT 1293 about 1500 cm.-^ from the position of the monomer band. This shift corresponds to a frequency of the order of 5 X 10^^ sec.-^, or to a sweep of the excitation over 10'^ pigment molecules during the life-time (if the latter remains of the order of 10 ~^ second). The polymeric molecules of this type show strong resonance fluorescence, in agreement with the theoretical prediction; a striking result, because resonance fluorescence is usually associated with vapors of low pressure, consisting of free atoms or simple molecules, and not with complex mole- cules or condensed systems. A, + A, DISTANCE Fig. 32 8. Potential energy curves of two atoms with no bonding in the ground state and exchange bonding in the excited state (example: Hej). Obviously, nothing similar to Scheibe's. polymers exists in chloroplasts : the red absorption band of chlorophyll is about as broad, if not broader in vivo than in vitro, as if no uncoupling of intramolecular vibrations from electronic excitation has occurred. Furthermore, the fluorescence of green cells is weak and not of the resonance type. These facts seem to preclude the interpretation of the red band shift in vivo as due to a "fast" resonance migration of excitation energy through a closely packed chlorophyll layer. They caused Franck and Teller (1938) to reject the "fast" energy migration mechanism as a possible basis of the "photosynthetic unit" concept; it was mentioned above that they have also rejected the "slow" mechanism because in their estimate it did not provide a sufficiently wide range of energy migration. 1294 THE PIGMENT FACTOR CHAP. 32 It is not quite certain, however, whether band sharpness and resonance fluorescence are necessary attributes of energy migration of the "fast" type. In the first place, the narrowness of the absorption band is predicated not only on the uncoupHng of electronic excitation from intramolecular vibra- tions, but also on a strict selection of permitted transitions from a broad electronic excitation band. The simplest case of energy resonance is that between two atoms, Ai + A2. Their mutual potential energy as a function of distance can be repre- sented by the lower curve in figure 32.8. If one of the two atoms is ex- cited, the resonance between the structures Ai + A2 leads either to attrac- tion or to repulsion, giving rise to two excited electronic levels, Ei and E2. The first absorption line of the separated atoms, hv^, is thus shifted to the red {i. e., to the smaller energy quanta), the amount of the shift being de- termined by equation (32.4). The second absorption hne, leading to the repulsion curve, is shifted to the short-wave side of the original band, and may be a "forbidden" line. If this concept is extended from two atoms or molecules to three, four or more resonating systems, the addition of each new link Avill cause an in- crease in the number of levels, until a practically continuous band of energy levels will be formed. (This situation is similar to that in the series Na, Na2 . . . Na metal, with the difference that, in a metal, the multiplicity of levels is due to the exchange of electrons, and not excitons.) Under cer- tain conditions, the probability of transition by light absorption will be high only to the lowest excited states, thus giving a sharp absorption and fluorescence band of the lowest possible frequency {i. e., w4th the maximum possible shift toward the red from the position of the absorption band of a single molecule). The sharp selection of the permitted electronic transitions appears to be valid for symmetry conditions prevailing in Scheibe's one-dimensional polymers; but under different symmetry conditions a broader electronic band could possibly arise, to replace, in the resonating system, the vibra- tion-broadened but intrinsically sharp electronic transition in the mono- meric molecule. A closer study is needed also to decide whether the trans- fer of energy from the electronic system to the vibrational degrees of free- dom— including those of the system as a whole, rather than of individual molecules — can be effectively prevented, by rapid fluctuation of electronic excitation, in all two-dimensional or three-dimensional resonating sys- tems, or whether this uncoupHng, too, can be fully effective only under special spatial relationships, such as may exist between the electronic ex- citation and molecular vibrations (or, at least, a certain type of moleculal vibrations) in one-dimensional polymers. As to the lack of, or weakness of, fluorescence, this can be caused, in a ENERGY MIGRATION AND THE PHOTOSYNTHETIC UNIT 1295 large resonating system, not only by a residual coupling with inter- or intramolecular vibrations, but also by an efficient quenching of fluorescence by a relatively small number of molecules of an impurity — a "trapping" of electronic excitation in a "potential hole," in which it can then be dissi- pated in the usual way by conversion into vibrations. Of some relevance here are the experiments on the absorption spectra and fluorescence of certain pigments (particularly chlorophyllides) in the crystalline state, described in chapter 37C, section 3. These experiments showed that, in chlorophyllide crystals and mono- layers, the red absorption band is shifted by as much as 80 m/x (up to 2000 cm.~^; cf. table 37C.II) toward the longer waves. Theoretical calcula- tions, using the actual density of the chromophores in the crystal, indicated that a shift of approximately this extent is to be expected in consequence of the interaction between closely packed chromophores (Kromhout 1952; Jacobs, Holt, Kromhout and Rabinowitch 1954; cj. also Heller and Marcus 1951). In contrast to calculations of Forster et at., in which only the energy exchange between two molecules was considered, these estimates were based on simultaneous consideration of "virtual dipole" interactions be- tween all pairs of molecules in a cubic lattice. The calculations were carried out for an isotropic three-dimensional lattice. The chlorophyllide crystals have, however, a layered structure (illustrated by fig. 37B.9) ; and molecular interaction can be expected to be much stronger within each layer than between them. (A difference is likely also between two directions within each layer.) This expectation is confirmed by the observation that the spectra of monomolecular layers of chlorophyllide show almost as wide a red shift as those of crystals. The predominantly two-dimensional interaction must lead to a more rapid saturation of interaction energy with distance than in a three-dimensional system (since the integration of the resonance energies has to be carried out over circular belts, rdr, instead of spherical shells, rHr). Con- sequently, the relative contribution of the nearest neighbors to the total resonance energy of a molecule is greater in the layered structure than in an isotropic three-dimensional system. This has a bearing on the ob- servations— described in chapter 37C — of the saturation of the band shift with increasing size of the microcrystals. Obviously, as a crystal begins to grow, the interaction effect must at first increase with the number of mole- cules in it; saturation will be reached when new molecules, added on the surface, are so far from an average molecule inside the crystal that they do not contribute significantly to its excitation energy. For an energy de- creasing with inverse third power of distance (such as the mutual energy of two dipoles), the contribution of circular belts of increasing diameter to 1296 THE PIGMENT FACTOR CHAP. 32 the energy of a molecule in the center dechnes with the square of the radius. If the ring of nearest neighbors contributes the interaction energy AE, the second ring will contribute (approximately) AE/4, the third one, AE/9, and so on, the total for an infinite number of rings being (7rV6)A£;. The contribution of rings beyond the Nth. one will be / 1 1 or <5% for A^ = 12, and <1% for A^ > 60. The interaction energy of the layer should thus be >99% saturated— and the band shift should practi- cally cease— when the radius of the layer has grown to sixty molecular diameters; the shift will reach the limit of rehabihty of our present meas- urements at 12-15 molecular diameters. The crystals for which band shift saturation was observed, according to figure 37C.16, contained about 10^ molecules; even if correction is made for scattering, and the true shift assumed to end at about 730 mfx, this limit still corresponds to crystals containing as many as 10^ molecules each. Assuming equal development in three dimensions (which is admissible for the "small", as contrasted to the "large" microcrystals ; cf. fig. 37B.8), such crystals consist, roughly, of 100 layers of 100 X 100 molecules each- somewhat in excess of the dimensions over which the interaction energy should be 99% saturated in a single two-dimensional layer. However, as the distance from the central molecule grows, the role of the crystal layer to which this molecule belongs becomes less dominant, and this should cause the energy saturation to be approached more slowly. Spectral effects similar to those observed in chlorophyll derivatives must occur also in other molecular crystals, if the absorption bands of the molecules are intense enough to cause strong interaction. This subject was analyzed theoretically by Davydov (1948), specially in appUcation to hydrocarbons (anthracene, naphthalene, etc.), whose absorption spectra have been described by Obreimov, Prikhodko and co-workers (cf. Prikhodko 1949). Similar to the chlorophyllide crystals, naphthalene crystals are thin, monochnic sheets. Their absorption spectrum (for light falling normally to the sheets) depends on polarization, since electric oscillations can be excited parallel to one or the other of the two crystallographic axes located in the plane of the sheets. Some bands occur with only one (or predominantly one) polarization; others wdth both of them. Two of the latter can be correlated with known bands of naphthalene vapor, shifted toward the longer waves (by 500 and 2000 cm.-', respectively). Their vibrational structure is similar to that of the vapor bands. Bands leading to three other excited levels observed in crystals have no parallel in the vapor spectrum. One of them is about equally strong with both polarizations, the other two are strongly polarized. These three electronic levels, considered as "crystal levels," rather than "molecular levels," combine with a set of vibrational quanta not encountered in vapor, which must be attributed to the lattice as a whole. ENERGY MIGRATION AND THE PHOTOSYNTHETIC UNIT 1297 One could suggest that bands leading to the "crystal levels" in naphthalene crj^stals are analogous to the sharp absorption bands of Scheibe's polymers, and that energy quanta absorbed in these bands are "communal property" of the lattice. In contrast to this, quanta taken up in bands leading to "molecular levels" can l)e considered as be- longing to individual lattice points, and capable of migration only by the "slow" reson- ance mechanism. From the point of view of the state of chlorophyll m vivo, the spectra of chlorophyll monolayers are of special interest. Of the two types of such layers (Jacobs ct al. 1954), the "optically dense" (probably, bimolecular) layers show a band position entirely different from that of chlorophyll in the living cell; but the "optically thin" (undoubtedly, monomolecular) layers show the band in about the same position as the living cells (675 m^; cj. table 37C.II). Two questions must be answered, however, before this coin- cidence is accepted as significant. Is the band position in the monolayer due to chromophore interaction, or to the binding of chlorophyll molecules to water? And, similarly, is the band position in the living cell due to chromophore interaction, or to the attachment of chlorophyll to other cell constituents (e. g., water, or proteins)? If the red shift were as wide as in chlorophyllide monolayers (or in "dense" chlorophyll layers), its attribu- tion to interaction between chlorophyll molecules would be much safer, since such shifts are unlikely to arise from solvation. Despite this uncer- tainty, it is at least a legitimate working hypothesis to assume that the band position in vivo is due mainly to the interaction between chlorophyll molecules within monolayers (which they probably form on protein discs). The width of the absorption bands in chlorophyll crystals and mono- layers make it likely (although, according to p. 1294, not quite certain) that no "fast" migration of excitation energy occurs in them; while the absence of fluorescence (or its shift into the infrared?) sets a rather low limit for the possible extent of "slow" resonance migration. It will be noted in this connection that the extensive band shift in crystals and monolayers, although it indicates a "stabilization" of the "red" excitation level by about 2000 cm.-^ as compared to isolated chlorophyll molecules, can be adequately explained, in the first approximation, by virtual dipole interaction between the chromophores in the lattice, without recourse to quantum mechanical resonance phenomena, and therefore offers no evi- dence concerning possible excitation energy migration in the lattice. In vivo, on the other hand, the small but measurable fluorescence yield leaves, as estimated on p. 1289, just enough leeway for resonance energy migration to satisfy the needs of a modest "photosynthetic unit." The red band shift in vivo may be the composite result of virtual dipole interac- tion between pigment molecules in a monolayer, the attachment of pig- ment molecules to proteins and other cellular constituents, and a small 1298 THE PIGMENT FACTOR CHAP. 32 resonance stabilization, associated with a limited migration of excitation energy. Returning now to the problems in connection with which the concept of the "photosynthetic unit" was first introduced on p. 1280 — namely, the al- ternative between a "preparatory" or a "terminal" process Hmiting the maximum flash yield, we note that although the "unit" was first invented to be able to attribute the flash yield limitation to the number of available re- duction sites (i. e., of substrate molecules available for transformation in a flash), it can be equally well combined with the assumption of a "terminal" limitation — e. g., by the amount of a "stabilizing" enzyme. In this form, the hypothesis does not contradict the observation that the "CO2 acceptor" seems to be present, in photosynthesizing cells, in a concentration of the same order of magnitude as that of chlorophyfl (and not <1% of it). As to the effect of carbon dioxide on chlorophyll fluorescence in vivo, this does not require, if energy migration occurs, the association of each chloro- phyll molecule with a carbon dioxide molecule. (A case in point is pre- sented by Scheibe's hnear poljoners, where, in a micelle containing up to 1000 dye molecules, the association of a single molecule with a quencher effectively quenches the fluorescence of the whole micelle.) To sum up, the role of resonance energy migration between chlorophyll molecules in the photosynthetic process is still uncertain. Theoretical es- timates show that excitation energy should have a chance to migrate, during an actual life-time of the order of lO-i*^ sec, over a small number (of the order of 10 or 100) of chlorophyll molecules, if the pigment is dis- tributed at random in the grana, and over a larger number (of the order of 1000) of molecules, if these are arranged in densely packed monomolecular layers. Whether this migration actually takes place, we do not know. To prove beyond doubt that energy migration does occur between chlorophyll mole- cules in vivo, and to determine its reach, would be of great importance for the understanding of the photochemical mechanism of photosynthesis; yet even after this proof one could not be certain that energy migration is the true cause of the kinetic phenomena (such as flash saturation) that had first led to the concept of a "photosynthetic unit." The now most plausi- ble interpretation of these phenomena is in terms of an enzymatic compo- nent ordinarily present in a concentration of between 1/300 and l/2400th of that of chlorophyll. Effective cooperation of this "hmiting enzyme" with all chlorophyll molecules could be achieved in three ways — by migra- ENERGY MIGRATION AND THE PHOTOSYNTHETIC UNIT 1299 tion of excitation energy from the chlorophyll molecules to the enzyme ("photosynthetic unit"), or by migration to the enzyme of photochemical products formed at the chlorophyll molecules, or by migration of the en- zyme itself, collecting its substrate from hundreds or thousands of chloro- phyll molecules, Hke a bee collects nectar from a plot of flowers. Even if the phenomenon of energy migration were well established, this would not prove that the other two are nonexistent or irrelevant. In the above discussion the analogy between "energy conduction" and electron conduction in crystals was mentioned. The kind of resonance required for the diffusion of energy is, however, different from that underlying the diffusion of electrons. The atoms of a metal are held together by electron exchange forces (the same effect that is responsible for the formation of hydrogen molecules, according to Heitler and London). In molecular lattices, the resonance effects are due to excitation energy resonances, as illustrated above by the example of a He2 molecule. In an H2 molecule, the degeneracy, which leads to the bonding, is due to the fact that the two electrons can exchange their places without a change in energy; in He2, formed from an excited He* and a normal helium atom, the degeneracy is due to the fact that the excitation energy may be ex- changed between the two atoms in the molecule. The first kind of resonance, extended to a system of many nuclei, leads to electron conduction; the second kind of resonance leads, in a similar case, to an energy conduction. If the electron exchange in a crystal is strong in the ground state, the crystal shows metallic conduction. If only excited atoms (or molecules) easily exchange their elec- trons, the crystal has insulating properties in the dark, and shows photoconductivity in light. Chlorophyll in the chloroplasts is certainly not a metallic conductor; but can it be a -photoelectric conductor? This could lead to a variation of the theory of the photosynthetic unit in which an exchange of electrons between molecules would replace the exchange of excitations. The primary effect of light on the "photosynthetic unit" would then be the same as that of ultraviolet light on an alkali halide crystal: to set an electron free. This electron then could diffuse through the unit until it meets a "reduc- tion center" that absorbs it, the whole process being equivalent to oxidation of a chloro- phyll molecule and reduction of a substrate anchored somewhere else in a "reduction center." However, the similarity between the absorption spectrum of molecularly dis- persed chlorophyll and that of the chlorophyll in the cell seems to preclude the possibility that the excited state of chlorophyll in vivo is a "conductivity state" (which would be entirely different from the excited state in vitro). Consequently this variation of the "photosynthetic unit" (proposed by Katz 1949) must be rejected. 6. Energy Transfer between Different Pigment Molecules Related to the problem of energy transfer between identical molecules (e. g., between several chlorophyll a molecules in the hypothetical "photo- synthetic unit") is that of energy transfer between different molecules, 1300 THE PIGMENT FACTOR CHAP. 32 (e. g., between chlorophyll a and chlorophyll h, or between carotenoids, phycobilins, and chlorophyll). It was repeatedly suggested in chapter 30 that the (by now well established) sensitizing action of so-called "acces- sory" pigments in photosynthesis may be based on the transfer of the energy quanta, absorbed by these pigments, to chlorophyll — perhaps, the only pigment that can (or, at least, the only one that does) act as the chemical "photocatalyst" in photosynthesis. Theoretically, the probability of the resonance transfer of excitation energy between different molecules by the "slow" resonance mechanism is determined (similarly to that between identical molecules), by the over- lapping of the fluorescence band of the primary light absorber (energy donor) and the absorption band of the energy acceptor. In a mixture of several chlorophylls this overlapping may be so wide that the probability of energy transfer between them can be of the same order of magnitude as for the energy transfer between two molecules of a single component. Forster (1947) estimated, for example, that the probability of energy migra- tion from 6 to a is about one half that from one a to another a. The transfer in the opposite direction — from a to b — is, however, about 300 times less probable (Duysens 1952). Band overlapping is considerable also between the phj^cobilins and the chlorophylls, since the fluorescence of the former pigments lies in the region where the latter ones absorb strongly. It is generally less extensive be- tween the carotenoids and the phycobilins or chlorophylls. Arnold and Oppenheimer (1950) first discussed the probability of reson- ance transfer of energy (a process they called "internal conversion") be- tween two different photosynthetic pigments, using phycocyanin and chlorophyll in blue-green algae as an example. They estimated, from crude observations, that 1 or 2% of the quanta absorbed by phycocyanin in Chroococcus are re-emitted as phycocyanin fluorescence. Postulating that the relative probabilities of fluorescence and energy dissipation in the chromoproteid should be the same in vivo as in vitro, and noticing that in the pigment extract this ratio is about 1:4 (i. e., about 80% of the absorbed quanta are dissipated, and 20% re-emitted), they concluded that dissipa- tion can account, in vivo, for only 4-8% of the absorbed quanta. This leaves 90-95% unaccounted for by either fluorescence or dissipation — and thus, they postulated, transferred to chlorophyll. Arnold and Oppen- heimer estimated that the transfer probability ("internal conversion coef- ficient") required for a transfer yield of 90-95%, lies within the theoretical limits one can calculate from the classical model of two coupled oscillators, by using the smallest and the largest plausible value, respectively, of the average distance between the molecules of the two pigments in the Chro- ococcus cell. ENERGY TRANSFER BETWEEN DIFFERENT PIGMENT MOLECULES 1301 In chapter 24, we mentioned the fluorescence experiments of Van Nor- man, French and Macdowall (1948) which made it plausible (but did not prove definitely) that a transfer of excitation energy from phycocyanin to chlorophyll actually does occur in the red alga, Gigartina harveyana. Sub- sequently, French and Young (1952) developed a double-monochromator permitting the determination of the fluorescence spectra of algae with monochromatic excitation. These experiments will be described in Chapter 37C (section 7). They indicate that the energy absorbed by chlorophyll in the blue-violet band, is not transferred to the phycobilins (probably because the chlorophyll molecule in the upper state of the blue-violet band is changed practically instantaneously, by internal conversion, into the ex- cited state reached directly by absorption in the red band, leaving the mole- cule with a quantum too small to be acceptable to phycoerythrin or phyco- cyanin). The quanta absorbed by phycoerythrin, on the other hand, are transferred to both phycocyanin and chlorophyll a, causing simultaneous cule with a quantum too small to be acceptable to phycoerylthrin or phyco- fluorescence of all three pigments (figs. 37C.41-47). The fluorescence spec- trum excited by 453.5 m^ (a frequency absorbed mostly by phycoerythrin), undergoes a characteristic change upon heating: the phycoerythrin band increases in intensity relatively to the chlorophyll band indicating that the resonance coupling, responsible for the phycoerythrin-sensitized fluores- cence of chlorophyll, has been destroyed or weakened by heating. Systematic studies of sensitized fluorescence in algae and bacteria have been carried out by Duysens (1951, 1952); these, too, will be described in chapter 37C. They prove an effective transfer of excitation energy from all pigments present in photosynthesizing cells to the one pigment with the lowest excitation level {i. e., one whose absorption band is located furthest towards the infrared). In purple bacteria, this ultimate energy acceptor is one of the several bacteriochlorophylls (BChl "890"); in green plants and algae, chlorophyll a; in red algae, it may be either chlorophyll a or a minor pigment with an absorption band located (in vivo) at about 700 ran — perhaps chlorophyll d. The efficiency of energy transfer from phycobilins or carotenoids to chlorophyll a (or to bacteriochlorophyll "890"), evidenced by sensitized fluorescence, parallels closely the contribution of these pigments to photo- synthesis. This provides a strong support for the hypothesis that the quanta absorbed by the "accessory" pigments are utilized in photosynthesis by being first transferred to the main "photocatalytic" pigment — chloro- phyll a. The observations of French and Young (chapter 37C, section 7) that in red algae only chlorophyll a fluorescence shows induction phenomena and a dependence of the yield on light-intensity (i. e., indirectly, on the rate of photosynthesis) also support this hypothesis. The following estimates of the efficiency of energy transfer in vivo were made by Duysens for purple bacteria: 1302 THE PIGMENT FACTOR CHAP. 32 Chromatium: BChl "800" -> BChl "850" -> BChl "890": 100% Carotenoids -* BChl "850": '-'40% Rhodospirillum (molischianum or rubrum): Carotenoids -* BChl "890" : ^-40% In the species R. riibrum — as contrasted to other purple bacteria — a striking difference between the absorption spectrum and the action spec- trum of phofotaxis was noted by Manten (1948), and interpreted as proof of phototactic inactivity of one of their most abundant carotenoids, spirillox- anthol. The action spectrum of photosynthesis of Rhodospirillum rubrum is in this respect similar to that of phototaxis (Thomas 1950) ; thus, spiril- loxanthol seems to be inactive also in photosynthesis. The efficiency of energy transfer to BChl "890" (as determined from fluorescence measure- ments) is, however, according to Duysens, about the same for spirilloxan- thol as for other carotenoids (namely, about 40%). Duysens suggested that two types of carotene-bacteriochlorophyll complexes occur in Rhodo- spirillum rubrum. The complexes containing spirilloxanthol (which pre- dominate in older cultures) are inactive (both in photosynthesis and in phototaxis), while complexes containing other carotenoids (mainly, rho- dopol), which predominate in younger cells, are photobiologically active; the fluorescence of BChl "890" is excited with the same efficiency by the carotenoids in both complexes. Later, Clayton found spirilloxanthol to be effective in phototaxis of another strain of Rhodospirillum rubrum; Duysens suggested that this may have been due to the use of much lower light intensities than those used by Manten. Thomas and Goedheer (1953) again compared the action spectra of photosynthesis and phototaxis in Rhodospirillum rubrum, and confirmed the earlier observation (Thomas 1950) that, while the absence of the spirilloxanthol band is common to both of them, the two spectra differ in that the action spectrum of photosynthesis indicates a higher ef- ficiency of the "active" carotenoids than the action spectrum of phototaxis. An explan- ation was suggested for this difference in terms of a two-fold photochemical function of carotenoids — first as energy suppliers to photosynthesis, and second as photocatalysts for the consumption of photosynthetic products (carotenoid-sensitized "photorespira- tion," cf. Emerson and Lewis, chap. 20, p. 568 and Warburg et. al., chap. 37D, section 4). The second process reduces the accumulation of photosynthetic products in violet (com- pared to red) light, and this in turn reduces phototaxis (which, in purple bacteria — but not in algae or higher plants — seems to be stimulated by the products of photosynthesis). In green algae (Chlorella), the following transfer efficiencies were calcu- lated by Duysens: Chlorophyll b -* chlorophyll a: 100% Carotenoids -* chlorophyll a: 44% The first figure was obtained by comparison of the yields of fluorescence excited by 670 and 650 m^u — i. e., by a frequency absorbed only by a, and one absorbed about equally by h and a. The second figure was derived from ENERGY TRANSFER BETWEEN DIFFERENT PIGMENT MOLECULES 1303 fluorescence yield at 480 mn (using the assumption that the energy taken up by Chi h at this wave length is transferred to a with 100% efficiency). The spectrum of fluorescence was found to be the same with excitation at 480 m/i (where h absorbs much light) and at 420 mn (where a is the main absorber). No fluorescence band was noticeable around 650 m/z, where the peak of the fluorescence of chlorophyll b in vivo should be located. Both observations confirm the quantitative transfer of energy from 6 to a. (We note, however, that table 24.1 lists several determinations of the chloro- phyll b fluorescence peak in Ulva and Elodea, at 655-657 van; but perhaps these photographic data are unreliable.) Energy transfer was also observed between the chlorophylls a and b in solution (about 1.2 X 10 ~' M in each pigment), by comparison of the intensities of chlorophyll a fluorescence excited at 429 mn (70% absorption by a, 30% by b) and 453 m^u (5% absorp- tion by a, 95% by b). The ratio of the two intensities was only 1.7 (instead of 14, as ex- pected in the absence of a transfer). This indicates a 40-50% efficiency of transfer from b to a. As shown in fig. 37C.23, a marked fluorescence of b itself is emitted in this case (because the transfer to a is not 100% effective). At 5 X 10~^ moIe/1. the transfer efficiency was down to about 35%, and at 1.2 X 10 ~^ mole/1, to about 5%. According to Forster's calculations (equation 32.6), the probabiUty of the transfer a — *- a is 50% at 7.7 X 10 "^ mole/1., if a life-time of 3 X 10 ~^ sec. is assumed; with a life-time of 1 X 10~' sec. (natural life-time 4 X 10~* sec; 25% fluorescence yield), this "critical" con- centration, [k]o, becomes 7.7 X 10 ~* X v^ ^^ 1 X 10 ~' mole/1. As mentioned above (p. 1300), the efficiency oi b ^-a transfer should be about one half of that between two molecules of chlorophyll a, or 0.25 at 1 X 10""^ mole/1. To bring it up to 50% the concen- tration should be increased by \/2 — 1.4, i. e. to about 1.5 X 10~^ mole/1., while Duy- sens' experimental value is about 1.2 X 10 ~* X 11 mole/1. (The probability of transfer increases, according to Forster, with the square of concentration; the above quoted estimates of Duysens— 50% at 12 X 10"* mole/1., 35% at 5 X 10"^ mole/1., and 5% at 1.2 X 10 ~^ mole/1. — do not agree well with this law, but claim no precision.) In brown algae, the transfer of excitation energy from fucoxanthal to chlorophyll a, first demonstrated by Button, Manning and Duggar (1943) and confirmed by Wassink and Kersten (1946) (cf. chapter 24, p. 814), was again observed by Duysens (1952). According to the latter, the efficiency of energy transfer (as measured at 500 m/z) is: Carotenoids (mainly fucoxanthol) — >■ chlorophyll: '^ 70% This value can be compared with the ratio of quantum yields of photo- synthesis at 500 and 660 m/i, which, according to Tanada (cf. fig. 30.9A), is approximately 0.8. Measurements of the action spectra of photosynthesis and of chloro- phyll a fluorescence in the blue-green alga Oscillatoria, also made by Duy- sens (1952), showed (similarly to the experiments on red algae, see below) that quanta absorbed by phycocyanin are transferred with considerable (perhaps > 90%) efficiency to fluorescent and photosynthetically active chlorophyll a; but that the greater part (55-60%) of chlorophyll a is pres- 1304 THE PIGMENT FACTOR CHAP. 32 ent in a form which is nonfluorescent and photosynthetically inactive. With excitation at 528 m/x (absorbed mainly by phycocyanin) , both phyco- cyanin and chlorophyll fluorescence are emitted and together account satis- factorily for the total fluorescence of the algae. With excitation at 420 m/x (absorbed mainly by chlorophyll a and the carotenoids, and sHghtly by phycocyanin), no chlorophyll a fluorescence band appears, but only a — relatively weak — fluorescence band of phycocyanin, and a broad fluores- cence band of unknown origin (chlorophyll dl) at 720-750 m^u. It ma}^ therefore by hypothesized that the "nonfluorescent" and photo- synthetically inactive chlorophyll a is associated with a minor pigment (chlorophyll (P.) which has an even lower excitation level and serves as a "sink" into which the energy of excited chlorophyll a disappears, producing fluorescence of the acceptor (chlorophyll dl), but no photosynthesis. A possible reason for the loss of energy conveyed to "chlorophyll c?" is that this pigment is present in a very small concentration. Because of this, energy trapped in its molecules has no chance of further migration, and may have httle chance to reach a "reaction center" (assuming that resonance migration of energy to a reaction center is a necessary step in photosynthesis!). A smaU number of "traps" can be sufficient to catch most of the quanta on their way from chlorophyll a molecules to the reaction centers. The spatial arrangement of the phycocyanin and the chlorophyll mole- cules in the cell may be such that the energy quanta absorbed by phyco- cyanin have a better chance to reach the reaction centers (even if they have to pass through some chlorophyll a molecules on the way) , than the maj or- ity of quanta absorbed by chlorophyll a itself. The chemical properties of the phycobilins and of chlorophyll are so different that they cannot be uniformly mixed in the cell (or a cell organ, such as a chloroplast, or a granum). (This puts considerable theoretical strain on any hypothesis which would assume a 100% effective resonance transfer of quanta from the phycobilins to chlorophyll; and yet, such a hypothesis seems to be indi- cated by the fluorescence experiments.) The contribution of carotenoids is small, both to the chlorophyll a fluorescence and to the photosynthesis of blue-green algae (c/. chapter 30, fig. 30.10A). Similar results were obtained with the red algae, such as Porphyridium cruentum, by French and Young (c/. above p. 1301). In evaluating their (and his own) measurements, Duysens concluded that 80% or more of the quanta absorbed by phycoerythrin are transferred to phycocyanin (the fluorescence of the latter being proportional to the sum of absorption by both pigments). The transfer from chlorophyll to the phycobilins is neghgible (<10%), even for excitation with X 420 m^. ENERGY TRANSFER BETWEEN DIFFERENT PIGMENT MOLECULES 1305 Both 430 and G80 mju quanta — absorbed mainly by chlorophyll a — are less effective in exciting chlorophyll a fluorescence in Porphyridium than 560 m/x quanta, absorbed mainly by phycoerythrin (or 630 m/z quanta, absorbed to a considerable extent by phycocyanin), the ratio being similar to that found in Oscillatoria (about 0.4). Experiments on another red alga, Porphyra lacineata, suggested that the energy absorbed by "non- fluorescent" and (according to Haxo and Blinks, cf. chapter 30, fig. 30.11, and to Duysens 1952) photosynthetically inactive fraction of chlorophyll a, are transferred to — and produce the fluorescence of — a minor pigment component, probably, chlorophyll d (identified in red algae by Manning and Strain, cf. p. 720). The fluorescence peak of this pigment Hes at 725 m/x. Duysens (1952) made an attempt to interpret all these experimental results cjuantitatively by applying Forster's theory of resonance transfer to mixed solutions of several pigments. He derived from Forster's equations the efficiency, E, of excitation energy transfer, E, from a molecule of type "/' to any of the molecules of type "A;" in such a mixture, as function of con- centration [/c], expressed in terms of a "critical" concentration, [k]^. The latter is the concentration at which the theoretical transfer probability for a single pair (7, k) would be 0.5, if all molecules were arranged in a simple cubic lattice — in other words, the lattice constant were equal to the "criti- cal distance," da, defined by Forster's equation (32.6). Following is Duy- sen's table: [k]/{k\f>: 0.05 O.il 0.27 1.00 >1 E: 0.3 0.5 0.75 0.96* 1 - 0.036 [^j" To apply the above table to the calculation of energy transfer in plant cells, Duysens had to make rough estimates of the concentrations of the dif- ferent pigments, their fluorescence yields (which determine the life-times available for transfer) and the "overlap integrals" between the absorption and fluorescence spectra of the several pigment pairs. For example, to estimate the probability of energy transfer from chloro- phyll h to chlorophyll a, in vivo, Duysens assumed

0.046 mole/1., the transfer efficiency from 6 to a should be, according to the table, >96%; the reverse transfer, from a to h, calculated in the same way, turns out (as mentioned on p. 1300) to be 300 times less probable, and thus experimentally undetectable. * If the transfer probability to a single neighbor is 0.5, that to any neighbor in a cubic lattice is 0.96. 1306 THE PIGMENT FACTOR CHAP. 32 For energy transfer from phycocyanin to chlorophyll a in blue-green algae,

90%, and the yield of fluorescence of phycocyanin would thus be reduced from 70% to 0.1%. Observations of Arnold and Oppenheimer indicated, however, a yield of the order of 5%; Duysens saw in this an indication that in blue-green algae chlorophyll and phycocyanin are sepa- rated in different structural units. This view is supported by the fact that phycocyanin can be separated from chlorophyll by fractional precipita- tion or centrifugation. Chlorophyll seems to be contained in (grana-like) "chromophores" (cf. p. 1741), while phycocyanin seems to be located in the extragranular plasma. This separation would increase the average dis- tance between a phycobilin and a chlorophyll molecule to something like 1000 A. ; offhand, this seems much too far to permit the observed, about 90% effective, energy transfer. Perhaps the difficulty can be over- come by assuming a migration of energy through numerous phycocyanin molecules in the cytoplasm, to the granum boundary, where it can be transferred to chlorophyll. Another possibility is that phycocyanin is located, in the main, not outside, but inside the grana in the protein layers, while chlorophyll occupies the interfaces between protein and the lipoidic layers. This assumption could bring the theoretical rate of energy transfer from phycocyanin to chlorophyll in line with the observed 5% residual fluorescence of phycocyanin ; but it calls for the further hypothesis that in the smashing of the cells, phycocyanin is leached out of the grana, while chlorophyll remains bound to their "skeleton." More recently this hypoth- esis has found support in the observations of McClendon and Blinks (1952) that the loss of phycobilins can be prevented by the addition of high-molecular substances to the medium {cf. p. 1754). Duysens estimated the probability of the transfer of energy from phycoerythrin to phycocyanin and chlorophyll in Porphyridinm cruentum, assuming all thi-ee pigments to be present in a common phase, and concluded that because of the greater band overlap the transfer to phycocyanin should be about five times faster, and thus occur to a prac- tical exclusion of that to chlorophyll. The transfer from phycocyanin to chlorophyll would than follow as a second step, with the efficiency estimated in the discussion of OsciUatoria. The overlap integral is negligible for the reverse transfers — from chloro- phyll to any one of the phycobilins. ENERGY TRANSFER BETWEEN DIFFERENT PIGMENT MOLECULES 1307 In purple bacteria the overlap between the fluorescence spectrum of Imcteriochlorophyll "800" and the absorption spectrum of bacteriochloro- phyll "850," as well as that between the fluorescence spectrum of bacterio- chlorophyll "850" and the absorption spectrum of bacteriochlorophyll "890," is considerable, so that, with

t ^ t V. t Fig. 33.2. O2 liberation during induction period in Ricinus leaf (after Blinks and Skow 1938). The first investigation in which the oxygen liberation during the induc- tion period was followed by a direct "differential" method was that of Blinks and Skow (1938-). They used the polarographic method described in chapter 25 (page 850). To avoid time lags, a small-surface platinum or mercury electrode was pressed against a thallus or a submerged leaf. With a polarization potential of 0.5 volt, the current was proportional to the oxygen concentration in the thin layer between plant and electrode. The capacity of this layer was so small that fluctuations lasting for less than 0.02 second could be recorded. The "time resolution" in Blinks and Skow's experiments was thus a thousand times higher than in the manometric experiments of Warburg, van der Paauw and Smith. Where the latter noted only a smooth increase in oxygen liberation in the first 2 to 5 minutes of illumination, the polarographic curves now revealed an initial "gush" of oxygen, which was over in a few seconds (fig. 33.2). 1320 INDUCTION PHENOMENA CHAP. 33 In the experiment recorded in the figure the lower surface of a castor bean leaf {Ricinus) was pressed tightly against a stationary mercury elec- trode. The oxygen was down to nearly anaerobic levels due to respiration and reduction by the electrode. The leaf was then illuminated (through the agar) with light of about 7500 meter candles during the shaded parts of the record : in a after a 10 minute dark period, in h after 34 seconds, etc., as shown in the figure. We note the immediate, brief gush of oxygen. Verti- cal lines are time marks, 1 minute apart; arrows indicate beginning of ZXISSSSJWS »-*»-tS4*ttitii<*llt |t-^-(^o^p-"j«-4^«fr J^-^.,^^^^.^!^^^^ Fig. 33.3. O2 liberation in light after dark intervals of the order of 1 sec. (after Blinks and Skow 1938). (a) 0.45 sec. light, 1.35 sec. dark; (b) 0.9 sec. light, 0.9 sec. dark; (c) 0.7 sec. light, 1.1 sec. dark; (d) later portion of c, with long dark period in last half of record. O2 production begins within the period of the galvanometer, and is nearly con- stant during the flash, or slightly higher at the beginning. Time marks are 0.2 sec. apart. Full height in d represents O2 content of water in equilibrium with air. illumination. (Similar records were obtained with a platinum electrode and with marine algae (e. g., Ulvd) without stomata or gas spaces.) In accordance with Warburg's and van der Paauw's data the inhibition begins to disappear after about 1 minute of illumination, and oxygen hberation is well under way toward the end of the second minute. Figure 33.3 shows that, even after dark intervals of only 1 second, the oxygen evolu- tion during the first half second of illumination is slightly faster than in the second half second; this may or may not be the first sign of the "gush" and subsequent inhibition OXYGEX EXCHANGE DURING THE SHORT INDUCTION PERIOD 1321 In figure 33.3 the duration of flashes is indicated by the paler parts of the record, flashes were produced by a revolving disc in front of an incandescent source (6000 meter- candles). Bhnks and Skow associated the oxygen gush with the "inverse induction," observed by Gaffron in the investigation of the influence of anaerobiosis on green algae (c/. chapter 6); and therefore emphasized that in their experiments, in consequence of respira- tion and cathodic reduction, an anaerobic state was reached at the end of the dark period in the layer between electrode and plant. However, the effects of anaerobic incubation usually appear only after one or several hours in an oxygen-free medium (c/. section 6 below, and chapter 6). Blinks and Skow observed the oxygen gush not only with the leaves of a higher land plant {Ricinus), but also with an aquatic plant (Potomageton) and an alga {Ulva). They concluded that the gush is a general feature of E 40 o O in CM O g. 30 o I- < en < 01 < a. a. < 20 10 - ~ / / / / (A) / / - 1 1/ 1/' 1 18 2° C. 1 1 J_ 10 20 30 TIME, min. 40 20 30 TIME, min. Fig. 33.4. Induction in Fucvs after 16 hrs. in dark at 5.6° and 18.2° C. (after Stee- mann-Nielsen 1942): (A) R = 0.45 mg. O./cm.^ hr.; (B) R = 0.25 mg. Oj/cm.^hr. the kinetics of photosynthesis. We will see below that Warburg and co- workers reached a similar conclusion, and ascribed to it far-reaching theoretical impUcations, but that other experiments indicate extreme vari- ability of the gush in duration and volume, and thus argue against its inter- pretation as an essential feature of photosynthesis. A more extensive, if less precise, study of induction in the evolution of oxygen was carried out by Steemann-Nielsen (1942) with the multicellular brown alga, Fucus serratus. He used Winkler's method of oxygen deter- mination in water. Steemann-Nielsen incubated the thalli for as long as 16-17 hours before exposure to light, in streaming sea water. Neverthe- less, the induction phenomena lasted for only about 10 minutes at 18° C. 1322 INDUCTION PHENOMENA CHAP. 33 and 20 minutes at 4°, and thus appeared to fall into the category of "short induction." The method used by Steemann-Nielsen (involving measuring periods of 2.5 to 5 minutes) was much too slow to reveal the oxygen gush; but it showed clearly the gradual increase in the rate of oxygen production, which lasted for 20-30 minutes at 4-6° C, and for about 10 minutes at 18° (in light of 23,000 or 2300 lux, in 2.5 X 10-» mole/1, bicarbonate solution). The same temperature was maintained during the incubation and the ex- posure. Figure 33.4 illustrates the results. 5 10 15 20 INTENSITY OF CONDITIONING LIGHT, klux Fig. 33.5. Initial rate of photosynthesis in Fucus in 23 klux as function of pre- vious illumination (after Steemann-Nielsen 1942). Average for first 5 min. (solid curve); initial rate extrapolated to zero time (broken curve). Steemann-Nielsen made observations on the repetition of induction upon stepwise transition from weaker to stronger light, a phenomenon previ- ously noted by McAlister in experiments on CO2 consumption (section 3 below). After Fiicus had reached the steady rate of photosynthesis in 2300 lux, an increase to 23,000 lux produced, in the first 5 minutes, only 65% of full rate; the latter was reached after 15 minutes additional induc- tion (at 5° C). Similar results were obtained with other transitions (7000 -^ 23,000 lux, 6°, 82% of maximum rate in the first 5 minutes; 1200 -^ 23,000 lux, 5.5°, 45% of maximum rate in the first 5 minutes). Fig. 33.5 shows the first measured rate in strong light as a function of inten- sity of the "conditioning" illumination. Steemann-Nielsen noted that the dotted curve (which represents the rate in strong light, extrapolated to the moment of light increase, as a function of conditioning intensity) is identi- cal with the light curve of photosynthesis, P = f(I) , of the same plant. He concluded that the "activation" of the dark-inhibited factors of the OXYGEN EXCHANGE DURING THE SHORT INDUCTION PERIOD 1323 photosynthetic apparatus in light proceeds only to the level sufficient for the maintenance of the maximum steady rate possible at the prevailing Hght intensity (this maximum being determined by factors not involved in induction). A similar conclusion was reached earlier by Gaffron and Franck (c/. part C). Steemann-Nielsen's observations of the time required for the prepara- tion of induction in the dark are iUustrated by figure 33.6. The curve showing the initial rate in light as a function of the length of the preceding 10 ^0 (0 Ao 100 uo mo 140 no 250 Fig. 33.6. Average rate of O2 liberation by Fucus serratus in first 5 min. of light of 23 klux at 4.8° C, as function of preceding dark incubation (after Steemann- Nielsen 1942). Abscissa, minutes dark after activation to Pmax. at 23 klux. Ordi- nate, per cent of Pmax. (net). dark interval appears to consist of two parts : About one half of the total induction loss is developed after 10 minutes in darkness; 50 minutes of additional incubation are required to reduce the rate by another factor of 1/2. A similar relation was found for the carbon dioxide exchange (Mc- Alister, fig. 33.9). This indicates the superposition of two different induc- tion-producing dark reactions: one being completed in about 10 minutes (at 18° C), the other requiring several hours. Fluorescence measurements (part B) also revealed two dark deactivation processes — one appeared to reach saturation in about 1 minute ; the second one was much slower. The duration of the dark period required to bring about a certain per- centage inhibition was found by Steemann-Nielsen to be independent of temperature, if incubation and illumination were carried out at the same temperature. Experiments in which only the incubation temperature was varied (while illumination temperature was kept constant) showed, on the other hand, that the absolute rate of the deactivation process was— as 1324 INDUCTION PHENOMENA CHAP. 33 expected — strongly dependent on temperature; its temperature coef- ficient appeared to be about 2. Steemann-Nielsen found, with Fucus vesiculosus, at 2.9° C, that the duration of the recovery period (~30 minutes) was not affected by absence of carbon dioxide during the first 10 minutes of illumination. If, however, the plants were deprived of carbon dioxide for as long as 40 minutes in light, the rate of oxygen production immediately after the readmission of carbon dioxide corresponded only to the level usually reached in 10 minutes after the beginning of illumination. Steemann-Nielsen saw in this an indication that Fucus contains an internal reserve of carbon di- oxide, sufficient to maintain the normal course of activation for about 10 minutes. These observations contrast with McAlister's report (c/. section 3) that "activation" of wheat in light can be completed in the absence of carbon di- oxide, so that the carbon dioxide uptake begins at full rate instantaneously after readmission of this gas. (Of course, carbon dioxide uptake by re- plenishment of an empty "reservoir" of bicarbonate in the cells is not im- mediately distinguishable from uptake by reduction, and could occur without the liberation of an equivalent volume of oxygen.) It is known (cf. Vol. I, chapter 19) that illumination of plants in absence of carbon dioxide causes photoxidation. The induction phenomena observed after a period of car- bon dioxide starvation must therefore be related to induction phenomena that occur after photoxidation (to be described in section 6). It will be noted that none of the experiments discussed so far has indi- cated the occurrence of a "second depression" of oxygen liberation, match- ing a depression often found in carbon dioxide uptake curves (cf. sect. 3), and the "second wave" of fluorescence (cf. part B). Franck, French and Puck (1941) therefore proposed an explanation of the second depression not requiring a parallelism between oxygen liberation and carbon dioxide absorption. Later, however, Franck, Pringsheim and Lad (1945) noted that a second depression actually did occur in the oxygen induction ciu'ves of ChloreUa, as determined by the trypoflavine phosphorescence-quenching method (cf. chapter 25, p. 851). This method can be used only under ana- erobic conditions; but Franck and co-workers considered these conditions irrelevant in this case and concluded that the experiments indicate a second depression of oxygen liberation occurring under the same conditions as the depression of carbon dioxide uptake. Warburg and co-workers used the two-vessel manometric technique to calculate oxygen and carbon dioxide exchange, in light and darkness, in minute-to-minute intervals (for a description of the method, cf. chapter 25, p. 848; cf. also discussions in chapter 29, pp. 1109-1112, and chapter 37D, section 4). OXYGEN EXCHANGE DURING THE SHORT INDUCTION PERIOD 1325 Warburg, Geleick and Briese (1951) studied dilute Chlorella suspensions, illuminated with a red light beam, superimposed on white "background" illumination just sufficiently to compensate steady respiration. These "liright periods" of a few minutes duration alternated with similar periods of illumination with the white "background" light alone ("dim periods"). A very strong positive induction was noted at the beginning of both the bright and the dim periods. A gradually subsiding "burst" of oxygen was found in the first 1-2 minutes of "bright" illumination, and a gradually Fig. 3.3.6A. Splitting of photo- synthesis into hght reaction and dark hack-reaction at 20 °C. (two-vessel measurements l>y Warl^urg, Geleick and Briese 1951). 100 (A. cells per vessel, compensating white back- ground light; 546 ni/u light added for first 3 minutes. 10.0% CO2. Pohits: average of measurements at cor- responding tim9s in 1(1 cycles. E E 3 - , D.6I ■ // (\ \ It / ^ r *■ // *" r \ X Lr '■ 0 7 1 1 Net photosynthesis y = 0.19 3 I MINUTES subsiding "gulp" of oxygen in the first 1-2 minutes of "dim" illumination. Fig. 33.6A shows this for a 3 minute light-3 minute dark cycle at 20° C. With the type of manometers used in this work, the precision of single minute-by-minute pressure change determinations was low; the points in the figure were therefore obtained by averaging the changes in the corre- sponding minutes of ten cycles. Warburg et al. interpreted the results of the type of those in figure 33.6A (more specifically, the steep initial slopes of the "bright light" segments) as evidence that oxygen liberation begins, after a few minutes of darkness, with a quantum requirement as low as unity. The average ratio Qp ( = A02/-AC02) calculated from the total gas exchange during ten three- minute periods of bright light, was about 0.85; this was considered as evidence that the burst was one of complete photosynthesis (for which the theoretical ratio is Qp = 1.0). The average ratio Qp{= AO./ACO.), calculated from the total gas ex- 1326 INDUCTION PHENOMENA CHAP. 33 change in ten three-minute "dim" periods, also was about 0.85; this was considered evidence that the "gulp" was due to a complete reversal of photosynthesis — i. e., oxidation of a substrate of the approximate reduction level of a carbohydrate by molecular oxygen. From the initial slope of the "dim light" curve segments, it was calculated that immediately after the cessation of "bright light," the rate of this back reaction (which we may call "antiphotosynthesis," to distinguish it from ordinary respiration), can be ten or more times that of steady respiration of the same cells (measured over extended periods of darkness). As reported in more detail in chapter 37D (section 4a), Warburg et al. interpreted the observed "bursts" and "gulps" of O2 and CO2 not as induc- tion phenomena, but as revelations of the hidden mechanism of photosyn- thesis, showing it to consist of a forward, photochemical reaction with a quantum requirement of 1 (one quantum per O2 produced and CO2 con- sumed), and a thermal back-reaction that consumes a large proportion of the products of the forward reaction. This proportion must be large enough for the liberated chemical energy (added, via some "chemosynthetic" mechanism, to the one quantum of light energy used in the forward process) to reduce the net yield of the cycle to a value compatible with the law of conservation of energy (=^2.8 quanta per O2 molecule). The postulated partial separation in manometric experiments of the forward photochemical, from the reverse thermal reaction implies that the back reaction has a rate constant as low as a reciprocal minute, so that the light-enhanced consumption of molecular oxygen needs a minute or two to get under way at the beginning of a light period, and is carried over for a minute or two into the following dim period, thus permitting tell-tale "bursts" and "gulps" to be caught by minute-to-minute manometric readings. Warburg and co-workers (1951) noted that activity bursts of the above- described type are not observed in carbonate buffers, and suggested that the back reaction is much faster in these media; later (1953, 1954) they found that ChloreUa cultures grown in a different way showed no activ- ity bursts of similar duration also in carbon dioxide solutions. Wide variations in the duration of the oxygen bursts and gulps was ob- served also by Damaschke, Todt, Burk and Warburg (1953) with a rapid electrochemical method of oxygen determination. The method was simi- lar to that used by Blinks and Skow, and the results were qualitatively similar to those described above on p. 1319. These measurements, too, showed a rapid oxygen burst at the beginning of illumination, and an oxygen gulp after its termination. (Fig. 33. 6B shows the recordings of gulps after bright periods of 3, 8 and 28 seconds.) OXYGEN EXCHANGE DURING THE SHORT INDUCTION PERIOD 1327 In contrast to the qualitative observations of Blinks and Skow, the re- cordings of Damaschke et at. were used to evaluate the absolute yield of oxygen production during the burst, and its net yield in a complete light- dark cycle; these calculations will be dealt with in more detail in chapter 37D. Depending on (unspecified) conditions, the duration of the "gulp" was found in these experiments to vary from a few seconds (as in fig. 33. 6B), to one, or even ten minutes. (Among widely varied conditions, one notes 30 seconds Fig. 33.6B. Polarographic recordings of Oo liberation by Chlorella (160 mm.' cells per 80 cc.) exposed for 3, 8 and 28 sec. of additional illumination, on top of continuous light sufficient to compensate respiration (after Damaschke, Todt, Burk and Warburg 1953). Time scale from right to left; gulps of O2 are seen to follow the exposures. 20% CO2, 20° C, X = 644 m^. the carbon dioxide concentration, which was 5%, 20% and even 54% in the several experiments illustrated in the paper.) The wide variation in duration (and volume) of the oxygen bursts and gulps observed by Warburg and co-workers in manometric, as well as elec- trochemical measurements in differently pretreated ChhreUa cultures, makes their interpretation as revelations of an intrinsic thermal component of the reaction sequence of photosynthesis rather implausible. This view is supported by evidence from other investigations described in this chapter, such as the polarographic studies of Brackett, in which no oxygen burst was noted at all (while the gulp was clearly noticeable, but much less prominent than in Warburg's curves) ; and Brown's mass-spectrographic data (in which the respiratory oxygen consumption of Chlorella was 1328 INDUCTION PHENOMENA CHAP. 33 found to be, as a rule, unchanged by illumination, while sudden, but rela- tively small, increases in respiration were occasionally noted after the end of an illumination period). A similarly bewildering variety of carbon dioxide induction phenomena will be described in the next section — ranging from a photochemical carbon dioxide gulp (van der Xeen, Warburg) to a photochemical carbon dioxide burst (BUnks, Emerson, van der Veen). +6 +4 ■t-2 0 -2 -4 g +2 = -2 .70 ^/ .60 .50 40 imr ;+8 t4 t2 0 -2 »-4 I +4 +2 O -2 E T 1 1 r 16 18 0 2 4 6 8 10 12 14 16 18 TIME IN MINUTES "T — I 1 — r n T — r — r 16 18 0 2 4 6 8 10 12 14 16 18 TIME IN MINUTES Fig. 33. 6C. Two-vessel manometric measurements (after Emerson and Chalmers 1954). O2 liberation by Chlorella in carbonate buffer, in vessel pairs of type H, hi (left) and H, hi (right) (cf. fig. 29.4A). Minute-by-minute pressure changes measured by twin differential manometers. The ratio of measured pressure changes varies in the two vessels on left, remains nearly constant on right, indicating synchronous reaction. All these results leave one with the impression that rapid fluctuations of gas exchange after the transfer of cells from darkness to Ught (or light to darkness) are "transients," caused by readjustments of the catalytic systems (and intermediate products) to new steady states (in other words, typical induction effects), rather than "tail ends" of steady activities in the preceding periods. This impression is further strengthened by new observations of Emer- son and Chalmers (1954), who were able to improve considerably the precision of the two-vessel manometric method by using a dou))le differen- tial manometer, making simultaneous readings with two cathetometers. Two matched light beams from the same source were thrown onto two OXYGEN EXCHANGE DURING THE SHORT INDUCTION PERIOD 1329 vessels, whose shape was carefully chosen to achieve the best possible synchronization of the gas exchange processes. The importance of the shape of the manometric vessels for the reliability of two- vessel determination was mentioned before on p. 1111. Slight differences in the effectiveness of stirring in the two vessels can become crucial in the analysis of transient phenomena. Emerson and Chalmers (1954) showed this by comparing the manometric data obtained in vessel pairs of different shape with Chhrella cells suspended in carbonate buffers (where only one gas is exchanged). With a vessel pair of the type used by Warburg and co-workers {H and /ii in fig. 29.4A), the course of pressure equiUbration was markedly different in the two vessels, as illustrated by the left side of figure 33. 6C; the synchronization was much better with vessels of the ''Emerson type" {H and ho in fig. 29.4A), as shown by the steadiness of the ratio of the pressure changes during the transition from dark to light and back, on the right-hand side of the diagram. The reason for this difference must be that shaking at a certain rate has a different effect for different distances between liciuid and ceiling. Furthermore, vessels "matched" for one gas, are not eo ipso matched for others. Emerson believes that, because of these experimental uncertainties, Warburg's quantitative analysis of manometric readings in the transient periods (and of the quantum yields derived from this analysis) are open to grave doubts; particularly uncertain is the determination of the carbon dioxide-oxygen ratio (which is the basis for the conclusion that the ob- served bursts involved photosynthesis and respiration as a whole). Emerson and Chalmers' measurements (1954) with vessels of better synchronous behavior confirmed that an "oxygen burst" does occur some- times in the first minute of illumination (as first noted by Blinks and Skow, and also indicated by the manometric measurements of Burk, Warburg et al., and by the electrochemical measurements of Damaschke et ah). However, this burst is not a regular feature of photosynthesis in Chhrella; nor does it usually have the volume required by Burk and Warburg's theory. Quantum requirements of 3 (quanta per oxygen molecule), or even less, sometimes can be calculated by comparing oxygen production at the height of the burst {i. e., in the first minute of illumination) with the peak oxygen consumption subsequent to the fight period; however, the latter peak is reached, according to Emerson and co-workers, not immedi- ately after the cessation of illumination, but several minutes later (fig. 33. 6D). The manometric method is too slow to catch gas bursts or gulps if they are not big enough to affect significantly the gas exchange in a whole minute. Therefore, Emerson's measurements can be used only as a check of Warburg, Geleick and Briese's conclusions concerning the volume and 1330 INDUCTION PHENOMENA CHAP. 33 general significance of the bursts. They cannot be compared with the polarographic data of Blinks and Skow, and of Damaschke, Todt, Burk and Warburg, whose instruments were able to record bursts and gulps lasting only a few seconds, irrespective of their volume. The transient gas bursts in light (and the equivalent gulps at the begin- ning of darkness) increase, according to Emerson's observations, with the number of cells in the vessel, and the carbon dioxide concentration, while the steady rate of oxygen liberation or consumption (reached after 5 or 10 z + 2 0 -I +-I 0 -I S 9 ^ .8 *"/ 7 c -eh ! +6 » o. +4 J =1 +2- «r ? 0 o f -2- "^ -4- (A ^ O " 1 1 1 1 T - "U_ H I : "^ :-. -J J J 1 M I 1 1 i 1 'dark LIGH- DARK 1 " }\Wl^ ' y M-r-^^_™_ n=co2 _ J-M-'- ■=02 y_ J 1 1 1 _1 1 — 8 12 16 20 TIME IN MINUTES 24 Fig. 33. 6D. Example of transient gas exchange phenomena measured with "synchronized" manometric vessels (after Emerson and Chalmers 1954). Shows O2 burst and CO2 gulp in first minute, CO2 burst in second to fourth minutes, steady photosynthesis, with Qp — 1 after the sixth minute. minutes in light, and only after a considerably longer time in darkness), is much less affected by these factors (more about this in the next section, since these transients have been studied more extensively in carbon dioxide than in oxygen exchange) . Van der Veen (1949^--) apphed another rapidly registering method — the measurement of heat conductivity — to the study of induction. In air, or nitrogen, this method is used to determine the exchange of carbon di- oxide independently from that of oxygen (cf. below, section 3); but if a hydrogen atmosphere is used, changes in both oxygen and carbon dioxide OXYGEN EXCHANGE DURING THE SHORT INDUCTION PERIOD 1331 content affect the heat conductivity. By absorbing carbon dioxide chemi- cally before the gas enters the "diaferometer," it becomes possible to meas- ure the oxygen exchange alone. Van der Veen (1949^) demonstrated in this way that in heat-pretreated algae the surviving carbon dioxide gulp and gush (after exposure to light and darkness, respectively, cf. next sec- tion) are unaccompanied by any oxygen exchange {cf. fig. 33.6E)— a sug- gestive example of the independence of the exchange of carbon dioxide from that of oxygen during the transient periods. The electrochemical data of Blinks and Skow, and of Damaschke et al, can be compared with the results obtained by a similar method by Brackett and co-workers. cS"! , ^ y N CO2 line 0 cS'g .O2 line Time Fig. 33.6E. CO2 gulp upon illumination (upward arrow) and burst upon dark- ening (downward arrow), of heat-treated leaves of Holcus lanatus (after van der Veen 1949^). No uptake or release of O2. Measured with a diaferometer. Olson and Brackett (1952) and Brackett, Olson and Crickard (1953 '-2) studied the induction phenomena in Chlorella with a polarographic oxygen meter that permitted readings every 10 seconds. They used dilute cell suspensions, illuminated simultaneously from two sides; the uniformity of illumination, achieved in this way, removed, according to their findings, some adventitious features of the induction picture. Brackett, Olson and Crickard (19530 found that the respiration of Chlorella underwent considerable changes in the first few minutes of dark- ness after exposure to light. The typical picture of this respiration induc- tion included a gulp of oxygen taking place in the first 10-20 seconds of darkness (probably related to the burst observed by Damaschke et al, but apparently smaller in volume), a minimum, a second, flatter maximum reached in 1-4 minutes (probably related to the respiration peak noted manometrically by Emerson et al), and a gradual decay to a steady level. (These observations are to be discussed again under "Photosynthesis and Respiration," section 3, Chapter 37, cf. figs. 37D.31 and 32.) During the illumination period, Brackett et al assumed respiration to change smoothly (and approximately logarithmically) from its level at the end of the pre- 1332 INDUCTION PHENOMENA CHAP. ;v.i ceding dark period to its level 0.5-1 minute after the beginning of the sub- sequent dark period, when the initial respiration burst has subsided. (This meant assuming, in contrast to Warburg, that the burst was something that happened entirely after the end of illumination, and not the "tail end" of enhanced respiration during the light period.) Brackett et al. used the \ r nx^AAWtA^'^f*. yK-'X* W*^» '** — •--K.v...^ 4- w ^ .11;;^ s n Fig. 33. 6F. Aerobic O2 induction in Chlorella measured potentiometrically in three successive 3 minute light periods with 9 minute dark intervals: dashed line, interpolated respiration; circles, measured points; crosses, points corrected for respiration (after Brackett, Olson and Crickard 1953^). Abscissa, time (min.). Ordinate, rate of oxygen evolution. SO interpolated time curve of respiration to correct the measurements of oxygen production during the light period (cf. figs. 33. 6F and 37D.32), and obtained in this way curves showing only the normal negative induction in light. The latter lasted between 10 seconds and 2 minutes, and was fol- lowed by a very steady photosynthesis for the rest of the light period. No evidence of an oxygen burst in the first seconds of illumination is visible on these curves. OXYGEN EXCHANGE DURING THE SHORT INDUCTION PERIf)D 1333 Figure 33. 6F, taken from the second paper (1953-) of the same authors, shows typical oxygen induction curves in three successive three minute light periods, interrupted by nine minute dark periods. The circles repre- sent luicorrected rates, the crosses, rates corrected for interpolated respira- tion (dotted lines below). After thorough dark adaptation, the initial rate of oxygen lil)eration in light may be zero (or almost zero, cf. the conclusions of Franck ct al. in section G), and induction may cover up to 3 minutes, as in the first segment of figure 33. OF. The initial rate becomes higher, and the induction loss declines, in repeated cycles. The induction loss of oxy- gen production appears complementary to Emerson and Lewis's carbon dioxide burst (cf. below section 3) ; in fact, it can be accounted for approxi- mately by assuming that, for each cjuantum of light diverted from the pro- duction of oxygen, one molecule of carbon dioxide is liberated. In these experiments, too, there is no sign of an initial burst of oxygen in light ("positive" induction, interpreted by Warburg and co-workers as evidence of a "one-quantum process of photosynthesis"). It seems cer- tain from Blinks' and Damaschke's polarographic measurements, however, as well as from Emerson's more recent manometric results, that such a burst does occur in some cases; however, its occurrence does not seem to have the generality (and therefore probably, also, the significance) at- tributed to it by Warburg, Burk and co-workers. Brackett, Olson and Crickard (1953^) observed a second, shallow mini- mum of oxygen liberation on some of their induction curves (about 1 minute after the beginning of illumination). This, too, reminds one of the shape of Emerson and Lewis's carbon dioxide burst (as well as of that of Aufdemgarten's and van der Veen's carbon dioxide gulp, both of which show two or more successive waves, cf. section 3). Brackett et al. plotted the oxygen liberation rate during the induction period (after complete dark adaptation) on a log (Po/P) = f{t) scale, and found that different runs showed different deviations from the logarithmic approach to the steady rate — curvatures, breaks in slope, etc., seemingly indicating complexity and variability even of the "normal" negative oxygen induction. The most significant evidence concerning Warburg's hypothesis of oxy- gen burst as evidence of strongly enhanced respiration in light was pro- vided by isotopic tracer studies of Brown and co-workers. Brown (1953) measured respiration during illumination, independently from photosynthesis, by using 0(18)-enriched oxygen {cf. chapter 37D, section 3). He found, with ChloreJla, either no change in respiration rate at all, or a sudden increase in respiration at the end of the illumination period. If this result were generally valid, it would make Brackett's method of inter- polation of respiration in light (figs. 33. 6F and 37D.32) inexact, affecting 1334 INDUCTION PHENOMENA CHAP. 33 slightly the shape of the induction curves calculated by the latter (fig. 33. 6F). Much more important is, however, the failure of Brown's measure- ments to support in any way Warburg and Burk's hypothesis of a strongly enhanced oxygen consumption in light, at least as far as Chlorella is con- cerned. Much stronger changes of respiration in light were observed by Brown and Webster (1953) in the blue-green alga Anabaena (cf. chapter 37D, section 3) ; they ranged (in dependence on Hght intensity and partial pres- sure of oxygen) from complete inhibition of respiration within a minute or two after the beginning of illumination (in 0.2-0.4% O2, and P ^=^ S.5R), to stimulation of respiration by a factor of two or more (in more intense light and higher oxygen concentration). Measurements of transient phenomena in cells of this type obviously could be entirely misleading if a constant (or continuously interpolated) respiration correction were applied to the pressure readings. It may be, however, that such extreme respira- tion changes in light can occur only in blue-green algae, in which the site of photosynthesis is not separated morphologically from the main site of cell respiration. Hill and Whittingham (1953) used the hemoglobin conversion to oxy- hemoglobin for rapid spectroscopic oxygen determination in the induction phase of Chlorella: the delay of response was thus reduced to <10 seconds, from about 1 minute in parallel manometric measurements. The course of induction was the same, corresponding to half-reactivation in 0.5 min- ute at 15° C. for dark incubation times of 13, 30 or 45 minutes; even 22 hours spent in darkness did not appreciably slow down the induction. Ex- periments in alternating light at different temperatures showed that at 7° C. induction losses were negligible after dark periods of 1 minute or 3 min- utes, but noticeable after 10 minutes darkness; at 16° and 25° C. they be- came significant after 3 minutes in the dark. The half-time of reactivation after a given short dark period was shorter at the lower temperatures. No positive induction (oxygen burst) could be noted in any of the runs. 3. Carbon Dioxide Exchange during the Short Induction Period Observations of short induction by carbon dioxide determination were first made by McAhster (1937, 1939, 1940) with wheat plants. He used the infrared spectrophotometric technique described on page 852. At first, the rate was calculated from carbon dioxide concentration measure- ments made at 0.5 minute intervals; later, it was recorded directly by a differential recorder. Figure 33.7 gives an example of curves obtained by the earlier, point- to-point method. The downward trend of the curves corresponds to liber- CARBON DIOXIDE EXCHANGE DURING THE SHORT INDUCTION PERIOD 1335 ation of carbon dioxide, the upward trend to its consumption. The induc- tion periods last for about 2 minutes (four points) at 31° C, and 3.5 minutes (seven points) at 12°. Figure 33.8 shows that, at a given temperature, the duration of the induction period is practically independent of light inten- sity between 214 and 1030 foot-candles {i. e., between 2000 and 10,000 lux), but that induction is absent at 44 foot-candles (about 450 lux). McAhster designated as "induction loss" (AP,) the amount of carbon dioxide that would have been taken up but for induction. It is determined, in fig- ures 33.7 and 33.8, by the intercept of the solid straight lines (which TIME Fig. 33.7. Induction in wheat (after McAlister 1937). correspond to steady assimilation) with the ordinates corresponding to the beginning of illumination. The inset in figure 33.8 shows the induction loss as a function of light intensity. Comparison of this curve with the light curve of photosynthesis, P = /(/), for wheat (fig. 28.1) shows that APi is approximately proportional to P, except for the lowest light intensi- ties (where APi disappears more rapidly than P). At the higher light intensities, the induction loss at room temperature is equal to the photo- synthetic production in approximately one minute of steady illumination. In a second investigation (1939), McAlister compared the induction losses with the amount of chlorophyll in the plant, and also investigated the relation between this loss and the length of the preparatory dark period. Figure 33.9 shows the induction loss in multiples of the chlorophyll con- tent (both in moles), plotted against length of the dark period. The in- 133G INDUCTION PHENOMENA CHAP. 33 duction develops rapidly as the dark time increases to little more than a minute; when fully developed, it accounts — at room temperature — for the loss of about one molecule of carbon dioxide for two molecules of chloro- phyll. (McAhster estimated that the induction losses of oxygen in the ex- Fig. 33.8. Effect of intensity of illumination on the induction period in wheat (after McAHster 1937). Numbers are rates of CO2 exchange by respiration or assimilation (true) in mm.' per 10 miii. First arrow, beginning of experiment; second, switch from dark to light; third, return to darkness. Dotted line shows return to initial position at end of experiment. periments of van der Paauw and Smith were of the same order of magni- tude.) At low temperatures, the induction loss may be three to five times larger (c/. fig. 33.7). A second induction-enhancing dark process becomes apparent when the dark period is extended to several hours ; this process leads to a stationary state after about 7 hours in darkness, and then accounts for an induction loss of about five molecules of carbon dioxide CARBON DIOXIDE EXCHANGE DURING THE SHORT INDUCTION PERIOD 1337 per molecule of chlorophyll. The duration of the induction period remains, however, of the order of magnitude of a few minutes, even after several hours of incubation. The analogy between these observations and the findings of Steemann-Nielsen on oxygen liberation (fig. 33.6), was pointed out before. The induction loss is, according to McAlister, larger in the carbon di- oxide-limited state than in the light-limited state, even when the final rate is the same in both cases. For example, the same steady rate of photo- synthesis prevails at 0.03% CO2 and 3000 foot-candles, and at 0.3% CO2 5 10 15 20 DARK REST, min Fig. 33.9. Induction loss in wheat vs. dark rest (after McAlister 1937). Broken line indicates "saturation" of the (short-) induction- preparing process. Inset shows continued growth and final satu- ration of a second, slow induction-enhancing reaction. and 1000 foot-candles; but the induction loss is much smaller in the second case. If the plant is preilluminated in a carbon dioxide-free atmosphere, and carbon dioxide is admitted afterward, its uptake begins without de- lay. (Related observations of Steemann-Nielsen on oxygen liberation were mentioned in section 2.) Another interesting observation of McAlister — also confirmed by Stee- mann-Nielsen's observations on oxygen liberation — is that, if the light intensity is raised in steps, each increase is followed by a new induction period. The sum of all carbon dioxide induction losses is approximately equal to the loss that would be incurred in a direct passage from darkness to full light. McAlister and Myers (1940) recorded changes in the intensity of chloro- phyll fluorescence simultaneously with the changes in the rate of carbon di- oxide consumption. The results will be discussed in detail in section B; some typical induction curves are shown in figures 33.21, 33.22 and 33.26. The upper curve in each of these figures refers to fluorescence, the lower 1338 INDUCTION PHENOMENA CHAP. 33 to absorption of carbon dioxide. As far as the latter is concerned, two or three distinct types can be noted. One (type I) is obtained with wheat plants in an atmosphere of low carbon dioxide content {e. g., normal air; cf. fig. 33. 21 A), and with Chlorella cells grown in an atmosphere rich in car- bon dioxide {cf. fig. 33.22A). It is characterized by a smooth increase in 0 I 2 3 4 5 TIME, min. 2 3 4 TIME, min. 2 3 4 TIME, min After 16 min. In dark After 4 min m dork After 1 min. in dark Fig. 33.10a. CO2 absorption by Stichococciis bacillaris at 19° C, as a function of time (after Aufdemgarten 1939). 0.32% CO2, varying light intensity. carbon dioxide consumption during the whole induction period (3-4 minutes). On closer examination, one notices a slight inflection near the origin of the curves (at about 20-30 seconds). Because of the delayed response of the carbon dioxide-recording device, this position of the inflec- tion probably means that the CO2 consumption begins with a "gulp" im- mediately upon illumination (as indicated by the dotted curves in fig. 33.21). CARBON DIOXIDE EXCHANGE DURING THE SHORT INDUCTION PERIOD 1339 Curves of a somewhat different type were obtained with wheat plants at higher concentrations of carbon dioxide (0.07 to 0.24%; no higher con- centrations were used). In these curves (c/. fig. 33.21B), the induction loss is considerably smaller than in the curves of the first type (as well as in the experiments used for the construction of the inset in figure 33.8). The full rate is reached in less than 1 minute ; and a secondary depression oc- curs between 1.5 and 2 minutes simultaneously with a second "burst" of fluorescence. An extreme form of this type of induction curves (type II) is illustrated by figure 33.22C', obtained with Chlorella grown and studied in 12 3 4 TIME, min. After 16 mm. in dark 12 3 4 TIME, min After 4 min. in dark 12 3 4 TIME, min. After 1 min. in dark Fig. 33.10b. CO2 absorption by Stichococcus bacillaris at varying temperatures (after Aufdemgarten 1939). 5000 lux, 0.32% CO2. normal air. Here, the total induction loss is normal; but the initial carbon dioxide "gulp" is much more pronounced than in figures 33.21A and 33.22A, and the inflection is consequently replaced by a distinct peak 20-30 seconds after the beginning of illumination. Characteristic in these curves is the parallel (rather than antiparallel) development of carbon dioxide consumption and fluorescence. (The registration of fluorescence occurs without a time lag; therefore the peak of fluorescence in figure 33.22C coincides approximately in time with the peak of carbon dioxide consump- tion, although the latter is registered 15 seconds later.) The induction curves of McAlister are closely paralleled by the curves obtained by Aufdemgarten (1939). As described in chapter 25 (page 853), Harder and Aufdemgarten (1938) developed a method of rapid carbon di- oxide assay, based on measurement of thermal conductivity of the gas 1340 INDUCTION PHENOMENA CHAP. 33 before and after passage through the plant chamber. Aufdemgarten studied by this method the induction in Stichococcus hacillaris. His re- sults are shown in figures 33.10 and 33.11. We find in them the same three features noted above in the curves of McAlister and Myers: an initial guly of carbon dioxide (registered, because of the time lag of the apparatus, about 15 seconds after the beginning of illumination), a gradual increase Fig. 33.11a. Time course of CO2 uptake by Stichococcus hacillaris (after Aufdemgarten 1939). (a to d), inorganic nutrition: (a) 4 weeks light, 30 min. dark, Kolkwitz solution, pH 7.6; (b) 4 weeks light, 30 min. dark, Eiler solution, pH5.8, 19° C; (c) 10 days dark, Kolkwitz solution, 18.1° C; (d) 10 days dark, Ei- ler solution, 18.9° C; (e) organic nutrition in daylight; 15 min. dark, 4 min. 6 sec. light. lasting for several minutes, and (in some curves) a secondary depression, D2, which is registered between 1 and 1.5 minutes after the beginning of illumination. As in McAlister's curves, this depression can be most clearly seen on curves in which the main induction effect is weak, such as c and / in figure 33.10a, and/ in figure 33.10b. The analogy between the carbon dioxide exchange curves of the type a in figure 33.10a (or type h in figure 33.11a) and the oxygen exchange curves of the type of figure 33.2 is obvious. Quantitative comparison is impossible, because the oxygen curve is an integral curve, while the carbon CARBON DIOXIDE EXCHANGE DURING THE SHORT INDUCTION PERIOD 13-41 dioxide curves are differential rate curves. Furthermore, in figure 33.2, the oxygen hberation is partly compensated by rapid oxygen consumption by cathodic reduction. If these factors are kept in mind, and the carbon dioxide curves are corrected for sluggishness, the similarity between the course of the oxygen liberation and that of carbon dioxide consumption becomes striking. It would be interesting to compare quantitatively not only the timing of changes, but also the volumes of oxygen and carbon di- oxide involved in them. In a rough measurement, Aufdemgarten (1939) found no marked deviation of Qp from unity during the first 10 minutes of illumination (19.5° C, 5000 lux, 0.6% CO2). Later, similar conclusions were reached and generalized by Warburg and co-workers, as described above in section 2; but measurements by (a) {b) Fig. 33.11b. Time course of CO2 uptake in Stichococcus bacillaris after varying periods of darkness (after Aufdemgarten 1939). 5000 lux, 1.0% CO2, 15 min. dark, 4 min. light, 1 min. dark, 4 min. light, (a) Kolkwitz solution, 18.4° C; (6) Eiler solution, 18.7° C. Emerson and co-workers, also reported there, revealed considerable minute- to-minute variations in Qp, even in cases in which the time course of the exchange of the two gases is cjUalitati vely the same ; extreme deviations of Qp from unity occur when the oxygen burst is combined with a carbon di- oxide burst (see below). The consumption of carbon dioxide during the gulp is independent of the dark period (between 1 and 10 minutes; cf. fig. 33.10a) as well as of temperature (between 13 and 25° C, cf. fig. 33.10b). This behavior is typi- cal also of the fluorescence burst {cf. part B). The oxygen gush has not been studied at different temperatures, but we may presume that it, too, is unaffected by temperature. If this is true, these three phenomena must be due to a straight photochemical reaction. On the other hand, the sub- sequent inhibition, which dominates the short induction period, depends on the length of the dark interval as well as on temperature, and must there- fore be associated with thermal reactions, both in its preparation (deactiva- tion) and in its liquidation (reactivation of the photosynthetic mechanism). In contradiction to other observers, Aufdemgarten found that the dura- 1342 INDUCTION PHENOMENA CHAP. 33 tion of the induction period was not independent of light intensity; it in- creased from 1.5 minutes at 1350 lux to 5 minutes at 10,000 lux (at 19° C). Aufdemgarten observed that the CO2 gulp disappeared in the presence of 0.001 mole/1, cyanide: we will discuss this observation in part B (Sect. 2e), together with the effect of cyanide on fluorescence (which has been studied much more thoroughly). The same observer found the gulp to be absent from induction curves obtained with higher plants; but this cannot be a general rule, since the gulp was observed by McAlister in ex- periments with both Chlorella and wheat. Perhaps the leaves used by Aufdemgarten had a high diffusion resistance, which prevented rapid fluc- tuations of the carbon dioxide exchange in the leaf from affecting markedly the concentration of this gas in the atmosphere. The gulp was also found by van der Veen (1949,^-^ 1950) in experiments with grass blades, conifer needles and some algae; but in Chlorella (and certain other algae, or at least in certain cultures of these algae) it is re- placed (or submerged) by a carbon dioxide burst, first discovered by Emer- son and Lewis (cf. below). In a second investigation (1939^), Aufdemgarten succeeded largely in eliminating the rapid oscillations (caused by unsteadiness of the apparatus) which marred the registration curves in figure 33.10. Using the same or- ganism {Stichococcus bacillaris) he now studied the effect on induction of pretreatment of the algae. He used: (a) organic nutrition in daylight, which produced cells full of assimilates, in the form of oil droplets; (b) and (c) inorganic nutrition in acid and alkaline medium, respectively; and (d) starvation (inorganic medium in darkness). The (rather unexpected) result (fig. 33.11a) was that the composition of the inorganic medium proved to be of more decisive importance for the carbon dioxide gulp than the alternative of nutrition or starvation. Ten days spent in darkness in an organic medium had almost no influence on the shape of the induction curve (compare a and c, b and d). On the other hand, the gulp was much more pronounced in cells incubated in the acid Eiler medium (pH 5.8) than in cells incubated in the alkaline Kolkwitz medium (pH 7.6) (compare a and b, c and d). It remains to be proved that acidity, and not other differences in the composition (such as the presence or absence of ammonia), was the decisive factor. This assumption is, however, favored by observations on the importance of acidity for anaerobic induction {cf. section 6). Figure 33.11a shows that, when the dark period was reduced to one minute, the difference between the induction curves in the two inorganic media disappeared, not, as one would be inclined to say at first sight, be- cause of disappearance of the carbon dioxide gulp, but rather because the CARBON DIOXIDE EXCHANGE DURING THE SHORT INDUCTION PERIOD 1343 strong decline of the inhibition effect as a whole makes the gulp unobserv- able. The experiments described so far in this section indicated, in general, a parallehsm between the course of induction as measured by carbon dioxide uptake, and the time course of oxygen liberation, as described in the first part of section 2, although, with the oxygen and carbon dioxide experiments performed by different observers and on different objects, only qualitative analogy can be asserted. Later in section 2 we reported more recent in- vestigations, indicating the occurrence of "positive induction" (oxygen gush) in the first minute of illumination ; we recall that Warburg and co- workers assumed that the same applies also to carbon dioxide consumption ("CO2 gulp"), but that Emerson and co-workers found this to be only oc- casionally correct (e. g., in the first minute of illumination in fig. 33. 6D); in other cases the oxygen gush was not accompanied by an equivalent car- bon dioxide gulp. We will now discuss experiments on the course of carbon dioxide uptake, which revealed strong "unilateral" carbon dioxide induc- tion effects, not paralleled by similar changes in oxygen evolution. (The discussion of some two-vessel manometric induction measurements in section 2, and of others in the present section is arbitrary, since all such measurements give both AO2 and ACO2; it is justified by the greater em- phasis laid in Emerson and Lewis's work on the "carbon dioxide burst," and in Warburg's work on the "oxygen burst.") Strong deviations of the photosynthetic quotient from unity during the induction period were first noted by Kostychev (1921). He found that, in the initial 5 or 10 minutes of illumination of leaves or algae in an atmos- phere containing 6% carbon dioxide, — ACO2 strongly exceeded AO2 (Qp = 0.21 — 0.79). According to Kostychev, the excess carbon dioxide absorbed in the first minutes of illumination is compensated by a reduced uptake later, so that the average value of Qp becomes unity after 15 or 20 minutes illumination. Blinks and Skow (1938) and Emerson and Lewis (1941) observed, in contrast to Kostychev, an initial deficiency of consum'ption (or even a liberation) of carbon dioxide, not compensated by a slower evolution (or outright consumption) of oxygen ; this means Qp values either > 1 or neg- ative. Blinks and Skow (1938^ used a rapid potentiometric method for the determination of the carbon dioxide exchange. They measured ACO2 by means of a glass electrode immersed in the algal suspension, or pressed against the surface of a leaf (c/. chapter 25, page 853). The evolution of carbon dioxide revealed itself by a decrease, and its consumption by an in- crease in pH. (Of course, similar effects could also be caused by formation or consumption of other acids.) The response was so rapid that light 1344 INDUCTION PHENOMENA CHAP. 33 flashes lasting only 0.04 second were recorded as distinct peaks on the pH curves. The unexpected result of Blinks and Skow's measurements was the ob- servation of a gush of acidity in the first moment of illumination. This gush was particularly strong after a prolonged dark rest, and could then be observed several times in succession — although with a declining strength — even after dark intervals of only a few seconds duration. Figure 33.12 shows three repetitions of the gush in a pond lily leaf (in this case, the dark intervals were of the order of 1 minute). I Light I Dark t if Q) 0) <3 Q) L. O d Q^ m «0 Cli t- u - o < UJ <^ 50 Li. O LU O q: UJ a X Experimental o Theoretical 1 2 3 4 5 10 15 20 30 40 50 LIGHT INTENSITY, x 1000 lux Fig. 33.13M. Relative induction loss as a function of light intensity (after Steemann- Nielsen 1942). shows a typical set of results. Steemann-Nielsen suggested that, at the higher temperatures, induction may be too brief to be discovered by meas- urements of 5 min. duration. (Li, 1929, noted that, at room temperature the steady rate of bubble evolution from a submerged plant was reached about 1 minute after transition from stronger to weaker light; but results obtained by this method are too rough for exact interpretation.) Steemann-Nielsen hypothesized that the relative inhibition observed in the fii'st moment after the light intensity had been reduced from a INDUCTION AFTER CHANGE TO LOWER LIGHT INTENSITY 1357 "saturating" to a "limiting" value may be independent of the specific value of the final light intensity, and ecjual to the "rate deficiency" in conditioning light (meaning by this the ratio of the actual rate to the rate Figs. 33.L3N and O. Initial inhibition in percent of photosynthesis of Fucus serratus at 2.3 klux, as function of the duration of "conditioning" to 23 klux (curve N), and of "deconditioning" in darkness after completed conditioning to 23 klux (curve O), 5° C. (after Steemann-Nielsen 1942). 1358 INDUCTION PHENOMENA CHAP. 33 that would prevail at the same light intensity if light saturation did not occur). To put it more simply, Steemann-Nielsen postulated that, upon a sudden decrease of light intensity, the quantum yield at first remains un- changed, and only gradually rises to the higher value characteristic of the weaker light. If generally valid, this would be a significant relation; how- ever, it was derived from only seven not very precise, measurements (five at 300 o 0) - (/) o o I a. 100 - 41.300 lux 40 2.200 lux _j_ _i_ 20 40 MINUTES 20 40 60 Fig. 33.13P. O2 induction losses of Cladophora insignis after upward and downward changes in light intensity (after Steemann-Nielsen 1949). Winkler method. 18° C, 5 X IQ-^ mole/1. HCO3, pH 8.1. variable conditioning light, /', and constant illuminating light, I", and two at constant I' and variable /")• In figure 33. 1 3 M, the curve represents the initial inhibition as calculated, on the basis of this theory, from the hght curve of photosynthesis of Fucus; the crosses represent the actually ob- served inhibitions. Steemann-Nielsen also carried out experiments in which conditioning to the higher light intensity was incomplete. The results are shown in figure 33.13N, in which the ordinates represent the yields in the first 10 minutes of weaker illumination as a function of the duration of preillumination with stronger hght. The curve shows a maximum after about 6 minutes conditioning. Figure 33.130 shows that a similar maximum cor- responds to a certain degree of "deconditioning" in the dark; it is reached about 12 minutes after the cessation of strong illumination. In other words, it appears that at some time in the course of conditioning to strong light, and again at some time during "deconditioning" in the dark, the photosynthetic apparatus passes through a state ca- INDUCTION AFTER CHANGE TO LOWER LIGHT INTENSITY 1359 pable of giving the highest initial rate in weaker light. It is noteworthy that, even in the maximum of both curves, the initial rate is about 25% below the steady rate at 2300 lux. Steemann-Nielsen saw in this an indication that these inhibition effects are due to the competing action of two independent factors, one activating and the other inhibiting the photosynthetic apparatus, rather than to variation in the intensity of a single acti- vating factor, the optimum value of which increases with light intensity. (In the latter case, the maxima of the cui-ves in figures 33.13N and O should be 100%.) By further analysis of the curves on the basis of these concepts, Steemann-Nielsen constructed a curve showing the time development of the inhibiting factor at 23,000 lux; the curve was sigmoid and showed a saturation after about 20 minutes (at 5.6 ° C). In section 7, we will describe observations of induction losses after a period of photoxidation, caused by various factors. One of them is ex- cessive light. It may be asked whether a relation exists between these ex- periments and the observations of Steemann-Nielsen. Offhand, it seems unlikely that significant photoxidation could occur at intensities as low as 7000 lux, even at temperatures close to 0° C. (In the experiments de- scribed in section 7, light intensities of the order of 100,000 lux were used.) We will see below, however, that Steemann-Nielsen tends to consider the analogy as significant. Steemann-Nielsen (1949) resumed the investigation of induction after reduction of light intensity, using this time Cladophora insignis. In this fresh-water green alga, induction losses of this kind could be observed only on rare occasions, despite an improvement in the technique of the Winkler method which permitted one-minute oxygen determinations. A number of curves were secured, however, which showed the effect. In one experi- ment, a change from 31 to 2.6 klux was followed by 14 minutes induction (at 10° C), and in another, a change from 10.7 to 2.2 klux, by 15 minutes induction (at 18° C). The effect was more pronounced after a change from 41 to 2.2 klux than after a change from 10.7 to 2.2 klux — although photosynthesis was saturated at both initial intensities; in agreement with this result, an induction loss was noted also upon transition from a higher to a lower intensity in the saturating range. Figure 33.13P shows a set of measurements illustrating these findings. In new experiments with two species of Fucus, the induction after light reduction could not be found (in contrast to earlier results with Fucus serratus) . Steemann-Nielsen concluded that photosynthesis includes one dark re- action, which can produce light saturation, but is not usually limiting and which involves reactivation of chlorophyll (somehow changed in the pri- mary photochemical process). Whenever this reaction becomes Umiting in the light-saturated state (because of some special metabolic conditions), a sudden reduction of light intensity finds a part of chlorophyll in the inac- tive state incapable of contributing to oxygen production. The initial 1360 INDUCTION PHENOMENA CHAP. 33 quantum yield of oxygen liberation in weaker light could then be the same as it had been in the conditioning stronger hght (as suggested in fig. 33.13M). Often, however, the initial rate deficiency is smaller; Steemann- Nielsen suggested that, in this case, light saturation was only partially de- termined by the rate of chlorophyll reactivation, and partially by another, chlorophyll-independent enzymatic factor, such as "catalyst B" of Franck and Herzfeld. In chapters 26 and 27, pp. 871 and 923, it was shown that cooperation of several reactions in the determination of the saturation level is to be expected theoretically if the "ceilings" imposed by them separately are not too different; on p. 1030, we discussed the possibility that a dark reaction involving chlorophyll and requiring about 20 seconds may impose a "ceiling'* on the rate of photosynthesis that is not too differ- ent from the ceiling imposed by a reaction of "catalyst B," which requires about 0.01 second. Steemann-Nielsen's explanation is in principle the same, but the postulated chlorophyll recovery is much slower— requiring up to 30 minutes for completion. This means that the chlorophyll "de- activation," postulated by Steemann-Xielsen, can be permitted to occur only once in a large number of primary photochemical reactions — perhaps one in about two hundred. (Steemann-Nielsen estimated 1 deactivation per 40 reduced CO2 molecules.) Steemann-Nielsen (1949) discussed the above-mentioned possibility that the same inactivation of chlorophyll may account also for the decline of photosynthesis in excessively strong light. His first experiments with Cladophora insignis spoke against such a hypothesis. He observed that, when the algae were exposed to a very strong light (160 klux) until the rate had declined by about 30%, a sudden drop in intensity to 21 klux did not reduce the rate to the very low value which was to be expected if the cause of inhibition at 160 klux were the same chlorophyll inactivation to which induction losses after exposures to 20-40 klux have been attributed. Later, however, Steemann-Nielsen (1952) found that, after one hour exposure of Cladophora to 100 klux (at 18° C), during which the rate had declined by about 50%, a sudden drop of light intensity to 3 klux (implying transition to the light-limited state) did reveal a strong inhibition, the re- covery from which required several hours. He concluded from this that exposure to excessive light affects not only a rate-limiting enzyme, which is not part of the photochemical apparatus proper (as suggested by Franck and French, cf. section 7 below), but inhibits the latter apparatus as well. The effect of a ten minute exposure to 100 klux proved to be the same as that of one hour exposure to 17 klux; it could be therefore suggested that inhibitions caused by exposure to excessive hght are due, at least in part (and if the exposure is not too long, perhaps predominantly), to the same kind of inhibition (photoxidation?) of the photochemical mechanism pro- GAS EXCHANGE DURING THE LONG INDUCTION PERIOD 1361 posed above as explanation of induction losses following transition from saturating (but not excessive) to limiting light. The time required for reactivation is about the same, whether a certain degree of deactivation had been reached by brief exposure to excessively bright light, or moderate exposure to moderately strong light; it becomes, however, much longer if deactivation by bright light had been pushed so far as to prorluco almost complete suppression of photosynthesis. 5. Gas Exchange during the Long Induction Period When Osterhout and Haas (1918) first described the induction of photo- synthesis, they left the plant {Ulva lactuca) in the dark overnight and ob- served, in the morning, a gradual increase in the rate of carbon dioxide consumption, which lasted for about 1.5 hours. 6 8 10 12 14 16 18 20 TIME.hr Fig. 3.3.14. Long induction in Fontinalis after 14 hours darkness for four plant speci- mens (after Harder 1930). 32 - :^o :/ ^.^^^^ Sun plonts /\. ^^^-^^..^^ 28 ^^-=^.;^:y-^ 26 -( 24 -■^ ^\ Shade plonts y^ / 22 - V-^/^^"'^-^ 20 1 1 . 1 . ■ . 1 . 1 . 1 i ..J 8 10 14 16 TIME.hr. 20 22 Fig. 33.15. Induction curves of sun-adapted and shade-adapted Fontinalis plants (18° C.) (after Harder 1933). Adaptation to sun: several months in an aquarium with southern exposure. Adapta- tion to shade: 6 days in shaded aquarium. Comparison with the results of Warburg, van der Paauw, McAlister and Aufdemgarten shows that what Osterhout and Haas observed must have been the long, rather than the short, induction; more precisely, it was a superposition of the two phenomena. Many of the subsequent in- duction studies also were made under conditions that must have led to the superposition of the two induction effects, often with the long induction 1362 INDUCTION PHENOMENA CHAP. 33 predominating. Because of the uncertainty of interpretation of this kind of measurements, we will describe them only briefly. Briggs (1933) kept the moss Mnium undulatum in the dark for 2 hours. After this, on the average, the induction periods in hght lasted for about 30 or 50 minutes, but the values scattered widely, and showed only a vague tendency to increase with increasing ight intensity, decreasing temperature and extended dark incubation. The e.xperiments of Harder (1930) have been quoted once before in chapter 26. Here, we are interested only in the part of Harder's complex time curves that can be at- tributed to the long induction. Figure 33.14 shows, as an example, the gradual rise of 14 12 10 Dork Light O 8 I- < cr. u m 6 o >- X o < cr 0 12 3 TIME, hr. Fig. 33.16. Induction after 40 days dark rest (after Gessner 1937). the rate of oxygen evolution by four Fontinalis plants after a dark rest of 14 hours. The induction periods in this case lasted from 4 to 6 hours. Later (1933) Harder found that, if plants adapted to weak light ("shade plants") were brought into stronger Ught, their photosynthetic efficiency declined at first, until they began to "readapt" themselves to strong light (c/. lower curves in fig. 33.15). We can look upon this decline as an ex- pression of "light injury" (to which shade-adapted plants are easily susceptible; cf. chapter 19, pp. 529, 532 and chapter 28, p. 995), and dismiss them as having no direct relation to induction, unless it is suggested that a complete theory of the changes that the photosynthetic apparatus undergoes in light and darkness should include "light injury" and "light adaptation," too {cf. section 4 above). In figure 26.8, we repro- duced Harder's scheme of the time course of photosynthesis corresponding to different ratios of conditioning and illuminating light. The uppermost curve in this figure was a pure induction curve, the lowest one a pure light injury curve, while the intermediate GAS EXCHANGE DURING THE LONG INDUCTION PERIOD 1363 curves were interpreted by Harder as the result of superposition of four processes: activation ( = induction), deactivation ( = light injury), re-adaptation to strong light (renewed rise) and "fatigue" (expressed in the final decline of the rate). Other authors who have studied the course of photosynthesis over long periods obtained less complex curves. Bukatsch (1935), for example, noted with green algae {Spirogyra, Zygnema, Mougeotia and Cladophora), after an extended dark rest, only a smooth increase in rate that lasted for about 1 hour, and was followed by constant oxygen production. Gessner (1937) often found, in experiments with higher aquatic plants {Elodea, Potomageton and Ceratophyllum), no long induction periods at all; the rate was constant from the first hour of illumination. (Two such curves were reproduced in figure 26.9.) Other curves of the same observer showed, however, an increase in rate during the first 1 or 2 hours of illu- mination, especially after a prolonged period of darkness. However, the induction never exceeded 2.5 hours, even after dark periods of 40 days (c/. fig. 33.16). Finally, it was mentioned in section 2 that Steemann-Nielsen (1942) observed, in Fucus, only induction phenomena of 10-30 minutes duration, even after incubation periods of 16 hours. Apparently, a long induction period after prolonged dark incubation is a fairly common, but not general, phenomenon; whether its occurrence is closely related to the kinetic mechanism of photosynthesis, or is more or less incidental, is as yet uncertain. Perhaps, vigorous circulation of the me- dium in the experiments of Gessner and Steemann-Nielsen had something to do with the absence of a prolonged induction period in their experiments. It was mentioned in section 1 that, even if no prolonged induction oc- curs after a dark incubation of several hours, the extent of the "short" in- duction is affected in a manner showing the superposition of a slow de- activating process upon the fast deactivating reaction responsible for induc- tion after a few minutes of dark rest. This twofold origin of short induc- tion losses is clearly shown both by Steemann-Nielsen's (1942) curve of losses of oxygen liberation by Fucus (fig. 33.6), and by McAhster's (1939) curve of losses of carbon dioxide uptake by Triticum (fig. 33.9). A special kind of induction phenomena has been recently described as following the transfer of certain algae from acid into alkaline media. In chapter 37D, we will describe the experiments of OsterUnd (1951, 1952) on the photosynthesis of Scenedesmus quadricauda in carbon dioxide and in bicarbonate solutions. He noted the occurrence of a long induction period (order: 30-50 minutes) in algae grown in acid, carbon dioxide-rich solutions, and transferred into carbon dioxide-poor bicarbonate solution before exposure to light; the same algae showed none, or only the usual short, induction if exposed to Ught in the original carbon dioxide solution. OsterUnd suggested as explanation a slow photoactivation of a factor in- 1364 - INDUCTION PHENOMENA CHAP. 33 volved in the utilization of bicarbonate in photosynthesis — either in the transport of bicarbonate through the cell wall, or in its chemical transfor- mation (in the latter case, this factor may be the enzyme, carbonic anhy- drate). Similar observations were made by Briggs and Whittingham (1952) with ChloreJla. They found, however, that induction continued if the illuminated algae were transferred from bicarbonate back into carbon dioxide solution, and therefore suggested a different explanation: that cells grown in a carbon dioxide-rich medium and therefore full of metabo- lites, suffer, when exposed to light in a solution that is low in carbon dioxide, "self-poisoning" by the formation of "narcotics" which cover the surface of chlorophyll, and that this causes a prolonged induction. This would place the observed phenomenon alongside certain other forms of "long" inductions, such as that after anaerobic incubation (c/. next sec- tion). OsterHnd (1952) could not confirm with his plant material the above-mentioned findings of Whittingham and Briggs with Chlorella. The long induction Scenedesmus quadricauda shows after transfer into bi- carbonate solution may therefore have a different origin from that observed under the same conditions in ChlorcUa. This is not implausible, because (as we will see in chapter 37D), Scenedesmus quadricauda appears to be much better adapted to utiUze bicarbonate for photosynthesis than Chlo- rella pyrenoidosa. 6. Influence of Anaerobiosis on Induction In chapter 13, we discussed the question whether small ciuantities of oxygen are necessary for photosynthesis, and answered it in the negative. (The question has been revived more recently by Warburg; but new experi- ments seem to confirm the above conclusion, cf. chapter 37D, section 2b). Nevertheless, there is no doubt that green plants deprived of oxygen during a prolonged period of darkness, and then exposed to light, often show considerable initial inhibition. Boussingault (18G5), Pringsheim (1887) and Willstatter and Stoll (1918) found that, after several hours of anaerobic incubation in the dark, the photosynthesis of higher land plants was prac- tically completely inhibited, and required a considerable time for recovery. Green algae were found to suffer, too, but usually to a lesser degree. From experiments of Franck, Pringsheim and Lad (1945) it appears that the maximum capacity of algae for photosynthesis is reduced by an- aerobic incubation, not to zero, but to a low, finite level, which may be two. ten or several hundred times lower than the saturation rate in the aerobic state. The level of inhibition depends on the nature of the plant and the duration and specific conditions of the anaerobic treatment. The speed of recovery in light depends — in continued absence of external oxygen supply INFLUENCE OF OXYGEN ON INDUCTION 1365 — on how long it takes for the oxygen produced by residual photcjsynthesis to "burn up" the accumulated (and continuously produced) inhibitors arising from the anaerol)ic metabolism. Whether this is at all possible is determined by the extent of competition of other oxygen-consuming side reactions, as well as by the exact rate of the residual oxygen production. In hydrogen-adapted Scenedesmvs, competition comes, for example, from the "oxyhydrogen" reaction (2 H2 -j- O2 —^ 2 H2O). If all the oxygen lib- erated by photosynthesis is swept away by a stream of oxygen-free gas (as was done in the experiments of Franck, Pringsheim and Lad), "anaero- bic inhibition" can be prolonged practically indefinitely, even in strong light, and reversible light curves of residual, anaerobic photosynthesis can be obtained. Some algae {e. g., Scenedesmus) acquire, simultaneouslj^ with anaerobic inhibition of oxygen production, the capacity for photochemical reactions that involve absorption or liberation of hydrogen (cf. chapter 6). It was the study of these phenomena that led Gaffron (1935, 1937 1-^, 1939'-, 1940) to the conclusion that the cause of induction generally lies in the de- activation of a catal3^st or catalysts concerned primarily with the liberation of molecular oxygen (Franck's "catalyst C"). In Scenedesmus, this de- activation is coupled with the activation (probably, by reduction) of a "hydrogenase," capable of transferring molecular hydrogen; in most other green plants, no hydrogenase is produced, and the only effect of anaerobic incubation is therefore an enhanced initial inhibition of ordinary photosyn- thesis. In purple bacteria, which have no oxygen-liberating enzymatic system to begin with, but which do contain an active hydrogenase, anaero- biosis does not affect the capacity for photochemical metabolism at all; often it even represents the only condition under which this metabolism is possible (cf. chapter 5). The question arises as to whether the persistent inhibition of photosyn- thesis observed under strictly anaerobic conditions is caused simply by a more complete deactivation (or even destruction) of the oxygen-liberating enzjrme (the autocatalytic regeneration of which by photosynthesis is thus delayed), or by another, independent phenomenon. It seems that both fac- tors play a role, so that "anaerobic inhibition" is the result of two processes — one affecting the activity of the oxygen-liberating enzymatic system, and the other, obstructing in a less specific way all catalytic agents, including chlorophyll and the "finishing catalyst" Eb (Franck's "catalyst B"). Perhaps the two inhibiting agents are identical, but widely different amounts are needed for the specific and the nonspecific "narcotic" action. (Perhaps the same hypothesis can be applied to "short" and "long" induc- tion under aerobic conditions.) Among experiments that demonstrate the production by anaerobic 1366 INDUCTION PHENOMENA CHAP. 33 metabolism of a general diffusible, acidic and reducing inhibitor, are those of Noack, Pirson and Michels (1939) and Michels (1940). They observed that neutralization of the medium as well as aeration can remove the anaero- bic inhibition, sometimes almost instantaneously. This is illustrated by 6 E o < o Bicarbonate buffer, Nj + Culture medium, pH 4, Nz +C02, 25° C ■*••'-*■— H— t-- -»*-+-*—* 0 120 240 360 480 TIME AFTER BEGINNING OF ILLUMINATION, min c o o CO < o in a. 8- Culture medium, pH 4, 25° C. o Quinone + Air l^:a>.v 120 240 360 TIME AFTER BEGINNING OF ILLUMINATION, min. Fig. 33.17. (a) Induction in Chlorella after 13.5 hours of anaerobiosis in acid or alkaline medium, (b) Quinone added after 60 minutes light, or air bubbled through solution 60-105 minutes after beginning of illumination (after Noack, Pirson and Mich- els 1939). figure 33.17, which shows the practically complete inhibition of photosyn- thesis of Chlorella by 13.5 hours of anaerobiosis in an acid nutrient solu- tion, and the absence of a similar effect in an alkaline buffer. Figure 33.17b shows that the inhibition can be removed by aeration in the dark (in the course of which a considerable quantity of oxygen is taken up by the cells), and, even more rapidly, by the addition of quinone. The chemical nature of the easily oxidizable, acid fermentation products indicated by Noack's experiments is as yet imknown. Assays were made for lactic acid but it was found in only relatively small quantity (in agreement with Gaffron's observations on Scenedesmus; cf. chapter 6, page 138). INFLUENCE OF OXYGEN ON INDUCTION 1367 Franck, Pringsheim and Lad (1945) found much less pronounced effects of alkali and quinone on anaerobic inhibition than Noack, Pirson and Michels. Franck attributed this difference to the use of streaming gas which kept the local oxygen concentration safely below the "reactivation" level even in alkaline medium. Noack's experiments, on the other hand, were made in a closed system, where even partial removal or neutralization of the inhibiting material may have been decisive in determining the balance between oxygen accumulation by residual photosynthesis and oxygen con- sumption by side reactions not contributing to the reactivation of photo- synthesis. According to Franck and co-workers, an important factor determining the extent of inhibition of a cell culture by anaerobic incubation is, beside duration of the incubation period (the difference between the results of Franck and of Noack may have been due, in part, to the latter's using a thirteen hour incubation period and the former, a three to five hour incuba- tion period), the concentration of the algae in the suspension. This indi- cates that inhibition is caused by a product capable of diffusing into and out of the cells. Since young cultures, placed in suspension media that previ- ously contained old, anaerobically incubated algae, often show only Uttle deterioration, one of the effects of anaerobic incubation (to which old cul- tures appear to be more sensitive) may be a change in the permeabiUty of the cells for general, "narcotic" poisons. It is obvious from the above that the induction curves observed after anaerobic incubation must depend on the balance of residual photosynthesis and various competing, oxygen-consuming processes and may therefore dis- play a great variety of shapes. This applies to the course of photosynthesis in the first minutes (or hours) after the beginning of illumination; the phenomena in the first few seconds of exposure, on the other hand (best revealed by the observations of fluorescence), are simpler under anaero- bic than under aerobic conditions. In general, anaerobic pretreatment makes the fluorescence start at a high level immediately upon illumination, thus eliminating all or most of the usual "fluorescence burst" (cf. fig. 33.47). Photosynthesis, on the other hand, begins under anaerobic con- ditions, at a low level (i. e., the initial oxygen burst and carbon dioxide gulp probably are absent). It looks as if, in this case, the inhibiting action is fully developed in the dark, and requires no initiating photochemical re- action. Emerson and Lewis (1941) found that the photochemical carbon dioxide gush, too, is absent in Chlorella after anaerobic incubation. As mentioned in sect. 2, the polarographic oxygen liberation curves of Blinks and Skow (fig. 33.2), even if they were obtained after exhaustion of oxygen in the layer between leaf and electrode, cannot be considered characteristic of anaerobic conditions. McAlister and Myers (1940) ob- 1368 INDUCTION PHENOMENA CHAP. 33 tained a number of carbon dioxide consumption curves in pure commercial nitrogen (0.5% oxygen), but here, too, no attempt was made to realize truly anaerobic conditions (which have to be maintained for some time prior to the beginning of illumination in order to obtain the typical anaero- biotic inhibition effects). Consequently, McAlister and Myers' ''low oxy- gen" curves reflected merely the known favorable effect on photosynthesis of a reduction in the partial pressure of oxygen (attributed in chapters 13 and 19 to the avoidance of photoxidation). The only effect of low oxygen concentration noticeable in these curves was a slight increase in rate, spread over most of the induction period. An investigation of the initial O o ZD o o a: Scenedesmus, 5 x 10"^ g. / 2.5 ml. 2 3 TIME, min. Fig. 33.18. Absence of induction in Scenedesmus after 3 hours anaero- biosis in Nj (after Shiau and Franck 1947). piCOi) = 20 mm. Fig. 33.18.^. Induction in Chlorella after 3 hours anaerobiosis in N2 (after Shiau and Franck 1947). course of photosynthesis under strictly anaerobic conditions, in a stream of oxygen-free nitrogen or hydrogen, was made by Franck, Pringsheim and Lad (1945), using the phosphorescence-quenching method. It was men- tioned before that reversible light saturation curves can be obtained under these conditions, the saturation level being from 50 to 0.1% of that under aerobic conditions, depending on the completeness of the incubation treat- ment. In some experiments, no induction at all was observed under anaerobic conditions, as far as the inertia of the apparatus permitted one to judge. This inertia was of the order of 20 seconds. Figure 33.18 gives an example of this behavior in Scenedesmus after 3 hours anaerobiosis. Fluorescence experiments (part B, section 2f) make it likely that 6ne/ induction changes actually occur in the first few seconds. After longer anaerobic pretreatment, curves with a clearly delayed as- cent were obtained; their shape indicated a "second wave" of inhibition about 1 minute after the beginning of illumination. Oxygen liberation INFLUENCE OF OXYGEN ON INDUCTION 1369 reached a constant level in about 2 minutes. This type of induction curve is correlated with the occurrence of sigmoid hght curves of steady photo- synthesis; they occur more often in hydrogen than in nitrogen, and are favored by carbon dioxide deficiency. With Chlorella, an even more pronounced second inhibition wave was observed (fig. 33.18A) ; it led to a minimum of oxygen production between 0.5 and 1 minute after the beginning of illumination. The "secondary induction loss" increases with the length of the dark period, although it was noticeable after only 1 minute in darkness (subsequent to a previous 3 hour incubation). It is enhanced by low temperature — at 0° C. a deep depression at 0.5 minute extended over as much as 4 minutes before the O O Q O CC Q- O Secondary wave Inverse induction ^Sfeody level 0 12 3 4 ■ TIME, min. Fig. 33.18B. Inverse induction in Scenedesmus in absence of added CO2 after 3 hours in pure streaming N2 (after Shiau and Franck 1947). 0° C, 12.5 X 10 -^ g. algae in 2.5 cc. oxygen production finally picked up. Poisoning by cyanide or extreme carbon dioxide deficiency acts in a similar way. These experiments first showed that the "second wave" of induction can affect also oxygen liberation. (It was previously observed only in carbon dioxide consumption, and in fluorescence.) Franck, French and Puck (1941) previously had suggested that the second wave may be caused by a "blockade" of the chlorophyll surface by accumulated intermediate oxidation products ("photoperoxides") not removed rapidly enough by the inhibited "deoxygenase." Such a blockade could increase fluorescence and prevent the uptake of carbon dioxide, but could not be expected to cause a decline in the rate of oxygen production. An alternative explana- tion of the second depression was therefore suggested by Shiau and Franck (1947) (part B, section 2). When carbon dioxide was absent or, more exactly, when only the small amount of carbon dioxide produced by fermentation was present, "inverse 1370 INDUCTION PHENOMENA CHAP. 33 induction" was observed by Shiau and Franck, following the secondary induction wave (fig. 33.18B). A similar and sometimes even stronger per- manent decline of oxygen production was produced by adding cyanide (10-^ M HCN, in the presence of 20 mm. CO2). The decline from the peak of oxygen production to the final steady level was more pronounced the longer the preceding dark period; after several hours in the dark, the rate in the first minutes of illumination was three to eight times higher than in the subsequent steady state. There is a parallelism between these results and the fluorescence induc- tion curves obtained in the presence of cyanide (fig. 33.39). In both cases, the initial part of the induction curves is unaffected ; a depression of photo- synthesis and enhancement of fluorescence appear only after the "second wave." In interpreting the phenomenon of "inverse induction" Franck and co-workers noted that it occurs under conditions of inhibited carbon dioxide supply (low carbon dioxide, presence of cyanide). The carbon dioxide supply deficiency is not so important at the beginning of illumination, after the carboxylation of the acceptor has had time to be completed in the dark. The insufficient rate of replacement of the carbon dioxide used up by photo- synthesis begins, however, to be felt after one or several minutes of illumina- tion. Some more recent observations of the effect of oxygen deficiency on in- duction can be mentioned here. Van der Veen (1949") found that the induction curves of Holcus lanatus did not change substantially by substituting (instead of air) nitrogen con- taining only 0.3% oxygen (with unchanged content of carbon dioxide), ex- cept that the preparatory dark period was lengthened. (The induction curve was, for example, the same after 30 minutes dark incubation in 0.3% O2 as after 2 minutes incubation in air.) This seems to indicate that the "de-adaptation" in the dark (which causes induction) is an autoxidation. When oxygen was removed altogether from the nitrogen atmosphere dur- ing the incubation period, the rate of steady photosynthesis of Holcus leaves remained low for hours afterwards. The effect was noticeable even after 1 minute anaerobiosis in darkness; after 30 minutes the steady rate was only a few per cent of normal. It could be completely restored by aerobic incubation in darkness. The capacity for carbon dioxide uptake and release by heat-inactivated Holcus leaves was also destroyed by anaerobic incubation. In Chlorella, no gas gulps or bursts of any kind appeared in an atmosphere of pure oxygen or hydrogen without carbon dioxide. A small amount of oxygen was produced in light in the hydrogen atmosphere. Higher plants, placed in pure hydrogen, produced an oxygeri gush, followed by a steady, slow oxygen production. Van der Veen suggested that this gush originated INDUCTION AFTER PHOTOXIDATION 1371 from the reduction of carbon dioxide, accumulated in the plant (by some fermentation reaction) during the dark period; the subsequent steady liberation of oxygen could be based on continued slow carbon dioxide production by the same fermentation process. Hill and Whittingham (1953) studied the induction phase in Chlorella and Scenedesmus by their spectroscopic method (c/. section 2) also under anaerobic conditions. They noted (in agreement with the earlier observa- tions by Franck and co-workers) that the induction period was more strongly affected by dark anaerobiosis in ChloreUa than in Scenedesmus. After anaerobic dark periods up to 80 minutes (at 15° C.) the half-time of reactivation was increased from 0.5 to 2-3 minutes, and after 17 hours, to about 7 minutes. In this case, as contrasted to that of aerobic incubation, the induction period is longer the lower the temperature; also, the reacti- vation is faster the stronger the illumination (while the duration of aerobic induction is independent of light intensity over a wide range, and even decreases in weak light). The anaerobic inhibition could be completely removed by addition of a small amount of oxygen in the dark, provided the anaerobic incubation had not lasted too long. Because they noted an effect on induction even after an anaerobic incubation of only a few minutes, Hill and Whittingham suggested a return to the Willstatter-Kautsky- Warburg hypothesis of direct oxygen participation in photosynthesis, in preference to the concept of Gaffron, Franck and others, according to which anaerobic inhibition is due to the accumulation of fermentation products and inactivation of enzymes (see in this connection, chapter 13, p. 327, this chapter, p. 1365, and chapter 37D, section 3). Olson and Brackett (1952) noted that anaerobic dark incubation of Chlorella leads to an initial total inhibition of oxygen evolution, which may last up to 2 minutes (after an incubation of 1-3 hours); after a shorter, or much longer, incubation period (-CI hour or 18-24 hours), inhibition does not occur. Once the period of total inhibition is over, the "light adaptation" curve follows the same time course as after aerobic incuba- tion. Franck, Pringsheim and Lad made use of the high sensitivity of the phosphorescence method to study oxygen liberation by single light flashes, and found evidence of both primary and secondary induction losses in the flash yield as well. The first flash produced less oxygen than the sec- ond one, and so forth. In Scenedesmus, the yield per flash soon became steady, while in Chlorella, a second depression was noticeable. 7. Induction after Photoxidation Since induction probably is a consequence of oxidation-reductions and autoxidations in the dark, we can expect characteristic induction effects to occur also after a period of photoxidation in light (brought about in one of the several ways discussed in chapter 19). Observations on these aftereffects of photoxidation were made by 1372 INDUCTION PHENOMENA CHAP. 33 Franck and French (1941). The}^ found that, if a carbon dioxide-starved leaf, after a period of photoxidation, was exposed to light in the presence of E E LJ O z < I o UJ cr Z) en CO UJ a: a. 15 10 ( 0 ' 7 = 80 X = 5.8 ^y^. Corrected 1 ^^ Dark __X--I.2 - 5 ^^^ - 0 ^ - -5 1 . 1 1 1 1 20 30 40 50 60 1 D (b) ^ 10 - ^ — - 5 - 1=80 ^X^orrected 7.3 Dork ""^ 0 ^« - -5 1 I 1 1 I.. _l 1 . 1 10 20 40 50 60 30 TIME, min. Fig. 33.18C. Photosynthesis in the presence of sufficient CO2, (a) before and (6) after photoxidation (after Franck and French 1941). Straight lines on right show increase in respiration after photoxidation. Corrected for this increased respiration, the final rate of photosynthesis, 7.3, is about equal to that before photoxidation, but is reached only 20 minutes after admission of COa in light. E E o Z) o o Q. >- X o 100 60 20 0 -20 -40 ■ 1000 f 28,000 f I 28,OOOJ 1000 f.-c I f-c,^ 28,000 f-c 'darkness 300 0 50 100 150 200 250 TIME OF ILLUMINATION, min. Fig. 33.18D. Progressive injury of cells by 28,000 foot-candles of light, and recovery in darkness (D) and in 1000 foot-candles {B and C) (after Myers and Burr 1940). carbon dioxide, 10 or 20 minutes were needed to recover its full photosyn- thetic efficiency (fig. 33.18C). The exact length of the recovery period INDUCTION AFTER PHOTOXIDATION 1373 depended on the duration and intensity of photoxidation. If the latter had gone too far, no complete recovery was possible. The photosynthetic capacity could be restored more or less completely also by a dark rest. In chapter 19, we described a second method for replacing photosynthe- sis by photoxidation: very intense illumination. A period of photoxida- tion induced in this way also is followed by an induction period upon return to photos3mthesis in more moderate light. Figure 33.18D, taken from the work by Mj^ers and Burr (1940), shows that the longer the cell suspension has been exposed to extremely strong light (28,000 foot-candles, or 300 klux), the slower and less complete is its recovery in moderate light (1000 0 10 20 30 LENGTH OF EXPOSURE TO 23,000 f.-c, min. Fig. 33.18E. Residual photosynthetic activity after varying ex- posures to 23,000 foot-candles (after Myers and Burr 1940). Rate 20 equals 100% for this batch of cells under 7500 foot-candles. foot-candles). Curve D shows that in this case, too, recovery occurs in darkness as well as in light. After a long exposure (3-5 hours) to intense light (> 10,000 foot-candles) no recovery is possible at all. Figure 33.18E shows the residual photosynthetic efficiency (at 7500 foot-candles) as a function of the duration of exposure to 23,000 foot-candles. Franck and French suggested that the aftereffects of a period of photoxidation have a double origin — the oxidation of some constituents of the enzymatic apparatus of photosynthesis, and the burning up of the reserves of intermediate products. The oc- currence of the first effect is clearly demonstrated by the fact — mentioned in chapter 19 — that photoxidation inhibits photosynthesis and not merely counterbalances it. The natural explanation of such a "catalytic" effect is the destruction (oxidation) of one or several members of the enzymatic system involved in photosynthesis. (Franck and French beheve that the most probable substrate of oxidation is Ea, the catalyst responsi- ble for the formation of the carbon dioxide-acceptor complex.) 1374 INDUCTION PHENOMENA CHAP. 33 However, Franck and French pointed out that the inactivation of "catalyst A" does not offer sufficient explanation for aftereffects lasting for 20 minutes and more, because from fluorescence experiments of Franck, French and Puck (1941) they con- cluded that Ea can be regenerated in less than 1 second. They therefore suggested that, in addition to the injury to Ea, photoxidation affects the capacity for photosynthesis also by burning the stocks of intermediate products of photosynthesis and respiration. In section 4 above, we described the experiments of Steemann-Nielsen which made him beheve that the induction which follows transition from saturating to hmiting Hght intensity is caused by partial inactivation (per- haps photoxidation) of a part of the photochemical apparatus — either of chlorophyll itself, or of an enzymatic factor associated with chlorophyll so closely that its inhibition makes the corresponding part of chlorophyll in the cell incapable of contributing to photosynthesis. Steemann-Nielsen also concluded that inhibition by excessively strong light (> 100 klux) may have, at least partially, the same origin. Longer exposures to excessive light may cause both deactivation of the photochemical apparatus (as sug- gested by Steemann-Nielsen), and the photoxidation of an enzyme (or enzymes) kinetically independent of chlorophyll, as suggested by Franck and French (as well as photochemical burning-up of intermediates of photo- synthesis and respiration). 8. Induction in the Photoreduction by Algae The algae (Scenedesmiis, Raphidium) that will liberate hydrogen after anaerobic incubation, if illuminated in an atmosphere of nitrogen, will absorb hydrogen if illuminated in an atmosphere of hydrogen (slowly in pure hydrogen, and much more rapidly in a mixture of hydrogen and carbon dioxide, in which the absorbed hydrogen can be utilized for the reduction of carbon dioxide). In chapter 6, this "photoreduction" of carbon dioxide was discussed as a significant variation of normal photosynthesis. In strong light, hydrogen liberation or photoreduction rapidly gives place to normal photosynthesis. If only the net pressure change is measured, very complex "induction curves" can be obtained (cf. Vol. I, fig. 14); the "positive induction" that characterizes many of these curves is caused by an initial liberation of hydrogen. Hydrogen absorption by anaerobically adapted algae also has an induc- tion period. But, contrary to ordinary photosynthesis, this induction period's duration increases with decreasing light intensity (cf. Vol. I, fig. 14). Franck and Gaffron (1941) suggested that "hydrogen induction" is due to the fact that carbon dioxide is first reduced at the cost of accumulated organic hydrogen donors (fermentation products), before the reduction at the cost of molecular hydrogen can get under way. If the disposal of ac- FLUORESCENCE INDUCTION PHENOMENA 1375 cumulated reducing intermediates is achieved by a simple photochemical reaction, it should require less time in strong than in weak light. This hypothesis links the induction period of hydrogen assimilation with the "hydrogen burst" produced by the same algae in carbon dioxide- free nitrogen atmosphere. In both cases, the first effect of light is oxida- tion (dehydrogenation) of the accumulated fermentation products, the hydrogen being either utilized for the reduction of carbon dioxide, or re- leased into the atmosphere. The assumption of "photosynthesis at the cost of fermentation prod- ucts" was also used by Franck, Pringsheim and Lad (1945) in the explana- tion of the cyanide-insensitive burst of oxygen production in the first minutes of illumination of anaerobically incubated algae. These authors have also contributed essentially to the understanding of the time curves of gas exchange in Gaffron's hydrogen-adapted algae, by showing that the return to normal photosynthesis can be postponed indefinitely by prevent- ing the oxygen evolved by residual photosynthesis from accumulating in the closed system. 9. Induction in Hill Reaction Clendenning and Ehrmantraut (1950) noted that no induction occurs in the liberation of oxygen by ChloreUa cells with quinone as oxidant; the same cells showed the usual induction period of about 5 minutes in bicar- bonate solution. No induction losses were found in the oxygen liberation by isolated chloroplasts with quinone or ferrocyanide as oxidants. The theoretical significance of these results will be discussed later. No induction was observed in the Hill reaction of chloroplasts by Hill and Whittingham (1953) with the hemoglobin method of spectroscopic oxygen determination. B. Fluorescence and Absorption Changes during the Induction Period* 1. Fluorescence Induction Phenomena in Leaves, Algae, Chloroplasts and Chlorophyll Solutions It was said before that fluorescence is one of the few if not the only property of chlorophyll that can easily be measured simultaneously with photosynthesis. All determinations of gas exchange, even by electrical or optical methods, are sluggish, since the gas has to be moved from the in- terior of the plant cell to the locus of measurement (or vice versa). Fluores- cence, on the other hand, is measured practically instantaneously; this * Bibliography, page 1431. 1376 INDUCTION PHENOMENA CHAP. 33 permits one to follow the very rapid changes in the photosynthetic appara- tus, such as must occur during the induction period. The recording of fluorescence has first led to the realization of the great complexity of the induction phenomena, particularly to the discovery of the developments that take place in the first one or two seconds of illumination. That the fluorescence of chlorophyll changes in a characteristic way during the induction period of photosynthesis was first noted by Kautsky and Hirsch (1931).* Kautsk}^ has since devoted a series of investigations to this subject (1931-1943). The reliability of his earlier observations (1931-1937) was limited by inadequate technique (which involved excita- tion by ultraviolet light, and visual estimation of fluorescence intensity). More recently, he has gone over to excitation by visible light and automatic registration of fluorescence, and more reliable data have been collected in this way, particularly by Kautsky and U. Franck (1943). Systematic determinations of the intensity of the chlorophyll fluorescence during the induction period have also been made by Wassink and co-workers (1939, 1942, 1944), McAlister and Myers (1940), Franck, French and Puck (1941) and Shiau and Franck (1947). Reviews of the subject have been written by Kautsky and U. Franck (1948), J. Franck (1949) and Wassink (1951). Nothing is known so far of changes in the fluorescence spectrum of chlorophyll, which are not impossible, if the chemical structure of chloro- phyll itself, or of its associates in the "photosensitive complex," under- goes changes in the transition from darkness to light and back. Correlation of measurements of fluorescence intensity with gas ex- change measurements would be much safer if experiments of both kinds were carried out with the same plant objects. Otherwise, there is a danger that, in attempting to present a comprehensive picture of induction phe- nomena, one may try to fit together pieces from several different jigsaw puzzles. A step in the right direction was taken by McAlister and Myers, who measured simultaneously the carbon dioxide absorption and the fluo- rescence intensity of wheat plants and ChloreUa suspensions. The analysis of the gas exchange leads to the conclusion that the most conspicuous induction phenomena observed under aerobic conditions can be formally explained by three processes: a preparatory dark reaction, which produces a "precursor," or "potential inhibitor" (or removes or in- activates a catalyst that, when active, prevents the accumulation of an inhibitor); a photochemical reaction, which "activates" the inhibitor; and a second, nonphotochemical process by which the inhibitor is removed. Special explanations are required for the occurrence of a second inhibition period, and for the carbon dioxide "burst" observed by Blinks and Skow * It is worth recalling here that French and Young found no induction of the fluo- rescence of phycoerythrin in vivo, indicating that the composition and environment of its molecules remain unchanged by illumination. FLUORESCENCE INDUCTION PHENOMENA 1377 and Emerson and Lewis. Anaerobic incubation accentuates tiie ordinary inhibition and, in addition, produces a second inhibiting factor that requires no photochemical activation. (The above summary was based on the assumption that the normal course of induction consists of a brief burst of uninhibited — but not en- hanced— gas exchange, followed by one main wave, and sometimes one or more secondary, waves, of inhibition; and that the only additional, uni- lateral effect superimposed on this standard sequence is the Emerson-Lewis carbon dioxide burst. More recent studies have shown that the true pic- ture is more complex. In particular, the burst of oxygen liberation and a^ 1 % COj \ I- 21 X 10* erg/cm^ sec \ A ^-^C Q >- t- z lij / (b) Fluorescence z r 1 1 1 1 1 UJ o 1 2 3 4 5 6 z LjJ TIME, sec o C [ UJ a: o _1 u. o Q , a' , / /^ C02 Uptake 1 = 8 8 X IO*erg/om^ sec. y \ ^ fo) )^ , i J _l L J .- L , \ 1 I B' 1 2 3 4 C ) 1 2 TIME, min. Fig. 33.19a. Course of fluorescence and photosynthesis in wheat during induction period (after McAhster and Myers 1940). Broken Unes show CO2 absorption curve corrected for slow response of the apparatus. TIME, mm Fig. 33.19b. Fluorescence vs. time curves at 23° C. of Hydrangea; excitation by 400- 600 m/i (after Franck, French and Puck 1941): (a) first half minute; (b) first 6 sec. carbon dioxide consumption in the first few seconds (or first minute) of illumination may appear larger than could be produced by uninhibited, steady photosynthesis during this brief period, and therefore seems to in- volve (similarly to the carbon dioxide burst) a sudden photochemical trans- formation (deoxygenation, carboxylation, decarboxylation) of some ma- terial or materials accumulated in the cell during the dark period. The following discussion of the induction effects in fluorescence was written with the simpler picture of a few years ago in mind. Li particular, it con- tains many references to "inhibition of photosynthesis" during certain in- duction phases, while we would now prefer to speak more specifically of inhibition of oxygen hberation, or inhibition of carbon dioxide uptake, or of both together, if this be the case, rather than of inhibition of photosynthesis as a whole.) 1378 INDUCTION PHENOMENA CHAP. 33 A sequence of three above-suggested processes is suggested also by the common shape of the fluorescence versus time curves found under aerobic conditions, which is shown in figures 33.19a and b. The fluorescence starts at the normal level, A, rises rapidly to a maximum, B (reached about 1 sec. after the beginning of illumination), and declines again to the initial level, CD, in about 1 min. These fluorescence curves are mirror images of the typical gas exchange induction curves (fig. 33.19a). Often, however, fluorescence measurements have led to induction curves of a somewhat dif- ferent type, shown in figure 33.20. Here, the "first wave" of fluorescence, B, is low and decays very rapidly, and the induction picture is dominated by a second wave, D, the maximum of which occurs from 3^^ to several minutes TIME, min. Fig. 33.20. Fluorescence curves of Chlorella with dominant second wave (after Wassink and Katz 1939). Buffer No. 9, 29° C, / = 1 X 10^ erg/cm.» sec. after the beginning of the light period. This wave appears to be related to the second depression of the gas exchange, although the latter has been usually encountered above only as a relatively minor disturbance, and not a dominant feature of induction curves (c/. figs. 33.10, 11, 13A,B,F,G). Apparently, the transformation of the photosynthetic mechanism that gives rise to the second wave of fluorescence does not always affect the gas exchange equally strongly, and may sometimes have no influence on this ex- change at all. Thus, while in fig. 33.21(B) the second wave of fluorescence coincides with a distinct depression in the curve of carbon dioxide absorp- tion, no such correlation is apparent in figure 33.22(C): the second wave of fluorescence (which appears here about 0.5 minute after the beginning of illumination, and does not decay until sometime beyond 4 minutes), does not interrupt the steady increase of carbon dioxide consumption. As a result, the fluorescence curve and the carbon dioxide absorption curve as- sume a parallel, instead of the usual antiparallel, course. Fluorescence measurements have shown that each of the two waves of fluorescence may possess a "fine structure," with features that reappear with remarkable persistence in curves obtained by different observers with different species. One such feature is a depression on the ascending part of the first wave, noticeable in figure 33.19b, and seen in more detail in FLUORESCENCE INDUCTION PHENOMENA 1379 figure 33.23 (where it is marked Z>i). Kautsky suggested that this de- pression becomes the main feature of the induction curves under anaerobic conditions, or in the presence of certain poisons (cf. figs. 33.33A and 33.39). (A) [Fluorescence C02 Consumption COj Consumption Fig. 33.21. Induction phenomena in wheat (after McAlister and Myers 1940). (A) Normal air (0.03% CO2), strong Ught (5 X 10^ erg/cm.2 sec.) after 40 min. light and 12 min. dark, 24° C. (B) High CO2 (0.36%) in air, strong light after 10 min. light and 10 min. dark, 24° C. Broken lines indicate approximate correction for time lag of gas exchange curve; time marks 1 min. apart. I (A) (B) (C) Fluorescence .J I i__L. COz Consumption Fluorescence __ L i__L_. COz Consumption Fig. 33.22. Induction phenoma in Chlorella (after McAlister and Myers 1940). (A) Grown in 4% CO2, studied in 0.24% CO2; 6 min. light, 10 min. dark. (B) Grown in air, studied in 0.33% CO2; 2 min. dark. (C) Grown and studied in normal air (0.03% CO2); 3 min. light, 40 min. dark. Broken curve indicates approximate correction for time lag of gas exchange curve; time marks 1 min. apart. A second, equally persistent feature is a "bulge" on the declining slope of the first peak (cf. lower fig. 33. 19b) . (Kautsky assumed the existence of two 1380 INDUCTION PHENOMENA CHAP. 33 depressions, the first of them marked A in figures 33.27 and 33.33B.) Sometimes this bulge develops into a maximum and can then l)e con- fused with the second main wave of fluorescence. Even while we do not believe, with Kautsky (1943), that each ripple on the induction curve indicates a new photochemical reaction in the main sequence of photosynthesis, the persistent occurrence of these details shows that they are not accidental, but indicative of a complex interplay of activation and deactivation processes. - Leaf squeezed 10 y<^. --.^ ^2 tu // Leaf wormed o // 2 II tu ■ j j o 5 ~Di j e so high that the rate of its photochemical activation ceases to depend on this amount {i. e., probably, until it occurs with a quantum yield of unity). UJ o z LJ O CO bJ o 13 1100- 900- 700- 500 20 40 60 80 TIME, sec. Fig. 33.25. Height of fluorescence peak (B in fig. 33.19) in relation to dark rest (after Franck and Wood 193G). (A) (B) (C) Fig. 33.2G. Induction in wheat in normal air and strong light (after McAlister and Myers 1940): (A) 40 min. light, 12 min. dark; (B) 2 min. dark; (C) 1-2 min. CO2 uptake. dark. Upper curves, fluorescence; lower curves. These predictions can be compared with the experimental results of Franck and Wood (1936), who found the height of B to rise Unearly with duration of the dark rest, until the latter reached about 60 seconds (c/. fig. 33.25). After this, the height of B continued to increase for a long time (as shown by experiments with plants that have rested overnight), but so much more slowly as to suggest an entirely different fluorescence-pro- moting i-eaction. This result bears obvious similarity to the effects of the 138G INDUCTION PHENOMENA CHAP. 33 duration of incubation upon the induction losses of oxygen and carbon di- oxide, illustrated by figures 33.6 and 33.9. Wassink and Katz (1939) found, similarly, with Chlorella, that the height of B can be maintained, in successive light flashes of 10 seconds dura- tion, if these flashes are separated by dark intervals of at least 2 minutes. The required intervals were much shorter in cells grown in glucose solution and having strong respiration than in cells grown in an organic medium and having w^eak respiration, thus confirming the surmise that the inhibitor is removed by an oxidation process. These results of Franck and Wood, and of Wassink and Katz, are in satisfactory agreement with determinations of Warburg, McAlister and others, which gave a value of about 1 minute for the dark interval required to repeat the induction curves of oxygen liberation and carbon dioxide absorption (c/., e. g., fig. 33.9). The observations of Aufdemgarten (figs. 33.10b and 33.11b) indicated a somewhat longer regeneration period of the "precursor" (>4 minutes), while McAlister and Myers noted (c/. fig. 33.26) that, in wheat in normal air, no burst of fluorescence at all oc- curred after 1 minute darkness, and only a faint burst was noticeable after 2 minutes. (b) Light Intensity Like induction losses of O2 and CO2, induction of fluorescence is not ob- served in weak light (fig. 33.27). (There are as yet no observations on lij o 2 liJ O (/5 LU tr o ID 10- a ft 1 b. •2 ■3 -4 .5 6 ^ gC "~-— ->..^ ~^ —■ ^**' ^ ■- / _!. . / - 1 1 1 1 1 . 10 20 30 TIME, sec. 40 50 Fig. 33.27. Fluorescence time curves of Ulva lactuca for different light intensities (after Kautsky and Franck 1943): first 50 sec. of illumination; 20° C; ordinary air. LIGHT INTENSITY, erg/cm.' sec. in Fig. 33.28. Fluorescence outburst relation to light intensity in Hydrangea (after Franck, French and Puck 1941). whether the CO2 gush, which occurs even in weak fight, is accompanied by variations of fluorescence.) With increasing light intensity, ascending FLUORESCENCE-TIME CURVES 1387 slope of fluorescence wave AB becomes steeper, as anticipated. In Hydran- gea leaves, according to Franck, French and Puck (1941), this increase continued until the light intensity has reached 20 X 10* erg/cm.^ sec, or approximately 40 klux. The height of point B, on the other hand, ceases growing much earlier, e. g., in Hydrangea at about 1.5 X 10" erg/cm.^ sec. (see fig. 33.28; cf. the "light saturation" of the carbon dioxide induction loss, illustrated by fig. 33.8). The maximum fluorescence yield, V 1- C/) n z z in LlI H 1- Z z liJ in O o Ii = 16 X 10* erg/cm^ sec. z UJ v o o \ to CD \ liJ llJ ' \ (T ! I = 0 8 X 10* erg/cm.^ sec. (T 1 \ O O 3 1 r) 1 ^ — _i ' ■" _l 1 a. 1 1 1 1 Ll ) — 1 1 1 0 12 3 4 0 0.5 1.0 1. TIME, mm TIME, sec Fig. 33.30. Fluorescence changes in Hy- drangea leaves after a sudden decrease of light intensity (after Franck, French and Puck 1941). At normal temperature, fluorescence first drops below the steadj^ level, then recovers slowly. Fig. 33.31. Fluorescence changes in Hydrangea leaves after sudden decrease of light intensity (after Franck, French and Puck 1941). At 0°, fluorescence requires about 1 sec. to decline to its steady level, showing the survival of the highly fluores- cent material present in intense light. which a ciualitatively similar picture was found for oxygen liberation; however, the "induction after transition to weaker light" lasted in Stee- mann-Nielsen's experiments for several minutes (at 4° C), as against only a second in figure 33.31. We have dealt so far only with the effect of light intensity on the first wave of fluorescence. Figure 33.27 shows that the second maximum also is shifted with increasing intensity, in a way which indicates that it, too, is due to a photochemical reaction that promotes fluorescence, and to a coun- teracting thermal reaction. However, the relations seem to be more com- plex than in the first maximum. For example, Wassink and Katz (1939) noted that, in the presence of cyanide (which prevents the final decay, DE in fig. 33.19), the ascending slope, CD, is affected not only by light inten- sity, but also by temperature {cf. figs. 33.39 and 33.40). Figure 33.41 indi- 1390 INDUCTION PHENOMENA CHAP. 33 cates that an "optimum" light intensity exists for the enhancement of the second maximum, D. (c) Temperature As mentioned before, Kautsky and Hirsch (1931) noticed that the "upward slope," AB, is independent of temperature, whereas the "down- ward slope," BC, is steeper the higher the temperature. Kautsky and Spohn (1934) measured this effect visually and found the period AB to be constant between 0 and 40° C, and the period BC to decrease from 800 seconds at 5° to 20 seconds at 35° (in Pelargonium zonale). \JJ o l ^^"^^..^^ 2 \ ~~-....^ 0° c. UJ \ ^^^ o \ ^^"■"'■■^-......^^^^^^ en \ UJ (T \ O ^v _J ^v.._ 23° C U- 1 1 1 1 1 0 I 2 3456789 10 TIME, min. Fig. 33.32. Effect of temperature on rate of decay of fluorescence outburst in Hydrangea (after Franck, French and Puck 1941). / = 44 X 10* erg/cm.^ sec. According to Franck and Wood (1936), the fluorescence decay, BC, is exponential at 25, 20 and 12° C, and can be attributed to a monomolecular reaction with a temperature coefficient of approximately 2. The curves ob- tained at low temperatures {e. g., 3°) show an accentuation of the second wave. According to Franck, French and Puck (1941), the decay in Hy- drangea lasts for 3 minutes at 23°, and for more than 10 minutes at 0° (fig. 33.32). The curves obtained by Kautsky and Marx (1937) with different leaves confirmed the enhancement of the secondary wave by decreasing tempera- ture. Leaves of different species showed analogous shapes at different temperatures, e. g., the curve of Piper amplum at 25° was similar to that of Ageratum mexicanum at 35°. Figure 33.33 shows the influence of temperature on fluorescence curves of Ulva lactuca. They confirm that decrease in temperature in- creases the heights of both the first and the second fluorescence peaks, and shifts them further from the beginning of illumination. At 0°, the second peak is shifted beyond the 3 minute registration limit. FLUORESCENCE-TIME CURVES 1391 The effect of temperature on the declining slope of D was studied also by Wassink and Katz (1939) with HCN-poisoned Chlorella. In this case, in contrast to fig. 33.33, maximum height of peak D was reached at the highest temperature (35°; cf. fig. 33.41). This can be explained by the fact that not only the decay, DE, but also the rise, CD, is ac- celerated by an increase in temperature. In nonpoisoned cells the influence of tempera- ture on the purely thermal decaj' process is more important than its influence on the combined (photochemical and thermal) activation process; maximum D therefore de- Fig. 33.33. Fluorescence time curves of Ulva lactuca at different temperatures, in presence of urethan, and in absence of oxygen (after Kautsky and Franck 1943): (A) first 3 sec, 10 lux; (B) first 3 min., 2.5 lux. dines with increasing temperature. In cyanide-poisoned cells, on the other hand, the decay, DE, is inhibited, and the enhancing effect of temperature on the rise, CD, is the only remaining effect. (d) Carbon Dioxide Concentration We have to distinguish between two phenomena: the effect of excess carbon dioxide (in concentrations which cause a "narcotic" poisoning of photosynthesis; cf. chapter 13, page 330), and the effect of low carbon dioxide concentration, in the range where this concentration limits the ef- ficiency of the photosynthetic apparatus. The influence of excess carbon dioxide was first noticed by Kautsky and Hirsch (1935). The activation curve, AB, was unaffected, but the decay BC, lasted for 2 minutes at 4% CO2, and 3 minutes at 10% CO2, instead of 1 minute at < 1% CO2. The inhibiting effect of carbon dioxide concentra- tions > 10% is confirmed by figure 33.34 of Franck, French and Puck. A sudden increase in carbon dioxide concentration, e. g., from 1 to 20%, dur- ing the decay (or after its termination) caused an immediate burst, followed by a renewed slow decay, of fluorescence. Figure 33.35 shows an almost complete disappearance of the first fluorescence wave of Ulva in concentrated carbon dioxide. The difference from the results of Franck, French and Puck (who found the height of peak B to be almost unaffected, even by 80% CO2) may be due to the use of a 1392 INDUCTION PHENOMENA CHAP. 33 different species, or of shorter dark intervals. (All curves in figure 33.34 were obtained after 1 hour dark rest, the decay of fluorescence in 80% CO2 being very slow. Earlier repetition might have given a result much more like that shown in figure 33.35.) o UJ o UJ o C02 -d- 80% Q- 20% 0 5% »• None A- 5% TIME, min. Fig. 33.34. Effect of excess carbon dioxide on decay of fluorescence outburst in Hydrangea (after Fianck, French and Puck 1941). 4 min. exposure after 1 hour dark. / = 1.7 X 10* erg/cm.^ sec; Hg lines 436, 492, 546 niju. ( O ) Control after the series, with no CO2. TIME, sec. Fig. 33.35. Effect of excess carbon dioxide on fluore.scence time curve of Ulva laduca at 20° C. (after Kautsky and Franck 1943). / = 40 m. c. (equivalent). The influence of carbon dioxide concentration in the range where it represents a strong "limiting factor" in photosynthesis (i. e., below ap- proximately 0.1%) is not shown clearly by figures 33.34 and 33.35. Ac- FLUORESCENCE-TIME CURVES 1393 cording to chapter 28 (page 1051), carbon dioxide limitation causes a shift toward lower light intensities of the transition from the "low" to the "high" steady fluorescence yield { ^2). In complete absence of carbon dioxide, one would expect the fluorescence yield to be equal to <^2 — 1-7 ^i even at the lowest light intensities. However, no difference between the steady yields at 0 and 0.5% CO2 appears in figure 33.34. Figure 33.35 also shows no difference in shape between the curves correspondmg to 0.04 and 5.4% CO2; but, here, the absolute fluorescence values cannot be com- pared, because of the arljitrary adjustment of the scale. (A) L J I I- Fig. 33.36. Induction phenomena in wheat in the absence of CO. (after McAUster and Myers 1940). (A) O2, 15 niin. hght, 10 min. dark. (B) N,. Figure 33.36A (McAlister and Myers 1940) agrees with figure 33.34 (Franck and co-workers) in that it, too, shows the first fluorescence burst to grow and decay normally even in the absence of carbon dioxide. In a nitrogen atmosphere, on the other hand, the decay of fluorescence after the burst is much slower in the absence than in the presence of carbon dioxide (c/. fig. 33.36B). This agrees with the hypothesis that removal of the in- hibitor formed (or activated) in the first few seconds of illumination occurs by a reaction with oxygen (which, in nitrogen, must first be produced by photosynthesis) . The occurrence of the complete first fluorescence wave in the absence ot carbon dioxide, demonstrated by figures 33.34 and 33.36A, recalls the ob- servation of McAlister that preliminary illumination in carbon dioxide- free atmosphere can eliminate the induction loss of carbon dioxide upon subsequent admission of this gas. Despite the "autocatalytic" character of the process by which the induction is liquidated, no complete photosyn- thesis appears to be required for this purpose. 1394 INDUCTION PHENOMENA CHAP. 33 The absence of induction losses upon transition from a carbon dioxide- free atmosphere to an atmosphere containing carbon cHoxide may be, however, not a general rule. At least, fluorescence has been observed to undergo a very characteristic change following an alteration of this concen- tration, by McAlister and Myers in wheat, and by Franck, French, and Puck in Hydrangea. As shown by figure 33.37A, the change consists of a dip of fluorescence, followed by a burst, a second dip and finally an ap- proacli to a new steady value (which may be lower than the value at the original, low concentration of carbon dioxide, if the latter was rate-limiting). CO2 from 003-40% COi from 04-40% o o LlI cr o 3 (B) TIME, min. Fig. 33.37. Effect of changes of CO2 concentration on fiuorescence. (A) Wheat; high Hght intensity (after McAHster and Myers 1940). (B) Hydrangea (after Franck, French and Puck 1941). According to figure 33.37A the effect disappears when the initial carbon dioxide concentration itself is saturating (e. g., 0.4%; cf. third curve). Franck and co-workers observed, however, a wave of fluorescence upon a transition from 1 to 20% CO2, and found that only the dip was absent if the initial concentration was saturating. A sudden decrease of carbon dioxide concentration (e. g., from 25 to 0.03%) also caused a dip, followed by a wave of fluorescence (c/. fig. 33.37B). Changes of carbon dioxide concentration often have a very pronounced effect on the second fluorescence wave. As mentioned in section A3, McAlister and Myers (1940) found in wheat one- wave curves in ordinary air, and two-wave curves, with a second maximum around 1.5 minutes in 0.1 to 0.3% CO2 {cf. fig. 33.21B). FLUORESCENCE-TIME CURVES 1395 The association of two-maximum curves with increased carbon dioxide concentration was confirmed by Franck, French and Puck (1941), as shown by figure 33.38, obtained with Hydrangea leaves in 1% CO2. In this case, the minimum occurs 1.5 minutes, and the secondary maximum 3 minutes after the beginning of iUumination. It seems that different species require different carbon dioxide concentrations to develop the second fluorescence wave — some show it at 0.1% CO2, others only at 1% or more. It has been mentioned above that, in ChloreUa, a very pronounced second maximum was observed by McAhster and Myers even in ordinary air (c/. fig. 33.22B and C). Kautsky and Franck (1943) found that Ulva lactuca shows no second maximum in carbon dioxide-free medium, and a pronounced second maximum in 0.1% CO2. 01 2 3 4 5 6 7 TIME, min. Fig. 33.38. Fluorescence curve of Hydrangea in 1% COz (after Franck, French and Puck 1941). A second maximum occurs at about 3 min. / = 0.82 X 10* erg/cm." sec. The occurrence of the second fluorescence wave at the higher carbon dioxide concentrations, and the fact that it is often associated with a mini- mum of carbon dioxide absorption, suggests association with the carbon dioxide gush. However, as mentioned before, there is no direct evidence that the gush is accompanied by changes in fluorescence. (c) Poisons Kautsky and Hirsch (1935) noticed that the decay of fluorescence is retarded by the presence of cyanide. Franck and Wood (1936) confirmed this and found that the decay period, BC, is lengthened despite a decrease in the height of peak B. According to Franck, French and Puck (1941), this effect becomes apparent only at very high concentrations of the poison (c. g., 2% HCN in air). A detailed study of the cyanide effect on two-maximum curves of ChloreUa was made by Wassink and Katz (1939). Figure 33.39 shows the 1396 INDUCTION PHENOMENA CHAP. 33 influence of variations of cyanide concentration. With sufficient cyanide present, the decay disappears entirely and the fluorescence becomes sta- bihzed at the level it had reached in the second maximum. (The influence on the first fluorescence wave is scarcely noticeable in this figure.) Wassink and Katz used the cyanide-poisoned cells for the study of the effect of temperature, oxygen and other factors on the initial part of the fluorescence curve, as- suming that eUmination of the final decay must make the analysis easier. Insofar as these measurements concerned the first wave of fluoi-escence, ABC, they are discussed in sections (c) and (f), because in this part of the induction curve the effects of cyan- ide are minor. Here, we will discuss some results concerning the "second wave." 18 UJ o z UJ o (/) LJ (T O ID 14- 12 10 0.33 - D -yi^I no5 no33 " — no2 ■ 0£l_^^____^ C 1 1 1 1 TIME, min. Fig. 33.39. Fluorescence-time relation in air at 29° C. as a function of inhibi- tion of photosynthesis by cyanide (after Wassink and Katz 1939). Per cent KCN shown on curves; 0.1 ml. added to 2 ml. cell suspension. Figure 33.39 shows that the decline of photosynthesis after the second maximum is much more sensitive to cyanide than the decline after the first peak. Even 5 X 10"'* per cent cyanide in solution (corresponding to 7.7 X 10~* mole/1.) had a strong effect, while 1.65 X 10"^ per cent (2.6 X 10~' mole /I.) eliminated the decay altogether. This compares with 2% HCN in air, or about 0.15 mole/1, in the equilibrated aqueous phase, which Franck and co-workers found to be the smallest quantity affecting the first wave of fluorescence. The two observations refer to different species, and it was shown in chapter 12 {cf. Table 12. V) that the sensitivity of plants to cyanide varies widely; how- ever, in this case, the variation is so extreme that it can be taken as indicative of differences between the reactions that bring about the fluorescence decay after the first and the second maxima. Figure 33.40 and 33.41 show the influence of light intensity and temperature on the ascending slope of the second wave, CD, in cyanide-inhiV)ited Chlorella cells. The sec- ond wave behaves as if it were caused by a combination of a photochemica reaction with at least two thermal reactions. At low intensities, the slope CD is proportional to / and independent of T; at higher intensities, a "saturation" is reached — the lower T, the earlier. At still higher intensities, the rate decreases again (apparently due to the de- velopment of a second minimum, near 1 minute). The retarding influence of phenylurethan on fluorescence decay was fir.st observed by Kautsky and Hirscli (1935). Figure 33.42 shows this FLUORESCENCE-TIME CURVES 1397 effect in Ulva lactuca, according to Kautsky and U. Franck (19-i3). At concentrations of the order of 10"^ mole/1., the first fluorescence wave is entirely eliminated, the fluorescence starts high, and a slight decline at the 18- y 16 z UJ o tr o 3 12 35° 29° 19° / / r = ioo TIME, min. Fig. 33.40. Fluorescence-time relations as a function of temperature at three different Ught intensities (after Wassink and Katz 1939). Gas phase, air; total inhibition of photosynthesis by cyanide. Order of curves in first peak, from top down: 2°, 19°, 29°, and 35° C. beginning of illumination replaces the normal increase. This decline lasts longer at lower light intensities (fig. 33.42B). The drop in ^ from the initial level to the steady level is greatest at the lowest light intensity and highest temperature. 1398 INDUCTION PHENOMENA CHAP. 33 20 40 60 80 INCIDENT INTENSITY 100 Fig. 33.41. Slope of second fluorescence wave as function of light intensity at various temperatures (after Wassink and Katz 1939). Gas phase, air; total inhibition of photosynthesis by cyanide. o z LlI O to UJ o UJ o (/) UJ o Z) TIME, sec. Fig. 33.42. Phenylurethan effect on fluorescence of Ulva lactuca (after Kautsky and Franck 1943). (A) Different phenylurethan concentrations; / = 40 m. c. (equiv.); 20° C. (B) Different light intensities (in equivalent meter candles); air; first 3 sees. (C) Different temperatures. FLUORESCENCE-TIME CURVES 1399 (/) Oxygen; Effects of Anaerohiosis Since Kautsky believed, at first, that fluorescence quenching by oxygen, converting the latter to a metastable form, is the first step in photosynthe- sis several of his papers were devoted to the study of the influence of oxy- gen on fluorescence during the induction period. However, the only defi- nite result, obtained in the very first investigation (Kautsky, Hu'sch and Davidshofer 1932), was that changes in the concentration of oxygen be- tween 0 5 and 100% have no marked effect on the fluorescence-time curves. Fluorescence rr Fig 33.43. Induction behavior of wheat under low (broken Une) and under normal 02 pressure (after McAUster and Myers 1940): 0.03% CO2, high light, after 30 min. dark rest. In chapter 24, we noted that an increase of oxygen concentration from 0.5 to 20% resulted in a certain decrease of steady fluorescence, probably related to the inhibiting effect of oxygen on photosynthesis. However, McAlister and Myers (1940) found that, during the induction period, not only fluorescence, but photosynthesis as well, were somewhat higher m nitrogen than in air (fig. 33.43). The explanation is uncertam, but we may recaU that inhibition of photosynthesis by excess oxygen requires time (c/. chapter 19, fig. 60), and therefore may be absent m the first min- utes of mumination. In other words, during the induction period, the quenching of chlorophyll fluorescence by oxygen {of. chapt. 23, section Ab) may be the main influence, while in the steady state the predominant eflect is that due to the photoxidative inhibition of photosynthesis. In any case, the effect is small. , This result destroyed the original theory of Kautsky, and caused tne 1400 INDUCTION PHENOMENA CHAP. 33 latter to shift the search for the effect of oxygen on fluorescence to very low- concentrations (in a hope of substituting a weakly dissociable complex, XO2, for free ox^^gen as energy carrier). Marked changes in the fluores- cence curves were, in fact, found in leaves almost completely deprived of oxygen; but whether these curves were caused, as alleged, by the absence of oxygen during the induction period (and not by anaerobic incubation) can be argued, since there is no way of depriving the cells of oxygen except by sweeping them with an oxygen-free gas for some length of time. 51.8 51.8 50 40 30 20 10 0 ■P^* — 1 -^^^ 1 »- 1 2I%02 1 1111 20 30 40 51.8 40 51.8 10 20 30 40 TIME, sec 51.8 10 20 30 TIME, sec. Fig. 33.44. Effect of O2 on fluorescence in Ageratum mexicana leaves (after Kautsky and Hormuth 1937). Kautsky (1939) said the "low Oa-curves" (fig. 33.44) cannot be at- tributed to anaerobic incubation for three reasons : first, because the sweep- ing out with nitrogen lasted only 60-90 minutes; second, because certain types of fluorescence ciu'ves were obtained at definite oxygen concentra- tions independently of the length of the incubation period; and third, be- cause the normal shape of the induction curves was restored within 1 minute upon the admission of oxygen. None of these arguments is con- vincing: The duration of anaerobic incubation needed to produce "after effects" varies widely from species to species (as mentioned in section AG) ; the degree — perhaps even the character — of the anaerobic metabolism may depend on oxygen concentration; and 1 minute may be sufficient time to burn up the fermentation products obstructing chlorophyll. Because of these considerations, the significance of the extensive collec- FLUORESCENCE-^IME CURVES 1401 tion of induction curves at different low oxygen concentrations, contained in the papers of Kautsky and co-workers, is uncertain. We will neverthe- less give a short summary of their results. Kautsky, Hirsch and Davidshofer (1932) and Kautsky and Hirsch (1935) first found that, in complete absence of oxygen, the fluorescence wave, ABC (fig. 33.19), disappears, and the fluorescence-time curve be- comes horizontal. Kautsky and Flesch (1936) observed that the slope AB first begins to flatten out in nitrogen containing less than 0.5% O2. To achieve complete suppression of the "first wave" the system had to be swept by pure nitrogen for much more than 1 or 2 hours. (a) (b) Norrnol gQO 0 I 2 TIME, sec. TIME, sec. Fig. 33.45. Fluorescence curves of Ulva lactuca at low O2 pressure (after Kautsky and Franck 1943): (a) variation of O2 at 20° C; (b) variations of temperature at 0.3% O2. / = 10 m. c. (equiv. ). Later Kautsky and co-workers (1936, 1937) found no substantial varia- tions in the fluorescence curves of Ageratum above 0.2% O2, but noted a slight change in 0.04% O2 and a strong change in pure nitrogen (estimated oxygen content, 0.0005%). After 2 hours of sweeping out with this gas, the fluorescence curve acquired the shape shown in figure 33.44. The parallelism between these results and the observations on the effects of anaerobic incubation on gas production (chapter 13, part A) is ob- vious; it seems natural to correlate horizontal fluorescence-time curves with anaerobic inhibition of photosynthesis. In transition from "aerobic" to the "anaerobic" fluorescence curves, the second fluorescence wave became prominent at a certain intermediary stage (0.2% O2, in fig. 33.44). An enhanced second wave was found also on certain intermediate temperature curves obtained at a constant, low value of oxygen. 1402 INDUCTION PHENOMENA CHAP. 33 "Anaerobic" fluorescence curves of a more complex shape were obtained by Kautsky and Eberlein (1939) and Kautsky and U. Franck (1943) with the green alga Ulva lactuca (cf. fig. 33.45). The incubation time was 100 minutes. A new feature, recognizable in the detailed figure 33.45a, is an inflection on the ascending branch, AB, which may be the first indication of transformation into the "anaerobic" fluorescence curve. (The latter has a maximum at zero time. ) No effect of oxygen was observed in Ulva lactuca between 10 and 80% O2; but, al- ready at 1% (at 20° C), the fluorescence wave was noticeably enhanced because of the delay in decay BC, and the initial fluorescence level was much higher than in air. TIME, min. Fig. 33.46. Fluorescence-time relations in air and N2 at 29° C, with and without cyanide (after Wassink and Katz 1939). Experiments on the effect of oxygen on the fluorescence curves of Chlorella were carried out by Wassink and Katz (1939) ; their results (fig. 33.46) were quite different from those of Kautsky. In oxygen-free nitro- gen, they found the fluorescence wave to be higher than in air, with decay BC taking more time, but decay DE accelerated. After 1 hour, the differ- ence between the two curves disappeared (probably in consequence of the oxygen production by photosynthesis). A similar family of four curves (air with and without cyanide, and nitrogen with and without cj'^anide) was given by Shiau and Franck, except that, basically, their curves were of type I, while Wassink's curves in figure 33.46 are closer to type II. (For example, in fig. 33.46 the second wave was marked even in the absence of cyanide, while in the curves of Shiau and Franck this wave only ap- peared when cyanide was present.) Most of the experiments of Wassink and Katz were carried out under complete inhibition of photosynthesis by cyanide, and consequent absence of the final fluorescence decay; the general shape of the curve was as shown in figure 33.39, and the authors studied the effect of oxygen concentration on maximum B and the second ascent, CD. Both were found to decUne sharply with increase in [O2] up to about 2%, and become FLUORESCENCE-TIME CURVES 1403 constant afterward. These results, even more than those of Kautsky and Eberlem indicate a real effect of the momentarij oxygen concentration on the fluorescence curves of Chlorella (whereas such effect seems to be almost nonexistent in the leaves of the higher plants studied by Kautsky, McAlister and Myers). The observations of McAlister and Myers (1940) on the more pro- nounced effect of carbon dioxide deficiency in nitrogen (compared with air) were mentioned, and their explanation was given, on page 1393. 0 10 in No Chlorella _i 1_ J I 1 L- in N, Scenedesmus TIME, sec. I 2 TIME, min. 3 (B) Fig 33 47. Anaerobic fluorescence induction curves of Chlorella and Scenedesrrms at 25° C." (after Shiau and Franck 1947): 3.0 X 10^ erg/cm.^ sec. (A) First 3 sec; (B) first 3 mins. Intensity units differ for the two algae. Shiau and Franck (1947) made extensive comparisons of fluorescence- time curves of Scenedesmus and Chlorella in air and in very pure nitrogen. The results obtained with the two species were similar. Figure 33.47 shows the development of fluorescence during the first 3 minutes. In general, the picture is similar to that in Wassink and Katz's figure (Fig. 33.46, curves for nitrogen and air). About 15 minutes of darkness were needed to repeat the anaerobic induction curve starting at the same high value. These observers, too, found that, if oxygen from photosynthesis is permit- led to accumulate, the difference between the two curves vanishes m 1404 INDUCTION PHENOMENA CHAP. 33 about 1 hour. The enhancing effect of anaerobiosis on the "second wave" is shown by figure 33.48. The curves obtained by Shiau and Franck in nitrogen and in air, in the presence of cyanide, also were mentioned on page 1402. Effects similar to those of cyanide could be brought about, as usual, by low temperature or by carbon dioxide deprivation. > 10 in N2 — in N2 2 3 INDUCTION TIME, min. Fig. 33.48. Anaerobic fluorescence induction curves of Chlorella at 24° C, showing the second wave (after Shiau and Franck 1947). Time in darkness shown on curves. Light intensities are: (a) to (d), 3.0 X lO^erg/cm.^sec; (e)2.2 X lO^erg/cm.^sec. Cells affected by long anaerobic incubation (as well as cells from old cultures) were found to be permeable to methylene blue, while young healthy cells were not stained by this dye. This points to cell permeabihty as one variable possibly responsible for variations in the shape of the induc- tion curves. The initial intensity of fluorescence of isolated chloroplasis was the same in nitrogen or air; there appears to be no inhibition of their photochemical activity by aerobic or anaerobic dark metabolism. The mechanism of the photochemical inhibition (revealed by the rise of ^ at the beginning of illu- mination) may be the same as in whole cells, but the practical absence of the subsequent decay in light, and its extreme slowness in the dark, indicate the inefficiency of the respiratory mechanism removing the inhibitor (which in this case apparently cannot overcome the continued production of the inhibitor in hght) . (g) Oxidants and Neuiralizers Related to the efl"ect of oxygen is that of "substitute oxidants," such as quinone or ferric iron compounds. If anaerobic inhibition is due, to a FLUORESCENCE-TIME CURVES 1405 large extent, to the accumulation of an acidic, reducing, fermentation product, this inhibitor could be removed by oxidation with oxygen or other oxidants, or by neutraUzation. Kautsky and Zedhtz (1941) noted that the fiuorescence-time curves of grana precipitates could be changed from the anaerobic type to the aerobic type either by aeration or by the addition of ferric oxalate or qui- none. (Whether these two oxidants are especially useful, because of their capacity to serve as oxidants in the "Hill reaction" of isolated or broken chloroplasts, remains to be seen.) Shiau and Franck (1947) investigated the effect of quinone or alkali on the fluorescence-time curves of Chlorella and Scenedesmus. In young healthy cultures they found no effect, in either air or nitrogen, except after 12 hours of anaeroboisis. After such long anaerobic incubation, the addi- tion of 10 ~' M quinone caused the induction to disappear and the steady state value of (p to drop considerably. Quinone still had an effect on

HX + A + CO2 The assumption that the primary products, not stabilized by catalyst Eb, react back was made before in chapter 28 (sect. B7) in the interpreta- tion of light saturation, and of the relation (or rather lack of systematic relation) between the saturation of photosynthesis and the yield of fluores- cence. The additional assumption made here is that the energy released in the back reaction splits A • CO2 into A and CO2. Complete inhibition of the catalytic mechanism somewhere beyond the carboxylation stage can thus explain why light causes the decomposition of the complex A • CO2 ; but it is more difficult to see why the recarboxyla- tion is so slow that the gush begins with a net quantum yield of almost unity, the photostationary state appears to be entirely on the side of de- carboxylation, even in very weak light, and the carbon dioxide uptake after cessation of illumination continues for as long as an hour. We recall that, according to the concept of saturation as a consequence of back reactions, which we considered most plausible in chapter 28, a large propor- tion (of the order of 1/8) of the light quanta not utilized for photosynthesis because of limitation by a finishing catalyst (in saturating light, this may mean >50% of all absorbed quanta) should bring about decarboxylation 1418 INDUCTION PHENOMENA CHAP. 33 of A -002. The latter should therefore proceed, in strong steady light, 10 or 100 times more rapidly than in the weak light used by Emerson and Lewis. Nevertheless, we do not observe a decarboxylation of the acceptor (unless the light is excessively strong and pick-up begins to become notice- able). This indicates that the recarboxylation is rapid enough to replace all the A -002 complexes destroyed — both those utilized for photosynthesis, and those merely decomposed into A and CO2. The pick-up experiments in fact show the recarboxylation time of the acceptor to be of the order of 10-20 seconds. In the case of the gush, the recarboxylation seems to be at least 100 times slower. One may recall in this connection that, according to figure 8.21, the rate of carboxylation in the dark was found to be unexpectedly slow also in experiments with radioactive carbon dioxide. In this case, however, explanations could be sought in the occurrence of carboxylations not re lated to photosynthesis; furthermore, the observed rates may be at least in part those of the replacement of carbon dioxide in a carboxyl group by C*02, rather than of the uptake of C*02 by a free acceptor. Similar explanations are not possible in the case of the gush. Its occurrence in light indicates close connection to the photosynthetic apparatus; since the gush is revealed by manometric measurements, it is not an exchange phenomenon. To sum up: The mechanism of the reversible carbon dioxide gush described by Emerson and Lewis is not clear, and its attribution to a shift of the carboxylation equilibrium of acceptor A, although plausible, remains uncertain. Since this section was first written, the question of the hypothetical carbon dioxide acceptor in photosynthesis. A, has been brought closer to direct experimental elucidation by continued studies of the uptake and distribution of radiocarbon in photosynthesis (to be described in chapter 36). These studies indicate that one — and perhaps even the only — carboxylation which is directly related to photosynthesis leads to phospho- glyceric acid (PGA). The carboxylation substrate is, however, not yet identified ; it seems likely that it is not a C2 compound (which could give a C3 acid by simple addition), but a longer-chain compound (a pentose) that breaks into two parts (e. g., C3 and C3) upon taking up a CO2 molecule. The reaction that leads to PGA appears to be a carboxylation coupled with a TPN-specific oxidation-reduction. Finally, the compound A, which gives PGA by coupled carboxylation and chain fission, is itself an intermediate product of photosynthesis (sinceit rapidly incorporates radiocarbon) . If this conclusion — which we have tentatively mentioned before — proves to be correct for all (or a large part) of the CO2 acceptor in the cell, this would mean the likelihood of a deficiency of this acceptor at the beginning of a INACTIVATION OF CATALYTIC SYSTEM AS CAUSE OF INDUCTION 1419 light period, and of an autocatalytic adjustment of its concentration in light to the level needed to maintain photosynthesis at the rate corre- sponding to the prevailing light intensity. The role of the carbon dioxide acceptor in the induction phenomena thus appears in a new hght, in par- ticular in the interpretation of those features of induction (such as the ap- proximate independence of its duration of light intensity) that point to a factor activated in light only to the level sufficient to maintain the rate of the over-all reaction sequence at the level permitted by other limiting reac- tions. 4. Inactivation of the Catalytic System as the Primary Cause of Induction. Gaffron-Franck Theory of Induction The assumption that the inactivation of one definite catalytic agent in the dark is the prime cause of the short induction period of photosynthesis was first made by Gaffron (1937, 1939, 1940), and later developed into a detailed theory by Franck and co-workers (1941, 1945, 1947, 1949). This theory represents the most comprehensive attempt to date to explain the induction phenomena, and we will give an account of it here, although in many details it is speculative, and new observations — particularly that of the autocatalytic formation of the carbon dioxide acceptor, and the consequences this may have for induction — may call for its re-examination. It must be kept in mind that the theory attempts to explain only the two waves of inhibition which characterize the short induction period and not the "bursts" and "gulps" of oxygen and carbon dioxide. (Franck has endeavored to interpret the latter in terms of invasion of the photosynthetic mechanism by respiration intermediates, as described in section 5 below.) Gaffron suggested that the cause of irreversible induction losses is the inactivation of the oxygen-liberating catalyst (or catalysts). We recall that kinetic researches, beginning with those of Blackman, Warburg, and Willstatter and Stoll, led Franck and Herzfeld to contemplate three major catalytic factors in photosynthesis — a carboxylating catalyst (Ea), a "fin- ishing" (or "stabilizing") catalyst (Eb) and one (or more likely, two; c/. chapter 6, page 133) oxygen liberating catalyst (Eq, and Eq in schemes 9.III, etc. That the carboxylase, Ea, is not rate-limiting during the initial period of induction is shown, for example, by the insensitivity of the gas exchange and fluorescence during this period to cyanide (cyanide is a specific inhibi- tor for Ea). Catalyst Eb, too, is not limiting at the beginning of induction. This has been demonstrated particularly clearly in anaerobic incubation experi- ments, where the yield per flash (determined by the available amount of 1420 INDUCTION PHENOMENA CHAP. 33 Eb, according to chapter 34) was found by Franck, Pringsheim and Lad to be unaffected by inhibition. Inactivation of Eb should affect immediately both carbon dioxide uptake and oxygen Hberation, but according to chapter 28 not the yield of fluorescence. The attribution of induction to the oxygen-liberating catalyst, the "de- oxygenase" (catalyst C in Franck's terminology; Eq rather than Eq in chapter 6 and chapter 9), is the remaining alternative; and it is sup- ported by several pieces of evidence. In the first place, anaerobic incuba- tion experiments with algae of the type of Scenedesmus have demonstrated directly that, after a dark anaerobic period (and, a posteriori, probably also after a dark period in air), the photochemical and catalytic mecha- nisms of photosynthesis are still intact, with the exception of the oxygen- liberating catalyst (since these algae are able to reduce carbon dioxide in light if hydrogen is supplied as substitute reductant). Further arguments for the oxygen-liberating catalyst as the primary cause of induction can be derived from experiments with hydroxylamine (Vol. I, page 311). The latter appears to be a specific poison for the oxygen-liberating enzyme system; its uniform effect on the rate in weak and strong light leads to the surmise that this system is formed (or acti- vated) by photosynthesis itself, and continuously deactivated by a dark reac- tion, so that its stationary concentration adjusts itself to the prevailing rate of photosynthesis. A catalyst of this kind obviously must be com- pletely deactivated in the dark, and is thus likely to cause induction effects. The "autocatalytic" formation of the "deoxygenase" explains the repeti- tion of induction losses upon each successive increase of the steady rate of photosynthesis (whether it is brought about by a change of light intensity, temperature or carbon dioxide supply). Since we have assumed that the inhibition of catalyst Eb does not affect fluorescence, one may ask why the same should not hold also for the inhibi- tion of the oxygen-liberating catalyst, since the latter, after all, occupies a similar "finishing" position in the scheme of photosynthesis. If we assume the validity of a scheme such as 7.IV, Eb and Eq assume exactly symmetric positions, and inhibition of either of them should lead to disappearance of the primary photochemical product (such as A -11002 or HX and A' -OH or Z) by back reaction. Inhibition of the second catalyst on the oxidation side, Eq, could, on the other hand, lead to the accumula- tion of somewhat more stable intermediates (designated by { OH } 2 or { O2 } in Vol. I). Franck suggested that these "photoperoxides," if not removed by Eo (or by the hydrogenase in hydrogen-adapted algae), tend to react with metabolic products (sugars?), converting the latter to substances capable of "narcotizing" chlorophyll (as well as Eb). In this way, the "finishing" catalyst, Eq, acquires the capacity of affecting indirectly the fluorescence of INA.CTIVATION OF CATALYTIC SYSTEM AS CAUSE OF INDUCTION 1421 chlorophyll — a change that can be caused directly only by preparatory cata- lyst. (This hypothesis already was used in chapter 24.) The totality of the events in the "first wave" of induction is attributed by Franck and co-workers (1941, 1945, 1947) to the photochemical produc- tion of an "internal narcotic." One may ask: Why introduce an un- known metabolite as an intermediate, and not assume direct action of the "photo-peroxides" on chlorophyll? The answer is: first, that a "block- ade" of chlorophyll by oxygen-precursors would be unlikely to lead to a reduction in the rate of oxygen liberation (although it may produce a delay in carbon dioxide uptake, and an upsurge of fluorescence) ; second, the strong dependence of the first wave on species, age, culturing and conditions that prevailed during the dark period points to the intervention of meta- bolic products. One puzzling question in connection with the "first wave" of induction is: How can a photochemical reaction lasting only 0.5 to 1 second, and thus permitting, at best, only one chlorophyll molecule in ten or fifty to absorb a quantum, bring about complete cessation of gas exchange, and an increase in fluorescence yield by as much as a factor of three? As far as the chemical inhibition is concerned, a plausible answer can be provided by reference to a catalyst the concentration of which is much lower than that of chlorophyll. Flashing light experiments led to the conclusion (c/. chapters 32 and 34) that this is true of the finishing catalyst, Eb- This catalyst appears to be present normally to the extent of about one molecule per 400-2500 molecules of chlorophyll. Franck and co- workers pointed out that the amount of the internal narcotic produced within a half second of moderate illumination is likely to be sufficient to inhibit all of catalyst Eb, and thus to cause practical cessation of photo- synthesis. If this picture is correct, then, in the first moment of illumination, car- bon dioxide consumption should begin at a comparatively high level, cor- responding to the full rate of the primary photochemical process, while the liberation of oxygen should begin at a rate determined by the amount of the active deoxygenase. Unless illumination is very weak (in which case the residual amount of the oxygenase suffices to maintain the initial rate, and no induction wave occurs), the initial rate of oxygen liberation is much lower than that of the primary photochemical process. As a result, inter- mediate oxidants accumulate, and within a second inhibit practically all Eb (through the intermediary of the oxidizable metabolite), and thus stop more or less completely both the formation of oxygen and the consumption of carbon dioxide. According to the hypothesis used before, respira- tion removes the "narcotic" poison — into which the metabolite produced in the dark is converted in light — and thus causes the inhibition wave to sub- 1422 INDUCTION PHENOMENA CHAP. 33 side rapidly in air, it extends over a much longer period under anaerobic conditions, where the only oxygen available is that produced by residual photosynthesis (c/. section 6). How can one, however, explain the strong effect of the first induction wave on the yield of chlorophyll fluorescence? How can such an effect be caused by a "narcotic" available (assuming it is produced with a quantum yield of 1) only to the extent of 1 molecule to 10 or 100 molecules of chloro- phyll? One is forced to assume either that 1 molecule of the narcotic can protect the fluorescence of 100 chlorophyll molecules, or else that the fluores- cence of the protected molecules is so much more intense than that of the unprotected ones as to raise the average yield of fluorescence by a factor of two or three. Both assumptions seem artificial. An explanation of the "second wave" of induction was first suggested in a paper by Franck, French and Puck (1941). They thought that, if the first wave is an indirect effect of accumulated photoperoxides (via an oxidizable metabolite), the second wave may be due to a direct effect of the same intermediates, which has to wait until their quantity had become sufiicient to "block" all chlorophyll. (Assuming a quantum yield of ~1 for the production of the photoperoxides, this should require from one half to several minutes, depending on the intensity of illumination.) This hypothesis could explain a "second depression" in the carbon di- oxide uptake curves, and also the "second wave" of fluorescence — but not interruption of the steady increase in oxygen production (since the oxygen "precursors" remain present in excess). Therefore, when Franck, Pring- sheim and Lad (1945) found a second wave also in oxygen liberation curves, a reinterpretation became necessary. This was provided by Shiau and Franck (1947); as mentioned before, their hypothesis was based on com- bination of normal induction with the depletion, in light, of the initially full reservoir of the reduction substrate, A-COo. The suggested interplay is illustrated by fig. 33.50. In A, curve a represents induction (more specifi- cally of fluorescence intensity) as it would be if the reactivation of the de- oxygenase were the only determining factor, and if the reservoir of reduc- tion substrates remained full. This curve shows the, at first slow, and then accelerated, decline of the "narcotization" of chlorophyll by the metabolic poison (the concentration of which follows that of the "photoperoxides," and thus, indirectly, the inactivation of the deoxygenase) . If one assumes — which is the salient point of the hypothesis — that the reduction sub- strates compete with the "narcotic" for adsorption on chlorophyll (as two adsorbable gases would for the surface of charcoal), then, with a lower con- centration of the reduction substrates, the "narcotic" would have a chance to occupy more chlorophyll, and a fluorescence-time curve of type 6 would result. If, at the beginning of illumination, the concentration of the reduc- ROLE OF RESPIRATION IN INDUCTION PHENOMENA 1423 tion substrate is high, and, after a few minutes of illumination, drops to a steady value (curve a in fig. 33.50A), the fluorescence (which depends on the ratio [narcotic]/ [reducible substance]), represented by curve c, will first follow curve a and then go over to curve b (fig. 33.50B), the transition being as indicated by the dotted curve. Whether this "second wave" of narcotization of chlorophyll will cause, in addition to a fluorescence wave, simultaneous inhibition waves in the uptake of carbon dioxide and the liberation of oxygen, may depend on o 1- ■ (A) 22 < 10 %^ a: m Z 3 8 :N \. c ^,,__...^/o ) O 00 OJ - AllSN o >. p, . ^^^ UJ 1— ^^^ . , . 1 z 16 P UJ-- o 4 2 0 i;:;^^__ iij o 14 UJ ^ '2 z ~l . ■ o o 1 2 5 LU DC 10 TIME.min. O UJ > 6 1- 5 4 Ul "" 2 0 TIME, min. Fig. 33.50. Explanation of the second wave by competition of reducible substances and inhibitor for chlorophyll (after Shiau and Franck 1947): (A) concentration vari- ation of narcotics (b), of photosynthetically reducible substances (a) and of their ratio; (B) theoretical fluorescence-time curves at two concentrations of photosensitive re- ducible substance, with transition curve shown as broken hne. whether these rates, at the moment when the narcotization wave appears, are limited by the amount of photochemically active chlorophyll, or the amount of active catalyst Eb- This may explain why during the second wave the usual antiparallelism between gas exchange and fluorescence is found often, but not always. According to this interpretation of the second wave, it should be en- hanced by all the factors that tend to depress the concentration of the re- duction substrate, [A-COa], such as low carbon dioxide concentration, low temperature and cyanide poisoning. This is confirmed by many ex- perimental curves, although the results obtained with variable [CO2] seem to be ambiguous (c/. p. 1339, 1381, 1391). Cyanide, in particular, clearly 1424 INDUCTION PHENOMENA CHAP. 33 enhances the second wave and delays its decay practically indefinitely (c/. fig. 33.39). We cannot recount here all the details of Franck's interpretation, by means of this induction theory, of all the variations in the shape of gas exchange and fluorescence-time curves uncovered in the experimental studies reported in parts A and B of this chapter. One general surmise of Franck should, however, be emphasized. He believes that the production of a "blanket" of a narcotizing metabolite on the surface of chlorophyll (and perhaps also on the surface of some catalysts) must be an important 12 3 4 TIME, sec. Fig. 33.51. Time course of concentration variation of two narcotics and their total concentration, which influences fluo- rescence intensity of chlorophyll (after Shiau and Franck 1947): (a) external inhibition; (6) internal inhibition. protective device developed by the plants to prevent destructive photo- chemical reactions (such as photoxidations) from being sensitized by chloro- phyll when photosynthesis is inhibited (for any possible reason). The peculiarities of the induction phenomena after anaerobic conditions were explained by Franck, Pringsheim and Lad by the assumption of the combined action of an "external" and the above introduced "internal" in- hibitor. The "external" inliibitor is the repeatedly mentioned diffusible acid and reducing material produced by fermentative metabolism and ex- creted into the medium. This material accounts for the effect of alkalies on anaerobic inhibition (Noack et al.), the importance of algal concentration (Franck, Pringsheim and Lad), the immediate high fluorescence yield at the beginning of illumination after prolonged anaerobic incubation, and many other characteristics of anaerobic induction. Variations in the premeabil- ity of cell membranes may play an important role in the inhibition phe- nomena caused by this factor, particularly in the different sensitivity of young and aged cells. ROLE OF RESPIRATION IN INDUCTION PHENOMENA 1425 Figure 33.51 shows the hypothetical changes in the amounts of the two inhibitors at the beginning of iUumination, and the resulting time course of fluorescence (the latter in agreement with the experiments). 5. Role of Respiration in Induction Phenomena The above analysis of the possible mechanisms of induction was based on the concept of photosynthesis as a separate system of chemical and photochemical reactions, fundamentally independent of other metabolic mechanisms in the cell. This has been a legitimate and useful concept, since it permitted attention to be concentrated on a minimum number of factors, and to ask questions conducive to meaningful experimentation. It is likely that, in the main, the chemical schemes and kinetic relationships based on this approach will prove vahd in the end. However, complete isolation of photosynthesis from other cellular processes undoubtedly was an oversimplification. More recently, evidence has begun to pile up con- cerning the mutual interplay of photosynthesis and other biochemical processes, above all, respiration. Instead of a separate, chemical struc- ture with one inlet through which carbon dioxide and water molecules enter, and one outlet through which sugar and oxygen escape — we now look on photosynthesis as a more open structure, with several connections to the ambient medium and to the other metabolic reaction sj^stems, on different reduction levels. Of course, even in the explanation of induction phenomena suggested in the preceding sections, Franck has assumed an interference of metabohtes with the smooth working of the photosynthetic apparatus, most prominently in his concept of "internal narcotics" (pro- duced, e. g., by fermentation, or by the action of excess photoperoxides on sugars), which were credited with inactivating the photochemical appara- tus by settling on chlorophyll, or on catalysts in permanent association with this pigment. The influence of such "internal poisons" appeared, however, as an inci- dental interference, which could only slow down or temporarily stop alto- gether the wheels of photosynthesis (although Franck also suggested that this interference may be providential when the danger exists that the photo- chemical apparatus, for lack of proper substrate, may begin to chew up it- self). The evidence for placing greater emphasis on the intrinsic character of cross-connections between photosynthesis and other metabolic processes came from two areas. Biochemical studies with isotopic carbon tracers have indicated that photosynthesis may have several intermediates which also occur in catabolic oxidative processes. It was found that several intermediates of respiration can participate re\'ersibly in the carbon dioxide 1426 INDUCTION PHENOMENA CHAP. 33 exchange with the medium (and not merely release carbon dioxide as inert final product of oxidation) ; and signs have been found that setting into motion or stopping the anabolic photochemical mechanism of photosyn- thesis affects the course and rate of the catabolic thermal processes more immediately than one could expect if their relation were due merely to the one process supplying substrates (sugars and carbon dioxide, respectively) for the other. These relationships will be discussed in chapter 36. The other reason for considering the interlocking of photosynthesis with other metabolic processes lies in kinetic findings, in particular the occur- rence of inverse induction, in which bursts of photochemical activity, con- siderably in excess of the steady rate, have been observed at the beginning of the illumination period (while bursts of nonphotochemical gas exchange have been observed in the initial periods of darkness, or reduced illumina- tion). As mentioned before, the concept of photosynthesis as a closed re- action sequence could account for inverse induction if it meant starting photosynthesis at full normal rate and then going into a temporary slump, or if it meant starting at a higher-than-normal rate of exchange of one gas (O2 or CO2) but lower-than-normal rate of exchange of the other one (to readjust the proper balance of the oxidized and the reduced forms of the photosynthetic catalysts and intermediates). It seems, however, that the various observed bursts and gulps of oxygen and carbon dioxide cannot all be explained in this way, and require the assumption that the pools of photosynthetic intermediates communicate with other metabolic pools and reservoirs, and that the levels to which these pools are filled depend on the rates of both the photochemical and the nonphotochemical metabolic processes. Every time one of these rates is suddenly changed, through an increase or decrease of illumination, all the pools and reservoirs are partly drained, or filled up, to a new stationary level, and this readjustment can be accompanied by evolution or absorption of either oxygen, or carbon dioxide, or both gases. This is obviously a generalization of the picture in which the filling of the pools of intermediates was considered as a matter of photosynthesis alone. Superimposed on these changes in the amounts of intermediates — such as various organic acids, perhaps also peroxides (or other intermediates in the exchange with molecular oxygen) — ^is the re- adjustment of catalysts, which may involve the oxidation or reduction of intermediate redox systems (such as the cytochromes, c/. above section B7). The attribution of gas "bursts" and ''gulps" in the first minute of illu- mination to the filling up of the photosynthetic pools by respiration inter- mediates was first suggested by Franck (1949) as interpretation for the rate measurements by Warburg and co-workers in intermittent hght. As stated before in this chapter, Burk and Warburg (1951) suggested that the ROLE OF RESPIRATION IN INDUCTION PHENOMENA 1427 bursts of activity they noted in the first minute of both hght and darkness (or "bright" and ''dim" hght) were revelations of steady high rate of gas exchange in hght — the high rate of anabohc photoprocess being revealed at the beginning of illumination, and the high rate of catabolic "anti- photosynthesis" at the beginning of darkness (where it takes seconds or minutes to be completed). We have concluded before that this interpreta- tion is not in agreement with the totahty of experiments, and that the ob- served bursts and gulps must be considered as transients rather than as "tail ends" of steady metabohsms before the change in illumination. Franck (1953) has attempted a quantitative analysis of the experiments described in Warburg's papers to show that these transients can be explained by assuming that respiration intermediates (of the reduction level of glyc- eric acid), accumulated during the dark period, become available for photo- chemical reduction in the first minute of illumination, and that this photo- chemical half-way reversal of respiration requires only two quanta (per one nonliberated molecule of CO2 and one nonconsumed molecule of oxy- gen). Because of the importance of this question for the controversy about the true minimum quantum requirement of photosynthesis, Franck's calculations will be given in more detail in chapter 37D (section 4d) . The absence of induction losses in the Hill reaction of Chlorella cells and isolated chloroplasts poses a question for the Franck-Gaffron theory of induction. It appears offhand that the simplest explanation of this fact would be to renounce the concept that the origin of the induction loss lies on the "oxidation side" of the photochemical process proper (e. g., in the inactivation of the oxygen-liberating enzyme), since processes there must be the same in photosynthesis and in the Hill reaction, and to seek this ori- gin on the "reduction side" of the primary photochemical process, where the Hill reaction differs from photosynthesis. Franck suggested, however, that this argument is not necessarily convincing, provided one attributes the bulk of the induction loss to a "narcotization" of photochemical ap- paratus by oxidation products formed in consequence of the accumulation of photoperoxides when the oxygen-liberating enzyme is inhibited. These "narcotics," he argued, may be unable to displace "substitute oxidants," such as quinone, from their position as hydrogen acceptors in contact with chlorophyll (and thus to inhibit temporarily the Hill reaction), while they are able to prevent the access to chlorophyll of hydrogen acceptors (such as phosphoglycerate) which must be reduced in photosynthesis. Although this explanation is not implausible, it must be acknowledged that the at- tribution of induction to the inhibition of the oxygen-hberating enzyme as the primary cause rests on circumstantial evidence. It could perhaps be revised without destroying the main ideas of Franck's induction theory, which include the inactivation, in the first moment of illumination, of some 1428 INDUCTION PHENOMENA CHAP. 33 essential photosynthetic enzyme, the consequent blanketing of chlorophyll by a "narcotic," and the removal of this narcotic by respiration. 6. Other Theories of Induction It is impossible to discuss here in detail the theories of induction suggested by other authors, particularly since they were mostly invented ad hoc, without relation to the totality of our knowledge of the kinetics of photosynthesis. The most elaborate specula- tions concerning the origins of the induction curves have been presented by Kautsky. They were based exclusively on the observation of fluorescence, in disregard of other as- pects of the induction phenomena, not to speak of general kinetics of photosynthesis in constant or intermittent light. In his first papers, Kautsky attributed induction to the interaction of excited chloro- phyll with molecular oxygen (which he considered — cf. chapter 18, page 514 — as the universal energy acceptor in dyestufif-sensitized reactions). After he himself had found that the fluorescence of living plants is insensitive to changes in oxygen concentration between 0.5 and 100%, Kautsky substituted, in the role of energy acceptor, an oxygen- acceptor compound, supposed to dissociate only below 1% O2 in the atmosphere. He attributed the first rise of fluorescence to the transfer of excitation energy from chloro- phyll to this compound (ilO-z), which was assumed to be converted into an activated form, fiOo* (perhaps a peroxide). ^Vhile this form accumulates, the fluorescence rises to a peak (because only nOo, and not UO,*, can act as a chemical quencher). In the second stage, fi02* reacts (thermally) with the substrates of photosynthesis, and is con- verted back into QO2; this leads to the renewed decline of fluorescence. In absence of oxygen, ilO^ is dissociated and the fluorescence wave disappears. Kautsky later added various additional assumptions intended to explain the de- tails of the fluorescence curves. An elaborate development of the theory was given by Kautsky and U. Franck (1943), who postulated, in addition to the photochemical acti- vation and thermal deactivation of the first energy acceptor, ^62, a sequence of three photochemical activations and thermal deactivations of another energy acceptor, which they made responsible for those features of the fluorescence curves that do not disappear, but, to the contrary, are accentuated in the absence of oxygen. They thought that the four successive photochemical reactions between excited chlorophyll and the two energy acceptors, which they beUeved detectable in the ups and downs of the fluorescence curves, must correspond to as many primary photochemical processes of photosynthesis, and thus explain why the latter may require 4 quanta. This explanation fails to deal with the basic problems. While attributing the varia- tions of fluorescence to accumulation and disappearance of various intermediates of photosynthesis, Kautsky does not even ask why these intermediates accumulate to a maximum and then disappear again (instead of assuming a constant level), or why the first peak is reached after an illumination period so short that only a few per cent of chlorophyll molecules can be excited. Later, Kautsky (1951) tried to bring his theory of induction in relation to Warburg and Burk's "new theory" of photosynthesis, which also ascribes to molecular oxygen an active role in the photosynthesis cycle. Smith (1937) suggested that the induction curves he obtained with Cabomha, as well as those found by Briggs for Mnium and by van der Paauw for Hormidium, can be explained quantitatively by the assumption that chlorophyll has to be activated photo- chemically, to Chi*, by a reaction the yield of which is proportional to (const. — [Chi* P), BIBLIOGRAPHY TO CHAPTER 33 1429 and is deactivated thermally by a reaction (with the substrate of photosynthesis) the yield of which is proportional to [Chl*]^ (this reaction leading to the final product of photosynthesis: Chi + hv —^ Chi*, Chi* + substrate —* Chi + product). No explana- tion was given for the occurrence of a square of concentration in both equations. In addition Smith's theory disregards the main objections against all induction theories based on the accumulation of thermal intermediates — the irreversible character of the induction losses. Briggs (1933) discussed two forms of an "inhibitor theory" of induction. In the first one, the inhibitor was assumed to be of the "narcotic" type, i. e., one which makes the sensitizer unavailable for photosynthesis by settling down on it. In the second theory, the inhibitor was assumed to be of the cyanide type, and to prevent the primary photoproducts from stabilization by the poisoning of an enzyme. Briggs derived, for these cases, equations representing the approach of photosynthesis to its final steady rate; but our present knowledge of the complexity of the induction curves makes us sceptical with regard to the value of all such theoretical functions, even if they were found to fit several experimental curves. Some more recent experimental studies (Mehler, p. 1568; Gerretsen, p. 1588; Arnon et al., p. 1537) drew attention to one additional possible source of "asymmetric" induction-photoxidation of ascorbic acid reserves. This may precede the utilization of water as reductant and delay the liberation of oxygen, while permitting carbon dioxide reduction to start immediately. (Initial reduction of "substitute reductants," instead of carbon dioxide, obviously must have the opposite effect, as repeatedly suggested above.) Speaking in general, with the gradual clarification of the biochemistry of photo- synthesis a tendency arises to inquire into the qualitative chemical sources of induction phenomena instead of analj'zing the induction curves in terms of a minimum nmuber of kinetic factors. This is inevitable and natural; but the considerable experimental work and ingenuity of interpretation invested in the stud.y of induction kinetics (and of the kinetics of photosynthesis in general) should not be considered as wasted (as some biochemists may be inclined to believe). They are fundamental contributions to- ward the general edifice of photosynthesis — and biochemistry in general — as an exact science. Bibliography to Chapter 33 Induction Phenomena A. Gas Exchange during the Induction Period 1865 Boussaingault, J. B., Compt. rend., 61, 493, 605, 657. 1887 Pringsheim, N., Sitzber. Akad. Wiss. Berlin, 1887, 763. 1918 Osterhout, W. J. V., and Haas, A. R. C, /. Gen. Physiol, 1, 1. Willstatter, R., and StoU, A., Untersuchungen iiber die Assimilation der Kohlensdure. Springer, Berlin. 1920 Warburg, 0., Biochem. Z., 103, 188. 1922 Kostychev, S. P., Ber. deut. botan. Ges., 39, 319. 1929 Li, Tsi Tung, Ann. Botany, 43, 787. 1930 Harder, R., Planta, 11, 263. 1932 van der Paauvv, F., Rec. trav. botan. neerland., 29, 497. 1933 Briggs, G. E., Proc. Roy. Soc. London, B113, 1. Harder, R., Planta, 20, 699. 1430 INDUCTION PHENOMENA CHAP. 33 1934 Emerson, R., and Green, L., J. Gen. Physiol., 17, 817. 1935 Bukatsch, F., Jahrh. wiss. Botan., 81, 419. Gaffron, H., Biochem. Z., 280, 337. 1937 Gaffron, H., Naturwissenschaften, 25, 460. Gaifron, H., ibid., 25, 715. Gessner, F., Jahrh. wiss. Botan., 85, 267. McAlister, E. D., Smithsonian Inst. Pubs. Misc. Collections, 95, No. 24. Smith, E. L., /. Gen. Physiol, 21, 151. 1938 Blinks, L. R., and Skow, R. K., Proc. Natl. Acad. Sci. Washington, 24, 413, Blinks, L. R., and Skow, R. K., ibid., 24, 420. Gessner, F., Jahrh. wiss. Botan., 86, 491. Harder, R., and Aufdemgarten, H., Nachr. Ges. Wiss. Gottingen, 3, 191. 1939 Aufdemgarten, H., Planta, 29, 643. Aufdemgarten, H., ibid., 30, 342. Gaffron, H., Biol. Zentr., 59, 288. Gaffron, H., Cold Spring Harbor Symposia Quant. Biol., 7, 377. McAlister, E. D., /. Gen. Physiol, 22, 613. Noack, K., Pirson, A., and Michels, H., Naturwissenschaften, 27, 645. 1940 Gaffron, H., Am. J. Botany, 27, 204. McAlister, E. D., and Myers, J., Smithsonian Inst. Pubs. Misc. Collections, 99, No. 6. Michels, H., Z. Botan., 35, 241. Myers, J., and Burr, G. O., J. Gen. Physiol, 24, 45. 1941 Emerson, R., and Lewis, C. M., Am. J. Botany, 28, 789. Franck, J., and French C. S. /. Gen. Physiol, 25, 309. Franck, J., French, C. S., and Puck, T. T., /. Phys. Chem., 45, 1268. Franck, J., and Gaffron, H., Advances in Enzymology, Vol. I, Interscience, New York-London, p. 199. 1942 Steemann-Nielsen, E., Dansk Botanisk. Arkiv, 11, No. 2. 1945 Franck, J., Pringsheim, P., and Lad, D. T., Arch. Biochem., 7, 103. 1947 Shiau, Y. G., and Franck, J., Arch. Biochem., 14, 253. 1948 Warburg, 0., Am. of Botany, 35, 194. 1949 van der Veen, R., Physiol Plantarum, 2, 217. van der Veen, R., ihid., 2, 287. Steemann-Nielsen, E., ibid., 2, 247. 1950 Clendenning, K., A. and Ehrmantraut, H., Arch. Biochem., 29, 387. van der Veen, R., Physiol. Plantarum, 3, 247. Kandler, 0., Z. Naturforsch., 5b, 423. Warburg, 0., and Burk, D., Arch. Biochem., 25, 410. 1951 Osterlind, S., Physiol. Plantarum., 4, 514. Warburg, O., Geleick, H., and Briese, K., Z. Naturforsch., 66, 417, Nishimura, M. S., Whittingham, C. P., and Emerson, R., Symposia Soc. Exptl Biol, 5, 176. 1952 Steemann-Nielsen, E., Physiol. Plantarum., 5, 334. Osterlind, S., ibid., 5, 403. Olson, R. A., and Brackett, F. S., Federation Proc, 11, 115. BIBLIOGRAPHY TO CHAPTER 33 1431 Briggs, G. E., and Whittingham, C. P., New Phytologist, 51, 236. 1953 Brown, A. H., Am. J. Botany, 40, 719. Brown, A. H., and Webster, G. C., ibid., 40, 753. Brackett, F. S., Olson, R. A., and Crickard, R. G., J. Gen. Physiol, 36, 529. Brackett, F. S., Olson, R. A., and Crickard, R. G., ibid., 36, 563. Franck, J., Arch. Biochem. and Biophys., 45, 190. Hill, R., and Whittingham, C. P., New Phytologist, 52, 133. Damaschke, K., Todt, F., Burk, D., and Warburg, 0., Biochim. et Biophys. Acta, 12, 347. Strehler, B. L., Arch. Biochem. and Biophys., 43, 67. Warburg, 0., Krippahl, G., Buchholz, W., and Schroder, W., Z. Natur- forsch., 8b, 675. 1954 Emerson, R. and Chalmers, R. V., Paper presented to Nat. Acad. Sci., April 1954. Emerson, R., and Chalmers, R. V., Paper presented at Intern. Botany Congress, Paris, July 1954. Rosenberg, J. L., J. Gen. Physiol., 37, 754. Gaffron, H., IVth Symposium of the Society of General Microbiology, Cambridge Univ. Press, pp. 152-185. 1955 Brown, A. H., and Whittingham. C. P., Plant Physiol, 30, 231. B. Fluorescence and Absorption during the Induction Period 1931 Kautsky, H., and Hirsch, A., Naturwissenschaften, 19, 694. 1932 Kautsky, H., Hirsch, A., and Davidshofer, F., Ber. deut. chem. Ges., 65, 1762. 1934 Franck, J., and Levi, H., Z. physik. Chem., B27, 409. Kautsky, H., and Hirsch, A., Biochem. Z., 274, 423. Kautsky, H., and Spohn, H., ibid., 274, 435. 1935 Kautsky, H., and Hirsch, A., ibid., 277, 250. Kautsky, H., and Hirsch, A., ibid., 278, 373. Knorr, H. V., and Albers, V. M., Cold Spring Harbor Stjmposia Quant. Biol, 3, 87. 1936 Franck, J., and Wood, R. W., /. Chem. Phys., 4, 551. Kautsky, H., and Flesch, W., Biochem. Z., 284, 412. Kautsky, H., and Marx, A., Naturwissenschaften, 24, 317. 1937 Kautsky, H., and Marx, A., Biochem. Z., 290, 248. Kautsky, H., and Hormuth, R., ibid., 291, 285. 1938 Kautsky, H., and Eberlein, R., Naturwissenschaften., 26, 576. 1939 Kautsky, H., and Eberlein, R., Biochem. Z., 302, 137. Wassink, E. C, and Katz, E., Enzymologia, 6, 145. 1940 McAlister, E. D., and Myers, J., Smithsonian Inst. Pubs. Misc. Collec- tions, 99, No. 6. 1941 Franck, J., French, C. S., and Puck, T. T., /. Phys. Chem., 45, 1268. Franck, J., and Gaffron, H., Advances in Enzymology, Vol. I, Interscience, New York-London, p. 199. Kautsky, H., and Zedlitz, W., Naturwissenschaften, 29, 101. 1432 INDUCTION PHENOMENA CHAP. 33 1942 Wassink, E. C, Katz, E., and Dorrestein, R., Emyinologia, 10, 285. 1943 Kautsky, H., and Franck, U., Biochem. Z., 315, 139, 156, 176, 207. 1944 Wassink, E. C, and Kersten, J. A. H., Enzymologia, 11, 282. 1945 Franck, J., Pringsheim, P., and Lad, D. T., Arch. Biochem., 7, 103. 1947 Shiau, Y. G., and Franck, J., Arch. Biochem., 14, 253. 1948 Kautsky, H., and Franck, U., Naturwissenschaften, 35, 43, 74. 1949 Franck, J., "The Relation of Chlorophyll Fluorescence to Photosynthesis" in Photosynthesis in Plants. Iowa State College Press, Ames, 1949, pp. 293-348. 1951 Wassink, E. C, Advances in Enzymology, Vol. XI, Interscience, New York-London, p. 91. 1952 Duysens, L. N. M., Tlwsis, Univ. Utrecht. 1954 Duysens, L. N. M., Nature, 173, 692. Duysens, L. N. M., Science, 120, 353. Duysens, L. N. M., ibid., 121, 210. C. Interpretation of Induction Phenomena 1918 Osterhout, W. J. V., and Haas, A. R. C, /. Gen. Physiol, 1, 1. 1932 Lubimenko, V. N., and Shcheglova, 0. A., Planta, 18, 383. 1933 Briggs, G. E., Proc. Roy. Soc. London, B113, 1. 1937 Gaffron, H., Naturwissenschaften, 25, 460. Gaffron, H., ibid., 25, 715. Smith, E. L., J. Gen. Physiol, 21, 151. 1939 Gaffron, H., Biol Zentr., 59, 288. Gaffron, H., Cold Spring Harbor Symposia Quant. Biol, 7, 377. 1940 Gaffron, H., Am. J. Botany, 27, 204. 1941 Franck, J., French, C. S., and Puck, T. T., J. Phys. Chem., 45, 1268. 1942 Franck, J., Am. J. Botany, 29, 314. 1943 Kautsky, H., and Franck, U., Biochem. Z., 315, 139, 156, 176, 207. 1945 Franck, J., Pringsheim, P., and Lad, D. T., Archiv. Biochem., 7, 103. 1947 Shiau, Y. G., and Franck, J., ibid., 14, 253. 1949 Franck, J., Arch. Biochem., 23, 297. 1951 Burk, D., and Warburg, 0., Z. Naturforsch., 6b, 12. I^utsky, H. ibid., 6b, 292. 1953 Franck, J., Arch. Biochem. and Biophys., 45, 190. Chapter 34 TIME EFFECTS. II. PHOTOSYNTHESIS IN INTERMITTENT LIGHT* This chapter calls for the same preliminary remark made in chapter 33. When it was first written, "induction" appeared to be merely a gradual rise of photosynthesis, after a dark period, from a low initial rate to a steady final level; and the effects of light intermittency appeared fully expHcable by the combined influence of induction losses (negative intermittency ef- fect), and a continuation in the dark, for a time of the order of 0.03 second, of the hmiting thermal reaction of photosynthesis (positive intermittency effect). This picture appears oversimplified now, after transient bursts and slumps of gas exchange have been shown to follow changes in the in- tensity of illumination. Combined with induction losses, these transients can make the time course of the exchange of carbon dioxide, or oxygen, or both, quite complicated. Furthermore, it now seems likely that photosynthesis in intermittent light may be affected by interaction with respiration and other cataboHc processes (which may be one of the causes of the above-mentioned "tran- sients"). Intermediates of the catabolic metaboUsm can be drawn into the photosynthetic process in a subsequent light period; in other words, in intermittent hght, illumination may reverse part-way some of the respira- tion begun, but not completed, during the dark periods (c/. chapter 37D, section 3). Still another general remark is appropriate here: In cell suspensions, "intermittency effects" are also inevitable in "steady" hght because of stirring, particularly in the case of dense suspensions illuminated by a nar- row beam of light. Diluting the suspension and spreading the illumination uniformly over the whole surface of the vessel minimizes the intermittency of the light to which each single cell is exposed. However, even in the ex- treme case when practically no mutual shading of the cells occurs, individual chlorophijll molecules still receive variable amounts of light depending on the momentary orientation of the cell (since a single chloroplast absorbs up to 50% of incident light in the absorption peaks of chlorophyll) . The closest approximation to uniform illumination can be obtained by using weakly absorbed (e. g., green) light, and a suspension layer containing, on the aver- age, less than one cell in the path of each hght beam. * Bibliography, page 1483. 1433 1434 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 After these preliminary remarks, we now proceed with the discussion of intermittent hght experiments. In a quantitative discussion of the intermittency effect, a basis must first be estabHshed for the comparison of yields in intermittent and con- stant light. Three methods of comparison have been used (c/. fig. 34.5): (a) The yield obtained during a certain total period of intermittent illumination (A'' light periods of t* seconds each, and A^ dark periods of ta seconds each) has been compared with the yield produced in the same total time, t = N{tci + i*), by uninterrupted light of equal intensity. We may call the ratio of these two yields the intermittency factor for equal intensity, and equal total time, in. This method of comparison has to be used, e. g., to answer the question: How will periodic interruptions of illumination by a rotating disc affect the yield of photosynthesis of a plant under a light source of constant intensity? (6) The yield produced during A'' light periods of t* seconds each can be compared \vith that of uninterrupted illumination of equal actual duration, Nt* seconds (and equal intensity). This comparison answers the question: Given a certain total amount of light energy of definite intensity, will it be better utilized for photosynthesis by dividing it into several exposures separated by dark intervals, or will it be used best in one continuous stretch? The ratio of the yields obtained in these two ways— the inter- mittency factor for equal intensity and equal total energy — will be designated by iiE- A simple relation exists between in and ijE, namely: (34.1) ill = iiE td + V The quantity that Briggs (1941) called the "yield" of intermittency was: llE ~ 1- (c) The yield obtained in intermittent light can further be compared with the yield produced by the same total amount of light energy dis- tributed uniformly over the same total time; the intensity of the uninter- rupted light is in this case smaller than that of intermittent light, in the ratio t*/{td -\- t*). This method of comparison answers the question: Given a certain amount of light energy to be used within a certain period of time, will it be more advantageous to distribute this energy evenly over the whole available time, or to concentrate it in separate exposures with dark intervals between them? The ratio of the yields obtained in this way can be called the intermittency factor for equal energy and equal time; it will be designated by ist- The relation between Iei and the other two intermittency factors is as follows : YIELD OF PHOTOSYNTHESIS IN ALTERNATING LIGHT 1435 t* (34.2) lEt = Piit = PiiEj^-T:ji Here, /S designates the increase in rate of photosynthesis brought about by an increase in the intensity of continuous hght by a factor of {ta + t*)/t. Equation (34.2) follows from the consideration that, if we first raise the intensity of continuous illumination by this factor, and then use a rotating sector to produce intermittent illumination with light periods of t* seconds duration, the result must be the same as that obtained by concentrating the whole energy of the original illumination in exposures of t* seconds each. Equation (34.2) shows that the factor Iei depends on the shape of the light curve, P = /(/), in continuous light. Section A of this chapter will deal with phenomena observed in inter- mittent light with equal light and dark periods, U = t*, which we will desig- nate as alternating light. In this case, its = 2 in (cf. equation 34.1), and one of these two factors (rather than the factor ist) is commonly used for the characterization of the intermittency effect. Section B will deal with flashing light {t* 0.5) if the periods ta and t* are very long or very short, and less than unity in the intermediate region. Long intervals (of the order of several hours) can improve the utilization of light energy because during the dark "rest periods" the plant can recuperate from the injury or ex- haustion that often follows a period of intense photosynthesis. Some phenomena involved in the natural adaptation of plants to the alternation * Bibliography, page 1483. 143G PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 of day and night probably belong to this class. Very short intervals (of the order of 1 second or less) also may cause an improvement of the energy con- version yield, because they allow the dark catalytic reactions of photo- synthesis to run to completion, restoring the photosynthetic apparatus to its full efficiency at the beginning of each new light period. In the intermediate range of frequencies — of the order of 1/min. or 1/hr.— alternating illumination can be expected to cause a depression of the yield (i/e < 1, in < 0.5), because dark intervals of this length permit the development of induction phenomena, which occupy most of the subsequent light periods. Thus, plotting Ije and in against log t, we can expect to obtain a curve of the shape shown in figiu'e 34.1. Short induction period -2-10123456 LOG / , sec Fig. 34.1. Expected trend of intermittency factors for equal dark and light periods. (In this curve, the "bursts" and "gulps" which complicate induction phenomena have not been taken into consideration. This is legitimate when the volume of these extra components of the gas exchange is con- siderably smaller than that of the induction losses.) The first observations of the actual effect of intermittent light on photo- synthesis were made by Brown and Escombe (1905). They reported that, under certain conditions, as much as one half, or even three quarters, of the total incident light could be taken away by a rotating sector without sig- nificantly decreasing the yield of photosynthesis (this means in ^^^ 1 and iiE ^ 2-4). Willstatter and StoU (1918) suggested that, since Brown and Escombe worked with strong light and a limited supply of carbon dioxide, their re- sults could have been due to the exhaustion of carbon dioxide in the im- mediate neighborhood of the chloroplasts during each flash and thus bear no relation to the intrinsic kinetic mechanism of photosynthesis. This inter- YIELD OF PHOTOSYNTHESIS IN ALTERNATING LIGHT 1437 Table 34.1 Intermittency Factors for CMorella in Alternating Light" (after Warburg 1919) t, sec 15 Ts 0.38 0.15 0.038 0.015 0.0038 0.0038'' 0.0038' i,E 1.14 1.36 1.46 1.56 1.77 1.72 1.96 1.88 1.0 " High light intensity, 25° C, [CO2] = 9.1 X 10"* mole/1. '' [CO2] = 136 X 10^-^ ' Low /. pretation caused Warburg (1919) to undertake new experiments on the effect of alternating light, in which care was taken to provide an abundant supply of carbon dioxide. He found that the intermittency effect occurs also under these conditions, where the explanation of Willstatter and Stoll cannot apply. Table 34.1 shows that the intermittency factor, Ije, is considerably larger than unity at [CO2] = 9.1 X 10"^ mole/1, (a concentra- tion high enough to make carbon dioxide limitation implausible), and even Fig. 34.2. Yellow cosmos (Cosmos sulphureus) grown with equal periods of light and darkness (after Garner and AUard 1931). Compare with fig. 34.1. at [CO2] = 136 X 10-^ mole/1. The last figure in the table shows that intermittency has no influence on the rate in weak light {ijE = 1-0; in = 0.5). This is understandable; in weak hght (more precisely, within the linear range of the light curves), the rate of photosynthesis is limited only by the frequency of the absorption acts; the catalysts can co})e with all (he intei'mediates produced by light without the formation of a backlog that could be utilized in the dark. In strong light, on the other hand, the 1438 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 photosynthetic apparatus could produce more if it were not for the slow- ness of certain enzymatic reactions — and these can be completed during the dark intervals. (Here, again, we neglect the fact that under certain conditions "bursts" of gas exchange have also been observed in weak light, where induction losses are negligible.) Padoa and Vita (1928) repeated Warburg's experiments with the water plant Elodea canadensis. They found ijE factors up to 2.71, with not less than five maxima (at 16, 80, 406, 650 and 887 alternations per second). Between these peaks, ijE values declined to 1. The reality of these several maxima and minima, not noted by other investigators, is very doubtful. The next step in the elucidation of the shape of Ije = f{t) curve was made in the well-known studies of Garner and Allard (1931) on the effect of intermittent light on the growth of plants. Figure 34.2, taken from their work, shows a striking minimum of the growth of potted plants of yel- low cosmos when the alternations occur about once a minute. The growth curve rises steeply on both sides of the minimum; 5 second intervals are almost as favorable as the "natural" intervals of 12 hours. It was often suggested that the basis of some intermittency effects in plant physiology may lie in the influence of intermittent light on photosyn- thesis. In confirmation of this, Portsmouth (1937) found that the growth curve of Garner and Allard runs closely parallel to the curve representing the yield of photosynthesis in relation to the frequency of light alternations. We recall that Willstatter and Stoll tried to attribute the intermittency ef- fects, described by Brown and Escombe, to an inadequate supply of carbon dioxide. Gregory and Pearse (1937) suggested a similar explanation for the growth curve of Garner and Allard. They thought that it may be caused by incomplete opening of stomata in intermittent hght, leading to carbon dioxide starvation. (Both the opening of stomata in light and their closure in the dark are not instantaneous; the ratio of their velocities determines the average aperture of the stomata in intermittent hght.) Portsmouth thought that the sluggishness of the stomata may provide a clue also to the decline of photosynthesis in alternating light. Gregory and Pearse measured the apertures of the stomata of Pelar- gonium zonale in alternating light, and found the slits to be particularly narrow when the periods of light and darkness were 5 seconds each. They thought this period to be sufficiently close to the minimum of the Garner- Allard growth curve to warrant the attribution of the latter to stomatal in- fluences. However, figure 34.2 shows that, at i = 5 seconds, the plant development proceeded quite satisfactorily. Furthermore, we mentioned above that Warburg had observed intermittency effects in Chlorella cells, where no stomatal effects are possible. More recently, Iggena (1938) found that a growth curve similar to that observed by Garner and Allard YIELD OF PHOTOSYNTHESIS IN ALTERNATING LIGHT 1439 with the higher land plants can be obtained also with stomata-free lower plants (green or blue algae). These plants, too, grew particularly slowly in light mth an alternation frequency of 1/min., and much better at t = 0.25 or 0.08 second, or t > I minute. More recent experiments on the growth of unicellular algae {Chlorella) in intermittent Hght gave a similar result (c/. part B, pp. 1476-1477). Fig. 34.3. Effect of intermittent illumination, with equal light and dark pe- riods, on photosynthesis in wheat plants (after McAhster 1937). Arrows indicate beginning and end of illumination. Points represent spectroscopic CO2 deter- minations in half minute intervals. We conclude from these experiments that the inertia of the stomata can be, at best, only a contributing cause of the inhibition of plant growth by alternating light with a frequency of the order of 1/min. Comparison of figure 34.1 with the figures in chap. 33 makes it Hkely that the main cause of this behavior is induction, which is almost fully developed after 1 minute of darkness, and permits only little photosynthesis in the first minute of subsequent illumination. Figure 34.3 shows that according to McAlister (1937) the average rate of consumption of carbon dioxide by wheat plants also is slowest at 1440 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 t = 60 seconds; it increases gradually when t declines to }qo second, or increases from 60 seconds to infinity. (The yields obtained at ^ = 120 and 300 seconds, not shown in figure 34.3, were intermediate between those obtained at 60 seconds and in continuous light.) Hormidium flaccidum. 19,6° C, 1500 lux 1 0 ^ ^tii 0.9 - \ /- ^ 0.8 - v^ (a) 07 1 1 1 1 1 1 1 1 15 20 60150 600 3600 TIME, sec. 0.7 I I I J L 60 150 600 3600 TIME, sec. 1.7 1.6 1.5 1.4 1.3 1.2 I.I Hormidium flaccidum, 20.5° C, 1500 lux, short intervals _L. _L 1.05 1.00 095 0.90 Stichococcus bacillans, 2II°C., 1500 lux, long mtervols (d) _L _L 600 3600 12600 TIME, sec. Fig. 34.4. Intermittency factors for two species of algae for equal dark and light periods (after Aufdemgarten 1939). 0.0085 0.06 0.2 Q5 5 TIME, sec. Factor ijt is determined in figure 34.3 by the ratio of the slope of the line representing the average consumption of carbon dioxide in intermittent light, and the slope of the corresponding line for continuous light, in is < 0.5 (iiB < 1) for / = 60 and 15 seconds, and > 0.5 (ite > 1) for the smaller values of t. At t = I^q second, ijt almost reaches unity {{k — 0.93, iip = 1.86). The findings of Warburg et al. (1951) that the gas exchange in alternating light, t = 1-3 min., results from a photosynthesis enhanced up to 3 times and respiration enhanced up to 10 times (compared to steady conditions), and the failure of Brown (1953) and of Whittingham (1954) to confirm them will be discussed in chapter 37D. THEORY OF ALTERNATING LIGHT EFFECTS 1441 Aufdemgarten (1939) measured the photosynthesis of Hormidium flaccidum in intermittent hght, using van der Paauw's gas flow method. His results are shown in figure 34.4a, the shape of which resembles closely that anticipated in figin-e 34.1 The minimum lies someAvhat above 1 min- ute (at t = 2.5 minutes); in some experiments, a second minimum was found at ^ = 10 minutes. In a Stichococcus bacillaris suspension, a flat minimum stretched from 1 to 2.5 minutes (fig. 34.4b). The factor Z/g rises sharply on the side of the short intervals as t declines to 0.06 second, then more slowly; at t = 0.0085 second, I/e reaches 1.7 (cf. fig. 34.4c). On the side of long intervals, z'/g rem.ains below unity up to ^ = 1.5 hour (cf. fig. 34. 4d). Experiments with excised leaves (Impatie7is parvijlora, Vitis vinifera, etc.) gave less regular curves, but they, too, showed a distinct minimum in the region of ^ = 5 minutes. 2. Theoretical Discussion of the Effect of Alternating Light There seems to be little doubt that the general shape of the Ije = f{t) curves in the region between 0.01 and 1000 seconds is strongly influenced by the interplay of two factors: the "Emerson-Arnold period," which causes the intermittency factor to be highest at alternations of the order of lOO/sec; and the induction period, which produces a minimum in the region between ^ = 1 and 5 minutes. The course of the curve above 5 minutes seems to reveal the influence of the "long" induction period (cf. fig. 34.4c). Whether "exhaustion" or "fatigue" effects produce a second hump, somewhere between t = 1 hour and oo (tentatively indicated in fig. 34.1), is not certain, and probably depends on special circumstances. A quantitative interpretation of the intermittency factors in the region where the induction period is the decisive factor is complicated by the fact that, in alternating light, a change of frequency affects both the light and the dark periods, and thus produces two antagonistic effects. A longer dark interval means a more complete preparation of the induction phenom- ena, while a longer light interval means more time for overcoming the initial inhibition. The first effect can be expected to prevail at alternation frequencies of more than 1/min. (where induction phenomena blanket prac- tically the whole illumination period), the second at alternation frequencies of less than 1/5 min. (where a further lengthening of the dark period can add only little to the induction losses; cf. fig. 33.6). The exact position of the minimum must depend on kinetic equations that govern deactivation of the photosynthetic apparatus in the dark and reactivation in light (for a dis- cussion of the kinetics of these reactions, see chapter 33, part C). The disappearance of photosynthetic production losses due to induction 1442 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 at ^-values below 1 minute should bring the factor Iib back to unity, but could not make it higher than 1. The experimental intermittency factor increases, however, far above this value, and approaches 2 at alterna- tion frequencies of the order of 100/sec. (c/. Table 34.1, and figs. 34.3 and 34.4c). This behavior becomes understandable if one assumes that short dark intervals can be efficiently utiUzed for the completion of the dark reaction that limits the rate of photosynthesis in strong continuous light. In the limiting case of very short intervals, the rate-determining catalyst will be as fully occupied during the dark interval as it is in light, leading to an intermittency factor of ^/B ^^ 2 {ijt ^^ 1). We may thus conclude, from the alternating light experiments, that the catalytic reac- tion that limits the rate of photosynthesis in strong light can continue for about 0.01 second after the cessation of illumination. According to the theory of Franck and Herzfeld, this reaction is the transformation of the intermediates produced by the photochemical proc- ess proper, which prevents them from reacting back. The catalyst that brings about this "stabilization" was designated by Eb in several reaction schemes presented in chapters 7, 9 (Vol. I), 24 and 28 (Vol. II, 1). As stated before, this hypothesis of Franck and Herzfeld is not bound to the specific reaction mechanism suggested by these authors (scheme 7VA), but can be used also in conjunction mth other reaction schemes. The fact that the catalyst Eb can work in the dark only for a limited length of time (about 0.01 second at room temperature), irrespective of the intensity of the preceding flash, can be understood if it is assumed that this catalyst acts on an unstable substrate. If the flash had produced more light products than the catalyst can handle at one time, only the batch that has become associated with the catalyst immediately after the light reaction is saved from back reactions and contributes to the final yield. We have already used this picture in chapter 32 (sect. 4) in explaining the maximum number of oxygen molecules that can be produced by a flash. (We have postulated that this number is determined by the number of available mole- cules of Eb; it may be either equal to Eb, or smaller by a factor of n, de- pending on whether Eb has to operate once, or n times — perhaps, four or eight times — to bring about the liberation of one molecule of oxygen.) Experiments in flashing light (to be discussed in section B) have per- mitted a more precise determination of the "working period" of Eb — about 0.02 second at 20° C. (c/. Table 34.11). If the intermittency ef- fect in alternating light were determined, in the region t < 1 minute, only by this catalytic action period (which we will call the Emerson-Arnold period) the factor ijn could exceed unity only for dark intervals of this order of magnitude. Instead, we find in Table 34.1 that ijE is higher than unity even for intervals as long as 15 seconds. THEORY OF ALTERNATING LIGHT EFFECTS 1443 The results of McAlister (fig. 34.3) are somewhat less extreme: the iiE values found by him are < 1 at 15 seconds, and practically equal to 1 at 5 seconds; but they reach 1.2 at 0.5 second and 1.6 at 0.1 second — periods which are still too long to be effectively occupied by the Emerson-Arnold reaction. Briggs (1941), too, found Ije = 1.6 for t — 0.6 second, a value even somewhat higher than Warburg's values in Table 34.1. Weller and Franck (1941) noted the need to explain the favorable effect of dark periods of the order of 1-10 seconds, and proceeded to repeat Warburg's experiments under a variety of conditions. In some cases, the Iie values remained below unity until t reached the order of magnitude of the Emerson- Arnold period; but in others, they exceeded 1 even at much longer intervals. Weller and Franck suggested that this "premature" rise of ijE occurs when a second catalytic reaction of photo- synthesis becomes rate-limiting, and thus influences the phenomena of intermittency. According to Franck and Herzfeld, the stabilizing cata- lyst, Eb, is usually limiting in strong continuous light, and in the presence of abundant carbon dioxide. However, under certain conditions, the limiting influence may pass partially or completely to other factors, par- ticularly those associated with the carbon dioxide supply. When the carbon dioxide concentration is low, or the diffusion path offers high resistance (as in the case of closed stomata), the limiting process may be the diffusion of carbon dioxide to the chloroplasts; in this case, dark intervals can be utilized for the re-establishment of the carboxylation equilibrium by diffusion (as this was first suggested by Willstatter and Stoll in the discussion of the results of Brown and Escombe). The length of the dark period required for this purpose must depend on specific conditions. Intermittency effects caused by slow diffusion are, however, unlikely to occur in unicellular algae suspended in buffer solutions (which were used in the Warburg experiments) and should be absent in the carbon dioxide- saturated state (since the diffusion supply can always be improved by an increase in the external concentration of carbon dioxide). However, a car- bon dioxide supply limitation of a different nature may occur even under these conditions if the quantity of the available carboxylating enzyme, Ea, is insufficient. In this case, the maximum rate of supply of carbon dioxide to the acceptor — and thus also the maximum over-all rate of photosynthesis — cannot be improved by an increase in the external carbon dioxide concentra- tion (as discussed in chapter 27, page 917) . If, because of E a deficiency, car- boxylation becomes rate-limiting in strong continuous light, dark intervals can be utilized for recarboxylation of the "denuded" acceptor. The time re- quired depends not merely on the action period of Ea, but also on its concen- tration. This difference from the case of limitation by a deficiency of Eb is caused by the fact that Ea acts on a stable substrate (carbon dioxide). 1444 PHOTOSNYTHESIS IN INTERMITTENT LIGHT CHAP. 34 and therefore can be used repeatedly during a single dark period (while Eb is supposed to act on unstable intermediates and is therefore available only once in each dark interval). Because of this difference, the frequency of alternations that gives the most favorable intermittency factor under the conditions of Ea limitation is not equal to one ''action period" of this catalyst, but is determined in a more complex way by the time that the available quantity of Ea requires to saturate the acceptor with carbon dioxide to such an extent as to give the maximum possible yield of photo- synthesis during the subsequent hght period. Franck has concluded, from the "pick-up" experiments described in chapter 8 (page 206, c/. also figs. 10 and 11 in chap. 33), that the time required to carboxylate all acceptor (once it has been totally decarboxylated) is about 20 seconds. This then must be the upper limit for the optimum frequency of alternations. (There certainly can be no advantage in extending the dark periods beyond the time required for complete re- carboxylation; while the simultaneous extension of light periods is disadvantageous since in longer flashes the A-C02 complex is more completely decarboxylated, thus caus-, ing an approach to the conditions of steady illumination and decreasing the favorable intermittency effect.) To obtain a lower limit of the most favorable frequency of alternations, one has to know the rate of processes by which the complex A-C02 is de- carboxylated in light. We presented in chapter 8 (page 167) and chapter 29 (p. 1086) evidence that led Franck to assume that each absorption act, whether it contributes to photosynthesis or not, may cause the decarboxylation of an acceptor molecule (because back reactions can result in a decomposition of A-C02 into free acceptor and carbon dioxide). If this is so (some difficulties of this hypothesis were mentioned in chap. 33), the rate of decarboxylation in light is proportional to light intensity, even in the satura- tion range (at least, until we come into the intensity region where the yield of fluorescence increases, as described in chap. 28B, thus revealing a decrease in the rate of the primary photochemical process). From the frequency of absorption acts (estimated on page 838) and the concentration of the acceptor molecules (estimated on page 204), we can deduce that, in light of approximately 10,000 lux, the velocity constant of the photo- chemical decarboxylation of A-COz complexes (partly by reduction and partly by dis- sociation caused by back reactions) is of the order of l/sec, and in light of 100,000 lux, of the order of 10/sec. 0.1 to 1 second must then represent the lower limit for favorable intermittency effects (i/E > 1 ) in the case of Ea limitation. Combined with the above- mentioned upper limit (20 seconds), these estimates define a region that in fact roughly corresponds to the range in which intermittency factors larger than 1 have been found by Warburg, McAlister, Briggs and Weller and Franck. This lends support to Franck's attribution of this anomaly to an inadequate quantity of the carboxylating catalyst, Ea. This interpretation implies that, in plants that show the anomaly, the rate-limiting proc- ess in strong continuous light also is the "preliminary" carboxylation reaction, rather than the "stabilizing" reaction catalyzed by Eb. This should be recognizable, e. g., by a greater sensitivity of the maximum rate to cyanide (c/. chapter 12, sect. Al). According to Weller and Franck, Hydrangea leaves generally behave as if they were deficient in enzyme Ea. It may be useful to point out that in the case of Ea limitation, the maximum obtainable value of Ije still is 2 (ijt < 1). Under no conditions can the average yield of photosynthesis in alternating light be higher than THEORY OF ALTERNATING LIGHT EFFECTS 1445 the rate of carboxylation; and the average rate of carboxylation in inter- mittent hght can at best approach (but never exceed) the rate of carboxyhi- tion in continuous hght. (Ea cannot work more efficiently in the dark than it does in hght, when practically all acceptor is maintained in the de- carboxylated state by intense photosynthesis.) Briggs (1941) also has discussed the efficient utilization for photosynthesis of dark intervals of the order of several seconds (in addition to those of the order of 10"^ second), and suggested two catalytieal components: one with a concentration approximately equal to that of chlorophyll, and a relatively long working period (of the order of several seconds) corresponding to the concentration of the carbon dioxide acceptor, A, and the worldng time of the carboxylating catalyst, Ea, in our hypothesis; and one with a con- centration about 500 times smaller and a working period of the order of 0.01 second (corresponding to concentration and working time of catalyst Eb in Franck's picture). Briggs suggested that the second catalyst is an intermediate between chlorophyll and the carbon dioxide acceptor complex {cf. the position of the system HX/X in some of our schemes, e. g., scheme 7.1 in Vol. I). However, the identification of the hmiting catalyst, Eb, with an intermediate oxidation-reduction system in this position is improbable, because of fluorescence phenomena {cf. chapter 28, part B). In part B of this chapter, when discussing more recent experiments with flashing light, we will again find evidence of the favorable effect of dark in- tervals of the order of 0.1-1 second on light energy utilization, and discuss several new attempts to interpret these results. It is particularly notable that Gilmour et at. (cf. section B7) found these effects also in the Hill reac- tion of chloroplast f rag-men ts, in M^hich carbon dioxide takes no part at all ; this seems to indicate that if two (or more) enzymatic reactions (of the type discussed by Franck and Weller, and by Briggs) are responsible for these complexities of induction and intermittency effects, both of them belong (or can belong, if more than two reactions are involved) to the part of the photosynthetic reaction sequence concerned with the photochemical oxidation of water and liberation of oxygen. Tamiya has attempted to show that all intermittent light results can be explained by a single enzy- matic reaction with kinetic characteristics different from those postulated by Franck and Herzfeld (on the basis of the data of Emerson and Arnold) ; for a discussion of his suggestion, see section B6 below. We see that, in all cases of catalytic yield limitation, the question asked at the beginning of ttiis chapter : whether an increase in photosynthesis can be achieved by regular interruption of illumination, e. g., by means of a rotating sector, must be answered in the negative. The factor ijt never ex- ceeds unity, and the factor irs never exceeds 2. (We have shown this foi alternaling light; we will see in part B that the same is true for the factor ill in flashing light, but that the factor ij^ may acquire, in such light, values much higher than 2.) Whether the same is true when the poorly understood phenomena of "injury" and "fatigue" come into play, is uncer- 1446 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 tain. It is conceivable, for example, that plants may produce more or- ganic matter in 12 hours of illumination followed by 12 hours of dark rest than they would in a uniform 24 hour illumination of the same intensity. The same may happen when photoxidation phenomena lead to a gradual inactivation of the photosynthetic apparatus (as may occur in an at- mosphere poor in carbon dioxide, or in light of excessive intensity, or in the ► irr < 2 Light Fig. 34.5. The three intermittency factors (for equal light and dark periods). Intensity and duration of illumination is shown by black areas; photosynthesis by shaded area. presence of excess oxygen; cf. chapter 19). In this case, the inhibition of photosynthesis has been found to grow autocatalytically, and it may be possible to prevent it by inserting dark recovery periods at appropriate intervals. Early in this chapter we also asked another question — whether the energy conversion yield can be improved by intermittency (i. e., whether the factor ist can become larger than unity). As far as enzymatic limita- tion in alternating light is concerned, this question, too, must be answered with a denial. To prove this we refer to equation (34.2). In weak hght (when the light curve is linear), /3 = 2 and in = 0.5; thus ist = 1- INTERMITTENCY FACTOR IN FLASHING LIGHT 1447 In saturating light, /? = 1, while i^ < 1, so that 4, is < 1. Thus, lEt can only vary between 0.5 and 1 . We thus conclude that, if a certain amount of light energy is available to be spent within a certain mterval of time, the best utilization of this energy for photosynthesis can be achieved by spreading it uniformly over the whole available period, rather than by using it in flashes. At least, this must be the case ^vith ravid alternations of darkness and light. In the mediMm range of alternations (one every minute) the same is obviously true (because induction losses are the main intermittency effect in this range). Only with slow alternations (e. g., one in several hours, or days) may there be a chance of obtaining Iei factors higher than unity. These conclusions apply to the continuous or intermittent illumination of the same plant, or the same cell in a suspension. Alternation can, how- ever unprove the utilization of a continuous, uniform light flux (such as that'from the sun) if the intervals in which one batch of cells completes its dark reactions are used to expose to the same flax another batch, e. g., by replacing the cells in the upper layer of an algal suspension with new ones at a suitable rate. As part of the plans for large-scale growing of unicellu- lar algae (for food, fodder or fuel), studies have been made of the possibility of using a turbulent flow of algal suspensions to create such favorable mter- mittencies of exposure; we will return to this topic on p. 1477. The relationships between the three above-used intermittency factors are represented graphically in figure 34.5, which needs no further explanation. B. Flashing Light* 1. Intermittency Factor in Flashing Light Separate variation of dark and hght periods enables one to find out more about the mechanism of the induction period and of the reactions going on during the dark intervals than can be derived from experiments m alternating hght with equal periods of darkness and light. We have en- countered the first example of their usefulness in describing how Warburg first determined the duration of the induction period and of the dark reac- tion that prepares it. Emerson and Arnold (1932) made an important contribution to the study of photosynthesis by introducing flashing hght, meaning by this light with very short illumination periods, separated by longer dark intervals. This method permits separation of the primary photochemical process (together with such rapid nonphotochemical trans- formations as are practically instantaneous, i. e., are completed withm <10-3 second) from all those chemical or physical components of photo- synthesis that require measurable time for completion. Two experi- mental methods have been used to obtain the required intense light flashes: * Bibliography, page 1483. 1448 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 condensor discharges through gas-filled tubes (Emerson and Arnold 1932), and appropriately shaped rotating sectors (in combination wth a constant light source of high intensity, e. g., high-pressure mercury arcs) (Pratt and Trelease 1938, Weller and Franck 1941, Tamiya and Chiba 1949). In the discharge tube technique, the flashes are shorter (of the order of 10"^ sec- ond) and more intense, but the rotating disc technique offers easier access to a wide range of integrated flash energies. In both tyi)es of experiments, the yield — determined by the usual tech- nique, such as manometric measurement — is the average yield for a large rmmber of identical flashes. Franck, Pringsheim and Lad (1945) were able to determine, by means of the phosphorescence-quenching technique, the oxygen production of a single flash. They used for this purpose photo- flash bulbs, which produce much stronger flashes than are obtainable by the other methods. The flashes last for about 0.04 second, and reach a peak intensity of 1.4 million lumen. It is easy to show that in flashing light (as in alternating light) the factor in must always be < 1 if the rate limitation is caused by catalyst deficiency. (The maximum yield in flashing light is reached when the rate-limiting catalyst is practically fully occupied for the whole duration of the dark intervals; it is then equal to the maximum jdeld in continuous light.) The factor irs (which, in alternating light, had a limiting value of 2) can reach much higher values in flashing light (c/. equation 34.1). For example, if in — 1, and t*/ta = 10-=^ {e. g., t* = 10"^ second and.^^ = IQ-^ second), iiE is approximately 1000. However, these high values of ijs are without real significance, for when the light periods are shorter than the "Emerson- Arnold period" the yield per flash depends only on the total energy of the flash (the time integral of its intensity) and not on these two factors separately. In other words, plants do not distinguish between a flash of a certain intensity that lasts for 10""* second, and a flash of tenfold intensity that lasts for only 10~^ second. Under these conditions, the two intermittency factors for equal intensity, in and ijE, lose their importance, and the only factor with which we need to be concerned is ist- These considerations appear not to apply to intense flashes lasting for several milliseconds or longer; in this case, the yield becomes a function of the duration of the flash, and is affected by dark periods of the order of 0.1 or 1 second. We mentioned this complicating effect in discussing the "alternating light" phenomena in part A, and will return to it below in dis- cussing the experiments and theoretical considerations of Weller and Franck, Tamiya, and Gilmour et at. We will also see that Burk and co- workers denied the validity of the photochemical "reciprocity law" in photosynthesis even for the very brief condensor discharge flashes. INTERMITTENCY FACTOR IN FLASHING LIGHT 1449 For low integrated light intensities, i. e., low values of the product (energy of a single flash X number of flashes per unit time), the factor iet must be unity (since in this case, all the absorbed light quanta can be utilized for photosynthesis, with the highest possible quantum yield, whether they are supplied continuously or in flashes). The factor t^t will decline below unity when the flash energy becomes so high that more intermediates are produced in a single flash than the available catalyst Eb can handle in one batch (since, according to our assumptions, the frac- tion of the intermediates that finds no free catalyst is lost by back reac- tions). Thus, if one plots the rate of photosynthesis in flashing light with 2 3 4 5 6 AVERAGE INTENSITY, f Fig. 34.6. Light curves for flashing light (after Weller and Franck 1941 ). The number of light flashes per second, 7, is the parameter in this set of curves. different frequencies of flashes against the integrated intensity (or, what is essentially the same, the average incident energy per unit time), one expects to obtain a picture of the type of figure 34.6, based on actual experimental results of Weller and Franck (1941). In this figure, the ratjo of yields in flashing fight and in continuous light with the same value of /, is the factor Iei. The figure confirms that this factor is unity for low values of /. It declines below unity, first for widely spaced flashes (where, in order to achieve the same average intensity of illumination, one has to use stronger flashes), and later for flashes that are more closely spaced. When the dark intervals, ta, approach zero, i. e., when the flash frequency becomes very high, the light curve for flashing light must become identical with that for continuous light. Figure 34.6 confirms that the yield in flashing light never exceeds that in continuous fight of the same integrated intensity; i. e., that the factor ist never is larger than unity. A theoretical proof of this rule was given in part A for alternating light; the same proof can easily be generalized to include intermittent light with any ratio of t* and ta. This proof is illustrated by figure 34.7. The shaded areas represent 1450 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 amounts of limiting catalyst actively engaged in photosynthesis (A) in saturating continuous light, and {B) in flashing light of the same integrated intensity, for two different flash frequencies (5i, B2). The average yield of photosynthesis, which is proportional to the sum of the shaded areas, is highest in A and lowest in B2. Instead of plotting the average yield of photosynthesis in flashing light against the average light intensity (as in fig. 34.6), one can plot the yield per flash, P against the energy of a single flash, I. Such "flash saturation curves" (fig. 34.8) can be constructed from the same data used in figure TIME Fig. 34.7. Photosynthesis in continuous Ught and flashing hght of the same inte- grated intensity, /, for two different frequencies. Flashes are "saturating," i. e., each produces enough intermediates to occupy all molecules of Eb. Photosynthesis begins at the saturation level (which is proportional to Eb ), and declines exponentially during each dark interval. Black rectangles, Ught. Shaded areas, photosynthesis (i. e., prog- ress of the limiting dark reaction). 34.6. At first, P is proportional to I and independent of r ( = t* + ta); this increase continues as long as practically all intermediates, formed during the flash, can be utilized by catalyst Eb during the subsequent dark period. With increasing flash energy, however, Eb, even if it is fully occupied dur- ing the entire dark interval, ceases to be able to transform all intermediates formed in the flash; therefore the flash saturation curves bend, one after another, and approach the horizontal. The saturation yield is propor- tional to the length of the dark intervals as long as the latter are very short compared with the working period, t^, of Eb, and thus can be fully utilized for the dark reaction catalyzed by this agent. When ta becomes commen- surate with ^B, full utilization of the dark intervals becomes impossible INTERMITTENCY FACTOR IN FLASHING LIGHT 1451 (because of the exponential decline of the yield; cf. fig. 34.7). Finally, after the intervals had become so long that practically all intermediates associated with Eb after the flash are stabilized before the next flash {to. 2 minutes), whether it is exposed to continuous light, or to 40 neon discharge flashes per second; in both cases the steady state was reached only after about 10 minutes of exposure. 2. Maximum Flash Yield Figure 34.10 shows the yield per flash as a function of the length of the dark intervals, according to the measurements of Emerson and Arnold £.r.j - y 2.0 ^J^^^^^^^ 1.5 - t 1.0 1 o Chlorella pyrenoidosa • Chlorella vulgaris 0.5 1 X 10-^ 1 1 1 1 ^ 5 10 15 20 25 DARK TIME (t^), sec. Fig. 34.10. Flash yield as function of dark time for Chlorella (after Emerson and Arnold 1932). 90 klux 31 klux 0.02 0.04 006 0.08 0.10 0.12 0.14 DARK TIME, sec Fig. 34.11. Yield per flash for two dif- ferent light intensities given for 4.5 msec, (after Weller and Franck 1941). (1932). The shajie of these curves is exponential, and they can be inter- preted as revealing a reaction of the first order (e. g., the monomolecular transformation of the complex I-Eb, where I stands for intermediate such as { HCO2 1 in chapter 7, or EbHC02 in scheme 28.11). If, at the beginning of the dark interval, the quantity of this complex is [I-Eb]o, the kinetic MAXIMUM FLASH YIELD 1453 law of monomolecular decomposition (identical with the law of radioactive disintegration) determines that, after time t, the residue of unchanged I'Eb complexes will be: (34.3) [I-Eb1 = [I-EB]oe-^A< where k^A i« the monomolecular velocity constant of the Emerson-Arnold reaction. The yield per flash, which we assume to be equal to the amount of I-Eb transformed during the dark interval, is: (34.4) P = [I-Eb]o - [I-Eb] = [I-Eb]o (1 - e-k^Atd) One half the maximum yield is obtained when e~''^'' = 3^, or: (34.5) Iea, this rate can also be expressed as NEb, where N is the number of flashes per second (sixteen in our example). In the poisoned state, on the other hand, the rate Umitation is given by the rate of supply of the substrate by the residual carboxylase, kEl. This applies to continuous light as well as to regular flashes. (Of course, the final products are liberated in the second case in bursts following each flash, as indicated by the peaked curves; but the total volume of these bursts is limited by the intake of carbon di- oxide, which goes on uniformly.) With grouped flashes, however, the EA-liniitation remains practically ineffective even in the inhibited state (since, in this case, the "ceihng" A;Ea = 0.43 mm.Vmin. is considerably higher than the [Eb ]-determined rate, 0.33 mm.Vmin.). The experiments mth flash groups provide the most convincing argument against the hy- pothesis of "substrate limitation" as an alternative to "hmitation by a finish- ing catalyst." 4. Influence of Temperature, Carbon Dioxide Concentration, Narcotics and Ultraviolet Light on Flash Yield Emerson and Arnold (1932^ found that decrease in temperature, though increasing required dark time, leaves P"'''" unaffected; fig. 34.14 shows I < a: LU Q. Q _i UJ >- llJ > I- < _l a: 600 500 400- 300 200 100 20 7° C^ ^-^:::::r— / ^/ y/^l°Z. ^ 1 1 1 1 1 1 1 QI6 ^ Q02 0.04 006 008 QIO 012 014 DARK TIME, sec. Fig. 34.14. Yield per flash curves for low and high temperature (after Weller and Franck 1941). confirmation of this conclusion by Weller and Franck (1941). This result appears natural, since lower temperature leaves unaffected both the quan- tity of the intermediates available after a flash, and the quantity of the INFLUENCE OF VARIOUS FACTORS ON FLASH YIELD 1461 catalyst available for their transformation; it only slows dovm the rate of this transformation. The maximum flash yield becomes dependent on temperature when, at the higher temperature, the duration of the flash ceases to be short com- pared with the working period of Eb (Franck, Pringsheim and Lad 1945). At 4.7° C, the shape of the curve is not as close to exponential as at 20.7°, but indicates a linear beginning. This change is similar to that caused UJ Z3 oc UJ Q. W (/) UJ X H Z >- W) o (- o X a. 0 20 40 60 80 CARBON DIOXIDE CONCENTRATION, m./l. Fig. 34.15. Effect of [CO2] on photo- synthesis in continuous and flashing light (24 flashes/sec.) (after Emerson and Arnold 1932). 10 12 2 4 6 8 DARK TIME, sec. Fig. 34.16. Course of dark reaction at two different concentrations of CO2 (after Emerson and Arnold 1932). by cyanide poisoning (fig. 34.12), and could therefore be quoted in sup- port of Franck 's hypothesis that, at low temperature, the cyanide sensi- tive carboxylation reaction (catalyzed by Ea) replaces the cyanide-resist- ant finishing reaction (catalyzed by Eb) as the main yield-limiting factor (cf. chapter 31). The effect of changes in the concentration of carbon dioxide on photo- synthesis in flashing light offers an interesting problem. As shown by figure 34.15, Emerson and Arnold (1932) found the "carbon dioxide curves" in flashing light to be similar to those in continuous light. However, half saturation occurred in flashing light (ta = 0.04 second) considerably earlier than in continuous light. This is natural, since carbon dioxide saturation must ensue as soon as the carbon dioxide supply reaction — which proceeds at a uniform rate throughout the dark periods — supplies enough material to cover the A • CO2 consumption of saturating flashes, which is consider- ably lower than that of saturating continuous light. Less easy to explain is figiu-e 34.16, which shows the flash yield in relation to the length of 1462 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 dark periods, for two different concentrations of carbon dioxide. The two curves appear to retain an approximately constant ratio at all ta values and not to approach a common limit — as we would expect if the maximum yield per flash were equal (or equivalent) to the available amount of the finishing catalyst, Eb. By using sufficiently intense flashes, it should be possible to produce enough intermediates to put all molecules of Eb to work, even if many of the absorbed quanta were lost because of the scarcity of carbon dioxide-acceptor complexes. The flashes used in the experiments of Emerson and Arnold were just strong enough to achieve flash saturation in the presence of abundant carbon dioxide ; perhaps, no true saturation was obtained in the carbon dioxide-deficient medium (despite the evidence of the last two points on the curve) . With the usual flashing-light technique (involving a long sequence of flashes) it is difficult to obtain flashes much more intense than those used by Emerson and Arnold, to see whether the two curves in figure 34.16 would converge at the higher intensities. Fig. 34.17. Induction in oxygen liberation by flashes after anaero- bic incubation (after Franck, Pringsheim, and Lad 1945). We mentioned, however, that much higher intensities of single flashes can be produced by exploding flash bulbs, as was done in the phosphores- cence measurements of Franck, Pringsheim and Lad (1945). Two points, however, have to be cleared before this method can be applied to the solu- tion cf our problem. In the first place, the phosphorescence technique re- quires absence of oxygen; under these conditions, the saturation yield in steady light may be only 1% of the normal, aerobic value. Is then the flash yield in the absence of air at all comparable with that obtained aero- bically? The answer is that it is of the same order of magnitude, but may be smaller by a factor of five or even ten (corresponding to Chlo/P = 10,000- 20,000, instead of the usual 2000). This shows that the poison that ac- cumulates during anaerobic incubation has much less effect on maximum yield per flash than on maximum yield in continuous light. This is con- firmed by the observation that the maximum flash yield is independent of the density of the algal suspension (which is decisive for the extent of INFLUENCE OF VARIOUS FACTORS ON FLASH YIELD 1463 inhibition in continuous light, cf. chap. 33, sect. A6). In describing an- aerobic induction we noted its twofold character: an extremely strong effect, attributable to the production of an acid, diffusible, metabohc poison, and a weaker effect, apparently due to enhanced production of the (nondiffusible, nonneutralizable) "internal narcotic" — perhaps the same that is also responsible for the induction losses in the aerobic state. Of UJ m 3 6 5 ;ligh Chi - '/ / - 1 1 1 1 1 1 3.0r 10 12 DARK TIME, sec Fig. 34.18. Flash yield as function of dark time for cells with low chlorophyll concentrations (circles) and high chloro- phyll concentration (dots). The ratio of the two concentrations is about 4:1; the saturation levels correspond to about 4 mm. 3 O2 and 1 mm.^ O2 per mm.^ cells, respectively (after Emerson and Ai-nold 1932). 2.5- 2.0 1.5 1.0 0.5 0.0034% urethone , IQ- 0 2 4 6 8 10 12 DARK TIME, sec. Fig. 34.19. Effect of phenylurethan on course of dark reaction (after Emer- son and Arnold 1932). these two aspects of anaerobic inliibition, only the second one appears to affect the oxygen production by single flashes. When several such flashes are produced in succession, the flash yield shows the typical short induction phenomenon, with or without a second minimum (fig. 34.17). The absolute value of the maximum flash yield under anaerobic condi- tions, although smaller than in air, is not so much smaller as to make the application of the anaerobic method to the problem of [CO2] influence on flash yield entirely unreasonable. The second point to be considered is the duration of the flash bulb ex- plosions. As mentioned before, it is 0.04 second longer than the Emer- son-Arnold period at 20° C. Therefore, the maximum yield obtain- able per flash could be somewhat higher than the amount produced by a single action of the available Eb (since the catalyst can act more than 1464 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 once in each flash). The relation is reversed at 0° where the Emerson- Arnold period is considerably longer than the duration of the flash. The experiments in which Eb is to be measured by the maximum flash yield with the help of flash bulbs must therefore be performed at low temperature. At 0° C, the maximum anaerobic flash yield of a suspension of Scenedes- mus was found to be 2.3 X 10~^ ml. in a carrier gas (nitrogen) without ex- tra carbon dioxide (beyond that produced by respiration and fermentation of the algae), and 3 X 10~* ml. in the presence of added carbon dioxide. The difference is so small as to suggest that (as expected in Franck's theory) the concentration of carbon dioxide has no effect on the maximum flash yield. (More precise confirmation obviously remains desirable.) At room temperature, the flash yield as determined by this method was not only somewhat higher than at 0° C. (this was mentioned before), but also was much more dependent on the carbon dioxide supply. Both observations are explicable by reference to the fact that the duration of the flash at 20° is longer than the working period of the yield-limiting catalyst. Emerson and Arnold (1932) found that the maximum flash yield is af- fected by chlorophyll deficiency (fig. 34.18). This dependence is another aspect of the problem discussed in chapter 32, section 2, in connection with the similar effect of fChl] on P"^^^- in continuous light. We concluded there — from the fact that the parallelism of [Chi] and p™"""- was absent in many plants, and not ahvays present even in Chlorella— that no direct relation exists between these two magnitudes, but that a depression of P'"*"' occurs when the decrease of [Chi] is brought about by treatment (such as iron starvation) that also reduces the concentration of other catalytic components of the photosynthetic apparatus, such as the usually rate-limiting catalj^st, Eb- The same hypothesis could explain the paral- lelism between [Chi] and the maximum flash yield, p™*^-. The effect of narcotics on p""^""-^ also noted by Emerson and Arnold (1932), and illustrated by figure 34.19, could be similar to that of carbon dioxide deficiency. By enveloping the sensitizer molecules, narcotics could cause a dissipation of the energy of a considerable proportion of the absorbed light quanta. In this case, a higher flash energy would be re- quired to produce flash saturation in the presence of narcotics, but the saturation level would be the same as in the nonpoisoned state. This is not clearly confirmed by figure 34.19. Explanation could be the same as suggested in the case of [CO2] — failure to reach true saturation in the inhibited state; but it is also possible that narcotics inhibit not only the sensitizer, but also the limiting catalyst, Eb (since their influence is not as specific as that of the "catalyst poisons"). In this case, the flash saturation yield will be affected qualitatively in the same way as the yield of non- saturating flashes. We recall that in chapter 12, part B, we made the same FLASH YIELD IN HEAVY WATER 1465 suggestion to account for the effect of narcotics on the rate of photosynthe- sis in strong continuous Hght. Brilliant and Krupinikova (1952^) studied the effects of dehydration on the yield of photosynthesis in constant and intermittent light. The observations of Arnold (1933) on the effect of ultraviolet light on photosynthesis in flashing light offer a similar problem. As described in chapter 13 (page 345), each quantum of ultraviolet light (253.6 m/x) ap- pears to "knock out" one catalytic "center" (these centers are not chloro- phyll molecules, since chlorophyll remains intact, but their concentration is approximately equivalent to that of the green pigment) . Arnold found that inhibition is proportionately the same in flashing and in continuous light, and that the full yield per flash cannot be restored by extending the dark intervals. The observation that the concentration of the sensitive cata- lytic centers is about equal to that of chlorophyll makes one think of the carbon dioxide acceptor, A, as the ultraviolet-sensitive target. If this is so, the inhibiting effect of ultraviolet light on the flash yield must be analogous to the effect of low carbon dioxide concentration. The effect on the flash yield of variations in [CO2], as well as that of the narcotics, could be explained without much difficulty on the basis of the theory of "substrate limita- tion" (c/. Arnold 1935) by assuming that at low [CO2I, or in the presence of narcotics, only a fraction of the normal number of substrate molecules are available for immediate reduction. Another possibility for solving this type of kinetic difficulty is by assum ing a morphological or kinetic "photosjmthetic unit" (Gaffron and Wohl" or "chlorophyll ensemble" (Tamiya) in which a number (say, 400-2000 of chlorophyll molecules are associated with a single "reduction center" or a single molecule of an enzyme, in such a way that the intermediate photo- products formed at these chlorophyll molecules can be further transformed only in this one reduction center or by this one enzyme molecule, and react back if they find this center or enzyme molecule occupied. (For a discus- sion of such kinetic mechanisms, see, e. g., Rabinowitch 1951, and Gilmour et al. 1953.) 5. Flash Yield in Heavy Water Pratt and Trelease (1938) found that substitution of deuterium oxide for ordinary water causes an extension of the Emerson- Arnold period to more than twice the original value, without affecting the maximum yield per flash (figure 34.20). The P = f{ta) curve is changed in a way reminiscent of the effect of cyanide — it is linear almost to the point of saturation. However, since the number of experimental points is small, this conclusion is not quite certain. 1466 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 This result can be interpreted in several ways. Pratt and Trelease,. considered two possibilities: deuterium oxide acting as a poison, and deu- terium oxide acting as a partner in the Emerson- Arnold reaction (with a velocity smaller than that of ordinary water). They considered the sec- ond alternative the more probable. Weller and Franck agreed, and said that the retarding influence of deuterium oxide may indicate that the reaction catalyzed by Eb involves the transfer of hydrogen atoms. (We have repeatedly suggested that it may be a dismutation.) If it should be confirmed that the shape of the P = f{td) curve in heavy water is similar to that observed in the presence of cyanide, this explanation will have to be 0.03 0.05 0.07 DARK PERIOD, sec. 0.09 Fig. 34.20. Yield in flashing light as function of dark period in H2O and D2O (after Pratt and Trelease 1938). changed. It is not necessary to revert to the other alternative of Pratt and Trelease and to assume that heavy water acts as a poison; the more plau- sible explanation is that the participation of deuterium oxide retards an- other dark reaction — not the usually limiting Emerson-Arnold reaction — to such an extent that it becomes rate-limiting. This reaction cannot be the carboxylation (as in the case of cyanide poisoning) since the latter in- volves no hydrogen, but it may be, e.g., a preparatory reaction on the "oxida- tion side" of the primary photochemical process. We will see (section 8) that a reaction of the latter kind actually is rate-limiting in so-called "adapted algae," where hydrogen is supplied by a substitute donor, in- stead of by water. NEW PLASHING LIGHT EXPERIMENTS 1467 6. Flashing Light Experiments Calling for Revision or Supplementation of the Emerson-Arnold-Franck Mechanism The above-described results of the flashing light experiments of War- burg, Emerson and Arnold, Kohn, Franck and Weller, Rieke and Gaffron, Clendenning and Ehrmantraut, and Ehrmantraut and Rabinowitch seemed to add up to a consistent picture, and to provide one of the few firm factual bases for the kinetic analysis of photosynthesis. It seemed to be well es- tablished that the normal maximum rate of photosynthesis in strong steady 25° P'"'"' = 7.3 X 10' — O 15° P""" = 5.3 X 10" 1° P'"""- = 3.7 X 10" 0.3 0.4 Vi sec. Fig. 34.21. Yield of flashes of saturating energy in Chlorella ellipsoidea as function of dark interval (after Tamiya and Chiba 1949). The yields are expressed in moles O2 per gram dry weight per flash; to express them in moles O2 per mole chlorophyll, multi- ply the ordinates by about 2 X 10*. light (Pmax! equal to about one molecule oxygen per molecule of chlorophyll every 20-30 seconds), can be factorized into a concentration factor of the order of Chlo/2000 (or Chlon/2000, with n a small number, perhaps 4 or 8) and a monomolecular rate constant of the order of 100 sec.~^ at 20° C. These two separate constants accounted for flash saturation as function of flash energy, and for flash saturation as function of the duration of dark intervals, respectively. A remaining difficulty was the enhancing effect on the (apparent, or true) light energy utilization (1^1 > 1) of dark inter- vals of the order of 1 second, or longer. Tentatively, this favorable effect of relatively low frequencies of alternation could be related to CO2 supply limitations, or to the "bursts" and "gulps" described in chapter 33 (which affect most strongly the gas exchange in the first seconds of exposure) . 1468 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 It could be hoped that a satisfactory explanation of these additional features of induction will also provide an interpretation of the additional intermittency effects, supplementing rather than replacing the original one. One could also hope that cell cultures showing no CO2 supply limi- tations, and Uttle or no "transients" would also show little or no compUcation of the Emerson-Arnold-Frank mechanism of flash saturation. o E Q. 4 25° P""" -. 7.3 xlO" 15° P"""" = 5 3 X 10" 70 pmax.^ 3-7^ iQ- 200 400 600 f, lux sec. 800 1000 1200 Fig. 34.22. Yield of flashes (with saturating dark intervals) in Chlorella ellipsoidea as function of flash energy (after Tamiya and Chiba 1949); for units, see preceding figure. (Closer approach to the extrapolated saturation level than shown in this figure is indicated by data in text of paper.) We now have to discuss several flashing hght studies which seem to call for a more thorough revision of the concepts derived from the above-enum- erated earlier observations. The most important of them is an investiga- tion by Tamiya and Chiba (1949). They used Chlorella ellipsoidea, a species similar to Chlorella pyrenoidosa of Warburg and Emerson. The cells were grown for 10 days in Knop's solution at 2-3 klux, and then incu- bated in the dark for 3 days. Its maximum steady photosynthesis at 25° C. in carbonate buffers was about 1.5 Aimole/sec. per gram dry weight (0.77 fjivaole at 15° and 0.38 jumole at 7° C). If one assumes a chlorophyll content of 5% (cf. table 25.1), this corresponds to an assimilation time of 33 sec. (at 25° C.) typical of normal shade cells {cf. table 28. V). Tamiya and Chiba exposed these cells, in buffer No. 9, to flashes of 0.6 to 8 msec, duration, obtained by means of rotating discs, on which the image of a 750 watt incandescent lamp was focussed. The maximum illumina- NEW FLASHING LIGHT EXPERIMENTS 1469 tion at the bottom of the suspension vessel was 120 kliix. By varying flash duration and lamp voltage, flash energies from 10 to 1000 lux sec. could be obtained. (According to p. 838, this should mean from 1.4 X 10^^ to 1.4 X 10'^ incident quanta per cm.^ per flash, or from one absorption act per about 500, to one absorption act per about 5, chlorophyll molecules; these estimates could be too high because the copper sulfate filter must have affected the spectral composition and thus also the average absorption of the light reaching the suspension.) - 2 o '■ e 1 = 21 lux sec. 0.6 - 0.4 - 0.2 02 0.4 15° 8 7° P= 0.5 X 10" 0.05 0 15 0.1 'rf, sec. Fig. 34.23. Oxygen jdeld per flash as function of dark intervals in Chlorella eilipsoidea for subsaturating flash energy (after Tamiya and Chiba 1949). The yield was measured manometrically in carbonate buffer No. 9, in 10 minute runs, with dark intervals from 0.015 to 0.60 sec, at 7°, 15° and 25° C. A suspension of 4 mg. algae covered an area of about 20 cm.'^ (about 10 ~* mole chlorophyll per cm.'^, indicating a probable absorption of <50% of incident Hght). The results differed from earlier data in three significant ways. (1) As shown in figure 34.21, the yield per flash increased with the dura- tion of dark intervals up to 0.2 sec. at 25° C, 0.3 sec. at 15° C. and about 0.5 sec. at 7° C. (for saturating flashes) — ten times longer than in table 39.11. (2) The absolute maximum flash yield at 25° C. was 7.3 X 10 ~* mole oxygen per gram dry weight ; with a chlorophyll concentration of about 5% this is equivalent to one molecule oxygen per 700 molecules chlorophyll — about three times the maximum flash yield observed in the earlier experi- 1470 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 ments. Fig. 34.22 shows the dependence of the yield on flash energy at saturating dark intervals; the "limiting yield" of Emerson and Arnold is indicated by a horizontal dashed line. (3) The maximum yield per saturating flash depended on temperature (compare fig. 34.22 with fig. 34.14). Tamiya and Chiba attributed the difference between their results and those of Emerson and Arnold to the use of higher flash energies, and sub- mitted that the light emission of condensor discharges, lasting only 10~^ second, was insufficient for saturation. They pointed out that their own results at low flash energies gave a picture similar to that of the earher observers (fig. 34.23), with a saturation level independent of temperature and reached after dark intervals of < 0.02 sec. at 25° C. and < 0.06 sec. at 7° C. However, the maximum yield in fig. 34.23 is only about one quar- ter that of Emerson and Arnold; with flash energies suflficient to equal the latter (about 100 lux sec), Tamiya's curves show a strong dependence of saturation on temperature. Furthermore, their explanation does not apply to the observations of Weller and Franck, who used flashes, produced by a 1000 watt Hg lamp and rotating sector, with energies up to 450 lux sec, and nevertheless found (in agreement with Emerson and Arnold) no dependence of the maximum flash yield on temperature (fig. 34.14). Tamiya (1949) suggested an interpretation of his experiments by a kinetic scheme in which the effect on the flash yield of back reactions in the primary photochemical apparatus was a function of temperature. This required no change in the mechanism of fight saturation postulated by Franck, but merely a change in the relative values of the several rate con- stants. Both Franck's and Tamiya's picture can be illustrated by scheme 28.11 (p. 1037) and the reaction sequence (28.41, a-e). (It is irrelevant for the kinetics whether the stabilizing reaction operates on the first reduc- tion product of carbon dioxide, AHCO2 in scheme 28.11, or the first oxida- tion product of water, AHO in the same scheme, or on some intermediate oxidation-reduction system.) Franck, in order to account for the inde- pendence of the maximum flash yield of temperature (observed by Emer- son and Arnold, and confirmed by Weller and Franck), suggested (in es- sence) that the relative values of k' (the rate constant of the back reaction in scheme 28.11), k^ (rate constant of the photoproduct-enzyme complex formation), and k[ (the rate constant of the transformation of this complex and reactivation of the enzyme Eb) are such that all products of the flash for which molecules of the enzyme Eb are available, are combined with this enzyme before they are lost by back reactions (A;' <^ keE\); but that by the time this first "batch" of photoproducts had been transformed, back reactions have taken care of all the remaining flash products, and no "second round" of the stabihzing reaction is possible {k' 0.1 sec. (needed to exceed the Emerson- Arnold maximum flash yield) a large part of the respiration intermediates, accumulated during the dark periods, is reduced in light, producing an oxygen burst and increasing the apparent yield of photosynthesis. In other words, one could suggest that light flashes efi"ectively inhibit respiration during the whole period of exposure to flashing illumination. However, no such effect was noticed by Brown (1953) in mass spectrographic study of respiration in light {cf. chapter 37D). Also, one does not see offhand why the same effect should not have been pro- duced also by instantaneous discharge flashes (after dark intervals of the same length). Refuge could be taken to the great variability which "transients" show in dependence on the metabolic history and nature of the cells ; but the effects of dark intervals of the order of one second on flash yield crop up somewhat too regularly for such an explana- tion. We can refer in this connection also to the findings of Gilmour et al. on chloroplast suspensions, and their kinetic interpretation in terms of a "reservoir" in which energy-rich photoproducts can be stored, to be re- leased for "finishing" after the flash. These authors suggested a mecha- nism of the filling of the reservoir which makes it ineffective in instantaneous flashes, but operative in flashes of the same integral energy but longer dura- tion {cf. section 7 below). An even more radical departure from the conclusions reached in the earlier flashing light experiments was suggested by Burk and coworkers (1951, 1952, 1953). They as- serted that not only Emerson and Arnold, but Tamiya and Chiba as well, have never even remotely approached flash saturation. The reaction mechanism postulated by Burk, Cornfield and Schwartz (1951, 1952) was in principle similar to that of Franck (or Tamiya): competition between a back reaction and a forward reaction for the primary photoproducts. Burk associated the competing back reaction with the "energy dismu- tation" mechanism. (This concept was first described in chapters 7 and 9, pp. 164 and 233, and later used by van der Veen, and by Burk and Warburg; the latter called it the "cyclic reaction," or the "one-quantum mechanism" of photosynthesis.) According to this concept, exothermal back reactions (such as reaction 28.41e on p. 1036, cf. scheme 28.11) are used in nature to store chemical energy for subsequent use as supplement to a quantum of light in the reduction of carbon dioxide. This special purpose of the back 1476 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 reaction does not affect the form of the kinetic equations, but only the assumptions concerning the relative values of the rate constants. Franck (and Tamiya) have assumed that the first forward reaction of the photoproduct (its association with the enzyme Eb) must be very fast compared to the back reaction (except when much of the enzyme Eb is occupied by the slow transformation of its complex with the photoproduct, with en- suing light saturation); this assumption was needed to permit effective utilization of quanta in weak light. Burk and Warburg, in their variant of the energy dismutation mechanism, postulated, to the contrary, that the back reaction must be at least twice as fast as the forward reaction (as needed to reduce the net quantum yield from 1 to the highest energetically possible value of about Va)- Burk denied the existence of a "bottleneck enzyme" other than chlorophyll (for a dis- cussion of this possibihty, see p. 1030); the maximum expected flash yield thus became Chlo/n, where n (^3) is the "energy dismutation factor", i. e., the number of quanta used for the net production of one molecule of oxygen. Burk thus reverted, in essence, to the early concept that in flash saturation each chlorophyll molecule must produce one molecule of oxygen. With n ^^ 3, this means a seven hundred times higher flash yield than obtained by Emerson and Arnold, or two hundred times greater than obtained by Tam- iya and Chiba. At first, Burk et al. (1951, 1952) implied that such high flash yields ac- tually are observable; a subsequent note (1953) indicated, however, that this was not the case. It was suggested that the failure to obtain the "theoretical" flash yields, P = Chlo/n, is caused by a "solarization"— the same phenomenon that reduces the rate of photosynthesis in excessively strong, steady light (cf. chapter 19, section A3). As mentioned before, Burk et al. thus rejected the concept that, with very brief flashes, the momentary intensity of illumination is irrelevant, and all that matters is the total energy of the flash. Instead, they postulated that, if the momentary intensity of illumination exceeds — even for a few microseconds — the intensity which produces solarization in steady light, the yield of photosynthesis declines in the same proportion as if this in- tensity were applied for an extended period of time. According to Burk et al. (1953), the threshold of solarization for Chlorella lies at about 50 klux; this means that the maximum flash yield obtainable with flashes of 10 ^sec. duration should be not much in excess of that corresponding to a flash energy of 0.5 lux sec. (which seems to be quite incompatible with observations), and flashes of 10 /^sec. should yield, at best, not much more than the amount of oxygen corresponding to a flash energy of 500 lux sec. (close to where Tamiya found flash saturation). This theory appears implausible (because "solarization," as described in chapter 19, is a cumulative rather than instantaneous response). It seems incapable of accounting quantitatively for the observations of Emerson and Arnold et al. ; and if it were the cor- rect explanation of Tamiya's observations, the latter should have noticed an "optimum" of flash energy. The theory appears as a reversion from Blackman's recognition of "limiting reactions" to the older concept of "cardinal points" (cf. chapter 26). A de- scription of the actual experiments of Burk and coworkers must be awaited before judg- ing whether they call for such drastic revision of our kinetic concepts. Intermittency factors E/j > 1 offer obvious inducements to use intermit- tent light in experiments on mass culturing of algae. Of course, with a light source of a given intensity ^ — c. g., the sun — no advantage could be ex- pected (on the basis of present kinetic knowledge) from the relation E/, > 1 if intermittency were achieved simply by blacking out the illumination for certain intervals (e. g., by means of rotating sectors) ; but if the same inter- NEW FLASHING LIGHT EXPERIMENTS 1477 ruption of illumination could be acliieved by placing another batch of algae in the path of the light while the first one is "digesting" the flash products in darkness, E/« > 1 would mean also E/b > 1, i. e., an increase in the utili- zation of solar energy (since, now, no hght energy will be wastefully ab- sorbed by black screens). If the Emerson-Arnold reaction were the only one determining the inter- mittency effect, the maximum improvement in the efficiency of hght utiliza- tion could be expected with hght periods lasting just long enough to excite — with the available hght intensity — one chlorophyll molecule out of between 250 and 2000 (to put all the enzyme Eb to work at the end of the flash), and dark periods lasting long enough to permit the Emerson-Arnold reac- tion to run to practical completion at the prevaihng temperature, but short enough to avoid induction losses afterwards. According to p. 838, in direct sunlight at noon ('-^85 klux), a directly exposed chlorophyll molecule absorbs a photosynthetically effective quantum about once every 0.08 second. Therefore, one molecule in 2000 will absorb a quantum in a flash of direct sunlight 40 ^sec. long, and one molecule in 250, in a flash of 300 fjLsec. duration. The dark period will have to last > 30 ^sec. (at 15-20° C.) to complete the Emerson-Ai'nold reaction. Finite optical density of a cell suspension will make the required length of the flash longer than 40- 300 /xsec; while the above-discussed (but not definitely understood) capac- ity of dark intervals ^ 0.03 sec. to enhance the yield of such "milHsecond flashes," may shift the optimum towards dark intervals much longer than 3000 Msec. Empirical studies of energy conversion yield as function of the "intermit- tency pattern" have been carried out on optically thin layers of Chlorella suspensions in connection with mass culturing of these algae, by Kok (1953) and Myers and coworkers (1953, 1954). Kok used rotating sectors which could be adjusted in width (varying tf/ta) and in speed of rotation (varying both tf and ta in the same proportion) . The curves given in his paper show that with a ratio ta/t/ = 4.5, a yield practically equal to that in constant hght of the same intensity could be achieved in artificial light of about 70 klux at 1500 r.p.m., (z. e., tf = S msec, ta = 13.5 msec). If this intermit- tency regime could be achieved by turbulent flow in a Chlorella suspension (instead of by wasteful black sectors), the gain in energy conversion (com- pared to nonstirred suspensions) would be by a factor of about 5. Since these cells were grown indoors and showed saturation in constant light at about 7 klux, still higher intermittency enhancement factors could be ex- pected to result from an increase in the ratio tg/tf above 10 (so as to reduce the average intensity of flashing illumination below the steady saturating intensity) . In fact, yields not significantly diff"erent from those in continu- ous hght could be obtained, with such "indoor" cells, at ta c^ 100 msec. 1478 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 and tf = 4: msec. These results confirm that dark intervals an order of magnitude longer than the Emerson-Arnold period of IQ-^ sec. can be ef- fectively utilized for photosynthesis after flashes lasting a few milliseconds (and having, in the above example, an integrated energy of the order of 70 X 4 = 280 lux sec). 7. Hill Reaction in Flashing Light Clendenning and Ehrmantraut (1950) studied the Hill reaction in flash- ing Hght, with quinone as oxidant, in whole Chlorella cells (cf. chapter 35, part C). A neon discharge tube (flash duration about 10 /zsec.) was used. The flash yields as function of the duration of the dark period are shown in 0.6 I o X 0.5 tn E E - 0.4 h I < UJ a. a u X o 03 - 0.2 - 0.1 - Photosynthesis 7 -/ o / - A ■—er' Quinone reaction -o ; i 1 1 L 5 10 15 20 DARK TIME, hundreths of seconds 25 Fig. 34.26. Effect of dark intervals on flash yield of photosynthesis and quinone reduction by Chlorella (after Clendenning and Ehrmantraut 1941). fig. 34.26; the curves follow a similar course for the quinone reaction and for photosynthesis (measured in the same Chlorella culture, but in bicarbon- ate solution instead of quinone-containing phosphate buffer). Saturation of both reactions requires dark intervals of 0.03-0.04 sec. at 10° C. The absolute flash yields in fig. 34.26 are 40-50% smaller in quinone than in bi- carbonate at the same flash energy. However, subsequent experiments by Ehrmantraut and Rabinowitch (1952) led to the conclusion that this difference was caused by the higher energy needed (in flashing as well as in steady light) to saturate the Hill reaction. Fig. 34.27 indicates that the saturation level probably is the same for both reactions — namely, about 12 mm.* oxygen per gram chlorophyll (i. e., one molecule oxygen per 2000 HILL REACTION IN FLASHING LIGHT 1479 molecules chlorophyll) per flash. In these experiments, the energy of the flashes was increased by raising the discharge voltage to 5.8 kvolt; in the higher range of energies the flash yield of photosynthesis appeared quite constant between 30 and 80 (rel. units), indicating true saturation (c/. above, section B2) ; no equally convincing proof of complete flash saturation of the Hill reaction was possible in the accessible energy range. In contrast to photosjoithesis the Hill reaction in Chlorella showed no induction loss in flashing Hght. 12 — * • • ^ ♦ "• ■* _ PHOTOSYNTHESIS .*-'' " % ' y 10 — / % I 1/1 _ ./• < / _i U. 8 — / a: /• a. / J 6 / • I / o o - / HILL REACTION ^ .• w4 — / o to / E Ee — A I I 1 , I , I , 1 W ' 0 20 40 60 80 100 RELATIVE LIGHT INTENSITY Fig. 34.27. Flash saturation of photosynthesis and quinone reduction by Chlorella (Ehrmantraut and Rabinowitch 1952). Gilmour, Lumry and Spikes (1953, 1954) made flashing hght experi- ments with chloroplast preparations from Beta vulgaris, using ferricyanide as oxidant, and an oxidation-reduction electrode as measuring device. The chloroplast suspension was stirred during the measurement. Flash illumination was provided by a 1000 watt incandescent lamp, and chopped by rotating sectors. The brightness at the vessel was up to 250 klux; with flashes lasting usually 2.8 msec, and in some experiments up to 16 msec, the flash energy was as high as 700, and sometimes even 4000 lux sec. The largest flash yields observed were of the order of four Fe(III) atoms reduced (corresponding to one O2 produced) per 3000 chlorophyll mole- cules— a yield considerably below the Emerson-Arnold maximum. How- ever, the maximum rate of Hill reaction in steady hght was, in Gilmour's 1480 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 preparations, only about 10% of the normal saturation rate of photosyn- thesis (for equal amount of chlorophyll). In this sense, the observed yields in very strong flashes were, similarly to those of Tamiya, much higher than expected from the Franck-Emerson-Arnold relation between the yields in constant and flashing hght. Reminiscent of Tamiya's obser- vations was also the fact that in (relatively) weak flashes (about 100 klux) the flash yield was in the "proper" relation to the rate in constant light (one O2 per 20,000 chlorophyll molecules per flash, or 10% of the Emerson- Arnold flash yield); complete saturation in respect to dark time was reached, in this case, after 0.04 sec. at 15.6° C, also in good agreement with the Emerson-Arnold observations on photosynthesis. With flash energies » 100 lux sec, on the other hand, the flash yields kept growing with flash duration far past the Emerson-Arnold Hmiting period (again in agreement with Tamiya's results). The main content of the work of Gilmour et al. was the analysis of the flash yield, P, vs. dark time, ta, curves under different conditions. In contrast to Tamiya, their interpretation was based on the assumption that the flash saturation curves of the Emer- son-Arnold type are real, and that a kinetic theory must account for them, as well as for the "Tamiya type" curves (and, more generally, for the effect of dark periods > 0.01 sec. on the energy utilization). At "low" flash energies (50-100 lux sec), the plots of A log P vs. Ata (increment in flash yield vs. increment of dark period) were simple straight lines, indicating, as mentioned above, a single rate-hmiting first-order dark reaction, with a half-time of about 0.016 sec at 6.9° C, and 0.0113 sec. at 15.6° C. — in good agreement with the constants of the Emerson-Arnold reaction in photosynthesis as reported by various observers (c/. Table 34.11). In some runs, a zero-order reaction was noticeable at the very beginning of the dark period. It was followed by an approximately temperature independent first-order reaction with a half-time of about 0.02 sec. (6.9- 33.3° C), and another zero-order reaction, with a rate constant of 4.3 X 10-3 at 6.9°, 4.7 X 10"^ at 15.6°, 2.5 X 10"' at 23.6° and 0.83 X lO"'' at 33.3° C. (mole Fe^Vmole chlorophyfl/sec). At the highest tempera- ture used, 33.3° C, the plot showed two sharply separated linear segments, the above-mentioned, almost temperature independent first-order reaction being preceded by a faster one with a half-time of 0.007 sec The latter may be identical with the temperature dependent first-order reaction (Emerson-Arnold reaction) which dominates the picture at the lower flash energies. The superposition of a zero-order limiting reaction upon a second-order one recalls Weller and Franck's suggestion that a zero-order process, by which the reductant is supphed to the photosynthetic system, may be the FLASH YIELD OF BACTERL\ AND HYDROGEX-ADAPTED ALGAE 1481 reason for the effect of long dark intervals on the flash yield in some or- ganisms (or under certain conditions, such as cyanide poisoning). One could extend the Weller-Franck hypothesis to the Hill reaction by assum- ing that the supply of the oxidant (ferricyanide, or an intermediate in its reduction) can become rate-limiting under certain conditions. Gilmour et al. offered a different concept. They suggested that the occurrence, after long and intense flashes of one (or two) zero-order hmiting reactions, together with a (temperature independent) first-order reaction different from that which determines the yield of short flashes, can be interpreted by assuming a "reservoir" in which a part of the energy-rich products, formed in the flash, can be reversibly stored, escaping the back reaction and permitting the "Emerson-Arnold enzyme" to operate more than once in the wake of a single flash. Gilmour et al. elaborated a chemical mechanism for the filling and emptying of the "reservoir" which could justify the postulate that the reservoir has no effect on the yield of brief flashes (or on the rate in continuous light) but increases the yield of "long" flashes of the same total energy. In essence, this mechanism amounts to a com- bination of the Franck-Emerson-Arnold reaction system with a side reac- tion feeding into a reservoir, from which the Emerson-Arnold system can draw after a flash. If the reaction feeding into the reservoir is enzymatic, the amount of photoproducts drained into the reservoir during the flash will depend not only on the integrated intensity of the flash, but also on its duration (as in Tamiya's model). The addition to the flash yield, provided by the stored photoproducts, ^vill then increase with the duration of the flash, and can be expected to depend also on temperature. The Emerson- Arnold reaction will nevertheless remain the bottleneck through which all photoproducts must pass, and which determines both the yield in steady light and the maximum yield of instantaneous flashes. Further details of the reservoir filling and emptying mechanism (includ- ing two catalysts, one of which is photochemically activated) were sug- gested by Gilmour et al. to account for the above-mentioned complex fea- tures of the flash yield-dark time curves ; experiments were also made (and interpreted by the same model) on the flash yield in chloroplast preparations partially inactivated by heat, cold, or ultraviolet light. 8. Flashing-Light Experiments with Bacteria and Hydrogen-Adapted Algae The only result available on the photosynthesis of 'purple bacteria in flashing light is the observation of Arnold, quoted by van Niel (1941) and recorded in Table 32.1, that the ratio P"'^''7BChlo is of the order of 1/400. This result can be interoreted as evidence that the limiting catalyst, E^. 1482 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 is present in these bacteria in the ratio of one molecule for 400 BChl molecules — or 400n, where n may be V4 or Vs- The duration of dark inter- vals required to achieve the full flash yield in bacteria — i. e., the constant /cea is as yet unknoAvn. However, it is likely that the rate-limiting reac- tion is the same in bacterial as in ordinary photosynthesis; and, if this is so, one sees no reason why the required dark intervals should be different. Confirmation would be of some interest. Rieke and Gaffron (1943) conducted some experiments on the photo- reduction of hydrogen-adapted Scenedesmus in flashing light. As discussed before, in chapter 6, the rate-limiting reaction is in this case the sup- ply of hydrogen by the hydrogenase system. Whenever the rate of the 15 10 15 TIME BETWEEN FLASHES x '/i2o sec. Fig. 34.28. Length of "stabilizing period" in photoreduction (after Rieke and Gaffron 1943). Rate of photoreduction as a func- tion of time interval between flashes in a group of two. primary photochemical process exceeds the maximum rate of the hydrogen supply, an accumulation of the intermediate oxidation products takes place, and causes a rapid "de-adaptation," i. e., return to normal photosynthesis. Limited hydrogen supply should have the same influence on flash yield as the Hmited carbon dioxide supply had in EA-deficient (e. g., cyanide-poisoned) algae. This j^rediction was verified by Rieke and Gaf- fron by means of experiments Avith flashes grouped in pairs. When uni- formly spaced flashes were used, the maximum yield of carbon dioxide reduction per flash was found to be approximately equal to that in non- adapted cells of the same species; but the dark intervals required for the full flash yield were much longer, since the supply of hydrogen during the intervals had to sufl5ce for the reduction of all intermediates produced by a flash. Otherwise, not only would the yield per flash be smaller (as in the case of cyanide-poisoned plants), but the accumulation of oxidation inter- mediates would have brought about immediate "deadaptation." When flashes were grouped in pairs — with the total number of flashes per minute unchanged — it was found that the interval between two flashes BIBLIOGRAPHY TO CHAPTER 34 1483 in a pair had to be > 0.025 second in order to obtain a full yield from both of them (c/. fig. 34.28). As described in the analysis of the cyanide experi- ments, this is an indication that the reaction which limits the yield of the first few flashes in a series, after a sufficiently long "interserial" interval, requires not more than about 0.02 second for its completion. This is the game order of magnitude as in ordinary photosynthesis. Thus, both the maximum flash yield and the rate of the reaction that determines the duration of the dark intervals within a series of flashes required for maxi- mum yield per flash, are not affected by adaptation. The slo^vness of the hydrogen supply in the adapted state merely leads to the necessity for inserting longer intervals between the flash series, in order to prevent deterioration of the flash yield after a small number of flashes. (The aver- age rate of the primary photochemical process has to be kept at such a level that the slow but steady hydrogen supply can keep pace with it.) Bibliography to Chapter 34 Time Effects. II. Photosynthesis in Intermittent Light A. Alternating Light 1905 Brown, H. T., and Escombe, F., Proc. Roy. Soc. London, B76, 29. 1918 Willstatter, R., and StoU, A., Untersuehungen ilber die Assimilation der Kohlens'dure. Springer, Berlin. 1919 Warburg, 0., Biochem. Z., 100, 230. 1928 Padoa, M., and Vita, N., Gazz. chim. Hal., 58, 647. 1931 Garner, W. W., and Allard, H. A., /. Agr. Research, 42, 629. 1937 Gregory, F. G., and Pearse, H. L., Ann. Botany, 1, 3. McAlister, E. D., Smithsonian Inst. Pubs. Misc. Collections, 95, No. 24. Portsmouth, G. B., Ann. Botany, 1, 175. 1938 Iggena, M. L., Arch. Mikrohiol., 9, 129. 1939 Aufdemgarten, H., Planta, 29, 643. 1941 Weller, S., and Franck, J., /. Phys. Chem., 45, 1360. ■ Briggs, G. E., Proc. Roy. Soc. London, B130, 24. 1951 Warburg, 0., Geleck, H., and Briese, K., Z. Naturforsch., 6b, 417. 1953 Brown, A. H., Am. J. Botany, 40, 719. 1954 Whittingham, C. P., Plant Physiol, 29, 473. B. Flashing Light 1932 Emerson, R., and Arnold, W., J. Gen. Physiol, 15, 391. Emerson, R., and Arnold, W., ibid., 16, 191. 1933 Arnold, W., ibid., 17, 135, 145. 1934 Arnold, W., and Kohn, H. I., ibid., 18, 109. 1935 Arnold, W., Cold Spring Harbor Symposia Quant. Biol, 3, 124. 1936 Kohn, H. I., Nature, 137, 706. 1938 Pratt, R., and Trelease, S. F., Am. J. Botany, 25, 133. 1484 PHOTOSYNTHESIS IN INTERMITTENT LIGHT CHAP. 34 1940 Emerson, R., Green, L., and Webb, J. L., Plant Physiol, 15, 311. 1941 Briggs, G. E., Proc. Roy. Soc. London, B130, 24. Franck, J., and Herzfeld, K. F., J. Phys. Chem., 45, 978. van Niel, C. B., in Advances in Enzymology. Vol. I. Interscience, New York, pages 263, 320. Weller. S., and Franck, J., J. Phys. Chem., 45, 1360. 1943 Rieke, F. F., and Gaffron, H., J. Phys. Chem., 47, 299. 1945 Franck, J., Pringsheim, P., and Lad, D. T., Arch. Biocliem., 7, 103. 1949 Tamiya, H., and Chiba, Y., Studies from Tokugawa Institute, 6, No. 2, 1. Tamiya, H., ibid., 6, No. 2, 43. 1950 Clendenning, K. A., and Ehrmantraut, H. C., Arch. Biochem., 29, 387. 1951 Burk, D., Cornfield, J., and Schwartz, M., Sci. Monthly, 73, 213. Rabinowitch, E., Ann. Rev. Phys. CJiem., 2, 361. Strehler, B. L., and Arnold, W., J. Gen. Physiol, 34, 809. 1952 Burk, D., Cornfield, J., and Riley, V., Federation Proc, 11, No. 796. Brilliant, V. A., and Krupnikova, T. A., Compt. rend. (Doklady) Acad. Sci. USSR, 85, 1383. Brilliant, V. A., and Krupnikova, T. A., ibid., 86, 1233. Ehrmantraut, H. C, and Rabinowitch, E., Arch. Biocliem. and Biophys., 38, 67. 1953 Burk, D., Hobby, G., Langhead, T., and Riley, V., Federation Proc, 12, No. 1, Abstract No. 601. Gilmour, H. S. A., Lumry, R., Spikes, J. D., and Eyring, H., Report 11 (July 1, 1953) to the Atomic Energy Commission. [Contract No. AI (ll-l)-82, Project No. 4 J. Kok, B., in Algal Culture, by J. S. Burlew, ed., Carnegie Corp. Publ. No. 600, Chapter II.6, p. 63-75. Myers, J., ibid., p. 37-54. Brown, A. H., Am. J. Botany, 40, 719. 1954 Gilmour, H. S. A., Lumry, R., and Spikes, J. D., Nature, 173, 31. Phillips, J. N., and Myers, J., Plant Physiol, 29, 152. PART FIVE ADDENDA TO VOLUME I AND VOLUME II, PART 1 * Chapter 35 PHOTOCHEMISTRY OF CHLOROPHYLL IN VITRO AND IN VIVO (ADDENDA TO CHAPTERS 4, 18 AND 19) Experiments witli chloroplast suspensions and products obtained by their dispersion and fractionation have narrowed (l)ut not quite bridged) the gap that had separated the photochemistry of chlorophyll in the living cell from the photochemistry of chlorophyll in solution. If this monograph were planned now, the part of it dealing with chloroplast-sensitized reac- tions would not have been tucked away among the unrelated and mostly wasted efforts described in chapter 4 ("Photosynthesis and Related Proc- esses Outside the Living Cell"), but would have found its logical place at the end of chapter 18 ("Photochemistry of Pigments in Vitro"), forming a transition to chapter 19 ("Photochemistry of Pigments in Vivo"). Be- latedly, we have adopted this plan in the present chapter, which thus con- stitutes an addendum to chapters 4, 18 and 19 of Volume I. A. Photochemistry of CHLOROPHYLLf in Solution* 1. Bleaching of Chlorophyll in Methanol (First Addendum to Chapter 18, Section A2) The reversible bleaching caused by illumination of oxygen-free chloro- phyll solutions in methanol, first observed by Porret and Rabinowitch, was described in Vol. I, p. 486. McBrady and Livingston (1948) have contin- ued the investigation of this phenomenon, which may provide clues to the function of chlorophyll in photosynthesis. Figure 35.1 gives a good illustration of the phenomenon. (The open circles represent the transmis- sion of the solution in light, the black dots its transmission in the dark.) McBrady and Livingston (1948) confirmed Livingston's earlier suspicion that the rate of the back reaction, and with it also the extent of bleaching in the photostationary state, can vary considerably from case to case— * Bibliography, page 1625. t Very little is known about the photochemistry of other photosynthetic pigments. According to Krasnovsky, Evstigneev et al. (19522), phycoerythrin solutions are much more stable in light than chlorophyll solutions; no reversible photochemical changes could be observed in them. 1487 1488 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 apparently under the influence of minor impurities. For example, the life- time of the bleached state in methanol solution was found by McBrady and Livingston to be of the order of 1 sec, in approximate agreement with the observations of Porret and Rabinowitch (1937), but about 100 times shorter than had been found in the earlier measurements of Livingston (1941). The life-time of the bleached state could be extended considerably by adding to methanol a small amount of carbon tetrachloride. Contrary to expectation, the rate of the dark back reaction (as well as the position of the photostationary state) was found to be insensitive to temperature changes, between 7.5° C. and 25° C. Chlorophyll h showed a slightly stronger reversible bleaching (and a somewhat slower irreversible bleaching) than chlorophyll a. Addition of 0.5 mole/liter allylthiourea, or 50% isoamylamine — sub- stances whose autoxidation is sensitized by chlorophyll — as well as the pres- ence of other redudants, such as phenylhydrazine or hydroquinone, had no 4 - 0> O 2 - 0 - c 3 O o . r..-- • • • • • • 1 1 1 1 1 50 100 150 TIME, seconds 200 250 Fig. 35.1. Reversible photobleaching of chlorophyll in oxygen-free solutions (after Livingston 1949): (•) dark; (O) illuminated. Ac = decrease in concentra- tions (mole/hter) of the red-absorbing component (assuming the reaction product does not absorb red hght at all). depressing effect on reversible bleaching (cf. p. 437). However, allyl- thiourea counteracted the enhancing effect the presence of carbon tetrachlo- ride had on bleaching, thus indicating that this enhancement must have been due to the presence of a reducible impurity in the ("reagent grade") carbon tetrachloride used. In contrast to earlier experiments, no enhance- ment of bleaching was found by McBrady and Livingston in the pres- ence of 10~® mole/liter /orwz'c acid; on the other hand, bleaching was en- hanced by the addition of 10 ~^ mole/liter oxalic acid (threefold increase of BLEACHING OF CHLOROPHYLL IN METHANOL 1489 steady-state bleaching, tenfold extension of the half-life of the bleached material). The back reaction was changed, by the addition of oxalic acid, from second to first order (in respect to the concentration of bleached ma- terial). An enhancement was produced also by the addition of methyl red (an azo dye whose reduction by phenylhydrazine is sensitized by chloro- phyll) ; 10~^ mole/liter methyl red increased the stationary bleaching by a factor of four. A particularly strong enhancing effect was caused by iodine (10~^ mole/liter) ; for example, in one experiment the steady-state bleach- ing increased from 0.2 to 26%, and the half-life of the bleached state from 0.5 to 20 sec. In this case, too, the back reaction became a first-order reac- tion. Irreversible bleaching was completely suppressed by iodine. In pure carbon tetrachloride, reversible bleaching was noticeable, but it was largely obscured by irreversible bleaching, which was much faster than in meth- anol, and continued for some time in the dark. (Such an after-effect was not observed in methanol, or in methanol-carbon tetrachloride mixtures.) The quantum yield of irreversible bleaching of air-saturated methanolic solution of chlorophyll a was estimated by McBrady and Livingston as 7f,., = 4 X 10~^ (c/. similar earlier estimates on page 497, Vol. I). They also estimated the quantum yield of reversible bleaching (from the photo- stationary state and the rate of back reaction, or from the initial rate of bleaching), with the results shown in Table 35.1. The rate of the reverse process may change from case to case, by a factor of 200 or more, but the rate (quantum yield) of the forward reaction was found to change much less — only from 1.3 X 10~* to 10 X 10~* in the examples given in Table 35.1. Table 35.1 Kinetics of Reversible Bleaching of Chlorophyll (after McBrady and Livingston)* Forward reaction Back reaction quantum yield, Solvent Order Rate constant 7 X 10* Methanol + oxalic acid 1 0 . 28 sec. ~^ 7 Methanol + ecu 2 2.8 X 10^ mole liter -i sec. -» 1.3 Methanol (pure) 2 "At lea.st 20 times higher" ~ 4 Methanol + I2 1 ? '--10 * Cf. below for data of Livingston and Knight. The values, jrev. = 10 X 10"'' and jrev. = 3 X lO"*, given earlier by Porret and Rabinowitch, and by Livingston, respectively, fall into the same range. (It is worth remembering that all these 7's are minimum values, since their calculation is based on the assumption that the reversibly bleached chlorophyll absorbs no red light at all.) In dealing with the mechanism of reversible bleaching, McBrady and Livingston did not go beyond discussion of the several possibilities enu- merated in Volume 1 (pages 489 and 514). One interesting new suggestion 1490 PHOTOCHEMISTRY OP CHLOROPHYLL CHAP. 35 was that the inhibiting effect of oxygen may be due to the catalytic acceler- ation by oxygen of the "de-tautomerization" (or deactivation of a meta- stable triplet state) of chlorophyll: (35.1) tChl ^'^ ) Chi a reversal of reaction (18.11a)— rather than to acceleration of an oxidation- reduction reaction which follows tautoraerization, as was assumed, e. g., in equations (18.12). The capacity of oxygen molecules to catalyze the destruction of tChl can be made plausible, according to McBrady and Livingston, by reference to the paramagnetism of the oxygen molecule, and to the probability that the long-lived state of chlorophyll, tChl, is a— mesomeric or tautomeric— triplet state and, as such, paramagnetic. McBrady and Livingston mentioned an alternative explanation of the oxygen effect (suggested by Franck) involving the formation of a "mol- oxide" of tautomeric chlorophyll: (35.2) tChl + O2 > tCh]-02 capable of either reacting with hydrogen donors (such as an amine, AH2) : (35.3) tChl-02 + 2AH2 > H2O + 2 A + Chi or decomposing into ordinary chlorophyll and oxygen : (35.4) tChl-02 > Chi + O2 Reaction (35.4) must be quite fast to be able to cause practically com- plete suppression of reversible bleaching; but reaction (35.3) must be even faster to account for the good quantum yield of sensitized photoxida- tion of amines. These two requirements seem difficult to reconcile. Livingston and Ryan (1953) later adopted this scheme in the interpre- tation of flashing light experiments, cj. reactions (g) to (i) in sequence (35.12A), and calculated rate constants for reactions (35.2) and (35.4); they did not discuss the rate constant of (35.3). McBrady and Livingston concluded that the earlier suggested mecha- nism of oxygen inhibition [reaction (18.12) ] combined with a reaction of the tautomer with impurities or with the solvent (as discussed on page 491, Vol. 1) still affords the most plausible scheme of the oxygen effect: (35.5a) Chi* > tChl (equation 18.11a) (35.5b) tChl + S > rChl + oS (equation 18.14) (35.5c) rChl + O2 >Chl+H02\ , , ,. . ,. ,,„,,, (35.5d) HO2 + oS > S + O2 jfiompare back reaction m equation (18.14) (here, S can stand for impurity, or solvent). The following mechanism was suggested for the effect of iodine on re- versible bleaching: BLEACHING OF CHLOROPHYLL IN METHANOL 1491 (35.6) tChl + I2 > tChl-Iz (analogous to moloxide formation) (35.7) tChM2 > Chi + I2 These two reactions lead to enhanced stationary bleaching (extended life-time of the complex tChl • I2) and make the back reaction a first-order reaction. Similar explanations could be suggested for the effects of oxalic acid on reversible bleaching. Further experiments on the reversible bleaching of chlorophyll were made by Knight in Livingston's laboratory (c/. Livingston 1947, 1948, Knight and Livingston 1950). The effects of solvent, temperature, oxygen concentration, chlorophyll concentration and of certain additions were ex- amined. The extent of steady-state reversible bleaching of chlorophyll a in oxygen-free methanol was found to be 1.10 rel. units at 13.3° C, 1.00 at 27.3° C. and 0.98 at 39.5° C. This small but probably real temperature coefficient indicates a very small activation energy of the back reaction. Experiments at three chlorophyll concentrations (1 X 10~^, 2 X 10~^ and 5 X 10~^ mole/hter) at 31° C, showed, quite unexpectedly, a slightly decreasing absolute bleaching: A [Chi] — 1.3, 1.0, and 1.0 X 10 ~^ mole/ liter, respectively. Taking into account the increase in absorption, Living- ston (1949) estimated that the extent of reversible bleaching at constant light absorption is inversely proportional to the square root of chlorophyll concentration. This implies that the rate of the back reactions is acceler- ated by an increase in chlorophyll concentration. It was mentioned before that in pure methanol the back reaction is of second order with respect to A [Chi], indicating that it is brought about by the encounter of two tChl-molecules : (35.8) tChl + tChl > 2 Chi or more generally, of two molecules produced by bleaching, e. g. : (35.9a) rChl + oChl > 2 Chi, or (35.9b) rChl + oS > Chi + S, or (35.9c) oChl + rS > Chi + S (where S can stand, as before, for a molecule of the solvent, or an impurity). None of these schemes involves the participation of normal chlorophyll molecules in the back reaction. An apparent acceleration of the back reac- tion by nonexcited chlorophyll molecules thus remains to be made plausible. In Vol. 1 (p. 515), a similar effect of increased pigment concentration ■ — the decrease in quantum yield of chlorophyll-sensitized reactions — was explained by the competition : ^^^■^^) ^^"^''l+Chl > oChl + rChl 1492 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 with the second reaction leading to recombination : (35.11) oChl + rChl > 2 Chi Livingston (1949) gave a somewhat different interpretation of the inhibiting effect of high pigment concentration on reversible bleaching: (35.12) tChl + Chi > 2 Chi However, this means that the back reaction ceases to be second order in respect to bleached chlorophyll — an implication not yet confirmed by ex- periment. It should also be taken into consideration that at higher chloro- phyll concentration, absorption becomes less uniform and is concentrated in a thinner layer of solution ; this favors second-order back reactions, such as tChl + tChl -^ Chl2 (in the same way as would an increase in light in- tensity). The observations (and speculations) concerning energy transfer between pigment molecules (chapter 32) and its role in the self-quenching of fluores- cence (chapter 23, section A5, and chapter 37C, section 4b) suggest another possibility — that increased chlorophyll concentration could diminish re- versible bleaching by dissipation of excitation energy in the course of its migration, as suggested in explanation of self-quenching. However, con- centration effects were observed by Knight and Livingston in a range (10~^ M) which is far below that where concentration quenching becomes noticeable (10"^ M); they must be therefore attributed to the deactiva- tion of metastable (and not of fluorescent) chlorophyll molecules. The only additional possibility derived from the energy migration concept is, then, that the deactivation reaction (35.12) may be caused by resonance rather than by actual kinetic encounters. (The probability of resonance exchange in the metastable state was mentioned in chapter 23, p. 785, and in chapter 32, p. 1290-1291.) That the back reaction is not simple was indicated by the previously noted effect of impuriiies. In a renewed study, the influence of traces of water was investigated. It was found that the presence of 2% water in a 2 X 10 ~^ mole/liter solution of chlorophyll a in methanol increases the steady-state bleaching by as much as a factor of three. It also accelerates considerably the rate of irreversible bleaching. In benzene as solvent, no reversible bleaching could be observed at all, and only a very slow irrevers- ible bleaching took place in the absence of oxygen. The presence of 1% methanol in benzene was sufficient, however, to produce as strong a revers- ible bleacliing as that which occurs in pure methanol. In carbon tetrachlo- ride (specially purified to remove the reductible impurity mentioned above) the irreversible bleaching is very strong. (Complete bleaching to a straw-yellow solution can occur within 3 min.; the product has an ex- BLEACHING OF CHLOROPHYLL IN METHANOL 1493 ceedingly high absorption peak at 400 mn, and only a weak band in the red.) Figure 35.2 shows the rate of irreversible bleaching and the extent of reversible bleaching as functions of oxygen pressure, [Oo]. The antiparal- lelism of the two developments — decrease in reversible bleaching and increase in irreversible bleaching — is easily apparent. In the neighborhood of 0.1 mm. O2, the irreversible reaction is comparatively slow and the re- versible reaction not yet completely hihibited. At 0.5 mm. O2, irreversible bleaching has reached its maximum rate, and reversible bleaching has been 0.6 - o Fig. 35.2. Reversible and irreversible photobleaching of chlorophyll as a function of the molality of dissolved oxygen (after Livingston 1949): (O) Rate of irreversible bleaching (right scale); (D) extent of reversible bleaching (left scale). reduced to almost zero. The low oxygen concentration sufficient to "satu- rate" the irreversible oxidation and to inhibit reversible bleaching can mean one of two things: either chlorophyll molecules associate with o.xygen molecules, the association being practically complete when the concentra- tion ratio is 1:1 (for evidence of such association, cf. Vol. I, page 460), or oxygen reacts with a long-lived active form of chlorophyll (such as tChl), which lives for a period of the order of 10 ~^ sec. Since, when irreversible bleaching is oxygen-saturated, its quantum yield is still very low, irrevers- ible bleaching must be initiated by a reversible step, with the back reaction of the unstable intermediate competing very successfully with the perma- nently bleaching second step of oxidation. Livingston (1949) suggests the reaction sequence tChl -f O2 -^ tChlOj; tClilO. -^ Chi + O2; tChlOa -}- Chi — > 0(yhl -f- Chi, where OChl stands for permanently oxidized chloro- 1494 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 phyll while oChl designated an unstable oxidation product, perhaps a free radical. Irreversible bleaching not only fails to be accelerated, but is even some- what slowed do^vn by heating (by about 15% when the temperature in- creases from 16 to 36° C). This indicates that the back reaction which re- verses the first step of oxidation has a somewhat higher temperature coef- ficient than the forward reaction with which it competes. In continuation of experiments by Rabinowitch and Weiss on the re- versible decoloration of chlorophyll solutions by ferric and eerie salts and the accelerating effect of ferrous ions on the back reaction in reversible photobleaching (page 464, Vol. I), Knight and Livingston found that re- versible bleaching of chlorophyll in methanol is enhanced by cerous chlo- ride, lanthanum chloride, and barium chloride (in concentrations from 2 X 10 ~^ to 2 X 10"^ mole/liter). The back reactions become first-order reac- tions, with a half-period of 7-8 sec. in the presence of the cerium salt, and 20 sec. in the presence of the lanthanum salt. With repeated illumination, the absorption at 650 m/x increased irreversibly, indicating the gradual for- mation of a product absorbing red light even more strongly than chloro- phyll itself. (In interpreting these results, one should recall the observa- tions of Rabinowitch and Weiss on the effect of salts on the reversible reac- tion of chlorophyll with ferric chloride.) When iodine was added to methanol, the back reaction was first order with a half-period of 25 sec. at 30° C. With increased temperature (23 -^ 47° C), the relative stationary bleaching in the presence of iodine increased by approximately 50%, de- spite the fact that the back reaction was accelerated by about a factor of 3 (still remaining a first-order reaction). Under all conditions, the extent of steady bleaching remained propor- tional to the square root of light intensity. It was found to be proportional to [I2] at concentrations up to 10 ~^ mole/liter; the half-life of the bleached state was, in this range, independent of [I2]. A chlorophyll-sensitized reac- tion of iodine with methanol appeared to be the cause why the bleaching effects were found to decrease with repeated exposure to light. In discussing these results, Livingston first reestimated, from figure 35.2, the quantum yield of irreversible bleaching in the oxygen-saturated state, and found jirr. — 4.5 X 10 ~^ in agreement with the earlier estimates (cf. Vol. I, p. 497). Maximum quantum yield of reversible bleaching was calculated by extrapolation as jrev. ^ 2.8 X 10~^ i. e., at least 6 times the value found in McBrady's work (table 35.1). In carbon tetrachloride, the quantum yield of irreversible bleaching was much higher — about 2.5 X 10 ~*, but this is still a low value compared to the quantum yield of re- versible bleaching. In the presence of iodine, the quantum yield of revers- BLEACHING OF CHLOROPHYLL IN METHANOL 1495 ible bleaching was only 7.8 X 10"^; in other words, the great enhance- ment of the stationary bleaching by iodine must be due entirely to a slowing down of the back reaction. When stationary reversible bleaching is as strong as it can be in the pres- ence of iodine (of the order of 10%), the bleaching effect of the photometric light beam cannot be neglected. Livingston showed that this influence was largely responsible for the observed deviations from the -y/l law. In summing up, Livingston pointed out that reversible bleaching ap- parently requires the presence of polar molecules. This parallels results obtained by the same investigator in the study of chlorophyll fluorescence (page 764) ; the latter, too, was found to require the association of chloro- phyll with polar molecules (alcohols or amines) . Livingston and Ryan (1953) continued the study of reversible bleaching of chlorophyll in two directions. In the first place, they attempted to identify the spectrum of the "bleached" state (its only previously known feature being reduced absorption in the red). For this purpose, mono- chromatic scanning beams (isolated by interference filters) were sent through a methanolic chlorophyll solution, illuminated by an immediately adjoining 1000-watt incandescent lamp (through a Corning 3-66 filter) in a constant temperature bath. Experiments of this type permit calculation of the product (concentration of changed chlorophyll) X (difference be- tween the absorption coefficients of changed and normal chlorophyll) for each scanning wave length; by assuming that the absorption coefficient, (xP, of the photoproduct is zero at the wave length where the bleaching ef- fect is strongest, a minimum value for the concentration of the bleached form can be calculated, and a consistent set of absorption coefficient ob- tained for all other wave lengths. (They will all be in error, by a constant factor, if the assumption aP = 0 was wrong.) The estimated absorption coefficients of the ''phototrope" (as the "re- Table 35.IA Average Specific Absorption Coefficients of Chlorophyll a and b and Their Phototropes, Chi a* and Chi b*, in Methanol, Determined for Bands Centered AT X„ (after Livingston and Ryan, 1953) T„. niM Ohio Ohio* Chi b Chi 6* 403 69.3 50.8 12.7 26.4 439.5 62.7 80.0 468.0 8.8 24.4 81.4 59.8 502 3.5 20.3 524.5 3.1 13.5 3.4 21.8 528 10.4 (0.0) 9.0 (0.0) 645 30.6 10.4 20.7 7.6 1496 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 versibly bleached" form was designated by Livingston) are shown in table 35.IA. The reduction product of chlorophyll, observed by Krasnovsky et al. in the reaction with ascorbic acid (c/, section 4) and the oxidation product observed by Rabinowitch, and by Linschitz et al., in reactions with ferric ions and quinone, respectively (c/. section 3), are similar to the phototrope observed in the reversible bleaching of chlorophyll in air-free methanol t X 10 sec Fig. 35. 2A. Decaj^ of chlorophyll h absorption at 468 m^ after a flash (Living- ston and Ryan 1953). Solid line, uncorrected observations; dashed line, correction for scattering of flash light; dotted line, corrected decay curve. which has (according to table 35. lA) a region of sharply increased absorp- tion in the middle of the visible spectrum (439-525 m^u for a; 525 m^ for h), and sharply decreased absorption in the red (528-645 m/n); however, the spectrum of the phototrope does not appear to be identical Math that of the two other unstable photoproducts (could it be a mixture of both?). The second set of measurements by Livingston and Ryan (1953) was made with four photoflash lamps surrounding the chlorophyll solution; a scanning beam — photomultiplier — oscilloscope system permitted changes in absorption with a resolving time of 0.1 msec, to be registered. Correc- tion for scattered flash light was made by subtracting the results obtained in air-saturated solution (where no reversible bleaching occurs) from those BLEACHING OF CHLOROPHYLL IN METHANOL 1497 obtained in air-free solution. Significant results were obtained mainly with chlorophyll b; figures 35.2A and 35. 2B show the decay curves of its phototrope as determined at X 468 and 524.5 m^u, respectively. The first curve is consistent with the assumption of a single phototrope (the same holds for 470.5 and 427.5 m^i) ; but the 524.5 m/x curve indicates the succes- sive formation of two photoproducts — one with decreased, and one with t X 10^ sec. Fig. 35. 2B. Decay of chlorophjdl b absorption at 524.5 m^ after a flash (Livingston and Ryan 1953). Dashed hne, uncorrected observations; soUd line, correction for scattered flash light; dotted Hne, corrected decay curve. increased absorption. The second, longer lived phototrope must be iden- tical with that known from steady light experiments (table 35. lA). Pre- viously described kinetic experiments indicated that this phototrope is a free radical, disappearing by a bimolecular reaction. Analysis of the decay curves in figure 35. 2B makes it likely that the short-lived intermediary product, not observed in steady light — at least, not in the (relatively) weak light which had been used for this purpose — disappears by a first-order reaction (simple monomolecular decay, or — more likely — "self-quenching" by encounters with normal chlorophyll molecules). Livingston and Ryan 1498 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 suggested that this product is the often postulated chlorophyll molecule in a metastable electronic triplet state. (It could also be a tautomeric triplet form; for the difference between the two, see pages 790-795.) The following set of reactions — most of which have been considered be- fore— was stated to account satisfactorily for the flash measurements: (35.12A a) Chi > Chi* (excitation) -hi. (35.12A b) Chi* > Chi (fluorescence) kt (35.12A c) Chi* > tChl (conversion to metastable state) (35.12A d) tChl + Chi > 2 Chi ("self-quenching" of meta- stable molecules) (35.12A e) tChl + S > oChl + rS (oxidation of tautomer to a radical) (35.12A f) rS + oChl > Chi + S (back reaction) (35.12A g) O2 + tChl > tChlOo Cor Chl-Oa) (quenching of the meta- (35. 12 A h) tChlOz + Chi > 2 Chi + O2 stable state by oxygen) (35.12A i) Chi -|- tChlOa > Chi -|- OChl (irreversible oxidation) Immediately after the flash, the solution was estimated to contain 65- 70% of chlorophyll in the metastable form, tChl, and 15-20% in the oxi- dized, free radical form, oChl. The relation between tChl and oChl then shifted in favor of the radical, as indicated by figure 35.2B (the radical alone is supposed to absorb at 524.5 mju). It will be noted, that in reaction sequence (35.12A), Franck's mechanism of the oxygen effect (equations 35.2, 35.4) was adopted. The radical for- mation was interpreted as reversible oxidation of chlorophyll (by the sol- vent, or by an impurity, not as reversible reduction (as in 35.5 b, c). This was done without special justification. Whether the stationary photo- bleaching in air-free methanol is a reduction or an oxidation (or a combi- nation of both?) remains open. The "concentration quenching" of tChl (reaction d) was included in sequence (35.12A) because of Livingston and Knight's observations of the effect of chlorophyll concentration on stationary bleaching (cf. above) ; by themselves, the flash data could be accounted for also by a simple mono- molecular deactivation, tChl — »■ Chi. Livingston, Porter and Windsor (1954) experimented with stronger flashes (~ 125 joule) and a synchronized flashing absorption beam, and found an aknost complete transformation of the chlorophylls and pheophy- tins into the metastable form in (Os-free) methanol, and even benzene; it decayed with a half-time of 2-5 X 10"^ sec. Reversible photobleaching of oxygen-free chlorophyll solutions was ob- served also by Linschitz and Rennert (1952) in glassy ether-isopentane- ethanol mixtures, at - 190° C. ; the change in spectrum (decrease of ab- PHOTOXIDATION OF CHLOROPHYLL 1499 sorption in the red and blue peaks, increase at 480-590 and 700-740 mju) was slight, and immediately reversible even at this low temperature, indi- cating that the photoproduct (metastable form of the chlorophyll molecule, or of a chlorophyll-solvent complex) reacts back monomolecularly without measurable activation energy. The light effect could be enhanced and stabilized by the presence of a quinone or imine (c/. next section). Similar experiments were made by Kachan and Dain (1951), who froze chlorophyll solutions in ethanol or ethanol-ether (1 :3) in liquid air and illuminated them with a 500- watt incandescent lamp, a mercury arc, or a spark. However, they found no effect of visible light; ultraviolet illumination caused the red absorption band to disappear in 20-30 min. The band returned upon melting; the reversible reaction was interpreted as an oxidation, with the electron detached by light and held by the solvent (and a proton transferred from chlorophyll to ether to form an oxonium ion, thus stabilizing the photo- product). No effect was observed, even with ultraviolet irradiation, if methanol was used as solvent. Uri (1952) used methyl methacrylate polymerization to indicate the formation of free radicals in illuminated chlorophyll solution. Polymeriza- tion was in fact observed in air-free 2 X 10~^ M solutions of chlorophyll in methyl methacrylate, or 10% methyl methacrylate in ethanol or pyridine (but not in benzene or acetone) exposed to red light. 2. Photoxidation of Chlorophyll (Second Addendum to Section A2 of Chapter 18) The reversible decoloration of chlorophyll ( in methanol) by ferric or eerie salts, first observed by Rabinowitch and Weiss, was described on p. 464. Evidence was presented there for considering this reaction a revers- ible oxidation-reduction, leading to an equilibrium. This equilibrium could be shifted back and forth by light (ef. page 488). Among the difficul- ties of this interpretation mentioned on page 465 is the fact that the reversal of the color change could be produced not only by Fe++ ions but also by nonreducing salts, even by sodium chloride. However, the much faster rate of the restoration of green color with FeCl2, the much lower ferrous salt concentration required for that purpose, the complete (immediate) reversi- bility of the reaction, and the possibility to repeat the light shift many times, were quoted there (and can again be quoted here) as supporting the hypothesis of reversible oxidation. The analogy with the chlorophyll- quinone reaction (to be described below) points in the same direction. Ashkinazi, Glikman, and Dain (1950) opposed this interpretation because they found that reversible color changes of chlorophyll can be produced also by nonoxidizing salts. They suggested that metal complex formation is responsible for such changes in all cases, including that of iron. Ashkinazi and Dain (1950) prepared a ferrous complex of chlorophyll by the action 1500 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 of ferrous acetate on pheophytin in acetic acid and found that in ethanol this compound is oxidized by air (presumably to a ferric complex) with a band shift from 645 to 610 m/x; upon illumination the band is shifted back to 645 m^ (presumably indicating reduction to the ferrous form). The chlorophyll-ferric ion reaction and the possible formation of metal compounds of chlorophyll (with magnesium still in the center of the mole- cule and the metal ion attached elsewhere— perhaps to the enol group in ring V) remains in need of further study. As mentioned on pages 465 and 492, the similarity between the reaction of chlorophyll with Fe+++, and the first, reversible stage of the "phase test" is suggestive; the first (reversible and light-sensitive) step in the conversion of chlorophyll to pheophytin (cf. page 493) also bears an outward similarity to these reactions. Reversible photobleaching (probably photoxidation) of chlorophyll by quinones was described by Linschitz and Rennert (1952). To block the thermal back reaction they carried out the irradiation at - 190° in glassy solvent (8 parts ether -j- 3 parts isopentane + 5 parts ethanol, by volume). The bleaching effect in red and blue, and the enhancement of absorption in green (480-590 mM) and far red (>700 m^), noted upon illumination of the oxygen-free chlorophyll solution {cf. above, section 1) is strongly en- hanced by the addition of 10 -^ or 10"* mole/1, of a quinone or imine; the bleached state survives in this case in the dark until the solvent is melted. Evstigneev, Gavrilova and Krasnovsky (1950) measured the quenching of chlorophyll fluorescence by various organic reagents {cf. table 23.IIIC, and chapter 37C, section 4c) in pyridine and ethanol, and their influence on the red absorption peak of chlorophyll {cf. chapter 21, page 647 and chapter 37C, section 2). They inquired whether these effects occur to- gether, and whether they are associated with accelerated photochemical bleaching of chlorophyll, but found no correlation. Thus, quinone, which is the strongest quencher of chlorophyll fluorescence in both ethanol and pyridine, has no effect on absorption and causes no photochemical bleaching in ethanol (and only a relatively slow one in pyridine). Ascorbic acid, which produces the fastest photochemical bleaching in pyridine (and is second only to oxygen in ethanol), has no effect on either fluorescence or absorption. Evstigneev et al. looked not only for the easily detectable progressive (slowly reversible or irreversible) photochemical bleaching of chlorophyll by quinone, but also for the more elusive rapidly reversible, stationary bleaching during the illumination (by sending a scanning beam through a strongly illuminated solution) ; but found no observable effect (compare, however, the above-described low temperature observations of Linschitz with the same system). They concluded that the reaction which leads to bleaching occurs not to the short lived, fluorescent chlorophyll molecules, but to the long lived, metastable ones (tChl), and therefore has REVERSIBLE PHOTOREDUCTION OF CHLOROPHYLL 1501 no relation to the quenching of fluorescence {cf. scheme 19.11 on page 54G; also page 788, and scheme 23.11). Calvin and Dorough (1948) and Huennekens and Calvin (1949) de- scribed the photochemical oxidation of certain chlorins to porphins by quinone, and reduction back to chlorins by phenylhydrazine. Evstigneev and Gavrilova (1953^) found that irreversible auioxidalion of chlorophyll a (in toluene) is accelerated by water and other polar molecules and slowed down by pyridine and other basic molecules (which favor the ^hoioreduction of chlorophyll). 3. Reversible Photoxidation of Bacteriochlorophyll Krasnovsky and Vojnovskaja (1951) noted that dissolved bacterio- chlorophyll photoxidizes in air; the oxidation is faster in alcohol than in pyridine. The alcoholic solution gives, after oxidation, a strong peroxide test with ferrous thiocyanate. Adding ascorbic acid, or hydrogen sulfide, regenerates the pigment; after 10 hours standing in the oxidized state the greater part of the pigment could still be regenerated in this way. Oxi- dants other than molecular oxygen (p-quinone, nitrite, nitrate, hematin) and reductants other than ascorbate or hydrogen sulfide (malic, pyruvic, succinic acids; thiosinamine) were ineffective. o-Quinone oxidized bac- teriochlorophyll, even in the dark, to a compound with a chlorin-type spectrum (with a strong band at 680 m/x), as mentioned by Schneider in 1934. Bacteriopheophytin is less easily oxidizable than bacteriochloro- phyll; its photoxidation, too, is reversed by ascorbic acid. The autoxidation of bacteriochlorophyll was interpreted by Krasnovsky as addition of Jxygen to a photochemically produced biradical. As with chlorophyll, iiu reversible dehydrogenation could be observed (the reaction with o-quinone is irreversible). 4. Reversible Photoreduction of Chlorophyll and Its Derivatives (Addendum to Chapter 18, Section A5) In chapter 16 (page 457) and 18 (page 505) we have discussed attempts to achieve reversible reduction of chlorophyll to a "leuco-chlorophyll." The conclusion was that no truly reversible reduction has as yet been achieved, and that chlorophjdl, as it is known in vitro, appears to be more inclined to undergo reversible oxidation, than reversible reduction. The above-described experiments of Livingston and co-workers support (or, at least, do not contradict) the hypothesis that the reversible bleaching of chlorophjdl is the result of reversible oxidation. Krasnovsky (1948M, on the other hand, has described experiments indicating a reversible photo- chemical reduction of chlorophyll, apparently to an unstable pink radical, by ascorbic acid. The reaction was observed in dry pyridine as solvent; 1502 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 it did not occur in ethanol. In a mixture containing 5 X 10 ~® mole/liter chlorophyll and up to 5 X 10^^ mole/liter ascorbic acid, illuminated by a 500-watt lamp through a red filter (and a layer of water), the chlorophyll fluorescence disappeared in 1-2 min. Phenylhydrazine also could be used as a reductant (instead of ascorbic acid). Some reversible reduction could be observed in alcoholic solution if a little pyridine or ammonia were added 1.000- N o 0.500 450 500 550 600 650 700 m^ Fig. 35.3. Reversible reaction of chlorophyll a with ascorbic acid in pyridine in light (after Krasnovsky 1948): (i) spectrum before reaction; {2) spectrum after reaction; (5) appro.ximate spectrum of the unstable reaction product (6 min. after darkening). to alcohol. No reaction was observed (in pyridine) with pyruvic, oxalic, malic or citric acids, ethanol, or hydroquinone as reductants. The dark back reaction required 2-3 hours. After complete recovery, the spectrum was very similar to that of original chlorophyll, especially when chlorophyll a was used. The chlorophyll h band at 470 m/x in an (a + h) mixture was weakened after the cycle, indicating, in Krasnovsky's first surmise, irreversible reduction of the carbonyl group (which distin- guishes chlorophyll h from chlorophyll a). Figure 35.3 shows the spectrum of chlorophyll a before the reversible reaction with ascorbic acid in air-free pyridine (curve 1), and after this re- REVERSIBLE PHOTOREDUCTION OF CHLOROPHYLL 1503 action (curve ^) . The slightly increased absorption in the green may indi- cate the formation of a small amount of pheophytin. Curve 3 is the ap- proximate extinction curve of the unstable pink product. Complete ab- sorption curves of reduced chlorophyll (cf. fig. 37C.10) were determined by Evstigneev and Gavrilova (1953^). They indicated two reduced species (perhaps an ion and a neutral semiquinone molecule). In the discussion, Krasnovsky pointed out that the reversible reduction of chlorophyll apparently requires the presence of a basic compound (such as pyridine) ; he attributed this tentatively to the greater stability of an ionic form of the free radical "mono-dehydrochlorophyll" (which is the probable immediate product of the reaction between light-excited chlorophyll and ascorbic acid). He suggested that this semiquinone is formed by hydrogenation of one of the conjugated double bonds in the "aromatic" porphin system. The following scheme of reversible reduction was suggested : (35.13) ^0=0/ + hi' (Formation of a "biradical" by light absorption: the C=C bond is as- sumed to be part of the conjugated system.) (35.14) N. . . HA + ^C— c/ > NH+ + A + \c— C"/ (Ascorbic acid, bound by a hydrogen bond to pyridine, cedes a proton to pyridine and an electron to the biradical of chlorophyll, thus forming mono- dehydroascorbic acid, and an anion of the monohydrochlorophyll radical.) In a follow-up reaction, the anion could combine with an H+ ion and the resulting neutral semiquinone dismutates into a quinone (chlorophyll) and a hydroquinone (didehydrochlorophyll = stable reduced product?); but in a basic solvent, such as pyridine, the H+ concentration may be so low as to make these steps improbable, and to limit the reversible reduction to the radical stage. The absence of reversible reduction in alcoholic solution was attributed by Krasnovsky to the rapid reoxidation of the neutral form of the semiqui- t H none, yC — C\^ (which can be formed because of the presence of H+ ions in ethanol). Krasnovsky suggested that reversible photochemical reduction of chlorophyll by ascorbic acid (a compound that is present in all green plants) may be a step in photosynthesis. If this were true, however, monodehydroascorbic acid should be able to liberate oxygen from water. 1504 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 If the interpretation of Krasnovsky's results as proving the possibiHty of a reverHihh photochemical reduction of chlorophyll is confirmed, as well as the conclusion, based on observations of Rabinowitch, Porret, Weiss, Livingston, and Linschitz that chlorophyll can undergo reversible photochem- ical oxidation, the green plant pigment emerges as a compound capable of both reversible reduction and reversible oxidation in light, a combination which could be important for its function in photosynthesis, particularly if the latter involves two sets of photochemical reactions — "photoxidations" and "photoreductions," with chlorophyll as an intermediate (c/. chapter 7). It appears that reversible oxidation of chlorophyll may require the pres- ence of polar molectdes (such as alcohol or water), while reversible reduction requires the presence of basic molecules (such as pj^ridine). According to Evstigneev and Gavrilova (1950), presence of Mg enhances the tendency for photoxidation, reduces that for photoreduction. Krasnovsky and Brin (1949) suggested that chlorophyll attaches itself to pyridine in the same way as does hemochromogen to a protein^ — by a link between the metal atom and a nitrogen atom in an imidazole ring; a type of complexing known to affect the redox potential of hemins. Krasnovsky, Brin and Voynovskaya (1949) investigated more systemati- cally the influence of solvent on the photoreduction of chlorophyll a and b by ascorbic acid. They found that in pyridine the height of the red ab- sorption peak was reduced in light by 90%, and that 80% of this bleaching was reversed in the dark. Dilution of pyridine with water decreased both the extent and the reversibility of bleaching. In aniline the red peak was reduced in light only by 22%, and only half of this bleaching was reversed in the dark. In ethanol the corresponding figures were 20 and 14%; in acetone, 29 and 19%. Addition of as little as 0.3% pyridine, imidazole, or histidine, to alcohol or acetone, increased the bleaching to 80%, with one- half of it being reversed in the dark. Twenty-six organic and two inorganic compounds were tried out as reducing agents in pyridine. Positive results were obtained with ascorbate, dihydroxymaleate, cysteine, phenylhydra- zine, and hydrogen sulfide; pyruvic acid (previously mentioned by Kras- novsky as giving results similar to those produced by ascorbic acid) was now listed among nonreactive compounds. Renewed investigation of chlorophyll h showed that its reduction by ascorbic acid in light, and reoxidation in dark, gives a product with absorp- tion bands at 693 and 432 mju, which is neither chlorophyll a, nor pheophy- tin a or b. (Conversion of chlorophyll b to chlorophyll a in this reaction was first suggested as a possibility, cj. above.) Evstigneev and Gavrilova (1950) noted that chlorophyll and Mg- phthalocyanine were reduced by ascorbic acid in light in pyridine solution REVERSIBLE PHOTOREDUCTION OF CHLOROPHYLL 1505 slower than the corresponding Mg-free compounds (but were photautoxi- dized faster than the latter in methanolic solution). Krasnovsky and Voynovskaya (1952) found that chlorophyll can be re- duced photochemically in pyridine also by sodium sulfide, but the reduction did not proceed beyond 25% (measured by the decline in the intensity of the red band). Krasnovsky and Voynovskaya (1949) found that reversible reduction can be obtained also with protochlorophijll (from pumpkin seeds). A re- versible chemical reduction of this compound was described (c/. chapter 37B) by Godnev and Kalishevich, who used Timiriazev's reagent (zinc + organic acid in pyridine). (For incomplete reversibility of this reaction, see chap. 18, p. 457, and chap. 37B, p. 1779.) Photochemical reduction (to a brown solution, with an absorption band at 470 m/x) was now obtained by illumination in the presence of ascorbic acid, at 8° C. The brown prod- uct was reoxidized after several hours in air, with (approximate) restora- tion of the original spectrum. If the illumination was prolonged (30 min.) and no red filter was used, a band at G75 mju appeared after reoxidation, indicating partial conversion of the porphyrin, protochlorophyll, to a chlorin (chlorophyll ?) ; with chlorophyll a, no conversion to a bacterio- chlorin was noted (which would have produced a band at 780 m/i). Krasnovsky and Voynovskaya (1951) made similar reduction studies with bacteriochlorophyll. As with chlorophyll, the reaction goes best in pyridine, but was clearly observable also in alcohol. The reaction is rapidly reversed in the dark, even without air. Bacteriopheophytin is reduced faster and further than bacteriochlorophyll. The unstable reduction product of bacteriochlorophyll is green, with an absorption band at 640 m^. No photoreduction w^as noted when malic, succinic, citric, or pyruvic acid, thiosinamine, or sodium thiosulfate was added instead of ascorbate. The possibility of reversible hydrogenation of bacteriochlorophyll indi- cates that the photoreduction of chlorophyll (or protochlorophyll) does not occur in an isolated double bond in ring II (or II and IV); Krasnovsky suggested that in both pigments hydrogenation disrupts one of the con- jugated double bonds, creating a free radical. Krasnovsky and Gavrilova (1951) compared the photoreduction of chlorophyll (a + h) by ascorbic acid and other organic acids in different solvents. No significant reaction occurred in acetone or ethanol; in pyri- dine, ascorbic acid alone gave rapid photoreduction (as described before). In dioxane, too, ascorbic acid was the only one of the tested compounds to react with chlorophyll, but in this solvent the reaction was irreversible; it occurred even in the dark, but was accelerated by light ; the product showed no characteristic absorption band at 525 m^i. The reaction did not occur 150G PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 in the presence of oxygen or quinone. Hydrogen sulfide and cysteine also reduced chlorophyll in dioxane irreversibly. Similar experiments with other dyestuffs (Mg-phthalocyanine, ribo- flavin, (8-carotene, safranin T, neutral red, and phenol-indophenol) showed that many of them (but not j8-carotene) also can be more or less reversibly reduced in light by ascorbic acid (some also by pyruvic acid) ; the photo- reduction depends on the solvent, being enhanced by the affinity of the lat- ter for protons. Krasnovsky and Brin (1953) again discussed the question of why bases promote the photochemical reduction of chlorophyll. Absorption studies (cf. chapter 37C, section 2) showed that some bases cause a new absorp- tion band at 640 m/x to appear in the chlorophyll spectrum (it becomes the dominant long-wave band in piperidine, cf. table 21. VII). However, the appearance of this band (which occurs only in polar solvents and must be ascribed to polarization and ionization of an acid group, such as the enol group in position 10 in chlorophyll) does not parallel photochemical activa- tion. The latter can be produced by a small amount of a base (such as phenylhydrazine) added to methanol solution without noticeable change in the absorption spectrum. Krasnovsky and Brin concluded that activation must be attributed to a reaction between pigment and base which is differ- Table 35.IB Influence of Bases on Photochemical Reduction (Measured by Decline of Absorption in Red) of Chlorophyll a and Pheophytin a (after Krasnovsky and Brin 1953) Chi a, % Pheo a, % Un- Restored Unre- Restored Medium reduced in dark duced in dark Pyridine 6 87 15 68 Piperidine 31 39 5 50 Nicotine 61 81 ■ — Pyrrole 81 81 Quinoline 93 94 72 86 Phenylhydrazine 4 62 — Ethanol 82 85 73 76 +pyridine (50)* 42 52 43 59 +piperidine (50)* 19 29 16 36 +nicotine (95)* 8 58 36 72 +urotropine (80)* 49 29 — — +quinoline (90)* 98 98 61 90 +aiyinine (satur.) 94 93 — ■ +phenylhydrazine (100)* 52 85 — — -1-ammonia (satur.) 9 26 — +KOH (0.1-0.01 N) 98 99 — — * Number in parenthesis means milligrams of base in 7 ml. solution. CHLOROPHYLL-SENSITIZED OXIDATION-REDUCTIONS 1507 ent from the acid-base interaction, and which is without marked influence on the absorption spectrum- — perhaps a coordinative binding of the base in the center of the molecule, reminiscent (as suggested by Krasnovsky and Brin 1949) of the hemochromogen-protein bond (known to change the oxi- dation potential of the hemochromogen). In chlorophyll the complexing may occur at the magnesium atom; it is not quite clear how a similar binding can occur in pheophytin (cf. table 35. IB). Speculations concerning the activating effect of organic bases on the photoreducibility of chlorophyll obviously need to be related not only to the absorption data (chapter 21, pp. 647-649, and chapter 37C, section 2), but also to fluorescence measurements (chapter 23, pages 766-771, and chapter 37C, section 4) and chemical evidence. One obviously deals here with events in at least two — if not more — sensitive centers in the chlorophyll molecule which can be affected separately or simultaneously. One of them may be the cyclopentanone ring V with its keto-enol tautom- erism; the other, the central magnesium atom. A correlation of all the pertinent evidence — derived from optical absorption spectra, infrared spectra, fluorescence, "allomerization," phase test, and photochemical activity — will be attempted in chapter 37C. Krasnovsky and Brin (1948) compared the "photochemical activity" (capacity to react with ascorbic acid in light) of various types of chlorophyll preparations. They found reactivity in organic solvents, oils, lecithin, and emulsions of lipoid solvents in water; also in colloids obtained by dilution of alcoholic solutions by aqueous detergents (anionic, cationic, or neutral), in similarly prepared chlorophyll-protein colloids, and in chloroplasts and grana suspensions prepared with the same detergents. No photo- chemical activity could be found in colloidal solutions prepared by dilution of alcohol with water (before or after coagulation by electrolytes); or in chlorophyll adsorbates on the proteins, zein or gliadin, and their colloidal solutions (for the position of the absorp- tion peaks in these preparations, see table 37C.IIIA). 5. Chlorophyll-Sensitized Oxidation-Reductions (First Addendum to Chapter 18, Part C) Capacity for reversible photochemical oxidation or reduction is impor- tant if a pigment is to serve as sensitizer for oxidation-reduction reactions — in the same way as capacity for reversible nonphotochemical oxidation or reduction is important for an oxidation-reduction catalyst. Only one chlorophyll-sensitized oxidation-reduction reaction in vitro was described in Volume I (page 513)— the oxidation of phenylhydrazine by methyl red. This reaction was discovered by Bohi in 1929, and studied quantitatively by Ghosh and Sengupta in 1934. They found that the quantum yield of methyl red bleaching can reach, or even exceed, unity. Livingston, Sickle and Uchigama (1947) pointed out that errors could have been caused, in Ghosh's and Bohi's work, by ill-defined nature of the chlorophyll prepara- 1508 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 tions, and by changes in the absorption spectrum of methyl red (which is an acid-base indicator), which may be caused by the presence of the base, phenylhydrazine. (The quantum yield was determined by photometry of methyl red !) Furthermore, in Ghosh's experiment, an induction period of varying duration was noted, wliich affected unfavorably the reliability of the rate measurements. Since light absorbed by methyl red could contribute to the reaction, Livingston used a red band X > GOO m/z. (Ghosh and Sengupta worked in the green, at X 435.8 m/x.) A green band (475-550 mju) was used for the determination of methyl red. The initial concentrations of phenylhydra- zine and methyl red (in methanol) were 5 X 10 ~^ and 1 X 10~^ mole/ liter, respectively. The reaction was found to be of zero order (i. e., the quantum yield remained constant) until about 80% of the dye was con- sumed. The concentration of chlorophyll was varied between 0.2 and 75 X 10-^ mole/liter (chlorophyll a), and 1 to 7.5 X 10"^ mole/liter (chloro- phyll h). The observed quantum yields scattered over the range from 0.09 to 0.15, showing no clear dependence on nature or concentration of the chlorophyll used. The reason why Ghosh and Sengupta had found 6-7 times higher quantum yields could not be explained. Experiments with a manometer, using an acetonic solution of allyl thiourea (0.5 mole/liter), containing chlorophyll (1-2 X 10 ~^ mole/liter) and oxygen, indicated that the quantum yield of this reaction (for which values up to 1.0 had been found by Gaffron) was up to seven times higher than that of the methyl red-phenylhydrazine reaction — thus indirectly confirming the low quan- tum yield of the last-named reaction given above. The manometric experi- ments also showed that no gas was produced in the reduction of methyl red by phenylhydrazine (liberation of nitrogen could conceivably occur in this reaction). Experiments showed that the occurrence of an induction period was entirely dependent on the presence of oxygen. From the amount of oxy- gen present in air-saturated methanol and the duration of the induction period, it appeared that the quantum yield of chlorophyll-sensitized autoxi- dation of phenylhydrazine (the reaction which probably takes place during the induction period) may be of the order of unity, similar to the quantum yield of the autoxidation of allylthiourea. In the green, at 435.8 m/x, the quantum yield of the methyl red-phenyl- hydrazine reaction was found, in preliminary experiments, to be about 0.1, independently of changes in methyl red concentration; and despite the fact that, in this case, a varying proportion of light was absorbed by methyl red and not by chlorophyll. This indicates that the reaction can oc- cur also by direct photochemical activation of the azo dye instead of by sensitization. CHLOROPHYLL-SENSITIZED OXIDATION-REDUCTIONS 1509 Attempts were made to substitute allylthiourea for phenylhydrazine as reductaiit in chlorophyll-sensitized reduction of methyl red, using ace- tone as solvent ; but no bleaching of methyl red was noted under these con- ditions. To check w^hether phenylhydrazine formed a complex with chlorophyll, the absorption spectra of chlorophyll a and h were measured in the presence of 0.05 M phenylhydrazine (in methanol). No change was noted in the spectrum of chlorophyll a; in chlorophyll h, the red band was somewhat broadened toward the longer waves. (C/. Watson's data, given below and in chapter 37C.) The views on the mechanism of the phenylhydrazine-methyl red reac- tion expressed in this paper were changed by Livingston and Pariser (1948), who studied more closely the dependence of the quantum yield on the concentrations of the reactants (phenylhydrazine and methyl red). Methyl red is known to occur in three colored forms, depending on acidity; it was concluded that only one of them — the intermediate one — takes di- rect part in the reaction. To determine the concentration of this form, ex- tinction curves were obtained for methyl red solutions in methanol (2 X 10-* M, containing from 0.4 to 10"* M HCl; and from 10 -« to 10 -» M NaOCHs). These curves were interpreted as due to the superposition, with varying relative intensities, of three absorption bands, with peaks at 406 (I), 491 (II), and 521 (III) van, respectively. The forms giving the bands (I) and (III), can be obtained in the pure state in strongly alkaline or strongly acid solution, respectively; but the intermediate form, which gives band (II), usually is present only in mixture with one of the other two forms. This intermediate form was interpreted by Thiel as a "zwit- terion": -OOC-C6H4 - N+=N— C6H4N(CH3)2 H However, this formula, with two opposite charges, may be correct for aque- ous solutions only; in methanolic solution, Livingston and Pariser con- sidered more likely tne existence of a neutral molecule, with a hydrogen bond between the hydroxyl and nitrogen : 0 H <^ \>_N=N-<^ ^(CH3),N The absorption peaks of the acidic form (III) are situated not too far apart in aqueous and methanolic solution (517 and 521 m/i, respectively) ; 1510 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP, 35 the bands of the other two forms are shifted strongly toward the red in methanol (from 406 to 447 m^u, and from 491 to 530 mju), with the result that what was the "long wave band" in water, becomes the "central band" in methanol. Livingston and Pariser tried to relate the quantum yield to the total concentration of the dye, or to the concentration of any one of its three constituents. It transpired that the simplest relationships are ob- tained if form II alone is taken into consideration as a reactant. The phenylhydrazine concentration (free base + hydrochloride) was varied between 0.02 and 1.0 mole/liter, the total methyl red concentration, from 0.24 to 6.10 X 10 ~* mole/liter, and the (calculated) concentration of "form II" from 0.085 to 6.10 X 10""* mole/liter. The quantum yields were found to vary between 0.10 and 0.49 mole/einstein. A maximum yield 7 ~ 0.5 can be explained by assuming a reduction of the dye in two steps. A single light quantum can form, at best, one molecule of the intermediate semiquinone; two semiquinone molecules then dismute into one molecule of the dye and one molecule of the leuco dye. From the effect of methyl red on the reversible bleaching of chlorophyll (section 1), Livingston concluded that the first step of the sensitized reac- tion possibly is the association of methyl red (in the form II) with the long- lived (tautomeric?) active form of chlorophyll, tChl, 35.15a) Chi + hv > Chi* (absorption) Ay (35.15b) Chi*— -^ Chi + hv (fluorescence, ^ 3%) -> tChl (tautomerization, ^97%) (35.15c) tChl > Chi (detautomerization) (35.15d) tChl + MR" '—^ tChlMR" (complex formation) (35.15e) tChlMR" — > Chi + MR" (complex decomposition + detau- tomerization ) The chlorophyll-MR" complex can be supposed to react with the reductant (phenylhydrazine, symbolized by PH2), transferring one H atom: (35.16) PHo + tChlxMR" ^^ PH + Chi + MRH" and forming two radicals, PH and MRH". These can be stabilized, either by dismutation, e. g.: (35.17) 2 MRH" > MRH2" + MR" (leuco dye + dye) or by dimerization : (35.18) 2PH > HP-PH CHLOROPHYLL-SENSITIZED OXIDATION-REDUCTIONS 1511 If this scheme is compared with the various mechanisms discussed in chapler 18 (Vol. 1), it appears as a variation of mechanism A/31 (page 515), exemplified in equations (18.33a-18.33d) for the case of molecular oxygen as oxidant. The alteration consists in the assumption that tChl associates with the oxidant in a complex, and in this form catalyzes the transfer of hydrogen from reductant to oxidant; while in its original form, scheme A/31 assumed that this catalysis occurs in two steps, tChl first transferring a hydrogen atom to the oxidant, and then recovering it from the reductant. The first mechanism is similar to an often postulated mechanism of enzymic catalysis, based on complex formation of enzyme and substrate ; the second one is more in line with the known mechanisms of nonenzymic oxidation- reduction catalysis. By permitting competition between the catalytic reaction (35.16) and the decomposition of the complex (35.15e), Livingston's scheme gives a possibility of explaining the dependence of the quantum yield on the con- centration of the reductant, [PHo]. To achieve a similar result in the original reaction scheme A/Sl, we would have to admit the possibility of a back reac- tion there, too. In scheme (18.33), e. g., a reaction oChl + HO2 -^ Chi + O2 would have to be postulated, reversing reaction (18.33b). Livingston and Pariser derived, from the above reaction scheme, an ex- pression showing the dependence of the average quantum yield, 7, on the concentration of phenylhydrazine (which remains practically constant during a run) and the "logarithmic mean" of the (strongly changing) con- centration of the reacting dyestuff : r<^^ iQ^ Tmt^i - [MR-Jo - [MR-] •^■^^•^^^ ^^^^^ ' ~ In [MR"]o - In [MR"] [MR"]o is the initial, and [MR"] the final concentration of the reactive form of the dye. The equation obtained for the mean quantum yield has the form ^Qron^ - - ^' V (WA^O [PH.] {k,/k; )[MR"] [6b.Z()) 7 - 2(k, + kj)^ 1 + {k,/k[) [PH2] ^ 1 + {kjh') [MR"] With the below-determined constants, and applied to momentary concen- trations of all components, this equation becomes i-ir. 91 , - n dR V 1 X 10-^ [PH2] ^ 5 X 10* [MR"] (35.21) 7 = 0.46 X 1 I 1 w in2 rPTj 1 X 1 + 1 X 102 [PH2] 'M + 5 X 10' [MR"] The factor i/2 comes, as mentioned before, from the dismutation of the half-reduced dye; the factor kt/{kt -\- k/) from competition between fluores- cence and tautomerization ; the factor ki/kf from competition between complex formation (35.15d), and the detautomerization of chlorophyll, (35.15c). As mentioned above, oxidation and reduction of tChl could be 1512 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 substituted for the complex formation and decomposition ; the assumption of a back reaction reversing oxidation is needed in this case to make the over-all yield dependent on the concentration of phenylhydrazine, as repre- sented in equation (35.21). Livingston and Pariser extrapolated the observed quantum yields to [PH2] = CO by means of the following equation : (35.22) '^(PHjlcD ^ V '^ hlPH,]) using for the ratio ki/k/ an empirically estimated value, 10^ liters/mole. This extrapolation changed the values of 7 significantly only for the lowest phenylhydrazine concentrations used — [PH2] from 0.02 to 0.05 mole/liter. i>« 1 / [mr"] Fig. 35.4. Quantum efficiency of chlorophyll-sensitized reaction of methyl red and phenylhydrazine in relation to concentration of methyl red in form II (after Livingston and Pariser 19-18). The values of I/7 for [PHo] = 0° were plotted against 1/[MR"] and, in accordance with equation (35.22), an approximately straight line was obtained (although the individual values scattered considerably). From the slope and the intercept of this straight line (fig. 35.4), the follow- ing ratios could be derived: (35.23) h/(kf + kt) = 0.92 h/k = 5.0 X 10< liters/mole The first value means 8% fluorescence and 92% tautomerization (a plan sible result as far as the yield of fluorescence is concerned), and thus an CHLOROPHYLL-SENSITIZED OXIDATION-REDUCTIONS 1513 "absolute" maximum quantum yield of 0.92/2 = 0.46. If a value for k/ is taken from earlier measurements of the life-time of the "long lived excited state" (reversibly bleached state) of chlorophyll, we have: (35.24) k; = 2.5 X 10-^ sec.-' (35.25) A-i ^ 10^ (liters/mole) This high value for the constant of a bimolecular reaction indicates a reac- tion with only a very small, or vanishing activation energy. In methyl red solutions, the life-time of the bleached state of chlorophyll was estimated (c/. section 1) as '--'1 sec. This can be considered as the life-time of the complex {tChl.MR"} (or, alternatively, as the life-time of chlorophyll oxidized by methyl red, oChl). This leads to a value of the order of 10^ for the constant kt, indicating that reaction (35.16) may re- quire an activation energy. Consequently, the quantum yield of the over- all reaction can be expected to show strong temperature dependence when [PII2] is low, and (35.16) is therefore the "bottleneck" reaction. Watson (1952) in a subsequent paper (based on his work in Livingston's laboratory) suggested a different interpretation of the mechanism of chlorophyll-sensitized reduction of methyl red by phenylhydrazine. His experiments dealt with the quenching of fluorescence by phenylhydrazine in a polar solvent (methanol), and its activation, by the same agent, in nonpolar solvents (benzene or heptane) (c/. page 768; Watson's experi- mental results will be described in chapter 37C, section 4c). From these experiments, and the strong effect of phenylhydrazine (PH2) on the ab- sorption spectrum of chlorophyll h (cf. page 699 and chapter 37C, section 2), Watson concluded that phenylhydrazine forms two different complexes with chlorophyll; its attachment to the chlorophyll molecule in one place leaves the absorption spectrum of chlorophyll a unchanged and activates fluorescence; its attachment in another place quenches the fluorescence (independently of whether the first place is occupied by another phenyl- hydrazine molecule or not). The calculated association constants were 1900 liters/mole for the fluorescence-promoting, and 16 liters/mole for the fluorescence-quenching complexing (in benzene) ; the first constant was the same, but the second was higher (58 liters/mole) in heptane, Watson pointed out that if these constants are correct then, under the conditions of Pariser's experiments, a large part of chlorophyll must have been as- sociated with phenylhydrazine. The symmetry of equation (35.20) per- mits substitution of the assumption of a primary reaction between PH2 and excited Chi in a pre-formed complex for that of the primary formation of a tChl-methyl red complex (e. g., 35.15e), and subsequent reaction of this complex with phenylhydrazine (e. g., 35.16). This mechanism, involving primary interaction of chlorophyll with the 1514 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 reductant, is in line with the observations of Krasnovsky and co-workers, which we will now describe. According to the experiments of Krasnovsky which were described above (section 3), chlorophyll can be reversibly reduced in light by ascorbic acid. In the presence of riboflavin, or safranin T, the latter dyestuffs are reduced instead of chlorophyll — presumably by reacting with the photo- chemically reduced chlorophyll. Because of the relative positions of their reduction potentials, Eo = 0.22 v. and 0.29 v., respectively (as compared with —0.05 V. for ascorbic acid), riboflavin and safranin T are not reduced by ascorbic acid in the dark. Reduction can be achieved by illumination a. E o >- (D UJ Q < O I- Q. O 0.20 Uj 0.10 nv ii\. A -ii\ 2 \ ' • + O2 1 1 .... 1 1 — 20 TIME, minutes 30 40 Fig. 35.5. Chlorophyll-sensitized reduction of riboflavin by ascorbic or pyruvic acid, and reoxidation in air (after Krasnovsky 1948). Curve i, ascorbic acid; 2, pyruvic acid. Broken curves, in ethanol; soUd curves, in pyridine. (•) Light off; (-(-02) air admitted. with light absorbed by the dyes themselves, or, in the presence of chloro- phyll, by illumination with red light absorbed by chlorophyll. In addition to ascorbic acid, pyruvic acid, too, can be oxidized in this manner. The reactions were carried out in ethanol or pyridine as solvent; the concentrations were : [dye] = 10-5 mole/liter [reductant] = 6 X IQ-^ mole/liter The change in dyestuff concentration was determined photometrically. The solutions were boiled in vacuum to remove oxygen, and illuminated 1-3 min. by focussed light from a 500-watt lamp. Admission of air accelerated the back reaction. The following mechanism was assumed for the non- sensitized reaction: CHLOROPHYLL-SENSITIZED OXIDATION-REDUCTIONS 1515 (35.26) 2 D + 2 AII2 ;f light dark ± 2 DH + 2 AH - ± (DH2 + D) + (AII2 + A) In the presence of sensitizer, the latter was assumed to serve as intermediate hydrogen acceptor. Figure 35.5 illustrates Krasnovsky's results. It indicates that rever- sibility is more complete in pyridine than in methanol ; Krasnovsky suggests as a possible reason the stabilization of the semiquinone DH by ionization DH ;^ ± D- -1- H- encouraged by the proton affinity (basicity) of pyridine. The reaction with pyruvic acid in methanol is at least partially reversible. CP o 0.75 Fig. 35.6. Chlorophyll-sensitized reduction of coenzyme I (DPN) by ascorbic acid (after Krasnovsky and Brin 1949). Chlorophyll (a + b) in pyridine with 10 mg. ascorbic acid and 1 mg. DPN; (1) spectrum before reaction; (2) spectrum after reaction (3 min. illumination in red light) without DPN; (5) spectrum after reaction (3 min. light) with DPN; (4) spectrum of reaction product (DPNH2?) (obtained by subtraction of 2 from 3). 0.50 - 0.25 300 350 400 Magnesium phthalocyanine also can be used as sensitizer, in the same concentration (10"^ mole/liter) as chlorophyll. Krasnovsky pointed out that in the sensitized reduction of one mole of safranin T by ascorbic acid under standard conditions as much as 16 1516 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 kcal of free energy must be accumulated, which is 40% of the energy of a red quantum, or 20% of the energy of two such quanta. Later, Krasnovsky and Voynovskaya (1949) observed that the same reaction of ascorbic acid with safranin T can be obtained also with proto- chlorophyll (from pumpkin seeds) as sensitizer. Krasnovsky and Brin (1949) described experiments which they inter- preted as indicating the sensitized reduction of the oxidized form of co- 4 8 12 16 20 TIME, minutes Fig. 35.7. Bleaching of chlorophyll (a + b)hy ascorbic acid in red light and its regeneration by riboflavin (5), safranin T (4), oxygen (2), DPN (3), and without added oxidant (1) (after Krasnovsky and Brin 1949). enzyme I (dipyridine nucleotide, DPN) by ascorbic acid, with chlorophyll as sensitizer. Here, again, aqueous pyridine must be used as solvent. In these experiments, 1 cc. of a 5 X 10"^ M solution of DPN (a 60% pure preparation) and 10 mg, crystalline ascorbic acid were added to 5 cc. of a 7 X 10 ~^ M solution of chlorophyll (a + h) in a Thunberg tube. After evacuation by an oil pump, the spectrum was measured by means of a Beckman spectrophotometer. The tube was then illuminated for 3 min. at 8° C. and the spectrum re-examined. Comparison of this spectrum after CHLOROPHYLL-SENSITIZED OXIDATION-REDUCTIONS 1517 illumination with the one obtained after similar treatment without DPN, showed a difference in the region 300-360 mix, which indicated the forma- tion of a compound with an absorption band at 345 mju (fig. 35.6)— the well-known position of the absorption band of DPNH2. The band at 525 mn, which appeared in illuminated chlorophyll-ascorbic acid mixtures in the absence of DPN and which was ascribed by Kras- novsky to reduced chlorophyll (HChl, or HoChl?), "sometimes" did not appear when DPN was present; this result was attributed to the high rate of the second reaction in the sequence (35. 27a, b) : (35.27a) Chi + AH, "—^ HjChl + A (or HChl + AH) (35.27b) HsChl + DPN (or 2 HChl + DPN) > DPNH2 + Chi (AH2 = ascorbic acid). To prove the occurrence of reaction (35.27b) directly, a mixture of chlorophyll and ascorbic acid was illuminated alone in an evacuated Thun- berg tube, and DPN (or other oxidants) was then added from a side tube. The disappearance of the absorption band at 675 m^ in light, and its re- appearance in dark after the addition of oxidants, is illustrated by figure 35.7. This figure shows that recoloration was far from complete — the optical density, which had dropped in light from 0.7 to 0.1, was restored by the oxidants only to a value of 0.45. The figure also shows that saf- ranin and riboflavin caused a much faster return of the color than did DPN, or air. These results are suggestive, and Krasnovsky's conclusion that reduc- tion of DPN by reduced chlorophyll is the link by which the photochemical process is tied, in photosynthesis, to the sequence of enzymatic reactions leading to fixation and reduction of CO2, is plausible; it fits well into the picture of the reaction mechanism of photosynthesis derived from C(14) experiments (chapter 36). However, just because of this suggestiveness and the crucial importance of the conclusions, much more rigorous experi- ments will be needed than those described by Krasnovsky and Brin, before the capacity of chlorophyll to sensitize the reduction of the oxidized form of coenzyme II, by hydrogen donors such as ascorbic acid (thus overcoming an opposing normal potential difference of about 0.24 volt, at pH 7), can be considered as proved. Spectroscopic proof of the formation of DPNH2 (based on a small difference between two high optical densities), as well as the proof of the reversible reduction of chlorophyll, may prove to be correct, but are as yet not quite convincing. The fact that the absorption at 675 m/i is restored to only one half its original value indicates a con- siderable irreversible change. Further development of these experiments, with better spectroscopic technique, supplemented by chemical separation and enzymatic tests, appears desirable. 1518 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 Krasnovsky and Brin (1950) surveyed different oxidants for the reoxi- dation of photochemically reduced chlorophyll. The following compounds, all with negative oxidation potentials, were found to accelerate the return of absorption in the red peak (in order of decreasing efficiency) : thionine (5 X 10-'' M), quinone (1 X 10-^ M), methylene blue (5 X lO"" M), phenol-indophenol (5 X 10"^ M) and dehydroascorbic acid (5 X 10'^ M). (The position of the latter in the series shows that the other enumerated oxidants, although they are known to react with ascorbic acid, reoxidize chlorophyll directly, and not through the intermediary of ascorbic acid.) Among compounds with positive potentials, not reacting with ascorbic acid in the dark, the order of effectiveness in the reoxidation of reduced chloro- phyll was: hematin (5 X lO"" M), NO3-(10-2 M), NO2-(10-2 M), Fe+++(10-^ M), Cu++(10~^ M), and air. Among compounds with posi- tive potentials, safranin T, neutral red, and Nile blue (all 5 X 10-* ilf) ac- celerated reoxidation; also 5 X 10"* M riboflavin, and 1 X 10^^ or 5 X 10-" M DPN (^0 = +0.32 volt) (c/. above). Since xanthin (^o = +0.37 volt) showed no influence, the authors concluded that the normal oxidation- reduction potential of the system chlorophyll-reduced chlorophyll is about +0.35 volt. (However, the potentials had been measured in water, while the observations of Krasnovsky and Brin were made in pyridine.) It is interesting to compare these results with those obtained with chloroplast suspensions; there, only oxidants with normal potentials below —0.1 volt were reduced with a good yield in air (c/. sections B4(d), (e) below); in the absence of oxygen, the reduction of compounds with normal poten- tials up to +0.1 volt could be observed; but compounds such as DPN, with E'o > 0.3 volt, were reduced only to a very slight degree, so that their re- duction could be ascertained only by ''trapping" with specific enzymatic systems (to be described in section B4(/)). Of course, in the Hill reaction of chloroplasts (as in photosynthesis) the reductant is water (£"o = —0.8 V.) and not ascorbic acid (^'o = -0.0 v.), making the (net) hydrogen trans- fer that much more difficult. Further studies are needed to find out whether a reduced form of chlorophyll, with the strong reducing power found by Krasnovsky in vitro, does play a role in photosynthesis, but is somehow prevented from displaying its full power in chloroplast suspensions. Per- haps, two types of reversible photochemical changes of chlorophyll occur in photosynthesis— one that permits it to acquire hydrogen from water (this capacity is preserved in chloroplast preparations) and one that permits it to transfer hydrogen to compounds with a normal potential >0.3 volt — this capacity being preserved in chlorophyll solutions in pyridine. This is, however, only a speculation, and one which calls for the minimum quantum requirement of photosynthesis to be 8 (c/. chapter 7). An alternative is that in vivo chlorophyll is able to transfer hydrogen directly from a system CHLOROPHYLL-SENSITIZED OXIDATION-REDUCTIONS 1519 with a normal potential close to —0.8 volt, to a system with a normal po- tential > 0.3 volt, perhaps with the assistance of high energy phosphate produced by partial recombination of the primary oxidation and reduction products (Ruben, Kok, van der Veen, Warburg and Burk). In any case, the relation between the "Krasnovsky reaction" of chlorophyll in solution, the "Hill reaction" of chlorophyll in chloroplast suspensions, and photo- synthesis is a most interesting photochemical problem. Of seven fatty acids tested, in a study by Krasnovsky and Brin (1950), only 10 ~^ M malic acid produced marked acceleration of reoxidation. Temperature had little effect on the rate — a result taken as confirmation of the hypothesis that the pink reduced form of chlorophyll is a semiqui- none, and, as such, able to react wdth a very small activation energy. (If this is the case, however, the low absolute rate of reoxidation becomes puzzling.) As long as Krasnovsky's assumption of a reversible oxidation-reduction of chlorophyll rests only on the observation of a (sometimes far-reaching, but never complete) restoration of the extinction coefficient in the peak of the red band, some doubt remains whether this assumption is correct and whether the reaction does not leave an irreversible change in the chloro- phyll molecule (in which case it could not serve as the basis for catalytic activity). It is of some interest, therefore, that Holt (1952) in checking Krasnovsky's experiments with ascorbic acid and quinone, was able to re- peat the bleaching and recoloration cycle three or four times with the same sample. True, the restored red band became weaker with each cycle; nevertheless, the result seems incompatible with the assumption that after a cycle all chlorophyll molecules that took part in it are left with a perma- nent change in their structure. It seems more likely that the reaction is basically reversible, but that a certain proportion of chlorophyll molecules that take part in it undergo irreversible side reactions which lead to the loss of absorption. It would be important to find a way to suppress these ir- reversible reactions in vitro as effectively as they appear to be suppressed in vivo. Krasnovsky and Voynovskaya (1952) compared the sensitization of oxidoreduction by chlorophyll, bacteriochlorophyll, and their pheophytins (in 10~* M solutions), using ascorbic acid and sodium sulfide (about 10 "^ M) as reductants, riboflavin or safranin T (about 10 ~* M) as oxidant, in 85% aqueous pyridine (since sodium sulfide is more soluble in aqueous solvents). No changes were observed in the dark in binary systems pigment-reductant in pyridine (a slow pheophytinization occurred in alcohol in the pres- ence of ascorbic acid); the pigment-oxidant binary systems and the mixtures oxidant- ascorbic acid also were stable, but sodium sulfide reduced safranin T and riboflavin in the dark, particularly in pyridine. Light had no effect on binary systems without the sensitizer. The binary system sensitizer-reductant reacted in pyridine in light as de- scribed in section 4. 1520 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 In ternary systems absorption bands of the sensitizers were not affected to more than 5-10% by 3 minutes illumination, while those of the oxidants were strongly reduced in intensity. The results are summarized in table 35. IC. Table 35.IC Pekcent Reduction of Different Redox Systems by 3 Minittes of Illumination WITH Chlorophyll, Bacteriochlorophyll or Their Pheophytins as Sensitizers (After Krasnovsky and Voynovskaya 1952) Chi o Pheo o BChl BPheo Redox system In In Eton C.HsN In In EtOH CsHsN In Ir- EtOH CbHsN In In EtOH CsHsN Riboflavin + ascorbate Riboflavin + Na2S Safranin T + ascorbate Safranin T + NajS 50 60 25 80 50 25 25 70 25 70 5 60 40-60 70 5 25 40-60 50 * Complete bleaching of bacteriochlorophyll although the latter shows no reaction with NazS in absence of riboflavin. A reduction of DPN could be observed with chlorophyll and ascorbic acid in pyri- dine (as described before) but not with bacteriochlorophyll. (As in fig. 35.6, the reduc- tion of DPN was derived from increase in the absorption of the illuminated solution around 340 m^- ) mv 300- 400 500- 600 Fig. 35. 7A. Changes of redox potential of chlorophyll and pheophytin in pyridine upon repeated illumination and darkening: (A) 10 "^ mole/1. Chi (a -f b), 0.8 X lO'^ mole/1, ascorbic acid; (B) 10~'» mole/1, pheophytin, 0.6 X 10~2 mole/1, ascorbic acid (after Evstigneev and Gavrilova 1953*). Although bacteriochlorophyll acts on the whole Hke chlorophyll, its reduced form reacts back so much faster that no determination of its absorption spectrum could be made. Krasnovsky also noted that while bacteriochlorophyll can be photochemically reduced in vitro by sodium sulfide, it does not react with other reductants used by purple bacteria, such as malic acid or propanol; he suggested that these hydrogen donors must be acted upon by appropriate dehydrogenases before their hydrogen becomes avail- able for photochemical transfer. CHLOROPHYLL-SENSITIZED OXIDATION-REDUCTIONS 1521 Evstigneev and Gavrilova (1953^) measured the photogalvanic effect in an illuminated solution of chlorophyll and ascorbic acid in pyridine. In the dark the oxidation-reduction potential of the solution is determined by the ascorbic acid-dehydroascorbic acid system. In light the potential rises, first rapidly, then more slowly, as shown in figure 35. 7A; in the dark it returns almost to its initial value. (Here again we note that the reaction can be repeated several times with the same sample!) The widest poten- tial change observed was about 0.29 volt with chlorophyll a (or a + 6), 0.25 volt with chlorophyll b, 0.35 volt with pheophytin (a + 6), 0.33 volt with Zn-pheophytin and 0.29 volt with Mg-phthalocyanine. Since the dilute (lO"* M) chlorophyll could not have oxidized more than about 1% of the 0.01 M ascorbic acid, and any change in potential caused by this oxidation must have been toward more negative values, the rise of potential actually observed in the illuminated solution indicates a high sensitivity of the elec- trode to the presence of a small amount of reduced chlorophyll. (The situa- tion is similar to that in thionine-ferrous ion photogalvanic cells, described by Rabinowitch in 1940; there, too, a shift of tlie oxidation-reduction equilibrium in light leads to the electrode potential becoming more nega- tive, i. €., the electrode responds to the reduction of the dye-leucodye sys- tem more sensitively than to the oxidation of the Fe+VFe+' system. The main potential-determining process in the illuminated chlorophyll- ascorbic acid solution must be the transfer of electrons from reduced chloro- phyll to the positive electrode, and their return to dehydroascorbic acid via the negative electrode; i. e., an electrode-catalyzed back reaction in the photochemically displaced chlorophyll-ascorbate equilibrium. Evstigneev and Gavrilova pointed out that the value -fO.35 volt (the maximum photogalvanic potential obtained in these experiments) is the same as was derived by Krasnovsky and Brin (1950) from the capacity of photoreduced chlorophyll solutions to reduce various oxidants (cf. above). Evstigneev and Gavrilova noticed that after light has been switched off, the photogalvanic potential disappeared within a minute, while the red absorption band of chlorophyll was not fully restored until 30-40 minutes later; they concluded that the electrochemical effect is caused by a par- ticularly unstable, intermediate form of reduced chlorophyll. This seems to call for a parallel experiment on the rate of disappearance in the dark of the bands at 518 m/x and 585 mn, ascribed by the same authors (1953^) to the ionic and the neutral form of a semiquinone, respectively (cf. chapter 37C, section 2). Evstigneev and Terenin (1951) made experiments on the photovoltaic effect (Bec- querel effect) with electrodes (Pt, graphite) coated with chlorophyll, phthalocyanine or pheophytin. With chlorophyll in the presence of air they found a positive photoeffect (coated electrode became more positive in light) requiring about 1 min. of illumination 1522 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 to reach a stationary value and decaying in about 1 min. in darkness. Reductants de- creased the positive effect or even changed its sign. If a rise of potential to a level 0.35 volt above the normal potential of ascorbic acid (about —0.04 volt at pH 7) could be maintained in the steady state of a photochemical reaction, chlorophyll-sensitized reduction by as- corbic acid of compounds with normal potentials of the order of +0.3 volt (as described by Krasnovsky and Brin, cj. above) would be plausible. However, this low photogalvanic potential corresponds to a wide shift of the concentration ratio of a very dilute oxidation-reduction system, Chi/ rChl; and as soon as the reductant in this system finds an oxidant with mm. Fig. 35. 7B. Changes in redox potential of chlorophyll in pyridine: (A) 10-* mole/1. Chi (a + 6), 0.6 X 10-^ mole/1, ascorbic acid; (B) same + 10-* mole/1, safranin T; (C) same + 4 X 10 -* mole/1, riboflavin (after Evstigneev and Gavrilova 1953). which it can react, the concentration of the reductant must decline and the potential go up ("photogalvanic depolarization"); the reaction will be slowed down or stopped altogether. This effect of hydrogen acceptors on the photogalvanic potential actually was noted by Evstigneev and Gavri- lova. Figure 35.7B shows that the presence of 4 X lO""* mole/liter ribo- flavin reduces the potential drop to less than one-half of its original value. In other words, the photogalvanic potential cannot be taken as a measure of the effective reducing power of an illuminated redox system. What oxidants can be reduced in light, and at what rate, is a problem of kinetics, and not, or only partly, of equilibrium (or pseudoequilibrium) potentials. This may be a proper place for a general remark concerning the use of oxidation-reduction potentials in the discussion of photochemical mecha- nisms. Let us assume that the action of light is to displace an oxidation- CHLOROPHYLL-SENSITIZED OXIDATION-REDUCTIONS 1523 reduction equilibrium (say, between two intermediate redox catalysts, A/AH2 and B/BH2). What can be said about the secondary oxidations and reductions which can follow this displacement? In the first place, one has to remember that "reducing power" (say an accumulation of BH2) arises simultaneously with "oxidizing power" (say, an accumulation of A), so that to speak of only one of them (one often hears of the "reducing power" of illuminated chloroplasts) means postulating that the systems A/AH2 and B/BH2 are either separated spatially immediately after the photochemical transfer of H from A to B; or else that, because of enzymatic specificity, the next system in the reduction sequence (say, C/CH2) "sees," chemically speaking, only the reduced system B/BH2 and not the simul- taneously present oxidized system A/AH2 (an analogous assumption must be made on the "oxidation side" of the primary photoprocess). The specificity of enzymes and the heterogeneous structure of the photo- synthetic apparatus make one — or both — of these important postulates not implausible. The next question (selecting arbitrarily the "reducing power" rather than the "oxidizing power" for further consideration) is: what secondary reductions can be achieved by the system B/BH2, in which the ratio [BHo]/ [B] has been increased by light? Let us assume that the normal potential of the pair B/BH2 is E^; if the logarithm of the concentration ratio [BH2]/ [B] has been shifted to 5 {~> 1), the redox potential of the pair becomes E = Eq -\- 0.03 5. On paper, by increasing 5, E^ can be increased in- definitely; e. g., with b = 4^, E^ = E^ + 0.12 volt, and so on. In reality, however, the possible contribution of this concentration term to the reduc- ing capacity of BH2 in the steady state is very limited, because an extreme value of the concentration ratio (such as 10*: 1) cannot be maintained when a reaction is in progress. This is particularly true of a photochemical reaction with a high quan- tum efficiency. Such a reaction is only possible if no (or relatively few) quanta are wasted in the primary process ; and if this process includes the transfer of hydrogen (or electrons) from AH2 to B, there must be enough B present at all times to accept the proffered H atoms (or electrons). In other words, the steady-state concentration of B cannot drop low in light if this light is to be effectively used, and this means that no extremely high photostationary ratios [BH2]/[B] are permitted, and the oxidation-reduc- tion potential of this system in light cannot be much different from ^o- To this consideration one must add a second one : to prevent exhaustion of B in light, the absolute rate of reaction of BH2 with the next hydrogen (or electron) acceptor, C, must be sufficiently high. To use a concrete ex- ample, if B is called upon to accept one H atom every second, and if one- half of the total B -\r BH2 must be available in the oxidized form to ensure 1524 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 that no photochemically supplied H will be wasted, BH2 must react with C within one second to maintain the ratio [BH2]/[B] at the required level. This puts a limit on the admissible activation energy of the reaction BH2 + C -»- B + CH2. If this reaction is endothermal, the minimum value of the activation energy is the heat of reaction, A//; and if the normal potential of the system C/CH2 is significantly more negative than that of the system B/BH2, the reaction BH + C -* B + CH is likely to be endothermal by a similar amount. Furthermore, if this reaction has no activation energy in excess of Ai/, the back reaction, CH2 + B ->- C + BH2 will be extremely rapid, and there will be little chance for any CH2 formed to escape it by stabilizing processes. The net result of all these considerations is that only very limited help can be expected, in a photochemical reaction against the gradient of chemical potential, from the concentration term in the free energy relationship: E = E,+°^ log R?il n [Ox] What is needed for the success of the photochemical reaction is the capacity of the primarily light-activated molecule to utilize its energy for direct reduction of a compound with a potential negative enough for all subsequent steps in the reaction sequence to go "downhill" (on the AH scale). In more concrete terms, if there is a primary photochemical oxidant, B, its reduced form must be capable of carrying on the reduction of whatever has to be ultimately reduced — be it CO2, RCOOH, or various Hill oxidants— by exothermal reactions. (These reactions may be made exothermal by cooperation of several BH2 molecules, i. e., by an "energy dismutation" mechanism — for example, with the aid of high-energy phosphates.) Uri (1952) noted that the yield of polymerization of methyl methacrylate, photo- sensitized by chlorophyll (cf. above, section 1) increased enormously upon addition of ascorbic acid or urea, indicating a great increase in the photostationary concentration of free radicals. Reducing salts (ferrous sulfate, ferrocyanide) inhibited rather than stimu- lated polymerization. Baur and Niggli (1943) in the last of the papers of the series reported in chapter 4 (section A4(fe)) had claimed that chlorophyll, dissolved in geraniol or phytol, and emul- sified in a solution of methylene blue in dry glycerol, could sensitize the reduction of cir- culating carbon dioxide to formaldehyde and liberate oxygen (from glycerol?). Lin- stead, Braude and Timmons (1950) were unable to confirm these results. Gurevich (1948) described still another chlorophyll-sensitized oxidation-reduction: the reduction of o-dinitrohenzene by phenylhydrazine. (For his experiment on the use of o-dinitrobenzene as oxidant in Hill reaction, cj. part B, Sect. 4(g).) He used ethanolic extracts from nettle leaves. 5 cc. of "dark-green but transparent" extract were mixed with 0.5 cc. of phenylhydrazine (10% free base in ethanol) and 0.5 cc. o-dinitrobenzene. One half of this solution was illuminated, at 15-30°, 15 cm. from a 300-watt lamp, (H2O filter) for 30 min.; the other half kept in darkness. The formation of o-nitrophenyl- hydroxylamine was proved by addition of concentrated ammonia, leading to dark-violet CHLOROPHYLL-SENSITIZED AUTOXIDATIONS 1525 coloration. No nitrophenyl hydroxylamine was formed in controls without chlorophyll or phenjihydrazine. Ascorbic acid could be used as hydrogen donor instead of phenyl- hydrazine. (This reaction occurs spontaneously in alkaline medium, but not in neutral solution. ) Similar results were obtained with eosin, erythrosin B and some other fluores- cent dyes as sensitizers. Gurevich (1949) later found that oxidation of phenylhydrazine by o-dinitrobenzene can be catalyzed in the dark by "chlorophjdl hemin." The latter was obtained from an alcoholic leaf extract, by successive conversion to pheophytin and introduction of iron (by means of heating with ferric acetate). The green crystalline product, added to a saturated solution of o-dinitrobenzene in ethanol, also containing phenylhydrazine (free base), catalyzed the reduction of dinitrobenzene to o-nitrophenyl hydroxylamine, which can be recognized by its violet salts. The same result can be achieved with Fe^Oj as catalyst, but the catalytic effect of the "chlorophyll hemin" does not depend on con- tamination with inorganic iron salts. Pariser (1950) and Wcigl and Livingston (1952-) studied the chloro- phyll-sensitized reduction of an azo dye (butter yellow) by ascorbic acid — a reaction of the same type as was studied by Krasnovsky and co-workers. According to Pariser, the maximum quantum yield of this reaction in meth- anol is about 0.5. Weigl and Livingston used deuterated ascorbic acid to see whether any deuterium could be found in chlorophyll after the reaction with butter yellow is over. To avoid isotopic exchange with the solvent, dioxane was used instead of methanol; this reduced the quantum yield to a very low value. After about ten deuterium atoms were transferred to the oxidant for each chlorophyll molecule present, analysis of chlorophyll for deuterium content revealed <4% of the amount to be expected if one D atom were to get stuck in chlorophyll in each reduction act. Experiments on deuterium exchange between chlorophyll and heavy water (Weigl and Livingston 1952^) showed that deuterium could not have been lost from chlorophyll during the chromatographic purification; it thus seems that, in this particular oxidation-reduction reaction, the primary photochemical process is not the transfer of hydrogen from excited chlorophyll to the oxi- dant (to be replaced later by hydrogen from the reductant). The primary act can be the transfer of hydrogen from the reductant to chlorophyll (as suggested by Krasnovsky's experiments) — at least, if one is permitted to assume that the same hydrogen is later transferred, by a thermal reaction, to the oxidant, without having been first pooled with other hydrogen atoms in the sensitizer. 6. Chlorophyll-Sensitized Autoxidations (Second Addendum to Chapter 18, Part C) In Volume I (chapter 18) a number of chlorophyll-sensitized autoxida- tions were described. Of these, only one was investigated quantitatively — the ethyl chlorophyllide-sensitized autoxidation of allyl thiourea (Gaffron 1526 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 1927, 1933, cf. Table 18.11). This reaction appeared to have a hmiting quantum yield of about 1, reached when the concentration of the oxidation substrate was sufficiently high and that of the sensitizer sufficiently low. Since one of the reagents— oxygen — is present only in small concentration in the liquid phase, shaking of the reaction vessel is needed to avoid yield deficiency due to local exhaustion of oxygen. The autoxidation of thiourea (substituted for allyl thiourea since it shows a less pronounced dark reaction) with ethyl chlorophyllide as sensi- tizer was reinvestigated by Warburg and Schocken (1949), with a view on using this reaction in a manometric actinometer. Several solvents were tried out. Pyridine was found to be the most satisfactory, acetone (used in Gaffron's work) being too volatile, and dioxane producing complica- tions (autoxidation quantum yields >1, with efficiency increased by traces of copper, and decreased by cyanide) conceivably caused by peroxide forma- tion. Attempts to substitute other reductants, such as hydroquinone, for thiourea, gave no positive results. In addition to ethyl chlorophyllide, protoporphyrin could be used as sensitizer. The green pigment is suitable for the work in red and blue, the red one for the work in green and blue; mixed together, they give almost complete absorption throughout the visible spectrum. A vessel 14 ml. in volume was used, containing 200 mg. thiourea and 2 mg. crystalline ethyl chlorophyllide (or 10 mg. crystalline protoporphyrin) in 5 cc. pyridine. It was illuminated for periods of the order of 15-30 min. through a flat bottom (area, 9 sq. cm.) with light from a large monochroma- tor (beam cross section, 2 sq. cm.). The vessel communicated with a ma- nometer filled with pure oxygen, and was shaken during illumination. Quantum yields of oxygen consumption obtained in this way are tabulated in Table 35.11. Table 35.11 Quantum Yields of Oxygen Consumption in Sensitized Autoxidation of Thiourea (after Warburg and Schocken 1949) Absorption, O2, 7. molecules O2 X, m/x microeinsteins micro moles per quantum 640-660 1.98 1.82 0.92 640-660 0.99 1.02 1.03 505-525 1.14 1.10 0.96 420-450 0.36 0.42 1.16 585-615 1.90 1.79 0.94 585-615 0.95 0.96 1.01 420-450 0.44 0.46 1.04 640-660 1.65 1.97 0.89 640-660 0.825 0.78 0.94 640-660 0.83 0.83 1.00 585-615 1.64 4.55 0.95 505-525 0.88 0.84 0.75 CHLOROPHYLL-SENSITIZED AUTOXIDATIONS 1527 It will be noted that 7 values show a decline with rising light inten- sity; Warburg and Schocken recommended not to use light intensities much in excess of 1 X 10 ~^ einstein absorbed per 15 min. in 5 cc. In part, the decrease can be due to oxygen exhaustion in the illuminated bot- tom layer, which may occur at higher hght intensities despite shaking. The nature and mechanism of the chemical change taking place in this actinometer remains uncertain. Up to 2.5 moles oxygen were consumed per mole thiourea after prolonged irradiation, while not more than 2 should be used if the reaction were : (35.27) CS(NH2)2 + 2 O2 > CNNH2 + H.,S04 About 75% of the amount of sulfuric acid to be expected according to this equation was actually found; qualitative test for cyanamide formation was positive. 17% of the thiourea consumed was recoverable as urea, perhaps because of hydrolysis of the primarily produced cyanamide: (35.28) CNNH2 + H2O > CO(NH2)2 While reacting with a quantum yield close to 1, the system continues to fluoresce strongly, which should account for the loss of probably not less than 10% of the absorbed quanta. That sensitization does not compete with fluorescence indicates (c/. Vol. 1, page 483) that it proceeds by the intermediary of a "long-lived active state" such as the "tautomer," tChl, which is formed in about 95% of all cases of excitation. Warburg and Schocken, having observed that thiourea will react with oxygen in ultraviolet light without sensitizer, concluded that the most likely mechanism of sensitization is transfer of excitation energy from the chlorophyllide to thiourea. However, the mechanism of sensitized photo- chemical reactions is generally different from that of nonsensitized reac- tions because the reaction substrate (thiourea) has no way of accepting a "red quantum," either directly by light absorption or indirectly by transfer from an excited chlorophyllide molecule. The more likely mechanisms of sensitized autoxidation are, therefore, those involving intermediate revers- ible chemical changes of the sensitizer, such as were described in Vol- ume I, pages 514-521 and in section 5 above. Whether the long-lived acti- vated chlorophyllide molecule, tChl, reacts first with thiourea or \vith oxy- gen remains unknown. It also remains a matter of speculation whether the chlorophyllide molecules enter into association with thiourea molecules (or oxygen mole- cules, or both) in the dark. According to table 23. HID, thiourea does not quench the fluorescence of chlorophjdl; and the same is probably true of chlorophyllide fluorescence. Association, if any, must therefore be of the type not affecting fluorescence ; in other words, the photochemical reaction 1528 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 in the complex must be delayed until after the conversion of excited chloro- phjdhde into the metastable form, tChl. Another possible mechanism is analogous to that suggested by Livingston for the sensitized methyl red- phenylhydrazine reaction (cf. section 5 above) ; in this scheme no reaction occurs until the complex of tChl with one reactant encounters the other re- actant (cf. equations 35. 15c and d). In any case, the reaction probably involves a one-step oxidation- reduction, followed by dismutation, or another short chain of elementary steps, bringing the quantum yield up to a value close to 1 (which implies the utilization of two oxygen atoms, i. e., the transfer oi four electrons, or hy- drogen atoms, by a single quantum). Because of the likelihood that the over-all reaction is composed of several successive elementary steps, in- volving intermediate free radicals or other unstable products, the quantum yield could be sensitive to factors such as the nature and purity of the sol- vent, concentration of the reactants, light intensity and temperature. In some more recent papers, Burk and Warburg (1951) abandoned the quest for a quantum yield of 1.0 in the actinometer, since they found that in pyridine, with ethyl chlorophyllide as sensitizer, a quantum yield of close to 1 required not only in a relatively low light intensity, but also the pres- ence of a certain amount of impurities (such as piperidine) in the solvent. In pure pyridine the newly observed quantum yields were between 0.8G (0.1 )ueinstein per minute per vessel) and 0.69 (3 jueinsteins per minute per vessel) . It was stated that by using crystallized pheophorbide (instead of chlorophyllide) and a large volume actinometric vessel (80 ml. instead of 7 ml. liquid), a constant quantum yield of 0.70 could be obtained for light absorptions up to 4 jueinsteins per minute. Still more recently, Warbiu'g and co-workers (1953) again assumed a quantum yield of 1.0 for a pheophorbide-thiourea actinometer of 120 ml., up to a light flux of 1 /ieinstein/min. Schenck (1953) found that an intermediate in the oxidation of thiourea in the actinometer — formed to an extent up to 81% of the oxidized thio- urea—is the sulfinic acid NH=C(HS02)— NH2. Among new qualitative data of the photochemical action of chlorophyll we can mention Pepkowitz's (1943) observations of the photochemical destruction of carotene in the presence of chlorophyll. Because of the dependence of the rate of destruction on the amount of chlorophyll present, it is suggested that chlorophyll takes part in the re- action, and not merely sensitizes it. B. Photochemistry of Chloroplast Preparations* (Addendum to Chapter 4, Part A) In chapter 4, Volume I (page 61) we have described in brief the produc- tion of oxygen by leaf macerates and dry leaf powders in light, first noted * Bibliography, page 1G27. PHOTOCHEMISTRY OF CHLOROPLAST PREPARATIONS 1529 by Friedel, and later described by Molisch and by Inman ; and the enhance- ment and stabihzation of this phenomenon which Hill (1939, 1940) had achieved by providing an oxidant, such as ferric oxalate. As anticipated in chapter 4, the study of the "Hill reaction" has proved one of the most promising approaches to the analysis of the mechanism of photosynthesis. Little doubt now remahis that the "Hill reaction" represents a part of photosynthesis which can be reproduced with nonliving material — al- though as yet only with complex colloidal systems obtained by mechanical disintegration of cells. This part consists of photochemical oxidation of water leading to the liberation of oxygen, undoubtedly with the participation of an enzymatic system closely linked to, and surviving with, the photochemical apparatus. The part lost in the preparation of the chloroplasts is that concerned with the use of carbon dioxide as acceptor for the hydrogen taken away from water. Only relatively strong oxidants, such as ferric salts or quinones, can be used as hydrogen acceptors in the Hill reaction with a good quantum yield. True, it has been found possible to couple this reaction, through the intermediary of pyridine nucleotides, to enzymatic systems permitting the reduction of pyruvate to lactate, or its reductive carboxylation to mal- ate, or (with the help of ATP), the reduction of phosphogly eerie acid to phosphoglyceraldehyde. However, so far, this nearest approximation to photosynthesis outside the living cell could be achieved only with a very low quantum yield ; there- fore it remains an open question whether this reaction in vitro represents a significant approach to the reconstruction of the actual mechanism of photosynthesis in vivo. The interpretation of the Hill reaction as oxidation of water is supported by oxygen isotope tracer studies. In chapter 3 (Vol. I, page 54) we described tracer experiments with heavy oxygen which directly demonstrated that all oxygen evolved in normal photosynthesis originated in H2O (and none in the oxidant, CO2). In the case of the Hill reaction, many oxidants — such as Fe[(CN)6]"'"' — contain no oxygen at all; furthermore, the ratio of the amount of oxygen liberated to the amount of oxidant reduced agrees with the assumption that oxygen comes from the oxidation of water. An exception is chromate, where oxygen could conceivably come from the anion, and where the amount of liberated oxygen, as found in Holt and French's experiments, was much smaller than stoichiometrically expected. Holt and French (1948^) made a mass spectroscopic analysis of the oxygen evolved in the Hill reaction from normal water and from water enriched in 0(18), using Hill Electrolysis Oxidant reaction of water 0 . 02 M ferricyanide 0.39 0.38 0 . 02 ilf ferricyanide 0.39 0.38 0 . 02 ilf ferricyanide 1.4 1.3 0.02 M ferricyanide 0.84 0.83 0 . 0067 M quinone 0.62 0.61 0.0035 M dichlorophenol-indophenol 0.57 0.62 0.00221 MK2Cr04 0.52 0.49 1530 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 for comparison oxygen produced from the same water by electrolysis. The results are shown in Table 35. Ill; they confirm that in all studied cases of the Hill reaction — including that with chromate — the oxygen came from water. (In the experiment with chromate, pH 8.3 borate buffer was used to retard the isotopic exchange of oxygen between water and chro- mate.) Table 35.III Isotopic Composition of Oxygen Evolved bt Illuminated Chloroplasts (after Holt and French, 1948^) 100[O(18)O(16)]/[O(16)O(16)l oxygen produced by Water used Normal Enriched 1. Can Carbon Dioxide Serve as Oxidant in Hill Reaction? Claims of having obtained oxygen evolution from water by isolated chloroplasts, with simultaneous reduction of carbon dioxide, if not to a carbohydrate, at least to formic acid, have been made by Boichenko (1943, 1944). Since these mvestigations have only been published in Russian, we will describe them in some detail, despite the fact that techniques used were rather primitive and the results not too convincing. Boichenko further claimed that chlorophyll preparations can reduce carbon dioxide not only in light, "photosynthetically," at the expense of water, but also in the dark, "chemosynthetically" (with hydrogen as re- ductant) . Boichenko thought (quoting Lubimenko) that under natural conditions the medium surrounding the chloroplasts has a very low value of rH- — in other words, is strongly reducing. Since photosynthesis produces both oxidation products (O2) and reduction products (carbohydrates), the reducing properties of the medium can be understood if the oxidation product (oxygen) is removed much more efficiently than the reduction products (sugai's) ("removal" may mean translocation or a chemical change, such as polymerization of reducing sugars to sucrose or starch). The accumulation of reducing compounds around the chloroplasts is thus a consequence of photosjoithesis, and, as such, more likely to inhibit than to stimulate it. Boichenko considered, to the contrary, that a low value of rH is a -prerequisite for photosjTithesis, and attempted to imitate CAN CARBON DIOXIDE SERVE AS OXIDANT IN HILL REACTION? 1531 nature by suspending chloroplasts in an artificial medium of low rH. To simulate natural conditions also in respect to pH and osmotic pressure, she followed the advice of Kuzin, and used as medium a solution of a reducing sugar (glucose, fructose) in satu- rated solution of basic magnesium acetate. What she called "chloroplasts" was a preparation obtained simply by shredding leaves of clover with scissors, washing in a 0.5 mole/liter glucose or galactose solution, and passing the suspension through a paper filter under suction. The filter paper carrying the dark-green precipitate was cut into strips, and immersed into the above- mentioned reducing medium, to which bicarbonate was added. The pH of the solution was 7.5-8.5. Boichenko claimed that if the medium contained 0.1% reducing sugar, giving rH <9.9 (estimated by means of redox indicator dyes), oxygen was evolved, and carbon dioxide used up, upon illumination of the green strip. The results are sum- marized in Table 35.IV. The table shows that with fructose and, to a lesser extent, with glucose, the rH decreased and the pH increased upon exposure to light, and oxygen was liberated. With hydrosulfite and glycolaldehyde, the pH and rH changes were in the opposite direction, and no oxygen evolution was observed. Table 35.IV "Photosynthesis" with Chloroplast Films on Filter Paper (after Boichenko 1943, 1944) rH pH O2 evolution Medium Before After Before After No sugar >14.4 14.4 8.0 8.0 None 0.1% maltose <14.4 <9.9 8.0 8.0 None 0.1% glucose 9.9 5.2 8.0 8.3 Slight 0.1% fructose <9.9 <5.2 8.0 8.5 Strong 0.1% glycolaldehyde 5.2 >14.4 8.0 7.5 None 0.1% hydrosulfite <5.2 25.2 8.0 8.0 None The fact that the rH of hexose solutions did not increase upon illumination was taken by Boichenko as proof that these sugars were not used up, and she concluded that they must have served as catalysts for the oxidation of water by carbon dioxide — which was also indicated by the increase in pH. She recalled in this connection the observation of van Niel and Gaffron that purple bacteria and "adapted" green algae can use organic hydrogen donors to reduce carbon dioxide. Inverting the reaction sequence suggested by them (which involved water as the primary, and organic compound as ultimate hy- drogen donor), she suggested that organic H donors (such as fructose) serve as primary reductants (even in normal photosynthesis!), and that the oxidation of water (and libera- tion of oxygen) are caused by a secondary reaction between the oxidized sugars and water. It hardly needs pointing out that using carbohydrates as primary hydrogen donors to reduce carbon dioxide in the photochemical stage of photosynthesis leaves to dark stages a reaction which must require all the energy of photosynthesis, namely, the oxidation of water to oxygen by the oxidation product of a carbohydrate (such as a gluconic acid). The hypothesis is therefore utterly implausible. Later, Boichenko (1947, 1948) described the "chemosynthetic" activity of the same 1532 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 chloroplast-impregnated filter paper strips: their capacity to catalyze carbon dioxide reduction by hydrogen in the dark. She found that these chloroplast preparations con- tain a "hydrogenase" capable of taking up molecular hydrogen and using it for the reduc- tion of various substrates (O2, CO2, etc. — cf. the reactions of the hydrogen bacteria, Vol. I, page 116, and of "adapted" algae, Vol. I, page 130). The tests were made by observ- ing the decoloration of methylene blue (r/. Vol. I, page 131) by the chloroplast-covered paper strip in an atmosphere of hydrogen; the uptake of hydrogen was confirmed by manometric measurements. The activity was highest with those species (Trifulium repens, Chenopodiurn album) whose chloroplasts did not easily disintegrate into grana. Instead of methylene blue, oxygen could be used as hydrogen acceptor: with 0.0- 0.67% O2 in the air, from 75 to 288 mm.' of oxygen were taken up l)y 25 mg. of chloro- plasts, together with from 0 to 200 mm.' hydrogen. At 0.6-1.5% O2 in the atmosphere, the ratio AH2/AO2 was 1.06-1.17; at [O2] > 2%, the uptake of hydrogen declined rapidly. Boichenko interpreted this as evidence of deactivation of the hydrogenase by oxygen (cf. Gaffron's observations, Vol. I, page 132). When [O2] approached 10%, the uptake of oxygen also ceased. The gas exchanges shown in Table 35. V were observed when carbon dioxide and hydrogen were offered to the chloroplast films in the dark. According to this table the uptake of CO2 occurred without (or with only a slight) uptake of oxygen (without which no chemosyn thesis had been observed in bacteria and algae!). Since the ratio AHj/ ACO2 is close to 1.0 (rather than 2.0, as in Gaffron's adapted algae), the reaction reminds one not so much of "chemosynthesis" of carbohydrates, as of the synthesis of formic acid from H2 and CO2 by E. coli (Woods, cf. Volume I, pages 185, 208). The uptake of CO2 and H2 was notetl even in the absence of oxygen, but was stimulated by its presence, even though only little oxygen was taken up. (It thus looks as if, when both CO2 and O2 are present, the former is used as oxidant in preference to the latter; a remarkable selection, considering the difference of oxidizing potential. However, a similar situa- tion exists in photosynthesis, where, too, photoxidation gets under way only when the cells are deprived of carbon dioxide.) Table 35.V Gas Uptake by 25 Mg. Chloroplasts on Filter Paper in Darkness, in Hj -|- CO2 (after Boichenko 1947, 1948) C02in atmosphere, % Gas exchange AH2 — AO2, mm.' - ACO2, mm.' — AHj, mm.' ( AO2 + ACO2) 2.0 0 250 238 0.99 2.7 0 338 413 1.22 3.0 0 375 388 1.03 0.6 25 75 100 1.00 1.6 13 200 200 0.99 3.0 0 375 263 0.70 Boichenko measured the effect of CO2 concentration on the rate of gas uptake ( ACO2 + AH2) and found that the uptake (by 25 mg. chloroplasts) increased from 12.5 mm.3 in 5 min. at 0.6% CO2 to 375 mm.' at 6% CO2, but declined rapidly at the higher carbon dioxide pressures (e. g., to 75 mm..» at 10% CO2). The ratio AH2/ ACO2 remained constant (1.0 to 1.1) up to 6% CO2, but only carbon dioxide was taken up at 10% CO2. The temperature coefficient of the gas uptake ( ACO2 + AH2) was Qio ^ 2 between 20 and 35° C.; above 35° C., the hydrogenase was destroyed. The rate of the chloroplast- catalyzed reduction of methylene blue by hydrogen appeared to be independent of tem- perature. (Table 35.VI.) CAN CARBON DIOXIDE SERVE AS OXIDANT IN HILL REACTION? 1533 Table 35. VI Temperature Effect on CO2 Reduction and Methylene Bltie Reduction by Chloroplasts on Filter Paper Strips (after Boichenko 1947, 1948) Temperature, A(H2 + CDs), Relative rate of MB ° C mm.' reduction 20 125 15 25 175 16 30 250 16 35 363 17 A very peculiar observation was made: when the reaction was continued long enough, it reversed itself, and both hydrogen and carbon dioxide were rapidly liberated again! This occurred very early at temperatures above 35° C; consequently at 35° (and even at .30° C.) no complete consumption of hydrogen or carbon dioxide could be achieved. At 25° C, only hydrogen came out again, and at <25° C, no reversal of the reaction took place at all. Boichenko associated these peculiar phenomena with the thermodynamic reversilMlity of the reaction H2 + CO2 <=* HCOOH, but it is thermody- namically impossible for a reaction to proceed first in the one and then in the other di- rection. That formic acid actually was formed in these experiments was confirmed by chemi- cal tests (reduction of AgNOs and HgCl2). The reducing power resided in the film, and not in the solution; the reductant (formic acid?) thus must have been present in "bound form." The amount of HgCl2 reduced agreed with the (manometrically determined) amount of hydrogen taken up by the chloroplast film. Boichenko saw in these experiments an imitation, with isolated chloroplasts, of the first step in the reduction of carbon dioxide in photosynthesis, but both the reliability of the experiments and their interpretation are doubtful. Vinogradov, Boichenko and Baranov (1951) applied C^^ tracer technique to the products of carbon dioxide fixation by Boichenko's chloroplast preparations in the dark, with H2 as reductant, and found the tagged products entirely extractable by hot water,. with up to 90% of it precipitable by barium chloride; about 75% of the precipitate con sisted of uronic acids, the rest were highly carboxylated nonreducing acids. Boichenko and Baranov (1953) studied C" uptake by the same preparations in light under anaerobic conditions. In the presence of hydrogen the uptake was 0.01 relative units in the dark anil 0.032 relative units in light (as against 0.067 and 0.015 relative units in the presence of 1.25% oxygen). In pure nitrogen the uptake was negligible (0.0003 relative units in light, 0.004 in dark). The authors saw in these results the indi- cation that the dark "chemosynthetic" fixation of CO2, which requires oxygen, is re- placed in light by "photoreduction," similar to bacterial photosynthesis, which occurs only under anaerobic conditions. Only a fraction of C*02 fixed in light was found in carboxylic groups — an observation which was considered evidence that carbon dioxide was actually reduced (and not merely taken up in a carboxyl). Franck (1945) noted a stimulating effect of carbon dioxide on oxygen liberation by chloroplasts in the absence of added oxidants (i. e., on the Friedel-Molisch phenomenon), and thought at first that this effect indi- cated the capacity of chloroplasts to use carbon dioxide as oxidant, albeit with a very low efficiency. Subsequent experiments with C(14) by Franck and Brown (1947) caused him to abandon this interpretation. 1534 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 Franck (1945) determined small amounts of oxygen by the quenching of phos- phorescence of dyes adsorbed on silica gel (page 851 ). Figure 35.8 shows the time course of oxygen liberation by a chloroplast preparation, both in the absence and in the presence of carbon dioxide. The oxygen liberation reaches a maximum immediately after the be- gmning of illumination; after about 0.5 minute, it begins to decline. After a few min- utes of this decline, a steady level is reached, and the oxygen liberation remains at this O X E E UJ tr (/) if) UJ cc a. 80 - 60 'l\ ■o o JC w 40 \ O V *i 01 ♦- 3 C E *2 ?0 r \ lO r V ■^ t /-"^^ / J Illumination /illumination 1-/ 1 1 n / 1 1 1 1 3 4 5 0 TIME, nninutes 80 ' n 60 40 " \ O a> Q. je w o T3 90% activity preserved after 2 hrs. at 0° C, about 25% after 10 min. at 35° C.) . At pH 5.0 or 7.5 the losses were greater (>35% after 2 hrs. at 0° C. and >80% after 10 min. at35° C). Kiimm and French (1946) found that when leaves were preilluminated for several hours before maceration they not only showed a higher photo- chemical efficiency (as described in section 2 above), but also kept better in storage. However, Clendenning and Gorham (1950^) noticed no such sta- bihzing effect in experiments with chloroplasts from Swiss chard. 100 > o < Fig. 35.9. Stabilization of chloroplast activity by methanol (after Milner, French, Koenig and Lawrence 1950). Zubkovich and Andreeva (1949) looked for any parallelism between the decay of photochemical activity and changes in the absorption spectrum of chloroplasts, their fluorescence, the concentration of chlorophyll, and its association with proteins. As the activity of sugar beet chloroplasts de- creased with time (e. g., from 43 mm.* O2 per hour per milhliter of suspension to 9 mm.3 after 2 days at 2-3° C; or to 6 mm.* after 2 hours at 30° C), the total chlorophyll content was found unchanged (0.49 mg. per cm.*); but the part of it extractable by 60% acetone (which is a measure of the chlorophyll attachment to protein), declined from 37% to 25 and 29%, respectively. The shape of the red absorption band was unchanged, and the suspension still fluoresced. With material from Phaseolus vulgaris, no change was noted also in the percentage of chlorophyll extractable with aqueous acetone, even after the photochemical activity had dropped, after 1548 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 Storage, to 50 or 20% of the initial value. The results show that the loss of activity must be due to the deterioration of enzymatic components and not to changes in the chlorophyll-protein complex. Some success has been achieved with chemical stabilization. Milner, French et al. (1950) noted a preserving effect of sucrose and of ethylene glycol, Holt, Smith and French (1950) studied the effect of various concen- trations of propTjlene glycol at 0° C. and pH 6.5. The best results were ob- tained with a 10% solution ; the activity loss in this medium was about 100 > t- o < _i < 2 75 50 25 V Chloroplasts ■o- Dispersion o— -i_ 3 4 DAYS Fig. 35.10. Inactivation of chloroplast fragments and of their colloidal dis- persions (after Milner, French, Koenig and Lawrence 1950). one half of that without propylene glycol. Diethylene glycol, n-propyl alcohol, isopropyl alcohol, glycerol (10%) and ethanol (10-50%,) had only a weak effect, or no effect at all. Strychnine, which stimulates the dye re- duction by chloroplasts (Table 35.XI), had no preserving effect. According to Arnon et al. (1954), whole chloroplasts prepared in ethyl- ene glycol show no C*02 fixation in Hght (in contrast to those prepared in 0.5 M glucose or 0.35 M NaCl) . A much stronger stabilizing effect than that of propylene glycol is produced by methanol (fig. 35.9). This was noted by Milner, French et al. (1950) when studying fractional coagulation of colloidal chloroplast dis- persions. They observed that not only did these sols withstand consider- able methanol concentrations without coagulating, but their photochemical efficiency was stabilized. PREPARATION, PRESERVATION AND ACTIVATION OF CHLOROPLASTS 1549 Colloidal chloroplast dispersions containing 15% methanol can be stored at —5° C. without freezing; they then retain 50% of their initial activity after 10 days. Ethanol in equimoleciilar quantities had the same stabihzing effect at 5° C; but a lesser one at higher temperatures. The more concentrated dispersions kept better in 15% methanol than the dilute ones. The loss of activity is fastest in preparations the absolute efficiency of which is high (fig. 35.10) ; in line with this, the effect of meth- anol is much stronger in colloidal chloroplast dispersions than in crude sus- pensions of chloroplast fragments. The most effective stabilization was achieved by macerating the leaves in 15% methanol, maintaining this composition of the medium in all sub- sequent operations, and keeping the temperature down to —5°. Hy- drosols obtained in this way retained up to 50% of the activity of crude suspensions — a much smaller loss than is commonly suffered by dispersion in the absence of methanol. A new, promising approach has been opened by McClendon (1944) and McClendon and Blinks (1952) Mith the finding that chloroplasts of red algae can be prevented from losing the phycobilins, and their Hill activity preserved, by preparing them in high -molecular solutions, e. g., containing 0.4-0.8 g./ml. of Carbowax 4000 (or 8000). The deterioration of chloroplast material is accelerated by illumination, particularly in the absence of oxidants (Aronoff 1946^; Milner, French et al. 1950; Arnon and Whatley 1949''^). To what extent photochemical deactivation is prevented when appropriate oxidants (such as ferricyanide or quinone) are present, so that the Hill reaction can take place, is not clear. A protective effect of a sensitization substrate on the sensitizer is a well- knowai phenomenon {cf. Vol. I, chapter 19). Warburg and Luttgens (1946) noted it with chloroplasts and quinone, but Holt, Smith and French (1950) reported that the several chloroplast fractions obtained by frac- tional centrifugation of a colloidal suspension of chloroplastic matter, lost their photocatalytic capacity in light, in presence as well as in the ab- sence of the ferric oxalate-ferricyanide mixture. Spikes et al. (1950) gave a curve for the inactivation of spinach chloro- plasts for the photoreduction of ferricyanide by prolonged illumination. (c) Activation by Anions Warburg and Luttgens (1944 2) first noted that after the capacity of spinach or beet chloroplast suspensions to evolve oxygen with quinone has been lost by dialysis, it could be restored by potassium chloride. They referred to the chloride ion as a "co-enzyme" of the Hill reaction, and found it to play a similar role also when ferric salts were used as oxidants instead of quinone. However, this stimulation is not specific for chloride. War- burg and Luttgens (1946) themselves found that bromides had a simi- 1550 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 lar, if slightly weaker, effect; iodide and nitrate had a small effect, while fluoride, rhodanide, sulfate and phosphate were without influence. Arnon and Whatley (1949^) noted that the importance of chloride de- pended on the nature of the oxidant used. For example, the total oxygen yield obtainable from a certain quantity of quinone was reduced from 100% of the stoichiometric equivalent in the presence of 0.01 M KCl, to 36% without KCl; with ferricyanide, the reduction was from 100 to 19%; but 100 KNO. ^0 a NaF „^ o — Q — o — * ,§--9::r:2 — •— * — • • control 6 10 14 TIME, minutes 18 22 Fig. 35.11. Activation of Hill reaction in chloroplast preparations from chloride- free cells by salts (after Arnon and Whatley 1949'^). with phenol indophenol, the change was only from 100 to 90%. However, chloride proved equally important for obtaining a high initial rate of the reduction of all three oxidants. Arnon and Whatley (19492) pointed out that CI" is not a necessary plant nutrient; it is therefore unlikely that it is a natural component of the photosynthetic apparatus (as suggested by Warburg and Liittgens). They grew sugar beet and Swiss chard in chloride-free nutrient solutions. PREPARATION, PRESERVATION AND ACTIVATION OF CHLOROPLASTS 1551 The plants grew well (and, implicitly, photosynthesized normally), but the chloroplasts obtained by the disintegration of their leaves (which, as expected, had no measureable Cl~ content) evolved only very little oxygen upon illumination in the presence of quinone. Addition of protoplasmic fluid from the same plants (which, too, was chloride-free) did not activate them for the Hill reaction, but addition of potassium chloride brought it under way. Bromide had the same effect, while effects of KNO3 and KI were much weaker, and NaF had no effect at all (fig. 35.11). Sulfate, phosphate, thiocyanate and acetate also were without influence. The con- centration of chloride required to fully activate chloride-free chloroplasts t 60 > ,/ Xc Dispersion in mefhanol Cone " — 25 % - 20% 0 < N 01 \ " < _; 0 40 cc 0 _j X u u. 0 K 1 5 % ^^^ 30% 0% 20 1 1 1 0.05 0.10 MOLAR KCl 0.15 0.20 Fig. 35.12. Effect of chloride on photochemical activity of chloroplast dispersions (after Milner, Koenig and Lawrence 1950). was about 7 X 10 ~^ mole/liter — an amount which could not have been left undiscovered in CI --starved plants, or lost in chloroplasts extraction. Arnon concluded that chloride is not a natural component of the photosyn- thetic apparatus, but is required to prevent photochemical deactivation of separated chloroplasts. The (almost) irreversible inactivation of chloro- plasts caused by preillumination of chloride-free chloroplast suspension in the absence of an oxidant could in fact be prevented by the addition of chloride. The chloride exercised a certain protective effect also on the deterioration of chloroplasts in the dark (see Warburg and Liittgens (1946); French and Milner (1951) were unable to stabilize disintegrated chloroplasts by the addition of phosphate buffer and KCl, as recommended by Warburg). Holt, Smith and French (1950) found that the deactivation of chloroplast dispersions in light (which, in their experiments, occurred in presence as well as in the absence of an oxidant) could not be reversed by chloride. 1552 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 Milner, Koenig and Lawrence (1950) investigated the effect of added salts on colloidal chloroplast dispersions ; they found that in the presence of methanol, addition of certain salts markedly increased the photochemical ef- ficiency of the material, but that no such activation was noticeable in the absence of methanol. Figure 35.12 shows that in water, quantities of KCl up to 0.2 mole/liter had no effect. (If [KCl] was >0.2 mole/liter, de- activation resulted; precipitation followed at about 1 mole/liter.) When KNOj ADDITION OF SALT TO DISPERSION IN 20% METHANOL v/A K-PHOS. V////////A LiCI ///////////////////A KBr '////////////////////////////A NaCI W/////////////////////////////A NH4CI V//////////////////////////////A KCl ■^'•;>^!'i.^;c■^;^:•:■:■^x■:■t■^>>:•:-:■:■:•^>>c«xx.:.;<x^ NQoSOa '//////////////////////////////////////A'///////////A///////\ H4)2S04 v////////////////v/////////////////////////////////////''y/^^- K2SO4 v////////////////////////////////////////////////////////////////////^^^^ 1 1 1 0 80 20 40 60 7o INCREASE IN ACTIVITY WITH 0.05 M SALT Fig. 35.13. Activation of chloroplast suspensions by different salts (after Milner, Koenig and Lawrence 1950). 25% methanol was present, addition of 0.1 M KCl doubled the photo- chemical efficiency. However, it must be taken into account that the chloroplast dispersion used had been only \i as active as the suspension from which it was prepared ; thus, KCl merely restored a part of activity lost by dispersion. Reactivation was accompanied by turbidity; after- wards, the active material could be separated by mild centrifugation. In other words, activation was associated with visible coagulation of par- ticles. The activation effect was strongest in the most concentrated dis- persions. It was not permanent : after 24 hrs. the activity was back to the level it had before the addition of KCl. The loss of activation was even faster when NH4CI had been used instead of KCl. Phosphate buffer (0.01 M) produced a passing activation similar to that caused by other salts. PREPARATION, PRESERVATION AND ACTIVATION OF CHLOROPLASTS 1553 Fig. 35.13 shows that, according to Milner, Koenig et al. chloride is not the best "reactivator" of chloroplast dispersions; it is exceeded by sulfates (found inactive by Warburg and Liittgens, and Arnon and What- ley). However, the activating effect of (NH4)2S04, for example, was ex- tremely short-lived, and was soon replaced by inactivation. Milner, French et al. also investigated the effect of some cations, vnih. and without addition of "versene" (ethylene diamine tetra-acetic acid, a chelating agent). In these experiments, versene, which is an acid, was neutralized by KOH to 7>H 6.5. The potassium salt of versene alone had a marked reactivating effect — probably similar to that of other potassium salts. Addition of versene to a divalent salt solution, which in itself had an activating effect, produced no important changes; but its addition to a 5 X 10 ~^ M CUSO4 solution (which, by itself, was strongly poisonous) leads to effective reactivation. (No reactivation was possible in 3 X 10 ~^ M CUSO4.) Versene also reactivated dispersions inactivated by HgCb. Ferrous sulfate had no activating effect. Gerretsen (195O0 described a strong enhancing effect of Mn + + ions on the change of redox potential, which occurs in crude chloroplast suspensions from Avena upon exposure to light (without added oxidants) under aerobic conditions. He suggested an interpretation in terms of manganous ions playing a role in the enzymatic mechanism of the liberation of oxygen. Clendenning and Gorham (1950^) found that the initial rate of the Hill reaction with quinone was two or three times higher with crude suspensions than with separated chloroplasts, and that repeated washing reduced it still further; only about }4> of the loss caused by washing could be restored by the addition of potassium chloride. Gorham and Clendenning (1952) investigated the effect of anions on chloroplast activity in more detail, and arrived at conclusions which dif- fered from those of both Warburg and Arnon. They grew spinach, Swiss chard, beet and millet in ordinary soil, and in chloride-containing and chloride-free nutrient solutions. Chloroplasts prepared from these plants were analyzed for chloride content, and their photochemical activity studied with various oxidants. The activity of crude suspensions was about the same in the three kinds of plants. Separation from cell sap and cytoplasm affected strongly the activity of preparations from plants grown in nutrient solutions — either with or without chloride — but material from soil grown plants could be washed extensively without inactivation. The activity lost by washing could be restored by anions, the order of effectiveness being Cl~> Br~ > CN~ > Hoagland's salt mixture (nitrate -{- sulfate -f phos- phate) > NO3- > I- > F- > CIO3- > BrOs- > IO3- > SO4— (no effect). Ascorbate and thiocyanate caused inhibition. Maximum restoration could be achieved with [Cl~] > 3.3 X 10~^ mole/1. Pre-illumination ex- 1554 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 periments did not confirm Arnon's hypothesis that chloride protects chloro- plasts from light injury, since to restore the photochemical activity of chloride-free chloroplasts it proved enough to add chloride after the pre- illumination. The chloride thus acts directly on the photochemical reac- tion. Its effect was found to be proportionately the same at all light in- tensities. The pH optimum of the Hill reaction was found to shift in the presence of chloride to higher values (from pYl 6.5 to 7.1 with Hill's mix- ture and 7.5 with quinone) (fig. 35.22E). Gorham and Clendenning em- phasized the similarity of the anion effect on the Hill reaction with the ef- fects observed in other enzymatic reactions, such as starch hydrolysis by dialyzed amylase or respiration of washed root discs, and suggested that the same explanation should apply to all of them. The fact that chloride stimulation of chloroplast activity occurs only at pR > 6 (fig. 35.22E) may explain why no chloride influence could be found in the Hill reaction (as well as in photosynthesis) of whole Chlorella cells (c/. part C below). To sum up, the nature of the salt effect on the Hill reaction is not yet clear; perhaps, at least in part, this effect is due to coagulation of smaller colloidal particles into larger ones which are more active photochemically {cf. next section) and less subject to deactivation {cf. section (b)). (d) Fractionation of Chloroplast Material Since the Hill reaction permits an important part of the photosynthetic apparatus to be separated from the living cell, biochemists would like to do with this part what they did so successfully with many other enzymatic systems— fractionate it by methods of protein chemistry and concentrate the photocatalytic principle in a small fraction ; or, if the catalytic system is complex, separate it into its components. So far, this undertaking has not been successful. Several difficulties are responsible for the lack of progress. In the first place, the chloroplast-protein material is not water-soluble. In other words, it does not fall apart, in contact with water, into more or less uniform building stones of macromolecular size. By mechanical frag- mentation, chloroplastic matter can be converted into a water-born sus- pension; and the fragm^ents can be disintegrated still further by one of the methods mentioned in section 3(a). Sols with colloidal particles of more or less uniform size can then be obtained by fractional centrifugation. However, there is no reason why such particles should consist of chemical or functional units; it is not even certain that they are identical in composi- tion and structure (and not merely in size) . This difficulty is common to all work with water-insoluble proteins. The chloroplastic matter, in addition to containing insoluble proteins PREPARATION, PRESERVATION AND ACTIVATION OF CHLOROPLASTS 1555 also includes a large amount — of the order of one third of the total — of lipoidic materials, including carotenoids and phospholipides (c/. Vol. I, chapter 14). It is likely — although it cannot be proved at present — that at least some of these materials are essential and that any fractionation which leaves them behind will lead to a loss of photocatalytic activity. The Hill reaction, although it must be simpler than complete photosynthe- sis, still is a complex process, partly photochemical and partly enzymatic. Its occurrence may be bound to the preservation of a definite pattern of pigments, prosthetic groups of one or several enzymes, proteins and lipoids. The "photocatalytic activity" of chloroplastic material usually is de- fined as the (initial) rate of the Hill reaction with a given oxidant, related to unit amount of chlorophyll. This activity can only be increased if : (a) the activity of the preparation as measured before fractionation was limited by an enzymatic component (i. e., if the measurement was made in satu- rating light, and with saturating concentration of the oxidant) ; and (b) if a fraction can be obtained in which this limiting enzyme is enriched in relation to chlorophyll. If chlorophyll and the limiting enzyme are enriched together, activity tests related to unit chlorophyll content will show no change at all. It is, therefore, desirable to check the results of fractionation also by measuring the photochemical activity per unit mass (or per unit amount of nitrogen, if only protein fractionation is considered). Following is a brief review of the chloroplast fractionation experiments. French, Holt, Powell and Anson (1946) made a preliminary study of the effects of various treatments (freezing, lyophylizing, disintegration by supersonic waves, and centrifugation) on the photochemical activity of a chloroplast suspension. They considered the results as indicating the possibility of successful fractionation of the chloroplast material and con- centration of active ingredients. Holt, Smith and French (1950) (c/. Holt and French 1949) attempted fractionation by high-speed centrifugation, acid coagulation, salting-out, and adsorption and elution. They used a spinach chloroplast suspension in 10% propylene glycol; it was disintegrated by supersonic vibrations. When the resulting colloidal dispersion was fractionated by centrifugation, the first precipitate, obtained after 5 min. at 7600 g. and presumably com- posed of the largest particles, was found to contain one half of the total N present; its activity (per unit weight of N) was lower than the average, while that of the supernatant was twice as high. Centrifuging the super- natant for 10 min. at 8600 g. precipitated one half of the remaining nitrogen, the activity per unit N remaining the same in the precipitate and in the supernatant. The latter was centrifuged 30 min. at 3200 g., again pre- cipitating about one half of the remaining N content. In this case, the 1556 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 sediment was 3.4 times more active per unit N weight than the original ma- terial, while the supernatant was only one half as active. The highest ratio of active material to total protein was thus found in the precipitate from the third centrifugation. However, an increase of specific activity by a factor of 3 or 4 is disappointing compared to the results obtained in experiments with water-soluble enzymes. Fractional coagulation by acid of the same (supersonically dispersed) chloroplast material produced, in the range pH 5.3-6.0, a series of precipitates which, when resuspended at pH 6.5, exhibited no marked difference in photochemical activity. Total precipitation at pH 4.6 led to complete loss of activity. Salting-out, e. g., with 30-40% Na2S04, also produced no marked enrichment. Detergents, such as dupanol and vetanol, which are of help in solubilization and frac- tionation of some proteins, destroyed photochemical activity. French and Milner (1951) reported that addition of 1 volume saturated (NH4)2S04 to 4 volumes of a dispersion of chloroplast material precipitated nearly all green material, leaving in solution some colorless proteins^not more than Vs of the total protein content. The resuspended precipitate showed only a slight photochemical activity, whether related to chlorophyll or to nitrogen content. A green precipitate could also be obtained by acidification, but its activity was similarly poor. Precipitates obtained by addition of salts to methanol-stabilized dispersions are more active than the starting ma- terial, but this appears to be due to activation by coagulation rather than to enrichment (cf. below). Attempts to coagulate dispersed chloroplast material with alcohols were unsuccessful; even in 95% ethanol, precipitation occurred only after several hours; the alcoholic solution was photochemically inactive. Simi- lar experiments with methanol later led French and co-workers to the dis- covery of the stabilizing influence of methanol on chloroplastic matter, which was described in section (c). Adsorption of the chloroplast dispersion on a column (Supercel) at pH 5.5 gave, according to Holt, Smith and French (1950), a green band, which could be partially eluted in neutral buffer or distilled water. However, the easily eluted fraction had the same activity as the noneluted residue. Ac- cording to French and Milner (1951), fractional adsorption on Fuller's earth or charcoal leaves, as the last, unadsorbed residue (a few % of the total) a dispersion that may be 2-3 times more active than the original ma- terial. Aronoff (1946) found that when "grana" from spinach chloroplasts were precipitated at 6700 g., the remaining supernatant, which was clear but showed a Tyndall cone, had a remarkably high photochemical activity with quinone (about ten times that of the precipitate, related to equal chloro- phyll amounts) ; its activity remained almost constant for about 2 hrs. at 19 klux. Dried in a lyophylizer and redissolved, the product kept its activity; PREPARATION, PRESERVATION AND ACTIVATION OF CHLOROPLASTS 1557 in contrast to the precipitated chloroplast, this material could be stored for weeks in the dry state, in the cold, without losing its activity. The dialysis of this supernatant into distilled water precipitated additional green ma- terial, but the solution still showed the chlorophyll band. The yellow color appeared to be due to a protein with absorption bands at 337 and 270 mju ; the yellow "prosthetic group" of this protein was not separable by dialysis. Spectrally, it appeared to be a flavone, but it did not show the oi-ange color upon reduction, which is characteristic of flavones. The high activity of the supernatant, obtained according to Aronoff's procedure, was not confirmed by Clendenning and Gorham (1950^, using Hill's solution. With quinone, they noted an increase of efficiency (related to unit chlorophyll amount) compared to the green precipitate by not more than a factor of 2. French and his co-Avorkcrs (Milncr et al. 1950, and Milner 1951) macer- ated leaves of Beta in a water-ice mixture in a Waring Blendor. The slurry was filtered through muslin and the filtrate centrifuged 1 min. at 12,000 g., all at a temperature not above 2° C. The sediment, which consisted of whole and broken chloroplasts, was resuspended in distilled water (2 mg. Chl/cm.^). Attempts to disperse this suspension by means of a colloid mill were unsuccessful. Ultrasonic wave generator (250 watt) was then used, with precautions to prevent the temperature from rising >5° C. The undispersed material was reprecipitated at 12,000 g. ; the green supernatant was clear, but showed a strong Tyndall cone. In this way, 5-10% of the original chloroplast material could be con- verted into a colloidal dispersion. Additional 10-15% could be dispersed by grinding the sediment from the last centrifugation with sand. The dispersion was about 50% less active photochemically (per unit chlorophyll amount) than the original material. A more complete dispersion could be obtained by forcing the chloroplast suspension through a small opening (a needle valve from an ammonia cylinder) under high pressure. 40 cc. of suspension were placed in a pre- cooled steel cylinder connected to the valve, and pressed through the valve by a 60-ton hydraulic press. Despite cooling with ice, the material came out at about 15° C. It was diluted to 0.5 mg. Chl/cm.^ and centrifuged at 12,000 g.; the dark-green supernatant now contained as much as 60-70% of the total chlorophyll, but its photochemical activity was only 25% of that of the original material. This loss was found to decline with the pres- sure used and the duration of exposure to this pressure. At low pressure, less material is dispersed, and less activity is lost (fig. 35.14). The supe- riority of the dispersion obtained by the valve method over that obtained by supersonics is illustrated by figure 35.15. Dispersions which were obtained by passage through the needle valve 1558 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 at 2000 psi sedimented slower than those obtained by ultrasonic waves; fractional sedimentation showed that the photochemical efficiency de- creased with decreasing particle size. No evidence was obtained in sedi- 100 80 UJ 3 < 60 - 40 20 \ Activity \ \ N Dispersed chlorophyl _i_ 5000 PSI 10,000 FOR DISPERSION 15,000 20,000 Fig. 35.14. Loss of activity (per unit chlorophyll amount) and efficiency of dispersion (percentage of chlorojjhyll dispersed by valving) as functions of pres- sure (after Milner, Lawrence and French 1950). 2 4 6 8 HOURS AT 12,000 x g. Fig. 35.15. Sedimentation of chloroplast dispersions obtained by supersonics and by the valve method (after French and Milner 1951). PREPARATION, PRESERVATION AND ACTIVATION OF CHLOROPLASTS 1559 mentation experiments of particles falling into discrete groups; rather the dispersion was completely heterogeneous, with random distribution of particles. About 25% sedimented in 15 min. at 20,000 g.; the rest in 10 min. at 60,000 g. About 90% had a molecular weight of 6-7 X 10«, 10% seemed to be smaller. Under the electron microscope, most particles ap- peared to be <2 m/x in size; however, a number of particles 8 m^ in size were observed, and some agglomerates of the 8 mju particles, about 25 m/x in size, also were noted. Such groups had not been noticeable in sedimenta- tion experiments, which were, however, made with a different batch of chloroplasts. Since the chlorophyll-bearing material of chloroplasts contains about 25% lipides, the effect of lipide-dissolving agents was studied by Milner et al. (1950), They shook a chloroplast dispersion with petroleum ether, extracting 2-3% of the total lipides. This caused the loss of up to one half of photochemical activity, which could not be restored by KCl. How- ever, when the lipide extract was evaporated and the residue, dissolved in a little ether, returned to the depleted dispersion, the activity was re- stored, and the material could be still further activated by chloride. The same effect could be obtained more simply by adding a little ether to the aqueous phase after the petroleum ether extraction, and removing the ether again by evaporation in vacuum. It thus seems that the effect of petroleum ether (as well as that of salts) is associated with changes in the size and colloidal structure of the active particles rather than with the removal (or addition) of specific chemical constituents. After the larger and more ac- tive particles had been broken into smaller less active fragments, addition of salts causes them to coagulate again and thus restores activity. Pe- troleum ether, by removing some of the hpoid "glue," may cause the par- ticle to fall apart into fragments. Addition and subsequent evaporation of ether may cause redistribution of the remaining lipides, bringing them from the interior to the surface, and thus cause renewed coagulation to larger units. French and Milner (1951) mention that some detergents were found to split chloroplast particles into smaller fragments without destroying their photochemical activity, but this method has not proved useful for prepara- tory purposes. All these experiments, however inconclusive, add to the impression that the photocatalytic activity of chloroplast preparations, as determined by the Hill reaction, is associated, not with one or two specific enzymatic components, but with a more or less complex structure which cannot be broken into small units without losing its activity. Perhaps this could be explained by the assumption — derived from kinetic experiments on photo- synthesis and the Hill reaction in whole cells — that one of the enzymes 1560 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 involved in both reactions is present in a concentration of about 1 mole- cule per 200-2000 molecules of chlorophyll. If the particles in a chloroplast dispersion contain 0.05 g. chlorophyll in 1 g. dry weight, have a density of about 1, and contain 50% water, then one molecule chlorophyll is contained in about 7 X 10"^ ju*, and one molecule of the Umiting enzyme, in about 1.5-15 X 10~^ M^ In particles formed by disintegration of grana, the chlorophyll concentration will be higher. It was estimated as 0.2 mole/ liter (or more) by Rabinowitch (1952); this corresponds to an average volume of 1 X 10 ~V^ (or less) per molecule of chlorophyll, or between 2 and 20 X 10~ V (or less) per molecule of the postulated limiting enzyme. Consequently, when particles in a grana dispersion become smaller than 10-^ M^ hi volume, or <0.01ju (10 mju) hi linear dimensions, an increasing number of them will contain not even a single molecule of the limiting en- zyme, and will thus be unable to contribute to the reaction at all. We have seen above that according to French, many of the particles in his dis- persion appeared to be <2 m^ under the electron microscope! The fact that chloroplast dispersions usually show a Tyndali cone in visible light merely indicates that the linear dimensions of some of the particles are > 100 ni/n; there may be enough smaller particles present to account for the drop in activity. Thomas, Blaauw and Duysens (1953) made a direct comparison of par- ticle size and photochemical activity of spinach chloroplasts, disintegrated by a magnetostriction oscillator. In air the disintegration led to the loss of about 50% of photochemical activity in 30 seconds, but in nitrogen the decrease in activity was only 5% after a whole minute of supersonic treat- ment. The active suspension, obtained in this way, was subjected to frac- tional centrifugation to obtain fractions of approximately uniform size. The distribution of particle sizes in each sample was observed under the electron microscope, and found to possess a sharp peak. Size measure- ments became unreliable below 3 m^ (limit of the resolving power of the electron microscope used), and above 15 m^ (where fractional centrifugation lost its effectiveness). Figure 35.15A shows photochemical activity {i. e., rate of Hill reaction in the light-limited state, with quinone as oxidant) as function of (mean) particle volume. Characteristic is the sudden decline in activity of particles when their average volume declines below 1.5 X 10-^ li\ Thomas and co-workers estimated that the particles became in- active when they contained less than between 40 and 120 chlorophyll mole- cules (rather than between 200 and 2000 molecules, as predicted above). The decline in activity mth diminishing size is less sharp in saturating hght [210 kerg/(cm.2 sec.)] than in limiting light. This is to be expected if particles below the "critical size" are not uniformly inactive but merely possessed of a statistical probability of being inactive, because some of them PREPARATION, PRESERVATION AND ACTIVATION OF CHLOROPLASTS 1561 contain no enzyme; while the over-all ratio chlorophyll : enzyme (and thus, in theory, also the limiting rate in strong light) remain unchanged. These results are suggestive, but will not be quantitatively convincing until more is known about the nature and composition of chloroplast frag- ments of different size. Are they pieces of protein lamellae with chloro- phyll attached to them? Are they irregular chunks of grana, containing pieces of both proteidic and lipoidic phases? Do they contain any of the proteins (or lipoproteids) of the stroma? Even the average chlorophyll content of the chloroplast material is uncertain unless it has been de- termined in an aliquot of the same sample as that used for photochemical studies. The uncertainty becomes as great as a factor of ten when we deal with chloroplast fragments selected by fractional centrifugation, or another " o 50 TOO Volume in &> xlO^ 150 Fig. 35.15A. Oxygen evolution by suspension of chloroplast fragments as function of their average size, in the presence of quinone (after Thomas, Blaauw and Duysens 1953). fractionation method. Particularly uncertain — and challenging — is the situation in the case of the chloroplast-free, phycocyanine-bearing blue- green algae, and of the chloroplast-containing, phycoerythrin-bearing red algae. What happens in the fragmentation of the chromoproteids? It seems that they are almost completely separated from the chlorophyll- bearing, insoluble fragments {cf. chapter 37A, section 3); the resulting suspension shows no Hill reaction (^^an Norman et al., 1948). McClendon and Blinks (1952) showed that the loss of phycobilins can be prevented by crushing the cells in a medium containing high-molecular compounds, such as Carbowaxes (c/. p. 1549); the material then shows well-sustained Hill activity (McClendon, 1954). The results obtained by electron microscopy and ultracentrifugation of dispersed chloroplast material from the higher plants, blue-green algae, and purple bacteria will be presented in chapter 37A. We will see there that the morphological picture is as yet far from clear, particularly in respect to the submicroscopic localization of the various pigments. The correla- tion between the structure and the photochemical activity of the various dispersions is a matter for future exploration. 1562 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 4. Different Oxidants (a) Ferric Salts It will be recalled that Hill first obtained sustained photochemical oxy- gen production from chloroplast preparations by adding cell extracts of indefinite composition, and later by adding complex ferric salts, such as ferric oxalate. Simple ferric salts, such as nitrate or perchlorate, are un- suitable because of their hydrolysis in the approximately neutral solution (needed for these experiments). Potassium ferricyanide was added by Hill to ferric oxalate in the assumption — based on his own observations — that Fe(CN)6~* ions will not take direct part in the photochemical reaction, but will oxidize (in successful competition with oxygen) the ferrous oxalate, produced by the reaction, back to ferric oxalate, thus enabling the photo- chemical reaction to proceed until all ferricyanide is exhausted [since ferro- cyanide ions — in contrast to ferrous ions — are not rapidly reoxidized to the Fe(III) level by molecular oxygen]: light, chloroplasts (35.29a) 2 Fe+s oxalate + HjO ^ 2 Fe+2 oxalate + 2 H+ + >^ Oj dark (35.29b) 2 Fe+2 oxalate + 2 Fe(CN)6-3 ^^ 2 Fe(CN)o-* + 2 Fe+2 oxalate (35.29) 2 Fe(CN)6-3 + H2O > 2 Fe(CN)6-' + 1^ O2 + 2 H + Holt and French (1946), on the other hand, found that ferricyanide can take part in the Hill reaction also directly, without the addition of ferric oxalate. Figure 35.16 shows oxygen liberation curves recorded by them with the complete Hill solution, and with the same solution minus some of its constituents: (1) Hill's complete solution: 0.02 M K3Fe(CN)6, 0.01 M FeNH4(S04)2, 0.50 M K2C2O4, 0.02 M sucrose, 0.17 M sodium sorbitol borate or 0.2 M phosphate, (2) same, minus FeNH4(S04)2; (3) same, minus K2C2O4; and U) same, minus both FeNH4 (864)2 and K2C2O4. The figure indicates that steady liberation of oxygen at about half the maximum rate observed with the complete Hill mixture can be obtained with ferricyanide alone. The rate maximum at pH 7-8 was there even when FeNH4 (804)2 was left out, indicating that it was not due to complex- ing of Fe+' ions with OH" ions (or some other initial step of hydrolysis). Hill and 8carisbrick had worked with chloroplasts from Stellaria and Chenopodium ; perhaps the spinach chloroplasts used by Holt and French contained enough ferric oxalate [or other organic salts of Fe(III) ] to make their addition unnecessary. (Liebich found 0.05% Fe in the dry matter of spinach chloroplasts; cf. Vol. I, page 377; Kohman found as much as 9% oxalic acid in the dry matter of spinach leaves.) DIFFERENT OXIDANTS 1563 A chloroplast-free Hill solution itself evolved some gas when illuminated with white light (250 f .-c.) ; to avoid errors due to this evolution, a nitrogen atmosphere and the presence of a 10% KOH solution in a side arm of the manometric vessel were used by Holt and French (1946). With these pre- cautions, all pressure changes observed could be ascribed to oxygen. The addition of inactivated (boiled) chloroplasts to Hill's mixture had no effect. 100 ° / TIME, minutes Fig. 35.16. Oxygen production by spinach chloroplasts immersed in various combinations of the constituents of Hill's mixture (after Holt and French 1946): (1) complete mixture; (3) 0.50 M K2C2O4 + 0.02 M K3Fe(CN)6; (5) 0.01 M Fe- (NH4)(S04)2 + 0.02 M K3Fe(CN)6; (4) 0.02 1/ K3Fe(CN)6. 15° C, pH 6.8, all in 0.17 A'' sodium sorbitol borate buffer, 0.28 mg. of chlorophyll per vessel. (6) Other Inorganic Oxidants Holt and French (1948) investigated a variety of oxidizing agents They used spinach chloroplast fragments in a phosphate buffer in nitrogen atmosphere. Manometric evidence of gas production (and chemical proof that this gas was oxygen) was obtained with chromate [which was earlier found inactive with live Chlorella cells by Fan et al. (1943)] and with m- vanadate (too little O2 for manometric study). Numerous other oxidants gave no oxygen at all, among them molybdate, bromate, chlorate, tetra- thionate, tungstate, hypochlorite, permanganate, nitrate, periodate, bis- muthate, iodine, arsenate, persulfate and perborate. An initial outburst of gas was obtained upon mixing chloroplasts \vith perborate or permanga- nate, but illumination led to no additional gas Hberation. 1564 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 The reaction with chromate was studied somewhat more closely. Oxy- gen was evolved only in light ; this evolution was prevented if the chloro- plasts were preheated to 50° C. for 15 min., showing that it involves an en- zymatic reaction. The total amount of oxygen evolved was only about one 150 130 no 90 E 70 E 50 30 / Light Dark .• — • — •- / Light. / / Added Hill's solution J_ 10 20 30 40 50 TIME, minutes 60 70 Fig. 35.17. Resumption of oxygen evolution by illuminated chloroplasts on addi- tion of Hill's solution, following cessation of evolution from chromate (6.65 X 10-6 mole KjCrOi; O2 equivalent; 111 mm.'; pH 7.3; 0.1 Af phosphate; 16° C; nitrogen atmosphere; 0.25 mg. chlorophyll) (after Holt and French 1948). half (45-75%) of that expected for the oxidation of water by Cr(VI), and the reduction of the latter to Cr(III) : (35.30) Cr04-2 + 5 H+ > Cr+=' + 2^ H2O + % O2 Figure 35.17 shows that the cessation of oxygen evolution, after about one half of the stoichiometric equivalent was produced, was not the result of damage to the photochemical system, since addition of Hill's mixture led to a resumption of oxygen liberation. On the other hand, oxygen libera- tion could not be renewed by the addition of fresh chloroplasts. Incuba- tion of chloroplasts in 2 X lO^^ jlf chromate solution for 30 min. in the dark did not inactivate them. The total fraction of chromate reduced was approximately the same for chromate concentrations between 1.5 X 10"=^ DIFFERENT OXIDANTS 15G5 and 3.5 X 10~^ mole/1, (c. g., in one set of experiments 70%); the initial rate also seemed to be independent of chromate concentration. (c) Oxygen as Hill Oxidant One of the puzzles of photosynthesis is how the photochemically trans- ferred hydrogen manages to avoid the abundantly available strong oxidant, molecular oxygen, in favor of the sparse and extremely unwilling oxidant, carbon dioxide. Only under certain abnormal conditions — such as carbon dioxide starvation, or excessive illumination (c/. chapter 19) — does photau- toxidation replace photosynthesis. It was suggested (c/. scheme 19.1, p. 544) that in these cases, oxygen acts as a "substitute hydrogen acceptor" replacing carbon dioxide, while the primary photochemical hydrogen trans- fer remains the same as in photosynthesis. However, as long as water is the reductant, the Hill reaction with oxygen as oxidant is a circular process, leading to no net chemical change, and therefore unrecognizable except by isotopic tracer experiments (photocatalysis of isotopic equilibration of free oxygen and water). For it to become observable by ordinary chemical methods, it has to lead to the oxidation of compounds other than water^ such as benzidine (page 528) or chlorophyll itself (page 537). This may occur by direct substitution of these reductants for water in the photochem- ical process (as suggested in scheme 19.1), or indirectly by the action of hydrogen peroxide formed as intermediate in photochemical hydrogenation of oxygen (see below). The isotopic tracer method was applied to the "hidden" Hill reaction (with oxygen as oxidant and water as reductant) by Brown (1953). Using water containing only 0^® and oxygen gas enriched in O''' and 0^^, he found that in light an exchange of the two isotopes took place at a rate comparable to that of the Hill reaction at the same light intensity. When chloroplasts deteriorated (by aging or phenanthroline poisoning) their capacity to photocatalyze the isotopic exchange of oxygen between H2O and O2 de- clined in the same proportion as their capacity for Hill reaction with qui- none as oxidant. An indirect but ingenious chemical proof that oxygen can serve as Hill oxidant with chloroplast suspensions was supplied by Mehler (1951''^, 1952). He used for this purpose the system ethanol-catalase, which can "trap" hydrogen peroxide (expected to arise as intermediate in the hydro- genation of oxygen). In the presence of ethanol, the oxidation-reduction reaction catalase (35.31) H2O2 + C2H4OH > C2H4O + 2H2O in which catalase acts as a "peroxidase," competes with the more common "catalatic" dismutation of H2O2 to H2O and O2. First, Mehler had to show 1566 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 that when the Hill reaction is carried out with other oxidants— quinone, Hill's mixture, or cytochrome c — no hydrogen peroxide is formed as an intermediate oxidation product of water. He found, in fact, that no acet- aldehyde is produced upon addition of ethanol and catalase to these com- mon Hill reaction systems. This is the most convincing confirmation to date of the surmise — made in chapter 11, p. 286 — that hydrogen peroxide is not an intermediate in the photochemical oxygen formation by plants. -90 - 20 40 60 80 MINUTES 100 Fig. 35.17A. Evolution and absorption of O2 by chloroplast material (300 /ig. chlorophyll in 2 ml. phosphate buffer pH 6.8) (Mehler 1951). O2 evolved in vessel 1 (5 X 10~« mole quinone); O2 consumed in vessel 2 (4 mg. catalase, lO"' mole ethanol; O2 first evolved, then consumed at double speed in vessel 3 (5 ^mole quinone, 4 mg. catalase, 10 ~^ mole ethanol). Vessel 4; catalase, ethanol, and 5 jumole hydro quinone. Mehler then could proceed with the demonstration that acetaldehyde does arise— as a reduction intermediate — if chloroplast suspensions containing ethanol and catalase are illuminated in the absence of added Hill oxidants, but in the presence of oxygen. Under these conditions, oxygen is con- sumed instead of being liberated, as in the usual versions of the Hill reaction (figure 35.17A). Two moles of acetaldehyde were found to be formed per mole of oxygen consumed, which is consistent with the reaction scheme (35.31A). 'different oxidants 1567 (35.3lAa) 2 H.O — ^^ 2 {H} + 2 {0H| (35.31 Ab) 2 {OH} > H.O + V2 O2 (35.31AC) 2 {H} + O2 > H2O2 catalase (35.3lAd) H2O2 + CaHsOH > C2H4O + H.O (35.31Ae) V2O2 + C2H5OH ^ C2H4O + H2O catalase The normal redox potential of the system O2/H2O2 is 0.27 volt (cf. table 11. 1), well within the region of other efficient Hill oxidants. Aerated chloroplast suspensions contain about 10~^ mole/1, oxygen; quinone and other common Hill oxidants need to be present in concentrations of 2 X 10~^ mole/1, or higher (cf. section 5(6) below) to ensure high efficiency of oxygen liberation. This indicates that oxygen is a highly effective com- petitor of the other Hill oxidants, and emphasizes how remarkable is the fact that it is not an effective competitor of carbon dioxide in live, aerobi- cally photosynthesizing cells. Mehler suggested that photoxidations in vivo (described in chap. 19) may be based on a Hill reaction with oxygen as oxidant (as suggested in Vol. I, p. 544), in which some of the hydrogen peroxide, formed as reduc- tion intermediate, is intercepted by peroxidases, causing it to react with the photoxidation substrates (in competition with its dismutation by catalase) . Mehler and Brown (1952) confirmed the above-suggested mechanism of the ''Mehler reaction" by showing mass spectrographically that the net consurnption of oxygen, according to equation (35. 31 A), is in fact the result of superposition of an oxygen evolution (eqs. 35.31Aa,b) upon a (twice as large) oxygen consumption (eq. 35. 31 Ac). For the purpose of this demonstration, the reaction was carried out in O^^ and O^^ enriched oxygen (in a helium atmosphere) . The experimental system also contained quinone, in addition to catalase and alcohol. In light, quinone was reduced first, causing 62^^ to be evolved, while the 02^** concentration remained con- stant. When practically all quinone was exhausted, the "Mehler reaction" got under way; both 62^- and 62^* were now consumed, but the second one much faster than the first. The curves showing the consumption of the two isotopic species were in quantitative agreement with predictions based on mechanism (35. 31 A). Mehler (195P) studied more closely the competition between quinone and oxygen as hydrogen acceptors in the Hill reaction. He found that in a chloroplast suspension containing both quinone and ethanol -f catalase, quinone was hydrogenated at the usual rate until the reaction was 90% complete, at which time the oxygen consumption began; remarkably 1568 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 enough the latter was now 2-3 times faster than in a qiiinone-free system. (In the latter, the rate of oxygen consumption in Hght is only V2 or Vs of that of oxygen liberation with quinone; while in the mixed system, after quinone had been nearly exhausted, the rate of oxygen consumption is about equal to that of initial oxygen liberation.) This stimulation of the photochemical alcohol oxidation by preceding quinone reduction was sus- tained even w^hen an equivalent of five times the total amount of ciuinone present had been oxidized. Checks showed that this catalytic effect is not due to the presence of hydroquinone. The stimulation is still present if alcohol and catalase are added 10-20 minutes after the quinone reduction had been completed. No similar stimulation follows the Hill reaction with f erri cyanide ; but 0- or p-naphthaciuinone have the same effect as benzo- quinone. No stimulation is brought about by simply incubating chloro- plasts in quinone solution in the dark. Reducing quinone in the dark by ascorbic acid also caused no stimulation if the two reagents were mixed before adding the chloroplasts, but it produced the same stimulation as photochemical reduction if ethanol and catalase were first mixed with the chloroplasts and quinone and ascorl)ic acid added afterward. The rate of oxygen uptake was enhanced by still another factor of two if more than the stoichiometric quantity of ascorbic acid was used. (It is uncertain whether the thus quadrupled rate of oxygen consumption equalled, or exceeded, the rate of photosynthesis by the same amount of chlorophyll, cf. section 5(6) below.) Only a slow photoxidation of ascorbic could be observed in il- luminated chloroplast suspensions in the absence of quinone. In the sys- tem (chloroplasts + quinone + ascorbic acid), one mole of oxygen was consumed in light. Adding only catalase caused a 50% inhibition of this reaction, but adding both catalase and ethanol lead to the above-described maximum stimulation of oxygen consumption. It continued until two moles of O2 had been used up per mole ascorbic acid, after which the reac- tion rate dropped to the level characteristic of the (quinone + ethanol + catalase) system without ascorbate. Similar effects were observed with glutathione; it, too, was photoxidized by chloroplasts slowly in the absence of quinone, and rapidly in its presence. Because of Gerretsen's observa- tions (cf. below), Mehler tried adding Mn++ salt to the above-described systems, and found that the only one on which it had an effect was (ox3'gen -f- ethanol -|- catalyse + chloroplasts) ; the rate of oxygen consumption by this mixture in light was doubled by the addition of 10~* mole/1. MnCla. The enhancing effects of quinone reduction and of Mn++ addition could not be superimposed upon each other. The effect of manganous ions was catalytic; ferrous ions did not produce it; instead, they themselves were rapidly photoxidized by chloroplasts in the presence of ethanol and cata- lase. DIFFERENT OXIDANTS 1569 Gerretsen (1950^) observed that crude chloroplast preparations from Avena, which slowly absorbed oxygen in the dark (c/. below section 5(c)), increased this absorption 2-3 times in light (instead of replacing it by oxygen liberation, as noted bv Hill and others with washed chloroplasts). Addition of glucose did not increase this photautoxidation ; it was, however, stimulated by asparagine. Addition of manganese salts — which Gerret- sen (1950') had found to increase strongly the effect of light on the redox potential of the same material — often produces a marked temporary increase in the rate of oxygen uptake in light. (d) Quinones An important new group of oxidants capable of serving in the Hill re- action was found by Warburg and Liittgens. They noted (1944) that chloroplasts isolated from spinach or sugar beet leaves can reduce p- benzoquinone — probably to hydroquinone — with the liberation of an equivalent amount of oxygen. A description of these experiments was given in a paper published in Russian (Warburg and Liittgens, 1946) and later reprinted in W^arburg's book Heavy Metals as Prosthetic Groups (1946; English edition, 1949). Warburg and Liittgens discovered the quinone reaction while studying the "respiration" of juices obtained by mac- eration of spinach leaves. The oxygen consumption of these juices de- clined with time; in an attempt to maintain it, various oxidation sub- strates were added, and it was noted that with hydroquinone (or pyro- catechol) the respiration was sustained in the dark, but markedly decreased in light. This was attributed to photochemical reversal of the oxygen- consuming process — i. e., to a (chloroplast-sensitized) photochemical oxida- tion of water by quinone: Chl, light (35.32) 2 C6H4O2 + 2H2O , 2 CeHeOz + O2 -52 kcal./mole dark Similar observations were made with some naphthaquinones, e. g., naph- thaquinone sulfonic acid. No change in oxygen consumption was noted in light, neither in the liquid fraction from the maceration of spinach leaves (with or without pyrocatechol or hydroquinone) nor in the chloroplast- containing fraction without added polyphenol; but in the chloroplast fraction to which pyrocatechol was added, illumination caused 50% in- hibition of the oxygen uptake (fig. 35.18); and if both pyrocatechol and hydroquinone were added, a complete inhibition of respiration or even a liberation of oxygen was observed in light. After it was first noted that chloride may be needed as a "co-enzyme" for the quinone reaction (c/. above, sect, (c)), 0.05% KCl was added by Warburg and Liittgens to all washed preparations. (If the centrifuged suspensions are not washed, natural chloride content is high enough to ac- tivate the photochemical reaction.) 1570 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 Purity of the quinone used is important. Commercial pure benzoqui- none was purified by distillation with water vapor and drying in vacuum. Dissolving the yellow crystals in pure water (even if carried out in the ab- sence of air) gives a strongly colored solution; it contains an impurity which inhibits the photochemical activity. To avoid this, quinone was dissolved in 0.01 N sulfuric acid. Illumination of quinone solution in the absence of chloroplasts with the blue-violet light absorbed by quinone causes darkening; no reaction can be observed afterward upon the addition T3 0) E E 50 40 30 20 10- - / - Dark / Light / - Y : / ^^ / 1 1 1 1 1 1 1 1 1 10 20 30 TIME, minutes Fig. 35.18. Effect of light on oxygen consumption by chloroplast preparations in the presence of pyrocatechol (after Warburg and Liittgens 1948). of chloroplasts (Clendenning and Ehrmantraut, 1950). A chloroplast suspension to which a pure, acid quinone solution was added at pH 6.5 in the absence of oxygen showed in the experiments of Warburg and Lutt- gens no gas exchange in the dark (except for occasional liberation of traces of CO2) ; upon illumination, it produces oxygen until the volume of the latter reached 80-90% of that calculated for the oxidation of water by the available quinone, according to equation (35.32), thus confirming the valid- ity of this equation (see figure 35.19). That the liberated gas was oxygen was confirmed by absorption in phosphorus. lodometric assay showed that when the photochemical gas DIFFERENT OXIDANTS 1571 liberation was ended, practically all quinone had disappeared from the solu- tion. Similar experiments with o-benzoqiiinone were unsuccessful because of the instability of this compound; but another orthoquinonoid compound, a-naphthaquinone sulfonic acid, could be reduced in the same way as p- benzoquinone. This, and the stimulation of the oxygen consumption in green extracts by the addition of pyrocatechol (section 5(c) below) makes it probable that o-quinones react similarly to p-quinones. 200 TIME IN LIGHT, minutes Fig. 35.19. Evolution of o.xygen from chloroplast suspension in quinone solu- tion in light (after Warburg and Liittgens 1946). 2 ml. suspension in 0.05 M phosphate buffer, pU 6.5, 0.05% KCI, 1 nM chlorophyll in vessel, 20° C, argon atmosphere. 2 mg. quinone in 0.2 ml. 0.01 A'' H2SO4 added at time 0. Since 18.5 X 10"" mole quinone could be reduced by a suspension con- taining only 0.2 X 10 ~" mole chlorophyll, the latter obviously could not have served as a reactant, but only as a catalyst; the same applies to all other components of the chloroplasts, none of which is available in such large quantities. As shown in equation (35.32) the oxidation of water by quinone is endo- thermal to the extent of 52 kcal./mole oxygen — somewhat less than one half as much as the oxidation of water by carbon dioxide. This energy must be supplied by light. 1572 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 Aronoff (1946) repeated Warburg's experiments with a variety of naph- thaquinones and anthraquinones and found that they reacted at a rate roughly parallel to their oxidation-reduction potentials — in other words, stronger oxidants were reduced faster than the weaker ones. (This com- parison refers to the saturation rate in strong light, not to quantum yield in weak light!) These results are in agreement with Holt and French's ob- servations on dyestufts (cf. Table 35.VIII). Wessels and Havinga (1952, 1953; cf. also Wessels 1954) prepared and investigated 28 quinones with a range of normal potentials (measured at pH 6.5) from —0.444 volt (tetrachloro-o-benzoquinone) , through -0.332 volt (p-benzociuinone), -0.181 volt (1,2 naphthaquinone), -0.086 volt (1,4 naphthoquinone), +0.020 volt (2-methyl-l,4-naphthoquinone), +0.135 volt (phthiocol), to +0.21 volt (chloranilic acid). They measured the effect of illumination on the redox potential of the chloroplast suspen- sions containing these quinones, and found, under aerobic conditions, a "photogalvanic effect" (change of redox potential in light) with all (17) quinones having normal potentials from -0.444 up to 0.058 volt (phen- anthrene quinone) ; and no photogalvanic effect for all (11) quinones with potentials above 0.058 volt. Under anaerobic conditions, 7 quinones, with normal potentials from -0.444 to +0.180 volt were studied, and a photo- galvanic effect was observed in six of them, with normal potentials up to +0.090 volt (2-hydroxy-l,4-naphthoquinone); no effect was observed only with sodium anthraquinone-2-sulfonate (Eo' = +0.180 volt). Similar results were obtained with quinonoid dyes (see below). (e) Organic dyestuffs Closely related to quinones are many organic dyestuffs that contain closed, conjugated double bond systems, and can be converted to "leuco- dyes" by the addition of two hydrogen atoms. In the leuco- dyes the conjugated system is either destroyed or, at least, reduced in length, wdth consequent weakening of absorption bands and their shift into the ultraviolet region. Many of these dyes form strong electrolytes, and their oxidation-reduction potentials can be easily determined. Holt and French (1948) first observed that a number of reversibly reducible organic dyes can serve as photochemical oxidants of water in the presence of chloroplast suspensions. Attention must be given to photo- chemical effects produced by light absorption in the dye, and to the sepa- ration of this absorption and its chemical effects from those of chlorophyll (or other chloroplast pigments). Quantitative experiments were made by Holt and French (1948) with the red dye phenol indophenol. No reaction could be observed in the dark, or with chloroplasts preheated to 50° C. Tests of the purity of the dye DIFFERENT OXIDANTS 1573 gave values as low as 55%; on this basis, the oxygen yields were 100 ± 4% of the theoretical value for the reaction (D = Dye) : (35.33) D + H2O > H2D + K O2 The rate decreased with increasing dye concentration (3.3 X 10""^ — * 3.3 X 10-2 mole/liter), probably because increased competition of the dye for light absorption brought the absorption by chlorophyll below the level needed for light saturation. With blue dyes, such as 2,6-dichlorophenol-indophenol, the absorption by the dye was so strong that the rate became too slow for manometric measurements. By using a very dilute solution, both in respect to the dye and in respect to chlorophyll (5 X 10-^ mole/liter), the occurrence of the dye reduction could be proved by ol^servation of the decoloration of the dye in light. l<]xperiments of this type were made with eight dyes at con- centrations of the order of 5 X 10"^ mole/liter or less, at pH 6.6 and 8.0, with the results shown in Table 35. VIII. Table 35.VII1 Dye Reduction by Chloroplasts (after Holt and French, 1948) -Eo' (volt)" at pH 6.6 at pH 9.0 Positive results (at pH 6.6 and 9.0) Phenol-indophenol 0.254 0.083 2,6-Dichlorophenol-indophenol 0 . 247 0 . 089'' o-Cresol-indophenol 0.217 0.089 Positive results (at pH 6.6 only) Thionine<^ 0.074 (-0.001) Negative results l-Naphthol-2-sulfoindophenol 0 . 147 0 . 003 Methylene blue 0.024 -0.050 Indigo tetrasulfonate -0.027 -0.114 Indigo disulfonate -0. 104 -0.199 " Throughout this book, oxidation-reduction potentials are given on Lewis and Ran- dall's scale on which strong oxidants have high negative potentials, cf. table 9. IV. '' Slow reaction. "= No oxygen evolution was found; reaction may thus be with a cellular hydrogen donor, such as ascorbic acid, rather than with water. The decolorization of thionine was incomplete, and reversed itself in the dark. This table shows, in agreement with Aronoff's findings {cf. above), that the photochemical effectiveness of the several oxidants (saturation rate of oxygen liberation in strong light) increases generally with their oxidation potential. However, the parallelism is not strict, and specific effects seem to occur, as shown by the inactivity of the l-naphthol-2-sulfoindophenol. It must be considered that the observed rate is the net balance of a photo- chemical forward reaction and a dark back reaction. The back reaction can involve either the final reduction product (such as the leuco dye), or a 1574 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 reduction intermediate (such as a semiquinone) . The thermodynamic equiUbrium hes practically always on the side of the dye and water, not on that of the leuco dye and oxygen. Therefore a back reaction is always possible — although it may be so slow as to be practically insignificant. The redox potential of the dye-leuco dye system is more likely to affect the rate of the dark back reaction than the quantum yield of the photochemical forward reaction. When the back reaction is fast, the Hill reaction becomes unobservable by ordinary methods. It may still be detectable by isotopic tracers, or by "trapping" the product chemically or physically, or by sensi- tive methods revealing a small shift of the oxidation-reduction equilibrium in light. We recall that in Hill's original experiments the rather fast re- oxidation of ferrous oxalate by oxygen had necessitated the addition of ferri- cyanide as "stabihzer." Leucothionine or leucomethylene blue, if they are formed in the Hill reaction together with oxygen, ^\dll be reoxidized even more rapidly. Leucophenol indophenol, on the other hand, is reoxidized much more slowly, thus permitting practically complete decoloration in light. Only when the back reaction is negligible does the rate of decolora- tion provide a true measure of the photochemical reactivity of the suspen- sion (i. e., of the quantum yield of the photochemical reaction). Additional quinonoid dyes were tried out by Holt, Smith and French (1951). Table 35.IX shows the results. Table 35.IX Indophenol Dyes Reduced by Illuminated Chlokoplasts (after Holt, Smith and French, 1951) Color at Dye (sodium salt) pH 6.5 pH 8.0 2,6-Dibromobenzenone-indo-3'-carboxyphenol 2,6-Dibromobenzenone-indo-2'-bromophenol 2,6-Dibromobenzenone-indo-3'-methoxyphenoI 2,6-Dibromo-2 '-methyl-5 '-isopropylindophenol Ben zenone-mdo-2 '-methy 1-5 '-isopropylphenol 2,6-Dibromobenzenone-indophenol Benzenone-3 '-methyl-6 '-isopropylphenol Red-violet" Blue" Blue" Blue" Blue" Pink* Blue" Red*- Blue" Blue" Blue-^ Blue" Blue" Olive green" " Means strong decolorization. '' Means little decolorization. " Means no decolorization. Macdowall (1952) illuminated a suspension of washed Swiss chard chloroplasts in distilled w^ater for 2 V4 hours wdth light of about 2 klux in the presence of nine dyes, and determined the degree of reduction at the end of exposure either photometrically or potentiometrically. The potentials reached with five oxidants are Hsted in table 35.IXA. The figures in table 35. IX A show that reduction was practically com- plete with the first three dyes, far-reaching with the fourth, and neghgible DIFFERENT OXIDANTS 1575 Table 35.IXA Normal Potentials of Five Dyes and the Potentials Measured in Chloroplast Suspensions Containing These Dyes before and after Illumination (after Macdowall 1952) 2,6-Dichloro- Indigo phenol-indo- Toluylene Cresyl Methylene disulfo- phenol blue blue blue nate ^o' (pH 6.5, 15° C.) -0.283 -0.169 -0.116 -0.060 -0.067 i; (before illumination) -0.435 -0.448 -0,496 -0.471 -0.470 E (after illumination) -0.071 -0.239 -0.046 -0.011 -0.009 with the fifth. The meaning of the potentials measured before the illu- mination is doubtful. The highest positive potential was reached with toluylene blue. A similar but more extensive study was made by Wessels and Havinga (1952, 1953; c^. Wessels 1954), who observed the decolorization in light and the photogalvanic effect (change of redox potential in light) of fifteen quinonoid dyes with normal potentials (pH 6.5, 30° C.) from —0.255 volt (2,6-dichlorophenol-indophenol) through —0.137 volt (toluylene blue), -0.028 volt (methylene blue), +0.098 volt (indigo disulfonate) to +0.269 volt (safranin T). They found, in air, photogalvanic effects for all 9 dyes with Ei^' values up to —0.028 volt, and none for all 6 dyes with more posi- tive redox potentials. Decolorization w^as observed for 7 dyes wdth poten- tials up to —0.137 volt, and not for the 8 dyes with more positive poten- tials. Under anaerobic conditions 10 dyes were studied, and photogalvanic effects were observed with all but 3 of them, with a dividing line between + 0.098 volt (indigo disulfonate) and +0.130 volt (indigo monosulfonate) . Decolorization could be noted, however, only for dyes with potentials up to —0.137 volt (as under aerobic conditions). (The potentiometric test is more sensitive to small shifts in oxidation-reduction equilibrium than visual observation of color changes.) The photostationary redox potentials, measured in illuminated solutions in which a photogalvanic effect could be observed, ranged from 0.00 volt to —0.10 volt, without obvious relationship to the E^ of the dye (or qui- none). (About the limited significance of such photostationary potentials, see p. 1522.) A study was also made by Wessels of Hill reaction with a mixture of two oxidants (quinones or quinonoid dyes). If both components were "good" Hill oxidants, both were reduced; if one was a "borderline case" (such as thionine, E^' = —0.0777 volt), it was hardly reduced at all. The presence in the mixture of an oxidant too positive for its own reduction did not afTect the reduction of a "good" Hill reductant, except in the case of phthiocol (which is known to inhibit the Hill reaction). 1576 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 These results, together with the earlier data of Holt and French (tables 35. VIII and 35. IX), and of Macdowall (table 35. IX A), establish rather clearly that, as far as electrode-active redox systems of the quinone- hydroqiiinone type are concerned, illuminated chloroplast suspensions can displace, by the transfer of hydrogen from water, the equilibrium of all of them whose normal redox potentials (at pll 6.5) are under —0.1 volt; if air is excluded (and the back reaction thus slowed down, because only the photochemically produced O2, or its precursors, are available for reoxida- tion), the displacement of the equilibrium in light can be recognized poten- tiometrically for systems with normal potentials up to +0.1 volt. The question whether free radicals are formed as intermediates in Hill reaction with quinones or quinonoid dyes was taken up by Uri (1952) and Wessels (1954) by inquiring whether this reaction has an effect on polymeri- zation of methyl aery late (Uri) or acrylonitrile (Wessels), or on the oxida- tion of benzene to phenol (Wessels), As mentioned in part A, Uri was able to confirm, by means of the polymerization test, the formation of free radi- cals in the photochemical reaction of dissolved chlorophyll with ascorbic acid; but neither he nor Wessels could obtain similar evidence for the presence of radicals in the quinone or dye reduction sensitized by chloro- plast suspensions. (/) Respiration Intermediates and Other Cellidar Materials The behavior in the Hill reaction of the compounds known to occur as intermediate oxidation-reduction catalysts in respiration is of particular interest, since these compounds could conceivably serve as links connecting the photochemical apparatus to an enzymatic system capable of reducing carbon dioxide in the dark. Among these intermediates, the most inter- esting ones are the well known "coenzymes" I and II (dipyridine nucleo- tide, DPN, and tripyridine nucleotide, TPN), and the more recently dis- covered "coenzyme A" ("thioctic" or "hpoic" acid). These catalysts have such high reduction potentials (—0.3 volt, cf. page 222) that their successful photochemical reduction would constitute a close approach — as far as energy utilization is concerned — to the reduction of carbon dioxide itself. More specifically, of the two steps in the reduction of carbon dioxide to the carbohydrate level (we imagine CO2 to be first incorporated in a carboxyl, RH + CO2 ^ RCOOH), the first one, the reduction of the car- boxyl group to a carbonyl group, requires a normal potential of around 0.5 volt; but the second one, the reduction of a carbonyl group to a hydroxyl group, needs only about 0.25 volt {cf. table 9. IV). Reduced pyridine nucleotides are capable of bringing about the second step by themselves, but not the first reduction step. The first one becomes possible if the car- boxyl is "activated" before reduction by conversion into a phosphate ester DIFFERENT OXIDANTS 1577 (e. g., by a "transphosphorylation" reaction with adenosine triphosphate, ATP). Reduction then leads to the degradation of a "high energy" car- boxyl phosphate to a "low energy" carbonyl phosphate, and the energy required for reduction is thus decreased by the energy difference between the two types of phosphate esters (about 10 kcal, corresponding to a shift of about 0.22 volt downward on the redox potential scale). We will return to these relationships below. Attempts to use pyridine nucleotides as ultinaate Hill oxidants ^^■ere unsuccessful (Holt and French 1948; Mehler 1951). Mehler found no reduction of DPN also when the dye dichlorophenol-indophenol (which reacts rapidly with DPNH.) was supplied as intermediate, although the dye was reduced in light practically completely. (Considering the wide difference between the normal potentials of DPN and the dye, this is not astonishing, cf. considerations on p. 1522). Similar results were obtained by Wessels (1954) when lie tried to use quinone, or dichlorophenol-indophenol, as intermediate catalyst to reduce oxidized glutathione, de- hydroascorbic acid, dehydroxyphenylalanine, riljoflavin or DPN. The reason for the negative result of attempts to reduce pyridine nucleo- tides in the Hill reaction could be a rapid back reaction of reduced pyridine nucleotides with intermediates in the oxidation of water (or with the final oxidation product, molecular oxygen). The surmise that a small amount of DPNH2 or TPNH2 is present in the photostationary state was confirmed by "trapping" these compounds with one of the enzymatic systems which use them for the reduction of pyruvic acid or other respiration intermedi- ates. This proof was given independently by Vishniac in Ochoa's labora- tory in New York, by Tolmach in Franck and Gaffron's laboratory in Chicago and by Arnon in Berkeley. Vishniac (see Ochoa 1950; Vishniac 1951; Ochoa and Vishniac 1951; Vishniac and Ochoa 1951) based his experiments on previous findings con- cerning reversible oxidative decarboxylation of malic acid (or c?-isocitric acid) to pyruvic acid (or ketoglutaric acid) in the presence of "malic en- zyme" (or isocitric dehydrogenase and aconitase), with TPN as specific hydrogen acceptor. These reactions normally proceed in the direction of decarboxylation and oxidation, but they can be reversed by the supply of excess TPNH2 and CO2, e. g. : malic enzvme (35.34) CO2 + CH3COCOOH + TPNH2 . " ^ COOHCH2CHOHCOOH + TPN pyruvate malate It was surmised by Ochoa, Veiga Solles and Ortiz (1950) that illuminated chloroplasts may be capable of converting TPN to TPNH2 and thus driving reaction (35.34) from left to right. The first proof of this was obtained by Vishniac and Ochoa, who showed that if pyruvate and bicarbonate are added to a chloroplast suspension in the presence of TPN, malic enzyme 1578 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 (prepared from pigeon liver) and Mn++ ions (which this enzyme specifically requires), malic acid is formed. This formation can be demonstrated after exposure to hght (but not in darkness) by means of a specific enzymatic test (hberation of CO2 by mafic decarboxylase). In a second test, tracer C*02 was employed, and practically all of the C^^ fixed in Hght was found in malate, with 75% of it localized in the /3-carboxyl group. Similar results were obtained ^\dth a-ketoglutaric acid and bicarbonate in the presence of chloroplasts, TPN and appropriate enzymes. Other pyridine nucleotide-specific reductions also could be carried out with the help of illuminated chloroplast preparations, including the reduction of pyruvate to lactate by DPN and lactic dehydrogenase: chloroplasts (35.34Aa) DPN + H.O — > DPNH. + V2O2 light (35.34Ab) CH3COCOOH + DPNH.2 > CH3CHOHCOOH + DPN pyruvate lactate also reduction of oxalacetate to malate in the presence of DPN and mafic dehydrogenase, reductive amination of a-ketoglutarate to glutamic acid by ammonia and DPN and reduction of fumaric acid to succinic acid in the presence of DPN and extracts from Escherichia coli (a reaction which can be brought about in the dark by molecular hydrogen) . All the above reductions (except that of fumarate) involve the hydro- \ \ / genation of the carhonyl group, C=0 (a C=C group is hydrogenated m / / ^ fumarate). As stated above, these reductions can be achieved by pyridine nucleotides without the assistance of high energy phosphate. The reduction of a carboxyl group requires a stronger reducing agent than TPNH2 or DPNHo (because, as mentioned on pages 215-219, C— O bonds are stabilized by accumulation at a single C atom). As was said before, in order to reduce a — COOH group by hydrogen available in reduced pyridine nucleotide, a high energy phosphate ester must be supplied, and its degradation to a "low energy" phosphate coupled with the reduction. (As an example, oxidation of glyceraldehyde to glyceric acid by DPN becomes reversible by oxidizing and phosphorylating a "low energy" triose monophosphate to "high energy" diphosphoglycerate.) By com- bining the reversal of this reaction with the (reversible) glycolytic reactions, it is possible to start with PGA, ATP and DPNH2 {i. e., DPN + chloro- plasts in fight), and end up with hexose diphosphate. Ochoa suggested (as did earfier Puben, Kok, van der Veen and others, cf. pages 1116-1117) that the high energy phosphate needed for this synthesis of hexose from PGA can be supplied, in fight, by reversal of a part of the photochemical reaction, e. g., by allowing some of the DPNHa formed in fight to be oxi- DIFFERENT OXIDANTS 1579 dized by molecular oxygen (perhaps via the cytochrome system) and stor- ing the oxidation energy in phosphate bonds (a coupling demonstrated by Lehninger) . In the above-described experiments of Vishniac and Ochoa the proof of the assumed photochemical reduction of pyridine nucleotides by water consisted in the demonstration of reduction and carboxylation of pyruvate (or other metabolic acids) in light ; a complete proof calls also for the dem- onstration of oxygen liberation in stoichiometric proportion. Furthermore, the relevance of these observations to the mechanism of photosynthesis depends on the yield of the reaction. As stated in chapter 4 in the discus- sion of earlier claims of photosynthesis in vitro, "everything is possible in photochemistry," provided one is satisfied with very small yields. Conse- quently, no reaction in vitro can be considered as a significant step in re- constructing photosynthesis outside the living cell unless its yield approaches that of natural photosynthesis. Vishniac and Ochoa (1952) were able to demonstrate, using chromous chloride as reagent, that oxygen was in fact formed in the stoichiometri- cally expected amount (specifically, 0.5 mole O2 were liberated per mole of lactic acid formed, in accordance with reactions (35.34Aa,b). The yield was, however, very low — corresponding to about one molecule oxygen per molecule chlorophyll every 3 hours or about 0.1% of the rate of photosyn- thesis in saturating light. Other above-enumerated enzymatic systems gave oxygen yields (calculated from the rate of formation of the reduction products) of one molecule oxygen per molecule chlorophyll between every 1.2 and every 60 hours. The light used was relatively weak — about 4 klux, but even with allowance for this fact, these rates are of different order of magnitude than those of photosynthesis. Tolmach (1951 ^•^) came to a study similar to that of Vishniac and Ochoa from a different side. The evolution of oxygen by illuminated chloroplast suspensions in the absence of specially added oxidants (chapter 4, page 62) points to the presence in plants of a "natural Hill oxidant" (or oxidants). These may, or may not, serve as intermediates in Hill reaction, or photo- synthesis, or both; in any case, it would be important to identify them. Hill had shown that the photochemical oxygen liberation by chloroplasts, without added oxidants, is very brief in washed chloroplasts, but lasts much longer if the chloroplasts are left suspended in the cell sap (or if a chloro- plast-free leaf extract is added to washed chloroplasts). Figure 35.19A shows the course of oxygen liberation which Tolmach was able to demon- strate by suspending a drop of a suspension of spinach chloroplasts in an illumination chamber traversed by a stream of pure nitrogen, and analyzing the gas leaving the chamber for oxygen by the highly sensitive phosphores- cence-quenching method of Pringsheim and Franck. 1580 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 Tolmach inquired whether the oxygen evolution can be enhanced by known metabohc intermediates; he found, in fact, that the addition to the drop of 1 X 10-2 mole/1, phosphoglycerate (PGA) or pyruvate (PA) in- 20 30 40 50 TIME (MINUTES) 60 70 80 Fig. 35.19A. Photochemical O2 evolution from 3 mm.^ of separated spinach chloroplasts (A) and from crude leaf juice (B,C), l)oth containing 3 ^g. Chi (Tolmach 1951). Phosphoroscopic determination of O2 pressure in N2 gas flow (9.6 cm.3 N2/min.), 10-12° C. creased the oxygen production by up to 16% of the stoichiometric equiva- lent of the amount added (assuming that each molecule of PGA or PA can consume 2 atoms hydrogen). Figure 35.19B shows the time curves of 20 30 40 50 60 T'MF (MINUTES) Fig. 35.19B. Photochemical O2 evolution from 2.5 mm.^ crude spinach juice with addition of 2.5 mm.^ water (A); and 0.02 M PGA (B) at 15° C. (Tolmach 1951). 9.6 cm.3 N2/min. Total O2 yields: 0.026 mm.» in (A), 0.095 mm.^ in (B). oxygen consumption of chloroplasts suspended in cell sap, with and without added PGA. The shape of the curves and the total additional oxygen liberation varied widely, apparently in dependence of the physiological DIFFERENT OXIDANTS 1581 state of the sample. This, and the failure of PGA to increase oxygen yield of chloroplasts which had been separated from the cell sap, indicate that PGA does not serve directly as a Hill oxidant, but contributes to the reac- tion in a more complex way — perhaps by dark reaction mth the reduced form of a "natural" Hill oxidant. Addition of adenosine triphosphate (ATP), glyceric acid or glucose to the chloroplast-bearing leaf juice did not increase the oxygen burst in light; but addition of TPN leads to strong stimulation of the oxygen production, as shown by figure 35.19C. At low TPN concentrations (e. g., 2.5 X 10~^ }0 40 90 «0 TIME (MINUTES) Fig. 35.19C. Photochemical Oo evolution from 1.5 mm.' chloroplast sus- pension (1.5 ng. Chi) to which were added: (A) 1.5 mm.' water; (B) 1.5 mm.' 1 X 10-' M TPN; (C) 1.5 mm.' 1 X IQ-^ M TPN. 10-12° C; 9.4 cm.' N2/ min. (Tolmach 1951). M) the amount of extra oxygen evolved was up to 17 times the stoichio- metric equivalent of the TPN added; at the higher concentrations (e. g., 5 X 10"* M) this ratio declined to or below 1, but the initial rate of oxygen production continued to increase with [TPN] (e. g., from 41 mm.^ O2 per milligram chlorophyll per hour in the absence of TPN, to 163 mm.* at 2.5 X 10-^ mole/1, and 282 mm.» at 5 X lO"" mole/1. TPN. Addition of TPN had no effect on oxygen liberation from washed chloro- plasts, indicating that this compound, too, entered into secondary oxida- tion-reduction reactions between the "natural" Hill oxidant and some reducible substances present in cell sap, and did not act directly as Hill oxidant (in agreement with the earlier observations of Mehler, and Holt and French). The above-suggested attribution of the incapacity of TPN to serve as such ultimate Hill oxidant, to rapid reoxidation of TPNH2, was 1582 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 supported by Tolmach's observation of a rapid (nonphotochemical) oxida- tion of added TPNH2 by spinach juice. At this point (following a suggestion by Vennesland and Conn), Tol- mach added to the chloroplast suspension the complete malic enzyme sys- tem (instead of TPN alone), to provide a "trap" for TPNH2. In unsep- arated, chloroplast-bearing cell juice, the addition of TPN, pyruvate, car- bonate, Mn + + salt and "malic enzyme" did not enhance the oxygen yield more than did TPN alone; but in precipitated, washed and resuspended chloroplast material, a small, but marked, stimulation was observed (figure 35.19D). If PGA was added together with the malic enzyme system, the extra oxygen burst was about double that with pyruvate alone. «■ ?25 O I z 2 UJ (£ W UJ (£ CL z u > o 2C 5 N^ SO TIME (MINUTES) Fig. 35.19D. Stimulation of photochemical O2 evolution from 3 mm.^ chloro- plast suspension (1.8 ^g Chi) containing 0.04 ^mole MnCl., 0.17 Mmole pyruvate, 0.4 Mg- TPN, 0.08 fimole glycylglycine buffer (pH 7.0), by addition (at 21 min.) of 2 mm.3 malic enzyme, and switching on the light (at upward arrow) (Tolmach 1951). Gas: 4.2% CO2 in N2; flow 8.8 cm.Vmin., 25° C. To complement the demonstration of oxygen evolution from the malic enzyme system by a demonstration of concurrent CO2 fixation, tracer C*02 was used. Some C*02 fixation could be observed in light even in the absence of malic enzyme; it was tentatively attributed to exchange reac- tions. Addition of malic enzyme increased the fixation by a factor of five, making it about equivalent to the extra oxygen production observed phos- phorometrically under the same conditions. The Chicago investigators did not attach to the pyruvate reduction, mediated by TPN in illuminated chloroplast suspensions, the same impor- tance as did Ochoa and Vishniac, who saw in it evidence of the actual mech- DIFFERENT OXIDANTS 1583 anism of photosynthesis. Tolmach was not certain whether the strong effect of TPN on oxygen hberation by crude leaf juice was due to intermedi- ate reduction of TPN (as in the pyruvate-mahc enzyme system), or to the conversion of some sap component into an effective Hill oxidant by TPN- mediated autoxidation (since he observed that addition of oxygen-saturated water to TPN containing juice also caused a new oxygen burst in light). The much greater volume of oxygen produced from crude leaf juice com- pared to washed chloroplast-pyruvate-bicarbonate-malic enzyme system indicated that only a small part of the former could be due to the reductive carboxylation of pj^ruvatc. Tolmach also pointed out that the successful competition of carbon dioxide with oxygen as hydrogen acceptor in photo- synthesis would be difficult to understand if the reduction of carbon dioxide were to be mediated by an intermediate (TPN) whose concentration in chloroplasts is very low (a value of 10~^ mole/1, was mentioned in Tol- mach's paper; however, subsequent determinations in Vennesland's labora- tory gave higher values) . According to Franck, efficient reduction in the presence of air Avould require TPN to displace the much more abundant O2 (about 3 X 10~^ mole/1, in cell sap equilibrated with air) from contact with chlorophyll. In fact, if one assumes, as Franck does, that a molecule of the oxidant must be associated with each chlorophyll molecule, only a primary oxidant with a concentration of the order of 0.05 mole/1, (in the grana) can function with a high quantum yield. If one assumes that effective energy exchange be- tween chlorophyll molecules reduces the number oi centers in which oxidant molecules must stand ready to accept the H atoms, transferred in the pri- mary photochemical process, by a factor of the order of 10- (or 10^), then a concentration of '~5 X 10~^ (or ^-^5 X 10 ~^) mole/1, may be sufficient to keep the reaction centers occupied; this could possibly — but not very likely — make TPN a successful contender for these sites. On the other hand, the lack of discrimination, characteristic of photochemical processes (because of the excess activation energy usually available in a quantum), should warn us from reading too much significance into the reduction — with a small yield- — of any reductant offered to the chloroplast system, partic- ularly of a reductant whose reduced form we happen to know how to trap very efficiently. Arnon (1951) also made experiments on the reductive carboxylation of pyruvic acid in the presence of malic enzyme and illuminated chloro- plasts. He demonstrated that the malic enzyme is present in the cyto- plasmic fluid left after the precipitation of chloroplast fragments by high- speed centrifugation, so that the complete photocatalytic sj^stem (except for TPN) can be obtained from one and the same plant. Figure 35.19E shows the photochemical evolution of oxygen from this system, whose de- 1584 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 tailed composition is given in the caption. The reaction rate in this figure corresponds to about 8 molecules oxygen evolved per molecule chlorophyll per hour in light of about 28 klux — about 5% of the saturation rate of photosynthesis. This yield, although considerably larger than that ob- served by Vishniac and Ochoa, is still quite small. Furthermore, accord- ing to curve I, the oxygen formation stops alter the evolution of only about 10~® mole oxygen from 10""^ mole pyruvate. 20r a 3 •a o fee O t6 IQ Fig. 35.19E. Curve/: O2 evolution from chloroplast fragments (0.5 mg. Chi) in light in the presence of 2 X 10 -« mole KCI, 2 X 10-« mole NaHCOj, 2 X 10-« mole MnClo, 1 X 10"^ mole li pyruvate, 6.5 X 10"^ mole TPN, and 0.2 ml. "malic enzyme" preparation from the same leaves. Total volume 3 ml.; 15.1° C, 28 klux. Curve II: same without malic enzyme. Curve III: same without TPN. (Arnon 1951.) Arnon and Heimburger (1952) added to the above-described study a chromato- graphic proof of the formation of tagged i-malate in the chloroplast-pyruvate-radiocar- bonate-malic enzyme system. The estimate of C *02 fixed in malate indicated a stoichio- metric equivalent of 4.5 mm.' oxygen; manometric measurement showed, in this par- ticular case, an evolution of 18 mm.^ O2 in the complete system and of 15 mm.' O2 in an aliquot from which malic enzyme was omitted, and in which practically no C" was fixed. The difference (3 mm.' O2) is in satisfactory agreement with the value calculated from C'^ fixation; but the very large O2 evolution in the blank (not shown by curve II in figure 35.19E) makes the experiment somewhat unsatisfactory. DIFFERENT OXIDANTS 1585 In evaluating the possible significance of these results for the mecha- nism of photosynthesis, Anion referred to the tracer experiments of Calvin et al. (to be described in chapter 30), indicating that malate is not normally an intermediate in the main reaction sequence of photosynthesis. Earlier in this section we mentioned the recently discovered "coenzyme A." Attempts to use this compound as ultimate Hill oxidant have not yet given positive results; whether it can serve as intermediate in the photo- chemical reduction of metabolites whose formation in respiration is coupled with the reduction of coenzyme A (such as the formation of acetate by oxi- dative decarboxylation of pyruvate) remains to be ascertained. An im- portant role was ascribed to this compound in photosynthesis by Calvin; but since this hypothesis was not based on photochemical experiments with chloroplasts, we will consider it in chapter 36 when dealing with vari- ous suggested chemical mechanisms of photosynthesis. More than halfway on the redox potential scale between the systems H2O/O2 (^0' = -0.81 volt at pH 7) and H,TPN/TPN (^0' = +0.282 volt at pH 7) lies the system ferrocytochrome c/ferricytochrome c (£'0' = —0.26 volt at pH 7). Despite this favorable position, oxidized cytochrome c was found by Holt and French (1948) not to produce oxygen in light in the presence of chloroplast suspensions. Holt (1950) reinvestigated this sys- tem, using a photometric method to observe the reduction of cytochrome c. He found that oxidized cytochrome c was reduced by chloroplasts (from Spinacia or Phytolacca amcricana) in light, but that the reduction was rapidly reversed in the dark. Experiments with added reduced cyto- chrome c in the presence of oxygen showed that the chloroplasts contained an enzyme ("cytochrome oxidase") capable of transferring electrons from reduced cytochrome c to oxygen. The oxidation was inhibited by cyanide, azide and carbon monoxide. The inhibition by the latter, however, could not be reversed by light as is the case with cytochrome oxidase in respira- tion. The presence of a cytochrome oxidase can explain the lack of oxygen production by illuminated chloroplast suspensions in the presence of oxi- dized cytochrome c; but it can not explain why practically complete reduc- tion of cytochrome c could be observed spectroscopically — unless this reduc- tion occurred not by hydrogen transfer from water, but by reaction with some cellular reductant. Mehler (1951-) also found rapid autoxidation of cytochrome c to follow chloroplast-sensitized photochemical reduction (revealed by spectroscopic observation). He attributed the failure of cj-anide to stimulate oxygen evolution from the cytochrome-chloroplast system (by poisoning the cyto- chrome oxidase) to a destructive effect on cyanide of concentrated spinach chloroplast suspensions (such as are used in manometric, in contrast to 1586 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 spectroscopic, observations of the Hill reaction). Mehler reported that Tolmach (working in the same laboratory) could observe photochemical oxygen liberation from the cytochrome-chloroplast mixture by using azide instead of cyanide as poison for the cytochrome oxidase. In chapter 37A we will describe the finding, in chlorophyll bearing plant tissues, of a new iron-porphyrin-protein compound, designated as ''cyto- chrome/" (Davenport and Hill 1952). It has a normal redox potential of — 0.37 volt (at pU 6-8; Eo' increases at the higher pH values by 0.06 volt per pH unit). Compared with the normal redox potential of cytochrome c (Eo' = —0.26 volt), cytochrome / is a stronger oxidant. Hill (1951) and Davenport and Hill (1953) pointed out that the difference between the normal potentials of cytochrome / and of the oxygen electrode at pH 7 (_0.81 volt) is 0.44 volt, or 1 1.2 kcal/mole. The free energy of the reac- tion: (35.34B) 2H2O +4Cyt/Fe + + + > 4 H+ + 4Cyt/Fe + + + O2 is therefore 44.8 kcal— or close to the energy of a red quantum. Daven- port and Hill suggested that this reaction may represent the primary photo- chemical process in photosynthesis (with further increase in reduction po- tential achieved by an energy dismutation mechanism, as repeatedly dis- cussed before). However, the transfer of four electrons by one quantum is an implausible mechanism. Furthermore, what matters for the possibility of a photochemical electron transfer is not the free energy change, AF, re- sulting from a similar transfer by the "slow" thermal mechanism, but the total energy change, AH*, required for an instantaneous transfer (with all nuclei held more or less rigidly in position, in approximate accordance with the Franck-Condon principle). Since these reactions involve a change in ionic charges, the two energy values, AF and AH*, may be quite different. (The AF of such reactions includes changes in the ion-dipole interaction with the medium— which are different for sudden and gradual transfer— and large entropy changes produced by liberation — or immobilization — of dipole molecules around the ions whose charges had been increased or de- creased.) Not much significance could therefore be attached to the simi- larity between the standard free energy of the thermal reaction (35.34B), and the energy content of a red quantum, even if it were not for the improb- ability of an elementary photochemical process involving simultaneous transfer of 4 electrons. We conclude that if cytochrome / molecules do serve as intermediates in photosynthesis, they must do so by accepting single electrons transferred from single H2O molecules by single quanta (which implies the loss of much more than one-half of the quantum energ\') ; alternatively, they could play a role in the thermal back reaction (which DIFFERENT OXIDANTS 1587 may be essential for "energy clismutation/' e. g., via the formation of high energy phosphate esters, as repeatedly discussed above) . No ox\'-gen liberation was observed in chloroplast suspensions with glutathione (Holt and French 1948), or with dehydroascorbic acid (Aronoff 1946, Holt and French 1948), or with oxalacetic acid and riboflavin (Mehler 195P). (It will be recalled that the last named compound was used ex- tensively by Krasnovsky for the dark reoxidation, in solution, of chloro- phyll after its photoreduction by ascorbic acid, cf. part A of this chapter.) This may be the place to mention attempts to use other intermediary metabolites, or miscellaneous biologically important substances, as Hill oxidants. Several of these compounds (including phosphoglyceric and pyruvic acid) were mentioned above as stimulating the photochemical oxy- gen evolution mediated by the ''natural Hill oxidant" in crude leaf juices, but causing no oxygen liberation with washed chloroplast fragments. Another oxidant of the same type is methemoglobin. In studying the "natural" Hill oxidant (first described in Hill's 1939 work, cf. page 63), Davenport (1949) noted that the initial rate of methemoglobin reduction by chloroplast suspensions was as high in the presence of a chloroplast free leaf extract as in the presence of ferric oxalate. On the other hand, methe- moglobin itself was not reduced photochemically by washed chloroplasts. Following this lead, Davenport, Hill and Whatley (1952) found that crude macerates of many leaves (cleared only by filtration through glass wool) re- duced considerable amounts of methemoglobin in light. Results of this type were obtained with leaves of Avena, Pisum sativu7n, Sambucus nigra, Chenopodium Bonns-Henricus and Strellaria media. Similar preparations from Calendula, Brassica and Tropaeolum were only weakly active, and those from Phaseolus muUiflorus and Centranthus ruber were inactive. The activity (related to unit chlorophyll amount) decreased with dilu- tion, e. g., from about 0.5 cm.* O2 per mg. chlorophyll per hour at [Chi] = 5.6 X 10-5 mole/1., to 0.15 cm. 3 at [Chi] = 1 X 10"^ mole/1. Figure 35.19F shows the loss of methemoglobin reducing capacity of chloroplasts upon washing (by two centrifugations) and its restoration by the addition of the supernatant from the first centrifugation, or of an aque- ous extract from previously acetone-extracted leaf powder (prepared as described on page 63). Similar aqueous extracts from acetone washed root material, or from (chlorophyll free) terminal bud material, were inac- tive. The bud extract inhibited the activity of the leaf extract, but the root extract contained no such inhibiting components. The "methemo- globin reducing" component of leaf extract was completely destroyed in 10 min. at 100° C, almost lost in 10 min. at 60° C, but remained almost unaffected after 10 min. at 50° C. It was gradually deactivated by ex- posure to air. It remained intact after overnight dialysis at 0° C, against 1588 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 10-2 M phosphate buffer (pH 7.4). It stayed in solution after precipita- tion by acidification to pH 5, and its activity was found unchanged after the pH was adjusted back to 7.4; but acidification to pH 4.5, or alkaliza- tion to pH 9.3, destroyed the activity, which could not be brought back even by immediate neutralization. The rate of methemoglobin reaction in the presence of the extract had a maximum at pU 8.3 (a similar optimum, pH 8.0, was found earlier for whole Stellaria chloroplasts in ferric oxalate solution, while optimum pH values of about 6.5 were found with quinone lOOr c 'x> o o o 100 time (sec) 200 Fig. 35.19F. Photochemical methemoglobin reduction (measured by oxy- hemoglobin formation) by 0.4 ml. chloroplast suspension from Chenopodium (1.2 g. leaves in 6 ml. 6% glucose, 0.033 M phosphate buffer, pH 7.4) (Davenport, Hill, and Whatley 1952). Whale muscle hemoglobin: 0.78 X 10"* mole /I. (A) crude suspension; (S) washed chloroplasts; (C) washed chloroplasts + 0.4 ml. first supernatant from washing; (D) washed chloroplasts + aqueous extract from acetone extracted leaf material, equivalent to 0.4 ml. supernatant. or ferricyanide as oxidant, cf. section 5(6) below). The rate increased with methemoglobin concentration, reaching saturation >1.4 X 10^* mole/ 1., it declined with increasing salt concentration in the extract (phosphate buffer or ammonium sulfate). Chloride ions appeared to have no effect on activity of the washed preparations. The reaction was inhibited by o- phenanthroline and urethan (cf. section 5(d)). Attempts to fractionate the — apparently proteidic — "methemoglobin-reduction-promoting" factor in leaf extracts, by fractional precipitation with phosphate buffer or ammo- nivmi sulfate, showed that activity was absent from the "globulin" fraction precipitated at half -saturation with the salts, and concentrated in the "albumin" fraction, precipitated only at complete saturation. Gerretsen (1951) suggested that ascorbic acid can play in chloroplast suspensions (and in live cells as well) the role of a substitute reductant (replacing H2O as hydrogen KINETICS 1589 donor). He based this hypothesis on the finding of an "induction period" in the poten- tionietric observations of photochemical processes in crude suspensions of Avena chloro- plasts. It could be ascertained that during this period the ascorbic acid content of the chloroplasts decreased. (Gerretsen suggested that photoxidation of ascorbic acid may be the cause of induction phenomena also in photosynthesis.) In agreement with this hypothesis, the "induction period" could be prolonged by tlie aildition of extra ascorbic acid. This hypothesis is equivalent to the assumption that chloroplasts complete, in light, a "Krasnovsky reaction" before they begin to carry out the "Hill reaction" and liberate oxygen. However, it is also possible that the photoxidation of ascorbic acid occurs by a mechanism of the type suggested by Mehler, i. e., via the formation of hydro- gen peroxide by Hill reaction with molecular oxygen as oxidant, and secondary' reaction of the peroxide with the substrate of photoxidation. (g) Miscellaneous Organic Compounds Some oxygen evolution was observed by Aronoff (194G) with salicylic aldehyde and benzaldehyde, but considerable chlorophyll bleaching oc- curred in this case. Some oxygen was also evolved with benzoyl peroxide; none with salicylic acid, fructose, methanol or butadiene monoxide. Gurevich (1947) found that illuminated chloroplast suspensions (from Primula, Stellaria or Atriplex) can reduce o-dinitrobenzene, NO2C6H4NO2, first to NO2C6H4NHOH, and then to NO2C6H4NH2. Since this reagent is a strong enzymatic poison, it was used in the form of a deposit on a filter paper strip suspended in the illuminated chloroplast suspension; in other words, it must have been reduced secondarily by products diffusing from the chloroplasts into the aqueous medium. 5. Kinetics (a) Methods for Measuring Reduction The common method for measuring the rate of the Hill reaction with any oxidants is the manometric (or chemical) determination of liberated oxygen. An advantage compared to photosynthesis is that only one gas is exchanged as a result of the photochemical reaction. The dark reactions also are less significant than respiration in living cells — at least with some chloroplast preparations and some oxidants. However, a correction for dark reactions, either consuming oxygen and liberating carbon dioxide ("respiration") or liberating CO2 without consuming O2 ("fermentation") often is needed {cf. section c) . With certain oxidants the determination of oxygen can be replaced or supplemented by convenient determinations of the oxidant. Physico- chemical methods are more valuable than chemical assaj^s because they permit the progress of the reaction to be followed without taking samples for titrimetric or gravimetric analysis. 1590 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 The use of organic dyes as oxidants invites the appHcation of photom- etry. French and Holt (1946) first used a visual method, determining the time needed for a given suspension to "completely decolorize" a certain quantity of dye. A complication arises when the absorption band of the dye overlaps that of chlorophyll; it is therefore most convenient to use dyes whose absorption peak is in the green. Photoelectric spectrophotometry permits more precise measurements than visual colorimetry. In this case, a narrow band or line can be used to measure the concentration of the dye in a region where the plant pigments absorb only weakly or not at all. Apparatus of this type has been described by Holt, Smith and French (1951) . They reported that, using the blue dye, 2,6-dichlorobenzenone-indophenol, the initial rate of the Hill reaction can be measured within 2 min., with as little as 0.05 mg. chloroplast material — a quantity which would yield only 0.1 mm.^ O2 during the same period. Careful and repeatedly checked calibration (dye concentration vs. photo- electric current) is required for reliable absorption measurements with photoelectric cells in spectral bands isolated by filters; this calibration can be avoided by the use of monochromatic light. Since some compounds may be reduced in light by chloroplast without liberation of oxygen, photometric measurements of the Hill reaction with untried oxidants must be checked to make certain that the reduction actu- ally occurs at the expense of water. Washing of chloroplast material re- moves water-soluble reductants (such as ascorbic acid), and thus makes the photometric method less subject to errors. Two electrochemical methods of measuring the Hill reaction are possible. One is applicable to oxidants whose reduction cause a change in acidity, for example : (35.35) H2O + 2 Fe(CN)6-3 > 2 Fe(CN)6-' + 3^ O2 + 2 H + A pH meter can be used to follow the course of this reduction. This method was first applied by Holt and French (1946). They checked whether any pH changes occurred in light in chloroplast-free solution of Hill's reactants (ferric oxalate + ferricyanide -f- ferric ammonium sulfate) and found that in white light (4500 f. c.) this mixture did produce some acid — probably by direct photochemical reduction of ferric oxalate; however, this reaction could be prevented by the use of red light. A similar method was also used by Clendenning and Gorham (1950) in their survey of the chloroplast preparations from different plants (sec- tion B2). With some species, strong acidification occurred in the dark (without liberation of oxygen). In a few cases, the pH drift in the dark was in the alkaline direction (cf. section (c) below). The absence of KINETICS 1591 such drifts must be ascertained when the pH method is used to measure the rate of the Hill reaction. Another potentiometric method for following the Hill reaction is the measurement of the oxidation-reduction potential. This is feasible when- ever the system contains one — and only one — electrode-active oxidation- reduction couple (such as ferrous ion-ferric ion, ferrocyanide ion-ferricyanide ion, or dye-leuco dye). This method was described by Spikes, Lumry, Eyringand Wayrynen (1950''-). They found it practicable with ferricya- nide as oxidant. However, the oxidation-reduction potential often showed considerable drifts even in the absence of oxidant. For example, when crude suspensions of cell content from sunflower leaves were illuminated, the redox potential of the medium changed rapidly. The drift continued for 12 hours. This may indicate that the Hill reaction was proceeding at the cost of a natural oxidant present in the cell juice (as described in the early experiments of Hill). Macerated cell material from spinach showed no such prolonged drift. Chloroplasts separated from the plasma and cell juice showed only a small change of potential with time during illumination without added oxidant. If the chloroplasts were boiled, the potential was not affected, even if ferricyanide was added. If, however, the chloroplasts were "live," a change of potential was observed in light, when a complete Hill solution, or ferricyanide alone, or quinone were added to them. The strongest change was produced when ferricyanide was used in a buffered solution (to prevent complications due to changes in pR). With this system, phosphate-buffered at pR 6.85, potential vs. time curves could be determined under a variety of condi- tions. Spikes et al. (1954) have developed a device to automatically record redox potentials in cell suspensions. The same method was also used by Wessels and Havinga (1952, 1953; c/. Wessels 1954) in experiments with different quinones and quinonoid dyes as Hill oxidants, and by Gerretsen (1950'-^, 1951), who combined the measurement of redox potential with that of pH changes in a study of light reactions in crude leaf juice. It will be noted that the displacement of an oxidation-reduction equilib- rium in light produces a nonequilibrium state in which one redox system (e. g., quinone-hydroquinone) is in the reduced, and another one (e. g., O2/H2O) in the oxidized state. Barring immediate effective spatial separa- tion of the two couples, an electrode immersed into the illuminated mixture, will follow — entirely or preponderantly — the change in the composition of the component with the greater electrode activity. The situation is simi- lar to that in the so-called "photogalvanic cells" (described by Rabino- witch) in which a platinum electrode, immersed into an illuminated mix- 1592 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 tiire of thionine and ferrous sulfate, responds to the reduction of the dye more than to the oxidation of ferrio salt (cf. part A, section 5 above). In following the Hill reaction with a redox electrode, one is actually measuring the photogalvanic effect in a system in w^hich only one of the two compo- nents is definitely known, and assumes that the other has no effect on the potential at all (e. g., because it remains confined in the chloroplasts and only escapes after the conversion of the oxidized form to molecular oxygen, which makes it "electrode inactive"). This is not necessarily true — some intermediates in oxygen liberation from water may diffuse into the medium, and may be electrode-active there. Another possible complicating factor is that in the presence of several Hill oxidants (e. g., of the "natural" hy- drogen acceptor together wdth the added one), the electrode may respond preferentially to one of them — and not necessarily to the one present in the highest concentration. An elegant method for the study of the Hill reaction is the measurement of isotopic oxygen exchange by means of a mass spectrograph. This method alone permits measuring, simultaneously, oxygen-liberating and oxygen- consuming processes. Its application by Brown and co-workers to photo- synthesis and respiration is described in chapter 37 D (section 3) ; it was applied to the Hill reaction by Mehler and Brown (1952) and Brown (1953). (6) Rate and Yield under Different Conditions We include in this section a summary of measurements of the rate of the Hill reaction under different conditions even though the reproducibility of these measurements is not too well assured. In particular, it has not yet been possible to maintain a constant rate of reaction in chloroplast sus- pensions for more than a short period of time (of the order of an hour at best, and often much less). The comparison of the efficiencies of the Hill reaction with different preparations and different oxidants is therefore best based on measurements of the initial rate, made in the first few minutes of illumination of a fresh preparation (or a preparation preserved without loss of activity as described in section 3(6) above). The most important kinetic constants of a reaction such as photosyn- thesis, or Hill reaction, are the maximum quantum yield as observed in weak light and the saturation rate in strong light. In studying the Hill re- action, attention has also been paid to the total amount of the added oxidant that can be utilized before the reaction comes to a standstill; however, this proportion has more to do with the chemical stability of the oxidant and its reactions with various components of the chloroplastic matter than with the efficiency of the photochemical apparatus. KINETICS 1593 Quantum Yield. French and Rabideau (1945) measured the quantum yield of the oxygen hberation by chloroplasts from Spinacia and Trades- cantia (with ferric oxalate as oxidant), and obtained figures ranging from a minimum of 15, up to 50 or more quanta per mole oxygen (c/. chapter 29, Table 29.X). The variations must have been due to differences in the state of the chloroplasts; subsequent experiments (sec. 2 above) made it appear possible that duration of illumination prior to the grinding of the leaves has been the most important factor. The quantum yield was < }/2 of that obtained in a parallel set of experiments with live Chlorella cells in carbon dioxide. In Chapter 29 we also tabulated the quantum yields found by Ehrman- traut and Rabinowitch (1952). Most of these were for whole Chlorella cells (c/. part C of this chapter) . Measurements were made, however, also with chloroplasts from Phytolacca, using quinone, Hill's mixture and ferri- cyanide as oxidant. The results were close to those obtained with live cells (I/t = 9 to 12, cf. table 29.XI). Since similar yields were obtained also for the oxygen production by Chlorella cells in carbonate buffers, they were considered as supporting the hypothesis that the primary photochemi- cal process in the Hill reaction is the same as in photosynthesis. Warburg (1952) arrived at entirely different conclusions. He measured the quan- tum yield of the Hill reaction in spinach chloroplasts with quinone as oxidant, at different wave lengths. His preparations (which he designated as "green grana," cf. section 2(6) above) gave I/7 = 65, 73, 85 and 101 quanta per O2 molecule for \ = 366, 436, 480 and 644 uifj., respectively. Warburg noted that these quantum yields were proportional to wave lengths, so that the energy yields (e on p. 1083) were constant (about 1%) throughout the spectrum (including the region of strong absorption by the carotenoids). He concluded that the Hill reaction is different from photosynthesis not only in that it has a very low photochemical efficiency, but also in that it has a X-independent energy yield, «, while photosynthesis has an extremely high efficiency and a X-independent quantum yield, 7. It will be noted that the disagreement is in this case opposite to that obtained in the study of the quantum efficiency of photosynthesis, Warburg's efficiencies being much lower than those found by other observers. He seems to have used chloro- plast preparations of low activity — as witnessed by their failure to give any reduction of Hill's mixture at all. Light Curves. Hill and Scarisbrick (1940) gave the earliest hght curve of the Hill reaction; it was reproduced in fig. 6 (p. 65, Vol. I), and showed saturation in the region of 20 klux. Aronoff (1946^) found that the rate of oxygen liberation from spinach "grana," with quinone as oxidant, was pro- portional to hght intensity up to 13 klux; (cf. fig. 35.21). Spikes, Lumry, Eyring and Wayrynen (1950) reported that the light curve for spinach chloroplasts, with ferricyanide as oxidant, is a rectangular hj'-perbola. Holt, Brooks and Arnold (1951) gave a light curve of the oxygen pro- duction by a chloroplast suspension from Phytolacca americana with an 1594 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 indophenol dye as oxidant. Its shape is similar to that for photosynthesis (fig. 35.20). Gilmour and co-workers (1953) measured hght curves for the reduction of ferricyanide by beet chloroplasts and found them to be rectangular hy- perbolas {i. e., 1/H is a linear function of 7). Their saturation level was rather low (corresponding to about 800 mm.^ O2 per milligram chlorophyll per hour, or a turnover time of 360 sec, cf. table 35.X). 100 80 - < a: UJ > I- < _i UJ 60 40 20 • • ( /• • / / / ^ L °" / '^ / / K 1 1 1 1 1 1 1 1 1 10 20 30 40 50 60 70 80 90 100 RELATIVE INTENSITY Fig. 35.20. Light curves of Hill reaction in chloroplast suspension: solid line, before irradiation; broken line, after irradiation with 253.7 raix (after Holt, Brooks, and Arnold 1951). Maximum Rate in Saturating Light. According to chapter 28 (section A5), the usual value of Pniax:/Chlo — the saturating rate of photosynthesis in strong light with ample supply of carbon dioxide, related to unit amount of chlorophyll — corresponds to about one molecule O2 liberated per ♦mole- cule chlorophyll about every 20 sec. (much higher values have been found only with aurea leaves). This corresponds to about 4000 mm.^ O2 per mg. chlorophyll per hour. Table 35. X gives in comparison* some of the maxi- mum (initial) rates of oxygen production by chloroplastic matter, as re- ported by several observers with different preparations and different oxi- dants. With a given chloroplast preparation, the oxygen production in saturat- ing light depends on the nature of the oxidant. This can mean that the rate-limiting process involves the supply (or the preliminary transforma- tion) of the specific oxidant prior to its participation in the photochemical reaction; or it may indicate a difference in the probability of back reac- KINETICS 1595 TABLE 35.x Maximum Initial Rate of Hill Reaction, H""^^- (in Mm.^ O2 per Mg. Chi per Hr.) Temp., Observer Preparation Oxidant Method ° C. r/max. Arnon, Whatley Chloroplast suspen- Quinone Mano- 2 1030 (I9491) sion from Swiss metric chard leaves; KCl Ferricyanide li a 800 added Phenol-indophenol it tt 730 Clendenning, Crude extracts from Hill's mixture Acidi- 50- Gorham leaves and algae metric 2500 (19502) from 80 species" Kumm, French Chloroplast suspen- Hill's mixture Mano- 5-10 320- (1945) sions from Tra- descantia flum- inensis [K2C2O4 0.5 M, Fe (III)-NH4- sulfate 0.01 M, metric 1500 Same, from K-,Fe(CN)6 0.02 It 570 Spinacia M, sucrose 0.2 M, borate 0.167 M] Holt, Brooks, Chloroplast suspen- 2,6-Dichloroben- Colorim- — 2200 Arnold (1951) sion from Phyto- lacca americana zenene etrio " Highest rates with lamb's quarters, millet, chard, lettuce, spinach, flax (cf. section B2 for details). * Cf. section B2 for comparison of species. tions between the (common) oxidation product (oxygen or a "photoper- oxide") and the (individual) reduction products. Back reactions do not by themselves lead to light saturation, but if the maximum rate in strong light is determined by the deficiency of a "finishing" or "stabihzing" catalyst, then a difference in the reducing power of the various reduction products can affect the turnover of this catalyst, and thus influence the saturation level. Aronoff (1946-) found three different hght curves (fig. 35.21), H = /(/), for the rate of oxygen liberation by "grana" from spinach chloro- plasts, as function of light intensity, for three different quinones. A dif- ferent yield in weak light, as well as a different saturation level, could be attributed to differences in the reducing power of the three hydroquinones ; but in this case the order of the curves should be the same in strong and in weak hght. The crossing-over of these curves, shown in fig. 35.21, remains unexplained. Effect of Concentration of the Oxidant. In photosynthesis, the effect of the factor [CO2] on the rate (discussed in chapter 27) was represented by a saturation curve, with evidence of inhibition at very high concentrations. The initial increase in rate with [CO2] appeared to be, to a considerable de- gree, the result of supply processes; we had to leave the question open whether supply effects can be eliminated, and an "intrinsic" CO2 saturation curve of photosynthesis obtained, reflecting the participation of carbon 1596 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 dioxide in the first step of photosynthesis, rather than its supply to the locus of this reaction. The same considerations apply to the Hill reaction; here, too, diffusion of oxidant may determine the shape of the function H = /[oxidant]. (In the case of Hill reaction in whole cells, the cell membrane is an impor- tant diffusion barrier.) High concentrations of oxidants produce inhibi- tion similar to that caused by high concentrations of carbon dioxide; with some oxidants inhibition seems to occur even before saturation is reached. Clendenning and Gorham (1950) investigated the effect of quinone con- centration of the rate of Hill reaction in separated spinach and wheat 80 Quinone . Anthraquinone I I L _L _L 40 60 80 % INCIDENT LIGHT 100 Fig. 35.21. Light curves of oxygen evolution by chloroplast fragments in light with different quinones (after Aronoff 1946). chloroplast suspensions (fig. 35.22A). Maximum initial rate was observed at 0.5-1 mg. quinone in 2 cc. (2-4 X 10 -- mole/liter). With 2 mg. quinone the initial rate was somewhat lower, but the total yield higher — the reac- tion proceeded steadily until about 90% of the stoichiometric oxygen amount was produced. With 4 or 8 mg. quinone the initial rate and the total yield both were much lower — an evidence of "self-inhibition" of the quinone reaction by excess quinone. It will be seen in part C that similar observations were made by Clen- denning and Ehrmantraut (1950) with live Chlorella cells. Holt and French (1946), using Hill's mixture (ferricyanide and ferric oxalate), observed (acidimetrically) no effect of [FeCye-^] between 5 X 10~^ and 60 X 10"^ mole/liter on the initial rate of the Hill reaction. Spikes, Lumry, Eyring and Wayrynen (1950), on the other hand, who meas- KINETICS 1597 ured the rate of reduction of ferricyanide by the redox potential method found that the initial rate declined with increasing [FeCye"^], from 1.30 to 0.12 mole ferricyanide reduced per mole chlorophyll per minute. Ap- proximately, this initial rate was proportional to [FeCy6~^]~S indicating a strong self-inhibition not found by Holt and French. On the other hand, the change in rate as the reaction proceeded followed closely the zero order law. Spikes et al. suggested, as possible explanation of this apparent con- 160 - ^-^•~A-A-A-A-£ 120 0 TIME, minutes Fig. 35.22A. Effect of quinone concentration on oxygen production by chloro- plast suspensions (after Clendenning and Gorham, 1950): (A) crude leaf mace- rate; (B) separated and resuspended chloroplast material. tradiction, that the ferrocyanide produced the same inhibition as ferricy- anide. With quinone, where Clendenning and Gorham found clear evi- dence of self-inhibition, Spikes et al. observed a proportional increase of initial rate with the concentration of the oxidant! Wessels (1954) also found that the reduction (he used the dye DCPI as oxidant) proceeds linearly with time, indicating no dependence of rate on the concentration, [DCPI] (cf. figure 35.22B). Maximum Yield and Back Reactions. What proportion of the theoretical amount of oxygen can be obtained in the Hill reaction with a given oxidant depends (a) on the extent of side reactions which destroy the oxidant, e. g.. 1598 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 by "dark" reduction or by photochemical reduction not involving water, and (6) on the rate of back reactions, which may lead to the establishment of a photostationary state short of complete reduction. Finally, deactiva- tion of chloroplasts may cause the reaction to stop before all oxidant has been exhausted. In the latter case, addition of new oxidant will not revive the reaction. With quinone, the main obstacle to 100% utilization of the oxidant seems to be its decomposition. Warburg and Liittgens (1946) reported yields of the order of 85-90%; Arnon and Whatley (1949i) reported "100%" with added 0.01 M KCl, and 36% without it (chard chloroplasts); Aronoff (1946-) could obtain only 35% of the theoretical oxygen amount from quinone. %dye reduced 10 12 U time in minutes Fig. 35.22B. Percentage reduction of 2,6-dichlorophenol indophenol in illuminated chloroplast suspension as function of time (Wessels 1954). Straight ascending segment indicates zero order reaction. With Hill's mixture, Holt and French (1946) obtained 93% of theoreti- cal oxygen yield (calculated for the initial amount of ferricyanide) after 1 hr. in an atmosphere of nitrogen ; upon longer illumination, oxygen was slowly consumed, probably through photoxidation. For the same reason, not more than 85% of theoretical yield could be reached in air, and the sub- sequent decline of oxygen pressure was much faster. With ferricyanide. Anion and Whatley (1949^) reported "100% yield" with 0.01 M KCl, and only 19% without added chloride. With phenol indophenol, Arnon and Whatley (1949^ reported a "100% yield" in the presence of KCl, and as much as 90% without added KCl. Earlier, Holt and French (1948) had reported yields of the order of 85% of the stoichiometric amount; however, the calculation was uncertain because of the low content of the dye in the commercial product. Clendenning and Gorham (1950^) found that, with both quinone or KINETICS 1599 Hill's mixture, 85-95% of the oxidant could be utilized for oxygen produc- tion in separated chloroplast fragments, and only 65-75% in crude sus- pensions containing plasma and cell sap, obviously reflecting oxidant losses by reaction with cellular material. Wessels (1954) considered more closely the second above-mentioned factor in "maximum utilization" of the oxidant — the influence of back reac- tions. This influence must depend on whether the experiment is carried out in an atmosphere of practically constant oxygen content in a closed system where the photochemically produced oxygen is permitted to accum- ulate, or in a stream of oxygen-free gas. It was already noted in Hill's w 15 20 25 30 35 40 45 50 55 w. time in minutes Fig. 35.22C. Effect of light intensity on change of redox potential in chloro- plast suspensions containing 2,6-dichlorophenol indophenol (^'c = —0.255 volt): solid hne, 15 and 12 klux; dashed line, 6 klux; dashed-dotted line, 2.3 klux; dotted line, 0.9 klux (Wessels 1954). early experiments that the reduction of ferric oxalate stopped short of com- plete reduction, depending on the partial pressure of oxygen. Incomplete decoloration of certain quinonoid dyestuffs also was observed before, and we mentioned several times that the failure to observe Hill reaction with many oxidants of low oxidizing power (such as the pyridine nucleotides) probably was due to the photostationary state being almost entirely on the side of reoxidation. Wessels observed, beside some cases of incomplete decolorization (e. g., of toluylene blue) , a large number of systems in which the photostationary redox potential indicated incomplete reduction. Of course, the potential, being a linear function of the logarithm of the ratio [reductant] [oxidant], never shows complete reduction (or oxidation) ; but when it differs from the normal potential by more than, say, 0.15 volt, its use for the calculation of 1600 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 the degree of reduction becomes doubtful. The potential may now be af- fected by some minor component or an impurity; and even if it does reflect correctly the state of the main redox system, this state may be maintained not by the (photochemical) forward and (thermal) back reaction, but by some side reactions (such as the reoxidation of the reduced form by a minor component) which are difficult to avoid even in pure solutions, not to speak of complex biological systems. For this reason, the stationary "photogalvanic" potentials measured by Wessels in various quinones and quinononoid dye systems can be used for kinetic speculations only when the -£ inmV 350 200 150 WO ss. add, o-pht N^. It ion Of inanthrol ne 'j;::^^:^^^ .^i^_. / dork 1 1 1 1 1 1 1 _ 20 25 30 35 iO i5 50 time in minutes 10 15 Fig. 35.22D. Back reaction induced by darkening or o-phenanthroline poisoning in chloroplast suspension in the presence of tokiylene blue (E'a = -0.137 volt) (Wessels 1954). concentrations of the two forms are not too different in order of magnitude. This is the case in the presence of oxygen, or in weak light, as illustrated by figure 35.22C. The normal potential of DCIP is -255 mvolt (on the "physicochemical" scale used in this book!); photostationary potentials of —285 mvolt and —225 mvolt thus correspond to 90% oxidation and 90% reduction, respectively. That the photostationary potential is in fact determined by a back reac- tion is illustrated by figure 35.22D, which shows how the reduction of toluylene blue {Eo = —137 mvolt) yields to reoxidation in darkness, or when o-phenanthroline is added, inhibiting the photochemical reaction. pH Effect. The optimum pH for the Hill reaction depends on the oxidant (and, according to Punnett, 1954, on whether one works with whole or broken chloroplasts) . With HilVs mixture (ferric oxalate + ferricyanide) Holt and French (1946) found a maximum initial rate at pH 7.6 at 3° C, and at pH 7.0 at 10° C. Clendenning and Gorham (1950^ also found for this oxidant an KINETICS 1601 optimum at pH 7.0; the rate depended, however, not only on the pR, but also on the chemical nature of the buffer. It was higher in 0.4 M sodium sorbitol borate buffer than in 0.04 Af sodium or potassium phosphate buffer, although the pH was the same (7.0) in both cases. Gorham and Clendenning (1952) found that the pH effect depends strongly on the presence of chloride (figure 35.22E). 1200 800 400 O 1200 800- 400 0-5 MIN. I0°C, \ WITHOUT \ \ 20°C. 8.0 1200 800 400 0 1200 800 400 0 H 5-IOMIN. IO°C. X fj^\. WITH \ KCL /^ ^ WITHOUT KCL 20°G. .y^"^' /^ Fig. 35.22E. Effect of joH on photochemical activity of washed, frozen and thawed chloroplasts in Hill's mixture, with and without 10"'' M KCl (Gorham and Clendenning 1952). 0.78 mg. Chi per vessel. With ferricyanide alone, Spikes, Lumry, Eyring and WajTynen (1950) found (by redox potentiometry) an optimum at pH 6.85, with sharp de- cline of the rate to almost zero at pH < 5 or > 7.5. With quinone, Warburg and Liittgens (1946) gave pH 6.2-6.5 as opti- mum, the rate declining to zero at pH 7.0; Amon and Whatley (1949), on the other hand, gave pR 7.0 at 10-15° C, and pB. 7.6, at 3° C. as pH optima for the same oxidant. 1602 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 Gorham and Clendenning (1952) found that the pH optimum lies at 5.0-6.5 without, and at 7.0-7.5 with added chloride. For chromate, Holt and French (1948) found pH 7.3 as optimum acidity; the initial rate dropped by }/3 of the maximum at pH 5.7, and by J^ of the maximum at pH 8.0. Effect of Temperature. Increase in temperature probably increases the maximum rate, JJ"^^^', but since it also accelerates the deactivation of 100 200 TIME, seconds 300 Fig. 35.23. Effect of temperature on the rate of acidification of chloroplast sus- pension in Hill's mixture in light; the chloroplasts had been in Hill's solution 3 minutes before illumination started (pH 6.8, 0.005 N KOH, 0.38 mg. chlorophyll) (after French and Holt 1946). chloroplasts, the increase may be overcompensated by the fast decay even in the initial rate measurement. French and Holt (1946) found that the stimulating effect of increasing the temperature from 3° to 20° or 25° C. was over after only 5 minutes of illumination (fig. 35.23). The tempera- ture coefficient of the initial rate (with Hill's mixture as oxidant) was Qio = 3.5 (3-15° C). Arnon and WTiatley (1949) found the highest rate of oxygen production at 15-20° C, dropping to equally low values at 10° C. and at 25° C. with all three oxidants used — quinone, ferricyanide and phenol indophenol. On the other hand, the total obtainable yield of oxygen was highest at KINETICS 1603 10° C; the loss at 25° C. was particularly high in the absence of chloride (when the photochemical deactivation of the chloroplasts is particularly- fast). Gilmour et at. (1953) found a linear relation between log P™'*''- and 1/T with ferri cyanide as oxidant, for the range 6-23° C. ; the slope corresponded to an activation energy of 7.4 kcal/mole. They reported that Bishop (1952) had observed, with the chloroplasts from the same species {Beta vulgaris) an activation energy of 10 kcal/mole. Concentration of Chloroplasts. Variations in the concentration of chloroplasts in the suspension could influence the rate of the Hill reaction in two ways. One is trivial: with increased optical density the light ab- sorption first increases almost proportionally with concentration, then rises more slowly, and finally approaches totality. The average absorp- tion per chlorophyll molecule is at first almost constant and then decreases steadily. The same must be true of the yield per unit chlorophyll amount in the lower part of the light curve, where light supply is the rate-limiting process. In saturating light, the rate per unit chlorophyll amount should be independent of chloroplast concentration ; but the higher this concentra- tion, the more light will be needed to reach this saturation. (These rela- tions have been discussed before, in chapters 25, 28 and 32.) A second, more significant concentration effect could be caused by ''poisoning" of the medium by products of respiration or fermentation of the chloroplast fragments; this inhibition (if it occurs at all) should be strongest in the most concentrated suspensions. Observations on the rate of Hill reaction in suspension of different con- centrations so far fall into the first, trivial category. Spikes et al. (1950) found a proportional increase in the rate of ferricy- anide reduction by spinach chloroplasts in a 0.25-cm. thick vessel at [Chi] values from 5 to 40 X 10^^ mole/liter in red light of "5000 lux" (this must mean intensity of white light before passing through a red filter, and seems to be too low a light for saturation, particularly in a concentrated suspension). Holt and French (1946) made the same observation in a much stronger light (4500 foot candles, or about 45,000 lux of white light before passing through a red filter) in a vessel about 0.5 cm. thick, and at chlorophyll concentrations up to but not in excess of 16 X 10 ~^ mole/liter. Clendenning and Gorham (1950^) found the rate of the reaction with Hill's mixture to be proportional to [Chi] up to 7.5 X 10"^ mole/liter. Wessels (1954) considered a different effect of chloroplast concentration — that on the final "photostationary" state (rather than on the rate of ap- proach to this state). Figure 35.23A illustrates the finding that while the initial rate of reduction of DCPI increases, as expected, with the concen- tration of chloroplasts, the photostationary state (as measured by the 1604 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 steady photogalvanic potential) remains the same. The kinetic signifi- cance attached to this result of Wessels, will be related below under "Kinetic Theories." Yield in Flashing Light. The data on the yield of the Hill reaction is flashing light in relation to flash energy and duration of dark intervals (Clendenning and Ehrmantraut 1951; Gilmour d al. 1953) were presented in chapter 34, section B7. Kinetic Theories. Kinetic analysis of the Hill reaction — of the type of that attempted in chapters 27 and 28 — is an even more uncertain undertak- ing than that of photosynthesis, because the deterioration of the activity of chloroplasts with time makes quantitative reproducibility of the data 10 15 20 25 30 35 iO 45 50 »• time In minutes Fig. 35.23A. Influence of chloroplast concentration on change in redox potential of suspension containing 2,6-dichlorophenol indophenol (^'o = —0.255 volt) (Wessels 1954). Chloroplast concentrations: soUd line, 4; dashed line, 2; dashed-dotted line, 1. questionable. In the case of live Chlorella cells (c/. part C below) there is no equally rapid deterioration, but the only oxidant successfully used so far, quinone, has a "self-poisoning" effect which complicates the kinetic picture considerably. A priori, the kinetic treatment of the Hill reaction should be simplified by the elimination of the complex carbon dioxide factor. However, the dif- fusion of the oxidant to the chloroplasts may sometimes become the limit- ing factor. Supply barriers are, of course, reduced by the greater aisper- sion of the material and the absence of cell walls ; but the necessity to use some oxidants in a very dilute form (this applies to dyes because of their strong light absorption; and to quinone — with live cells — because of its poisonous effect) increases the possibility of "oxidant limitation." Assuming the absence of such limitations, the parts of the discussion in KINETICS 1605 chapters 27 and 28 applicable to the Hill reactions are those on pages 1020- 1047, covering the effect of back reactions in the photochemical apparatus and the influence of "finishing" reactions of limited capacity, such as the enzymatic liberation of oxygen. An additional factor, which can play an important role in the Hill reac- tion, is the back reaction between the two final products, molecular oxygen and the reduced oxidant (ferric salt, or h;ydroquinone, or leucodye). In photosynthesis this kind of back reaction is the cellular combustion of car- bohydrates, an enzymatic reaction which is slow compared to photosynthe- sis, at least in strong light. In the Hill reaction many fully reduced oxi- dants are rapidly autoxidizable, so that a photostationary state is estab- lished in light (as described earlier in this section). The position of this state depends on light intensity, oxygen concentration, and the rate con- stants of the several forward and back reactions. One of the few attempts to date at kinetic analysis of the photostation- ary state of the Hill reaction, is that of Wessels (1954). He postulated the simple sequence of three reactions : light (35.35Aa,a') H2O + X . XH2 + HO2 dark (35.35Ab,b') XH2 + Q v QH2 + X (35.35AC) QH2 + 3^02 > Q + H2O where X is an intermediate redox catalyst common to all Hill systems (and probably to photosynthesis as well) and Q (for quinone or quinonoid dye) is a specific Hill oxidant. Comparison with reaction schemes such as 28. IB (page 1026) shows one important (and probably unjustified) omission — both back reactions, (35.35Ab') and (35.35Ac), are supposed to involve molecular oxygen, and no provision is made for reoxidation by oxygen precursors (systems Z/ZH2 or A'H0/A'0H2 in our earlier discussions, cf. for example reaction 28.21d in scheme 28. IB). The reaction sequence (35.35Aa,b,c) leads to expressions for the photo- stationary state which the reader can easily derive. These are of the form : [oxidant ] a [X ] [O2 ] '/2 + ?; [O, ] + c [oxidant ] [O. ] V2 (35.35B) R = [reductant] d/[Chl] Depending on whether the back reaction (35.35Ab') or (35.35Ac) pre- dominates, either the first, or the second and the third term in the numera- tor can be neglected. In the second case (if one assumes the concentration of the intermediate X to be proportional to that of chlorophyll), R becomes independent of both chlorophyll concentration and the concentration of the oxidant, in agreement with Wessels' experimental results (cf. e. g., figure 35.23A). In the second case, R depends on both [Chi] and [oxi- 1606 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 dant]. Wessels concludes that reoxidation occurs mainly by action of oxygen on the intermediate X, and not on the hydroquinone Q. Among the uncertainties of this derivation are, in addition to the neglect of intermediate systems "on the oxidation side," also the use of the concen- trations [X], [Chi], [ox.] and [red.] as if the system were homogeneous (while in fact, chlorophyll, and probably X as well, are structure-bound in the chloroplast fragments). Finally, the independence of R of [ox.] and [Chi] cannot be accepted as a general rule without more varied data. For mathematical elaboration of the importance of the last point, we must refer to Horwitz (1954^). We can only mention here the more com- plex kinetic derivations of Gilmour et at., intended primarily to account for observations in flashing light (c/. p. 1479). (c) Dark Reactions It was mentioned before that cell macerates and chloroplast suspensions undergo various reactions in the dark, which may involve molecular oxy- gen or added oxidants, and complicate quantitative investigation of the Hill reaction. Warburg and Liittgens (1946) found that press juices from spinach or beet leaves "respire" {i. e., consume oxygen in the dark); this oxygen con- sumption is not significantly affected by light. The suspensions used in this stage of Warburg's work were prepared by grinding the leaves in a meat grinder, and pressing the mash through cloth. Wliole cells and undissolved salts were precipitated from the sus- pension by centrifuging briefly at 1200 g. The supernatant was dark green, had a pH of about 6.5 and contained whole as well as broken chloroplasts. This green extract showed in a respirometer a rapidly declining oxygen up- take. Carbon monoxide (80%) reduced it by 50%. Since this inhibition was not light sensitive, it was considered indicative of oxygen transfer by a copper-containing oxidase (such as the cytochrome oxidase). This hy- pothesis was supported by the observation that the oxygen consumption increased upon the addition of the polyphenol, pyrocatechol. Even in the presence of this substrate, respiration declined with time; it could be further stabilized by the addition of excess hydroquinone. The oxidation- reduction chain, which is operative in the presence of both compounds, probably is p-hydroquinone^ f p-quinone + / > \ + o-quinone ■, /pyrocatechol + < + Cu "*■ oxidase ' ^Cu +2 oxidase + > [ + MO2 / I H2O (35.36) p-hydroquinone + x2 O2 > p-quinone + H2O KINETICS 1607 The pyrocatechol thus serves as a catalyst; the accumulated oxidation product is p-quinone, which is less toxic than o-quinone. The oxygen up- take continues uniformly until all hydroquinone is used up (fig. 35.24). The production of carbon dioxide (which, without added substrate, is about equivalent to the consumption of oxygen) is reduced by about 50% by the presence of pyrocatechol, and is close to zero if both pyrocatechol and hy- droquinone are added. This shows that hydroquinone has completely dis- placed the endogenous respiration substrate. When the suspended ma- terial was separated from the liquid by centrifuging (at pH. 6.5), oxygen 300 10 20 30 40 TIME IN DARK, minutes Fig. 35.24. Oxygen consumption by spinach chloroplast preparations in darkness (after Warburg and Liittgens 1948). consumption continued in both phases; but after centrifuging at pH 5 the "polyphenol oxidase" of the liquid phase was coprecipitated with the cliloroplast material (since now only the precipitate showed autoxidation) . Arnon (1949) studied the polyphenol oxidase activity of leaf material. He found that chloroplastic matter, separated from the cytoplasmic fluid, retains all the polyphenol oxidase activity of the original cellular material. Clendenning and Gorham (1950^ noted that in crude leaf macerates, freed only of heavy particles (by slow centrifugation), acid production upon addition of Hill's mixture occurs also in the dark. In some species the rate of this dark reaction was quite high — for example, in material from spruce and other gymnosperms it reached 4 moles acid per minute per mole 1608 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 chlorophyll. Leaf material from vegetables and cereals reacted much slower in the dark than leaf mash from trees, ferns and aquatics. Clendenning and Gorham (1950^) later studied the dark reactions of chloroplast suspensions in more detail. Acidification was faster at pH 7.5 than at pH 6.0. It required Fe+^ as ferricyanide or oxalate and was non- enzjonatic; it may involve reductants such as ascorbic acid, glutathione or tannins. Two other dark reactions also were observed : boiled sap from bean leaves produced carbon dioxide in the dark when mixed with Hill's solution— apparently by oxidative decarboxylation; crude chloroplast sus- pensions of the same species showed oxygen absorption on the dark (c/. the above-described observations of Warburg and Liittgens). A similar but less rapid oxygen uptake occurred with chloroplasts of other species. The autoxidation was heat sensitive in some species and not in others, indi- cating that both enzymatic and nonenzymatic reactions are involved. Gerretsen (1950-) also observed the ''respiration" of crude Avena chloroplast suspension in the dark. It was cyanide-insensitive (up to 0.04% HON), not stimulated by glucose, but enhanced by the amino acid asparagine; Gerretsen therefore concluded that the main cyanide-sensitive and glucose-stimulated respiration apparatus of the cells was destroyed in the maceration, while an accessory cyanide-insensitive mechanism, able to utilize amino acids, survived it. Mn++ ions had no influence on the rate of oxygen uptake by the chloroplasts in the dark (as contrasted to light). Gerretsen (1951) observed an increase in acidity of aerobic chloro- plast suspensions in dark, which he ascribed to CO2 formation by respira- tion; under anaerobic conditions a similar trend was observed and ascribed to fermentation. {d) Inhibition and Stimulation If Hill's reaction is brought about by a part of the photosjmthetic ap- paratus salvaged after the mechanical destruction of the cells, it must con- tain intact the photochemical mechanism and the enzymatic components (designated by Ec and Eo in chapters 7 and 9) involved in the liberation of oxygen from the primary photochemical oxidation product (designated there by Z, A'OH, or {O2}). Whether any enzymes which ordinarily oper- ate between the primary photochemical reduction product (designated as HX in chapter 7) and carbon dioxide are involved in the Hill reaction, we do not know. The substitute oxidants (such as quinone or Fe+^) may re- act with the primary reduction product HX directly, without the help of an enzyme (or they may even take part in the photochemical reaction proper, without the intermediary of the system X/HX which we have as- sumed to be tied to chlorophyll) . Certainly the Hill reaction does not need the carboxylating enzyme, Ea, which catalyzes the dark association of KINETICS 1009 CO2 with an acceptor (A or RH) prior to reduction. The "stabilizing" catalyst, Eb, whose existence was postulated by Franck and Herzfeld in their interpretation of the kinetics of photosynthesis, and which they made responsible for the light saturation of photosynthesis under normal condi- tions in constant as well as in flashing light, should not be involved in Hill reaction if its function is to stabilize the intermediate reduction products of ACO2 (symbolized by AHCO2, AH2CO2. . .), but it should be required if it acts on oxidation intermediates (or on reduction intermediates not de- rived from carbon dioxide). In Franck and Herzf eld's scheme 7.VA (Vol. I, page 164) this catalyst was supposed to act on both the reduction and the oxidation intermediates; but we have argued that this is implaus- ible because of the well-laio\vn specificity of enzymes. The experiments of Clendenning and Ehrmantraut on Hill reaction in live cells (c/. part C) indicate the probability that the same catalyst limits the rate in strong light of both photosynthesis and the Hill reaction ; this makes it likely that this catalyst operates on the primary photochemical oxidation product (or a primary reduction product not derived from CO2), and not on a reduction intermediate of carbon dioxide. We expect the Hill reaction to be sensitive to inhibitors which affect the enzymes Eb, Ec, Eq, but not to inhibitors which act on enzymes, such as Ea, involved only in the transformation of carbon dioxide. Cyanide was characterized in chapter 12 (Vol. I, page 309) as a specific poison of the' "carboxylase," E^. Hill and Scarisbrick (1940) found that the oxygen liberation of isolated chloroplasts in the presence of ferric salts is not inhibited by cyanide. Warburg and Luttgens (1946) said that they were unable to study the effect of cyanide on the Hill reaction with quinone as oxidant because (at 20° C. and pH 6.5) quinone oxidized HCN directly in the dark, with liberation of carbon dioxide. Aronoff (1946) found no ef- fect of cyanide on the photoreduction of quinone in "grana preparation" from spinach leaves. Ehrmantraut and Rabinowitch (1952) noted no inhibition of the quinone reaction in live cells at 10° C. and pH 6.5 when 0.005 M HCN was added immediately after the beginning of illumination (cf. part C below) . Macdowall (1949) found that >0.01 M KCN were needed to produce 50% inhibition of Hill reaction with phenol indophenol (Table 35.XI). Wessels (1954) found no inhibition of potentiometrically recorded Hill reaction (with quinone as oxidant) by 0.01 M KCN. Gorham and Clendenning (1952) noted that the presence of 10"' mole/1. KCN affects the i^ = / (pH) curves of the Hill reaction. It stimulated the activity of washed chloroplasts at pH 6.3-7.3, but inhibited it at other pH values. This influence is similar to that of other anions {cf. section 3(c)). 1610 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 Azide. An inhibition of the Hill reaction by sodium azide was ob- served by French, Holt, Powell and Anson (1946), Macdowall (1949) (c/. Table 35. XI), and Arnon and Whatley (1949), but not by Hill and Scaris- brick (1940) and Aronoff (1946). Clendenning and Gorham (1950) found that the effect of azide is dif- ferent with quinone and with Hill's mixture as oxidants. Wessels (1954) noted that NaNs reacts with quinone. Low concentrations (<6.10~' mole/1.) of azide caused only partial inhibition of oxygen liberation with quinone as oxidant. The strong inhibition, observed by Arnon and What- ley with quinone (50% inhibition at 6 X 10~^ mole/1.) must have been due to a reaction of the latter with the azide. With 2,6-dichlorophenol-indo- phenol (which does not react with NaNs) Wessels found only partial inhibi- tion by 6 X 10~^ mole/1. NaNs, and complete inhibition by 6 X 10"^ mole/ 1., in approximate agreement with Macdowall's data. Hydroxylamine. Hill and Scarisbrick (1940) reported that the Hill reaction is unaffected not only by cyanide but also by hydroxylamine — a result which seemed to disagree with the conclusion, reached in chapt. 12 (Vol. I, page 311) that hydroxylamine probably is a specific poison for the oxygen-liberating enzymatic system in photosynthesis. However, Hill's observations were confirmed by Aronoff (1946), who found no effect of hydroxylamine on photoreduction of quinone by fragmented chloroplasts from spinach leaves. French, Holt, Powell, and Anson (1946), on the other hand, noted an inhibition of the photoreduction of Hill's mixture by hydroxylamine; and Macdowall (1949, cf. Table 35.XI) found that 3 X 10 -Mf hydroxylamine inhibits the oxygen liberation by chloroplasts, with phenol-indophenol as oxidant, by 50%. The effect was about the same in strong and in weak light (see Vol. I, page 312 for similar observations in photosjnithesis) . Arnon and Whatley (1949) also found hydroxylamine to be a strong inhibitor for the photoreduction of quinone by chloroplasts. Clendenning and Gorham (1950) found that 10"^ mole/1. NH2OH in- hibited the photoreduction of quinone but not of Hill's mixture. Wes- sels (1954) noted that, similarly to azide, hydroxylamine reacts with quinone (and also with 2,6-dichlorophenol-indophenol, probably reducing the dye to a leucodye). To be able to observe the true effect of NH2OH on Hill's reaction, Wessels used toluylene blue, whose E'q is too high for it to be reduced by hydroxylamine. He found a strong, albeit not complete, in- hibition by 6 X 10~^ mole/1. NH2OH; the surmise that hydroxylamine must inhibit the Hill reaction was thus confirmed. The controversial re- sults obtained by other observers who used stronger oxidants, which react directly with NH2OH, must have been due to differences in the relative amounts of the reactants (perhaps also in temperature and schedule of the experiments). KINETICS 1611 0-Phenanthroline. Warburg and Luttgens (1946) found a strong inhi- bition of the photoreduction of quinone by o-phenanthroline, a compound that, similarly to cyanide, forms complexes with heavy metals, and also inhibits photosynthesis (c/. Vol. I, page 319) (50% inhibition by 4.3 X 10-« mole/liter, 100% inhibition by 8.5 X 10"^ mole/liter). Since Zn+2 ions form complexes with phenanthroline, Warburg and Luttgens suggested that an enzyme containing zinc as active metal may be involved in the Hill reaction, and thus probably also in photosynthesis. The phenanthro- line inhibition could be cured by the addition of zinc sulfate. The effect of phenanthroline on the quinone reaction was confirmed by Aronoff (1946) and Arnon and Whatley (1949). The latter also confirmed (1949^) the reversal of the phenanthroline inhibition by zinc ions, but showed that the same effect could be produced by ferrous ions and copper ions (which are required micro nutrients), as well as by nickel and cobalt ions (which are not) . Added in 1:1 stoichiometric relation to phenanthroline, nickel and cobalt produced a more effective reversal of inhibition than zinc ; the latter, in turn, was more effective than copper or iron. The assumption that an enzyme with a zinc ion in its prosthetic groups participates in photosynthe- sis thus remains speculative. Wessels (1954) noted a difference in the effect of phenanthroline on the Hill reaction with DCPI (almost complete inhibition by 6 X 10~^ mole/1.) and with quinone (6-10 times more poison needed for the same effect). He confirmed that the poisoning can be cured by Zn++, Co++ or Ni++ ions, but not by Mn++, Ca++ or Mg++. The phenanthrohne effect seems tc be due to its attachment to a specific spot in an enzyme (rather than to a chelation of free zinc ions, as suggested by Warburg), because extracting the chloroplasts with phenanthroline does not leave an impairment of their activity after washing. Chelating agents, similar to phenanthroline (a,Q:'-dipyridyl, 3,3'-dimethyl-2,2'-dipyridyl), do not inhibit the Hill reac- tion, while the strong chelating agent "complexon" (ethylenediamine- tetraacetic acid) even stimulate it. o-Phenanthroline inhibits the methemoglobin reduction by crude chloro- plast suspensions (50% inhibition at about 5 X 10"* mole/1.). Carbon Monoxide. Warburg and Luttgens (1946) found that the qui- none reaction is not inhibited by carbon monoxide (indicating the absence of iron-porphyrin enzymes in the enzymatic apparatus of the Hill reaction) . Narcotics. Hill and Scarisbrick (1941) and Warburg and Luttgens (1946) noted the inhibition of the reduction of ferric oxalate or quinone by phenylurethan (50% inhibition by 0.1 mg. in 1 cc. according to Warburg and Luttgens — approximately the same effect as on photosynthesis in Chlorella, cf. page 323, Vol. I). Aronoff (1946) obtamed only 60% inhibi- tion of the same reaction by a saturated phenylurethan solution; and 1612 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 Macdowall (1949) found that 2 X 10 ~' M phenylurethan are required for a 50% inhibition with quinone (table 35.XI). Arnon and Whatley (1949) found a higher sensitivity — nearer that observed by Hill and Warburg, than that observed by Aronoff and Macdowall. Wessels (1954) found incomplete inhibition by 3 X 10~' mole/1, phenylurethan of the photoreduction of benzoquinone, but a complete inhibition of that of 2,6-dichlorophenol-indophenol; the latter was 50% inhibited by 3 X 10"" mole/1. Davenport et al. (1952) observed 50% inhibition at 1.2 X 10~^ mole/1, phenylurethan (or 0.35 mole/1, ethyl- urethan) of the methemoglobin reduction in crude chloroplast suspensions. Warburg and Liittgens (1946) observed that the quinone reaction is completely inhibited in a saturated solution of octanol. Other Inhibitors. Macdowall (1949) made a study of the inhibition Table 35.XI Inhibition of Hill Reaction in Chloroplasts from Spinach or Swiss Chard with Phenol-Indophenol as Oxidant (after Macdowall, 1949) Concentration producing 50% inhibition, Inhibition occurs Inhibitor mole/liter in strong or weak light Poisons of Heavy Metal Prosthetic Groups Cyanide" >0 . 01 — Azide" 0 . 08 In strong light Pyrophosphate >0. 1 — Hydroxylamine" 3 X 10"" Equally in both Resorcinol 0 . 09 Equally in both o-Phenanthroline" 2 . 7 X 10~^ In both, but more pronounced in weak light Thiourea 0. 15 — Poisons of Sulfhydryl Groups* Dinitrophenol" 6 X 10~^ In both, but more pronounced in weak light lodoacetate'' 0.01 In strong light Heavy Metal Ions Cu"^^ '' IX 10~^ Much stronger in weak light Hg+2 '> 4 X 10"^ More pronounced in strong light Narcotics Phenylurethan 2 X 10~' Stronger in weak light Thymol* 2 X lO"' Equally in both Chloroform 0 . 03 Stronger in weak light Ether 0.5 Strychnine >1.0 Stimulation in strong, inhibi- tion in weak light " Cf. other data above. * Cf. Wessels' data below. KINETICS 1613 of the Hill reaction by a variety of agents; his results are summarized in Table 35.XI. Wessels (1954) found partial inhibition of the photoreduction of quinone and DCPI by 6 X IQ-" mole/1., and full inhibition by 6 X IQ-' mole/1. 2,4-dinitrophenol, in agreement with table 35. XI. In agreement with Aronoff (1948) and Clendenning and Gorham (1950), but contrary to French et al. (1946), he found no marked effect of sodium fluoride on the Hill reaction. Thymol inhibited quinone reduction by 50% in a concentra- tion of 3 X 10~^ mole/1. In contrast to Macdowall's data, iodoacetamide (3 X 10"- mole/1.), p-chloromercurihenzoate (6 X 10^* mole/1.) and p- aminophenyldichloroarsine (6 X 10"* mole/1.) (all three strong sulfhydryl inhibitors), were found without effect. Only slight inhibition was ob- served with 3 X 10-2 mole/1, nitrite. Phthiocol (6 X 10"^ mole/1.) in- hibited the reduction of DCPI, but not that of quinone; 0.001 M penicillin or Chloromycetin had no effect. Sulfanilamide and nicotinic acid were with- out effect even at 10 "^ mole/1. Of the cations studied by Wessels (all 10-* M), Zn++ was found to stimulate the reaction with quinone, but not that with DCPI; Mn++ stimulated both of them. Mg++ ions had no ef- fect, while Ca++ ions accelerated only the reaction with quinone. Cu++ ions inhibited strongly at 10-" mole/L, Hg++ ions even at 10"^ or 10"® mole/1, in agreement with table 35. XI. No effect was caused by oxidized glutathione, DPN (cf. section 4(/) above) or ATP (6 X 10 -" mole/L). Figures in Table 35. XI and the inhibition experiments discussed before, support the view that the cyanide- and azide-sensitive component of the photosynthetic mechanism is not involved in the Hill reaction, and is prob- ably concerned with the transformation of carbon dioxide. Hydroxyla- mine, o-phenanthrolin, and dinitrophenol are powerful inhibitors of the Hill reaction, affecting it either equally strongly in weak and strong light, or preferentially in weak light. Of the two heavy metals, cupric ions inhibit much stronger in weak light {cf. Vol. I, page 340 for a contrary observation by Greenfield on photosynthesis in Chlorella), mercuric ions in strong light. The narcotics act, as usual, at all light intensities; but phenylurethan and chloroform show a somewhat enhanced action in weak light. Strychnine (5 X 10"* M) strongly stimulated the Hill reaction in strong light (46 klux), but inhibited it slightly in weak light. In addition to the specific inhibitors listed in Table 35. XI, Holt, Smith and French (1950) and Macdowall (1949) observed also the influence of various organic reagents: sucrose, formaldehyde, acetone, ethanol, methanol, carbitol, isopropanol, n-propyl glycol, propyleneglycol, glycerol, toluene, dioxane; and of salts, such as MgS04, Na2S04(NH4)2S04 and CaCla. All had a more or less pronounced inhibiting effect. Complete 1614 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 inhibition was produced by 5% carbitol, 10% dioxane or 0.2 M CaCl2; 50% inhibition was produced in strong light by 0.18 M formaldehyde, 0.4 M sodium sulfate, 0.7 M acetone, 1AM ethanol, etc. No stimulation by small quantities of formaldehyde (as reported by Bose for photosynthesis, cf. Vol. 1, page 343) was found at concentration down to 10"^*^%. The suspensions used in Macdowall's experiments contained between 2 X 10~^ and 4 X 10~^ mole/liter chlorophyll. In checks with o-phenan- throline (2 X 10"^ mole/liter) and thymol (2.5 X 10"^ mole/liter) the inhibition proved to be approximately independent of chlorophyll concen- tration in these limits. Arnon et al. (1954) reported that with whole chloroplasts capable of (1) Hill reaction with quinone, (^) P* fixation as ATP* in light, and (3) C*02 fixation in Hght: 2,4-dinitrophenol (8 X 10-^ M) inhibits (5) and (2) much stronger than (1); iodoacetamide (2 X 10"^ M) inhibits (3) by 97% and has no effect on (2) ; and p-chloromercuribenzoate inhibits (3) strongly, (£) less strongly, and (1) hardly at all. Ultraviolet Light. Irradiation by ultraviolet light (253.6 mix) was found by Holt, Brooks and Arnold (1951) to inhibit the Hill reaction in chloro- plast fragments in the same way, as it does the Hill reaction of whole cells (cf. part C below) and photosynthesis (page 344). In this case, too, the logarithm of the residual rate is a linear function of the duration of irradia- tion, indicating a "first order" deactivation process (i. e., deactivation by single quanta of ultraviolet light, and not by combined effect of two or more quanta on a single molecule). The fact that the Hill reaction — in chloroplasts or whole cells — is inhibited by ultraviolet light to about the same extent as photosynthesis is an argument against the hypothesis, made in chapter 13 (on the basis of observations of inhibition of C* uptake in the dark by ultraviolet light), that the 253.6 m.ix sensitive factor in photo- synthesis is the carbon dioxide fixing enzyme. Activity was tested with the complete Hill's mixture and with ferricyan- ide alone, and the effect proved to be the same in both cases. This proves that photochemical decomposition of oxalate — which is known to occur in ultraviolet light — is not the cause of the inhibition of the reaction with Hill's mixture. Similarly to whole Chlorella cells, the irradiation of chloroplast prepara- tions with the line 253.6 mn does not cause any significant changes in the absorption spectrum, even if it leads to complete deactivation. This has been checked spectrophotometrically for the region 220-400 m/^; no change of color was noticeable after irradiation. The proportion of the photochemical activity which "survives" a cer- tain irradiation dose is the same whether the activity is tested in strong or in weak light. Ultraviolet light thus belongs to the group of agents (in- KINETICS 1615 eluding also hydroxylamine and narcotics) which affect photosynthesis more or less uniformly at all light intensities. (e) Survival of Photochemical Reductant in the Dark In the next chapter we will deal with the controversial question of whether illumination of live cells produces in the latter a strong reductant which survives long enough to permit the demonstration of its "reducing power" after the cessation of illumination. This controversy has extended also to the Hill reaction. Mehler (1951 ^) attempted to find a surviving reductant in illuminated (crude) spinach leaf juice. From the data of Calvin, and Fager, on live cells (c/. chapter 36) , coupled with estimates of the recovery of chlorophyll and other cell constituents in the preparation of this juice, he expected to find an amount of reductant equivalent to about 4% of that of chlorophyll. The suspension was illuminated in phosphate buffer in N2 atmosphere, pipetted into a solution of dichlorophenol-indophenol (DCPI), and the absorption at 610 m/x measured about 15 seconds later. No difference could be noticed between the effect on the optical density of chloroplasts illuminated for 3^, 1 or 2 minutes, or not illuminated at all. The tempera- ture of the experiment was not stated. Krasnovsky and Kosobutskaya (1952), who made practically identical experiments with chloroplast suspensions from Phaseolus, arrived at the opposite conclusion. They illuminated the suspension for 3 minutes at 0° C. in a vacuum Thunberg tube, added DCPI from a side tube 1-5 sec- onds after the end of illumination, and found the optical density at 600 m/x, measured 20-40 sec. after the addition of DCPI, distinctly different from that measured in a similar experiment with not preilluminated suspension. This difference (AD) varied strongly, depending on the ''physiological state" of the material; it was much smaller in a suspension of washed chloroplasts (<0.05) than in the crude leaf juice (up to 0.15). The AD remained unchanged after 3^, 1 or 5 minutes, indicating that the "reducing power" survived for over 5 min. in the dark (at 0° C. and in the absence of O2). Similar results were obtained with thionine, but AD was only one- half as large as with DCPI. A slow reduction of DCPI in the dark was found to be superimposed on the light reaction (or reaction of preillumi- nated chloroplasts) ; this dark reaction was attributed to the presence, in the juice, of dehydrogenases and hydrogen donors, such as ascorbic acid. Preillumination appeared to increase the quantity of the reductants in the chloroplast juice by 10-30%, equivalent to 5-8% of the amount of chloro- phyll present. The light-activated reductant did not seem to be a reduced form of chlorophyll, since the absorption spectrum of the latter was not 1616 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 changed by preillumination. Krasnovsky suggested that Mehler's results were affected by high temperature and presence of air, which favored back reactions. Gerretsen (1950^ believed he found potentiometric evidence that anaerobically preilluminated, crude chloroplast suspensions maintain in the dark, for more than an hour, a capacity to form a peroxide upon ad- mission of oxygen or addition of Mn++ salt. 6. Mutations Davis (1952) apphed induced mutations to the study of the relation of Hill reaction to photosynthesis. He subjected Chlorella cells to ultraviolet irradiation and isolated three strains which differed from the wild type by being unable to grow in an inorganic nutrient solution in light. These strains grew well, however, if supplied with glucose. They were green and contained chlorophyll. Two strains (322 and 349) did not evolve oxy- gen in light when suspended in a carbonate bicarbonate buffer, and did not take up carbon dioxide. One strain (332) liberated oxygen in light, but did. not take up carbon dioxide. Intact cells of strains 322 and 349 did not liber- ate oxygen when suspended in Hill's solution in the absence of external car- bon dioxide and illuminated. They thus seemed to have a block in the oxygen-liberating mechanism. Strain 332 liberated oxygen in light in the absence of carbon dioxide, and in the presence as well as in absence of Hill's solution. In the latter case, hydrogen from water must have been donated to some cellular component. Strain 322, which is unable to photosynthe- size, also has a subnormal respiration rate. Assuming single gene muta- tion, this correlation can be considered as indicative of the presence of two respiratory systems, one of which is associated with the chlorophyllous mechanism. With strains 322 and 349, the effect of light on respiration could be studied; experiments showed the absence of such an effect. The wild type was inhibited when grouTi in the presence of sufficient quantities of mutant cells. The inhibition could be caused either by chlorellin (c/., page 880) being produced in greater quantities by the mutants than by the wild type, or by compounds which accumulate within the mutant cells as a result of the blocked reactions and diffuse into the medium. C. Photochemistry of Live Cells* The first observation that a "Hill reaction" (i. e., sensitized oxidation of water to oxygen, with oxidant other than carbon dioxide) can occur in live cells was made by Fan, Stauffer and Umbreit in 1943 (Vol. I, page 541). They found that oxygen is liberated in light from carbon dioxide-free Chlorella suspensions supplied with ferric phosphate or other ferric salts. * Bibliography, page 1629. PHOTOCHEMISTRY OF LIVE CELLS 1617 However, back reactions (reoxidation of ferrous salts by free oxygen) caused the oxygen production to come to an early stop. Better oxygen yields could be obtained with various organic compounds containing a car- boxyl group, particularly hcnzaldehjde and acetaldehyde, as well as para- banic acid and nitrourea. Numerous other aldehydes, oximes, acids, quin- onoid dyestuffs, carbohydrates, and urea and its derivatives (for a list, see Vol. I, page 542) were tried but gave negative results. The greatest oxygen production was observed with benzaldehyde, but the interpretation of this reaction was somewhat uncertain, since a considerable dark reac- tion was noted which produced carbon dioxide. Conceivably, this reac- tion could be accelerated in light, and benzaldehyde, instead of participat- ing directly in the oxygen-liberating reaction, could be first photoxidized to carbon dioxide (or dismuted to reduced compounds, such as benzalcohol, and carbon dioxide) and the latter assimilated in the normal way. True, Fan et al. were unable to trap any free carbon dioxide by alkali ; but this evidence is never completely convincing because rapid intercellular reutiliza- tion of carbon dioxide may make its trapping by extracellular absorbers impossible. The highest observed rate of liberation of oxygen with benzaldehyde as oxidant was about 10% of the rate of photosynthesis. Warburg and Liittgens (1946; cf. Warburg, 1948) obtained the first reliable results with o-henzoquinone. In this case, too, a dark reaction oc- curred, which produced carbon dioxide; but its rate (about 0.02 cc. CO2 per 1 cc. of cells in 5 min.) was only about 2% of the rate of liberation of oxygen in light (0.5 cc. O2 per 1 cc. of cells in 5 min.). Aronoff (1946^ repeated Warburg's experiments with Scenedesmus in a nitrogen atmosphere (0.1% O2). He observed only slow oxygen evolution in light, and thought it to be limited by the rate of penetration of quinone into the cells. Clendenning and Ehrmantraut (1951) made manometric studies of the oxygen production by Chlorella with o-benzoquinone or Hill's mixture as oxidants. (No oxygen liberation was obtained with ferricyanide alone or with phenol-indophenol.) The manometer vessel, containing C02-free cells and the substitute oxidant, could be placed in homogeneous light side by side with a similar vessel containing identical cells in carbonate buffer; or the two vessels could be used alternately in the same position. In this way the yield of the Hill reaction could be measured in relation to the yield of photosynthesis under the same conditions (except for the difference in pH). Several con- clusions emerged from this comparison : Constancy of Rate, and Total Yield. With 10 mm.^ of cells and 1 mg. of quinone in 3 cc. (about 3 X 10"* mole/liter), the initial rate of oxygen evolution in continuous light was maintained for 30-40 min.; a total 1618 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 oxygen amount was obtained which corresponded to 75 ± 5% of the theo- retical equivalent of the amount of quinone added. Since white light was used, this low yield might have been caused by direct photochemical de- composition of quinone. An increase in the amount of cells did not change the yield. At the lower concentration of quinone (0.25 mg. or 0.5 mg. in 3 cc), the initial rate and the fuial yield were the same as with 1.0 mg. and exhaustion was reached correspondingly sooner — in 30 min. and 15 min,, 0 20 40 60 80 100 120 PHOTOCHEMICAL REACTION TIME, minutes 20 40 60 80 100 REACTION TIME, minutes Fig. 35.25. Effect of quinone concentration on photochemical oxygen production by Chlorella (concentrations in milligrams in 3 cc.) (after Clendenning and Ehrmantraut, 1951). (A) In white light (fluorescent Ught); (B) in red-orange light (X > 520 m/n). respectively. With 2 mg. quinone the initial rate and the total yield were markedly smaller, and with 4 mg. only very little oxygen was produced (fig. 35.25A) . Two phenomena seem to be involved in the failure to obtain a stoichiometric oxygen equivalent of the added quinone. One is the de- struction of quinone (by reactions with cell constituents; in blue- violet light, quinone is decomposed even in the absence of cells). That quinone actually is lost is indicated by the observation that if the reaction with 0.5 mg. quinone is continued until all oxygen production ceases, the addition of another 0.5 mg. revives it, and approximately the same oxygen amount (equivalent to ~75% of the extra quinone added) can be produced again. The results were improved by using red-orange light, only slightly absorbed PHOTOCHEMISTRY OF LIVE CELLS 1619 by quinone (figure 35.25B) . However, even in this case, very little oxygen was produced when 4 mg. quinone were added. This cannot be explained by direct photolysis of quinone, and indicates the occurrence of a second phenomenon — "self-inhibition" of the Hill reaction by quinone (also noted with isolated chloroplasts, c/. section 4(d) above. z o h- o 3 Q O cr a. UJ X o U- o < cr. Photosynthesis : O 20 40 60 80 RELATIVE LIGHT INTENSITY 100 Fig. 35.26. Light saturation curves for photosynthesis and the quinone reac- tion in whole Chlorella cells (after Clendenning and Ehrmantraut 1951). (•) net photosynthesis; (O) photosynthesis corrected for respiration. Maximum Rate in Strong Light. Figure 35.26 shows the light curves of photosynthesis and of the Hill reaction (with quinone as oxidant) for two aliquots of the same cell suspension. (10 mm.^ cells in 3.0 cc. and 3 X 10~^ M quinone.) The two curves approach the same saturation level in strong light, although more light is needed to saturate the Hill reaction. The rates were measured in the first 30 min. of illumination, following a 20-min. dark period after the addition of quinone. (We will see below that even brief preillumination of cells without quinone produced an inhibition.) Clendenning (1954) found that at low temperatures (0-15° C.) O2 evolution with quinone as oxidant far exceeds that by photosynthesis. This supports Franck's view (chapter 31) that at these temperatures car- boxylation becomes the rate-Hmiting reaction in photosynthesis. 1620 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 Quantum Yield. Figure 35.26 shows a lower yield of quinone reduction compared to photosynthesis in Chlorella in subsaturating light. This seemed to point to a lower maximum quantum yield of the former reaction ; Clendenning and Ehrmantraut (1951) found, in fact, about a 70% lower efficiency in the weakest light used. However, the percentual difference between the rates of the two reactions seems to have a maximum in the region of half-saturation, and to decline in stronger and in weaker light. Ehrmantraut and Rabino witch (1952) found that this difference disappears in both limiting cases — that of light saturation as well as that of light limi- tation. Their quantum yield measurements, carried out partly with filtered neon light and a Warburg-Schocken actinometer and partly with a mono- chromator and bolometer, were summarized on pages 1130-1131. The second method, which is more precise, gave an average quantum efficiency of 7 = 10 ± 1, practically identical with that found in parallel experiments for the quantum efficiency of the photosynthesis of the same cells in carbon- ate buffer No. 9. The actinometric data were slightly, but not significantly higher (cf. table 29.XI). These results, together with those obtained with chloroplast suspensions, support strongly the hypothesis that the primary photochemical process is the same in photosynthesis and Hill reaction and requires the same number of quanta for the transfer of one hydrogen atom, despite the fact that much less energy is needed to transfer it to a typical Hill oxidant, than to carbon dioxide. The relation between the light curves of photosynthesis and those of the Hill reaction in Chlorella — convergence in two limits, divergence in the mid- dle— is peculiar, and cannot be explained by a simple kinetic model. It is probably related to the "self-poisoning" of the Hill reaction by quinone, which may leave the saturation rate unchanged, and the effect of which on the rate in subsaturating light may depend on the duration and intensity of illumination (so that in measurements in weak light, self-poisoning has no time to develop itself). Yield in Flashing Light. The experiments on the oxygen liberation by Chlorella cells in quinone in flashing light already were described in chapter 34 (section B7). As illustrated by figure 34.26 (Clendenning and Ehrman- traut 1951), the dark intervals needed to obtain the maximum yield per flash (with "instantaneous" discharge flashes) were found to be the same for quinone reduction and for photosynthesis of the same cells in bicarbonate. Experiments with increased flash energy by Ehrmantraut and Rabino- witch (1952) indicated (figure 34.27) that the maximum yield per flash also is the same for both reactions, although it requires a higher flash energy in the case of quinone as oxidant (the same relation was noted also in continuous light, cf. figure 35.26). The significance of these findings — PHOTOCHEMISTRY OF LIVE CELLS 1621 which seem to support the concept of a common rate-Hmiting enzymatic process in Hill reaction and photosynthesis — was discussed in chapter 34C. Induction. Under the conditions when photosjaithesis in continuous as well as in flashing light had the often observed induction period of about 5 min., fully developed after about 5 min. of darkness (c/. chapter 33), oxygen liberation with cjuinone was found to begin instantaneously. This finding of Clendenning and Ehrmantraut (1951) already was described in chapter 33 (section A8) and its possible implications for the theory of the induction phenomena were discussed there (section C4). The same difference between induction in photosynthesis and quinone reduction by Chlorella was found also in flashing light. Inhibition. The Hill reaction with quinone in whole Chlorella cells is not inhibited by cyanide (0.0005 mole/liter) — a result that agrees with pre- viously reported observations on chloroplast preparations (section B5(d) above) , but which is nevertheless important because it proves that the Hill reaction in live cells does not proceed via the formation of free carbon di- oxide, followed by normal photosynthesis. With Hill's mixture, oxygen evolution by Chlorella cells in dark was inhibited by cyanide as effectively as photosynthesis — a confirmation of the suggestion (cf. below) that this evolution results from photosynthesis at the cost of carbon dioxide produced by photolysis of oxalate. The Hill reaction in live cells was found by Clendenning and Ehrman- traut to be inhibited by hydroxylamine (3 X 10"*^ gave 100% inhibition), fluoride (0.02 M, 65% inhibition), ethylurethan (1% solution gave 33% inhibition), iodoacetate (0.02 M, 100% inhibition) and — rather unexpect- edly— malonaie (35% inhibition at 6 X 10 ~^ M). No chloride effect was noted by comparing the Hill reaction in Chlorella cells grown with or with- out potassium chloride, or by adding chloride to the reaction mixture. Preilluminaiion Effect. The quinone reduction by Chlorella is inhibited by preillumination of cells in the absence of oxidant (Fig. 35.27). The inhibition occurs in red light as well as in light of shorter wave lengths, indicating that it is a chlorophyll-sensitized process. (It is not a direct photochemical transformation of chlorophyll — or, at least, does not reveal itself as such by a change in color.) The preillumination effect reminds one of a similar phenomenon ob- served by Warburg and Liittgens (1946) and Arnon and Whatley (1949) with isolated chloroplast preparations. (We recall, however, that no in- hibition by preillumination of chloroplasts was found by Holt and French, 1948.) It also makes one think of the observation reported in chapter 19 (Vol. I) that photoxidation can be induced in cells by illumination in the absence of carbon dioxide. However, inhibition by preillumination was observed by Clendenning and Ehrmantraut in nitrogen as well as in air. Clendenning and Ehrmantraut noted that preillumination of Chlorella, which reduced the rate of quinone reduction by 50%, did not affect the 1622 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 20 40 60 80 PHOTOCHEMICAL REACTION TIME, minutes 100 Fig. 35.27. Effect of preilluminating Chlorella cells and quinone on subsequent photochemical oxygen production by Chlorella (after Clendenning and Ehrman- traut 1951). total yield of oxygen liberated from a given amount of quinone. It would be interesting to laiow whether inhibition by preillumination affects equally the maximum rate of the Hill reaction in strong light and its quantum efficiency on weak light ; no such observations have been made The Hill reaction with quinone also is inhibited by the preillumination of quinone without the cells (usmg the blue-violet light absorbed by qui- none) . This illumination causes a darkening of the quinone solution, and apparently produces a strongly poisonous substance that inhibits the Hill reaction. The two preillumination effects, one due to the deactivation of cells, and the other caused by photochemical decomposition of quinone, are independent and additive. When the Hill reaction is measured with quinone as oxidant, in light from which blue-violet rays have not been excluded, one must e.xpect that the decomposition of quinone and the production of the poison will pro- ceed simultaneously with the main reaction (unless all quinone added is taken up by the cells and bound there in such a way as to be protected from photodecomposition). Available evidence does not show to what ex- PHOTOCHEMISTRY OF LIVE CELLS 1623 tent and how rapidly quinone — or any of the other oxidants used in the study of Hill reaction — are taken up by cells (or by colloidal chloroplast dispersions). Perhaps the drop in the rate and in total yield of the Hill reaction, which occurred when the amount of quinone was increased above 3 X 10 ~^ mole/liter, could be due to the incapacity of the cells to take up so much quinone and protect it from photochemical decomposition. Quinone is not only a self-inhibitor of the Hill reaction but it also in- hibits photosynthesis and respiration. Chlorella cells which have once been used for the study of Hill reaction are permanently incapable of photo- synthesis (at least in carbonate buffer — no experiments were made in phosphate buffer). Exposure to quinone (1 hr. in 0.08% solution) in dark- ness also destroys the capacity for photosynthesis. Respiration, too, is strongly but not completely inhibited by incubation with quinone. It thus seems that despite unchanged appearance the term "intact cells" should be used with caution in application to cells which have been exposed to quinone. The Hill reaction in Chlorella cells with quinone as oxidant also is in- hibited by ultraviolet irradiation (X = 253.6 m//). According to Holt, Brooks and Arnold (1951) the kinetics of this inactivation is similar to that observed in the study of photosynthesis or of the Hill reaction in chloroplast preparations. The absolute rate of inactivation also seems to be the same as in the case of photos3aithesis. Different Algae. The Hill reaction, with quinone or ferricyanide-ferric oxalate mixture as oxidant, can be observed also with blue-green algae Cylindrospermum (Clendenning and Ehrmantraut, unpublished). The yield appeared to be considerably lower than with Chlorella. No experi- ments have as yet been reported on the possibility of carrying out the Hill reaction with live higher plants. Different Oxidants. From the oxidants tested so far, o-benzoquinone appears to be the only effective one for whole cells; yet its photochem- ical instability makes its use desirable only in light with X > 500 m^t, and its poisonous properties preclude its use in any but very low concentrations. Because of these drawbacks, it would be useful to find another oxidant capable of penetrating into live cells which would be more stable and less poisonous. Clendenning and Ehrmantraut tried some of the dyestuffs that gave good results with chloroplast suspensions, such as phenol indo- phenol, but obtained no oxygen liberation with whole Chlorella cells; per- haps these dyes could not penetrate through the cell membrane. No posi- tive results were obtained also with chr ornate. Clendenning and Ehrmantraut (1951) reported oxygen evolution by Chlorella cells from Hill's mixture in light; but Ehrmantraut and Rabino- witch (1952) found that ferric oxalate underwent direct photolysis in cell- free solution (in the white light used by Clendenning and Ehrmantraut) at about the rate at which oxygen is liberated by cells in Hill's mixture. This indicates that the latter reaction is, photosynthesis utilizing the carbon 1624 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 20 18 34 30 26 - TIP i 34 • •,32 °OoOOo Light _L 30 60 MINUTES 90 120 Fig. 35.28. Isotopic tracer study of Hill reaction with quinone and oxygen in Chlorella cells (Brown 1953). 1 mg. quinone tipped into cell suspension in phosphate buffer as shown. Respiration poisoned at once. Only 02(32) evolved in light ( 18 klux) until 87% of quinone reduced; 02(34) taken up and 02(32) evolved afterwards in light as expected for Hill reaction with Oo as oxidant. dioxide produced by the decomposition of oxalate. This surmise is con- firmed by the absence of oxygen liberation in red light, and by the inhibi- tion of oxygen liberation by cyanide. Ehrmantraut and Rabinowitch (1952) also could not confirm the libera- tion of oxygen by Chlorella in the presence of benzaldehyde; the findings of Fan, Stauffer and Umbreit, mentioned at the beginning of Part C, must therefore be considered as in need of renewed study. Isotopic oxygen tracer experiments by Brown (1953) seem to indicate that oxygen can act as Hill oxidant in Chlorella cells— in other words, these cells function as photocatalysts for the exchange of 0^* between O2 and H2O (cf. section B4 (c)). In the presence of quinone (c/. figure 35.28) this isotopic equilibration in light does not begin until practically all quinone has been reduced (similar observations were described in part B for chloro- plast preparations) . In the absence of quinone, one has the problem of dis- tinguishing between (light-stimulated) respiration and the "Mehler reac- tion," as possible mechanisms of isotopic equilibration. The best evidence is obtained with species such as Anahaena, or Scenedesmus, in which respira- tion can be poisoned by cyanide without affecting the isotopic exchange in PHOTOCHEMISTRY OF LIVE CELLS 1625 500 450 400 350 - 300 - 250 200 150 60 MINUTES Fig. 35.29. Isotopic tracer evidence of closed-circle Hill reaction in live Anahaena cells in carbonate buffer No. 11. Respiration (1.73 mm.'/miu.) and CO2 reduction poisoned by cyanide (0.01 mole/1.). The only effect of light (6.5, 1.5, 26 klux) is 1:1 isotopic exchange of O between O2 and H2O (after Brown 1953). light (whose rate remains considerably higher than that of unpoisoned dark respiration, cf. figure 35.29). It seems most likely that the photochemical exchange mechanism is in these cells — and by inference, in other species as well — the one suggested by Mehler for chloroplasts: a Hill reaction leading to hydrogen transfer from H2O to O2, converting the latter to H2O2, followed by the catalatic decomposition of the peroxide. Horwitz (1954^) found that substitution of deuterium oxide for ordinary water reduces the rate of quinone reduction by Chlorella at all hght intensi- ties, and not only in the light-saturated state, as is the case with photosyn- thesis {c,J. p. 296). Bibliography to Chapter 35 A. Photochemistry of Chlorophyll Solutions 1943 Pepkowitz, L. P., J. Biol Chem., 149, 465. Baur, E., and Niggli, F., Heh. Chim. Acta, 26, 251, 994. 1626 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 1947 Livingston, R., Sickle, D. and Uchiyama, A., J. Phys. & Coll. Chem., 51, 775. 1948 McBrady, J. J., and Livingston, R., ihid., 52, 662. Livingston, R., and Pariser, R. L., /. Am. Chem. Soc, 70, 1510. Krasnovsky, A. A., Compt. rend. (DokJady) Acad. Sci. U.S.S.R., 60, 421. Krasnovsky, A. A., ibid., 61, 91. Gurevich, A. A., ibid., 59, 937. Gurevich, A. A., ibid., 64, 369. Krasnovsky, A. A., and Brin, G. P., Compt. rend. (Doklady) Acad. Sci. U.S.S.R., 63, 163. Calvin, M., and Dorough, G., /. Ain. Chem. Soc, 70, 699. 1949 Warburg, 0., and Schocken, V., Arch. Biochem., 21, 363. Livingston, R., "Photochemistry of Chlorophyll" in Photosynthesis in Plants, Iowa State College Press, Ames, Iowa, 1949. Krasnovsky, A. A., and Voynovskaya, K. K., Compt. rend. (Doklady) Acad. Sci., U.S.S.R., 66, 663. Krasnovsky, A. A., and Brin, G. P., ibid., 67, 325. Krasnovsky, A. A., Brin, G. P., and Voynovskaya, K. K., ibid., 69, 393. Huennekens, F. M., and Calvin, M., J. Am. Chem. Soc, 71, 4024, 4031. 1950 Knight, J., and Livingston, R., /. Phys. & Coll. Chem., 54, 703. Pariser, R., Thesis, Univ. of Minnesota. Linstead, R. P., Braude, E. A., and Timmons, C. J., Nature, 166, 557. Krasnovsky, A. A., and Brin, G. P., Compt. rend. {Doklady) Acad. Sci. U.S.S.R., 73, 1239. Ashkinazi, M. S., Glikman, T. S., and Dain, B. J., ibid., 73, 743. Evstigneev, V. B., Gavrilova, V. A., and Krasnovsky, A. A., ibid., 74, 315. Evstigneev, V. B., and Gavrilova, V. A., ibid., 74, 781. 1951 Krasnovsky, A. A., and Vo^'novskaya, K. K., ibid., 81, 879. Evstigneev, V. B., and Terenin, A. N., ibid., 81, 223. Krasnovsky, A. A., and Gavrilova, V. A., ibid., 81, 1105. Kachan, A. A., and Dain, B. J., ibid., 80, 619. Burk, D., and Warburg, O., Z. Naturjorsch., 6b, 12. 1952 Weigl, J. W., and Livingston, R., J. Am. Chem. Soc, 74, 4160. Uri, N., ibid., 74, 5808. Weigl, J. W., and Livingston, R., ibid., 74, 4211. Watson, W. F., Trans. Faraday Soc, 48, 526. Linschitz, H., and Rennert, J., Nature, 169, 193. Krasnovsky, A. A., and Voynovskaya, K. K., Compt. rend. (Doklady) Acad. Sci. U.S.S.R., 87, 109. Krasnovsky, A. A., Evstigneev, V. B., Brin, G. P., and Gavrilova, V. A., ibid., 82, 947. Holt, A. S., Paper at Gatlinburg Conference on Photosynthesis, October, 1952. 1953 Krasnovsky, A. A., and Brin, G. P., Compt. rend. (Doklady) Acad. Sci. U.S.S.R., 89, 527. Evstigneev, V. B., and Gavrilova, V. A., ibid., 89, 523. Evstigneev, V. B., and Gavrilova, V. A., ibid., 91, 899. BIBLIOGRAPHY TO CHAPTER 35 1627 Warburg, 0., Krippahl, G., Buchholz, W., and Schroder, W., Z. Natur- forsch., 8b, 675. Schenk, G. 0., Personal communication. Livingston, R., and Ryan, V. A., /. Am. Chem. Soc, 75, 2176. Evstigneev, V. B., and Gavrilova, V. A., Compt. rend. (Doklady) Acad. Sci. U.S.S.R., 92, 3S1. 1954 Livingston, R., Porter, G., and Windsor, M., Nature, 173, 485. B. Photochemistry of Chloroplasts 1939 Hill, R., Proc. Roxj. Soc, B127, 192. 1940 Hill, R., and Scarisbrick, R., Nature, 146, 61; Proc. Roy. Soc. London, B129, 238. 1942 French, C. S., Newcomb, E., and Anson, M. L., Am. J. Botany, 29, 85. 1943 Fan, C. S., Stauffer, J. F., and Umbreit, W. W., J. Gen. Physiol, 27, 15. Boichenko, E. A., Compt. rend. {Doklady) Acad. Sci. U.S.S.R., 38, 181. 1944 Boichenko, E. A., ibid., 42, 345. Warburg, 0., and Liittgens, W., Naturwissenschaften, 32, 161, 301. 1945 French, C. S., and Rabideau, G. S., J. Gen. Physiol, 28, 329. Franck, J., Rev. Modern Phys., 17, 112. Kumm, J., and French, C. S., Am. J. Botany, 32, 291. 1946 Holt, A. S., and French, C. S., Arch. Biochem., 9, 25. Warburg, 0., and Liittgens, W., Biokhimiya {Russ.), 11, 303; reprinted in Schwermetalle als Wirkungsgruppen von Fermenien, Berlin, 1946, and Heavy Metal Prosthetic Groups and Enzyme Action, Clarendon Press, Oxford, 1949. French, C. S., Holt, A. S., Powell, R. D., and Anson, M. L., Science, 103, 505. Aronoff, S., Plant Physiol, 21, 393. French, C. S., and Holt, A. S., Am. J. Botany, 33, 19a. 1947 Boichenko, E. A., Biokhimiya (Russ.), 12, 196. Gurevich, A. A., Compt. rend. (Doklady) Acad. Sci. U.S.S.R., 55, 263. Holt, A. S., and Macdowall, F. D., A.A.A.S. Symposium, Chicago, Dec. 1947. 1948 Brown, A. H., and Franck, J., Arch. Biochem., 16, 55. Holt, A. S., and French, C. S., Arch. Biocliem., 19, 368. Holt, A. S., and French, C. S., ihid., 19, 429. Boyle, F. P., Science, 108, 359. Boichenko, E. A., Biokhimiya (Russ.), 13, 219. Andreeva, T. F., and Zubkovich, L. E., Compt. rend. (Doklady) Acad. Sci. U.S.S.R., 60, 681. Van Norman, R. W., French, C. S., and Macdowall, F. D, H., Plant Physiol, 23, 455. 1949 Holt, A. S., and French, C. S., "The Photochemical Liberation of Oxygen from Water by Isolated Chloroplasts" in Photosynthesis in Plants, Iowa State College Press, Ames, Iowa, 1949. Zubkovich, L. E., and Andreeva, T. F., Compt., rend. (Doklady) Acad. Sci. U.S.S.R., 67, 165. Davenport, H, E., Proc. Roy Soc. (London) B136, 281. 1628 PHOTOCHEMISTRY OF CHLOROPHYLL CHAP. 35 Macdowall, F. D. H., Plant Physiol, 24, 462. Arnon, D. I., Plant Physiol., 24, 1. Arnon, D. I., and Whatley, F. R., Arch. Biochem., 23, 141, Arnon, D. I., and Whatley, F. R., Science, 110, 554. Arnon, D. I., and Whatley, F. R., Am. Assoc. Plant Physiol., New York Meeting, Dec. 1949. 1950 Arnold, W., and Oppenheimer, J. R., J. Gen. Physiol., 33, 423. Clendenning, K. A., and Gorham, P. R., Can. J. Research, C28, 78. Clendenning, K. A., and Gorham, P. R., ibid., 28, 102. Clendenning, K. A., and Gorham, P. R., ibid., 28, 114. Gorham, P. R., and Clendenning, K. A., ibid., 28, 513. Spikes, J. D., Lumry, R., Eyring, H., and Wayrynen, R. E., Proc. Nat. Acad. Sci., U.S., 36, 455. Spikes, J. D., Lumry, R., Eyring, H., and Wayrynen, R. E., Arch. Bio- chem., 28, 48. Milner, H. W., French, C. S., Koenig, M. L. G., and Lawrence, N. S., ibid., 28, 193. Milner, H. W., Koenig, M. L. G., and Lawrence, N. S., ibid., 28, 185. Milner, H. W., Lawrence, N. S., and French, C. S., Science, 111, 633. Gerretsen, F. C, Plant and Soil, 2, 159. Gerretsen, F. C, ibid., 2, 323. Ochoa, S., Veiga Solles, J. B., and Ortiz, P. J., J. Biol. Chem., 187, 863. 1951 Vereshchinsky, L V., Biokhimiya, 16, 350. Holt, A. S., Smith, R. F., and French, C. S., Plant Physiol, 26, 164. French, C. S., and Milner, H. W., Symposia Soc. Exptl Biol, 5, 232. Hill, R., ibid., 5, 222. Gerretsen, F. C, Plant and Soil, 3, 1. Arnon, D. L, Nature, 167, 1008. Mehler, A. H., Arch. Biochem. and Biophys., 33, 65. Mehler, A. H., ibid., 34, 339. Vishniac, W., and Ochoa, S., Nature, 167, 768. Tolmach, L. J., ibid., 167, 946. Tolmach, L. J., Arch. Biochem. and Biophys., 33, 120. Vishniac, W., Federation Proc, 10, 265. Holt, A. S., Brooks, I. A., and Arnold, W. A., J. Gen. Physiol, 34, 627. Vinogradov, A. P., Boichenko, E. A. and- Baranov, V. L, Compt. rend. (Doklady) Acad. Sci., U.S.SR., 78, 327. Clendenning, K. A., and Ehrmantraut, H. C, Arch. Biochem., 29, 387. 1952 Ehrmantraut, H. C, and Rabinowitch, E., Arch. Biochem. and Biophys., 38, 67. Smith, J. H. C, French, C. S., and Koski, V. M., Plant Physiol, 27, 212. Davis, E. A., Am. J. Botany, 39, 535. Bishop, N. I., Thesis, Univ. of Utah. Ochoa, S., and Vishniac, W., Science, 115, 297. Vishniac, W., and Ochoa, S., J. Biol. Chem., 195, 75. Davenport, H. E., Hill, R., and Whatley, F. R., Proc. Roy. Soc. Londo7i, B139, 346. Davenport, H. E., and Hill, R., ibid., B139, 327. BIBLIOGRAPHY TO CHAPTER 35 1629 Wesscls, J. S. C, and Havinga, E., Rec. trav. chim., 71, 809. Krasnovsky, A. A., and Kosobutskaj^a, L. M., Compt. rend. (Doklady) Acad. Sci., U.S.S.R., 82, 761. Warburg, O., Z. Naturforsch., 7b, 44.3. Fager, E. W., Arch. Biocheyn. and Biophys., 37, 5. Gorham, P. R., and Clendenning, K. A., ibid., 37, 199. Mehler, A. H., and Brown, A. H., ibid., 38, 365. Macdowall, F. D. H., Science, 116, 398. Rabinowitch, E., Ann. Rev. Plant Physiol., 3, 229. Uri, N., J. Am. Chem. Soc, 74, 5808. McClendon, J. H., and Blinks, L. R., Nature, 170, 577. 1953 Thomas, J. B., Blaauw, 0. H., and Duysens, L. N. M., Biochem. et Bio- phys. Acta, 10, 230. Wessels, J. S. C, and Havinga, E., Rec. trav. chim., 72, 1076. Gilmour, H. S. A., Lumry, R., Spikes, J. D., and Eyring, H., AEG Tech- nical Report No. 11, Contract No. AT (ll-l)-82. Brown, A. H., Personal communication. Boichenko, E. A., and Baranov, V. I., Compt. rend. (Doklady) Acad. Sci. U.S.S.R., 91, 339. Hill, R., Northcote, D. H., and Davenport, H. E., Nature, 172, 948. Punnett, T., and Fabiyi, A., ibid., 172, 947. 1954 Spikes, J. D., Lumry, R., Rieske, J. S., and Marcus, R. J., Plant Physiol, 29, 161. Horwitz, L., ibid., 29, 215. Punnett, T., Thesis, Univ. of Illinois. Tolbert, N. E., and Zill, I. P., /. Gen. Physiol., 37, 575. Arnon, D. I., Allen, M. B., and Whatley, F. R., Nature, 174, 39. Arnon, D. I., Whatley, F. R., and Allen, M. B., J. Am. Chem. Soc, 76, 6324. McClendon, J. H., Plant Physiol, 29, 448. Horwitz, L., Bull Math. Biophys., 16, 45. Wessels, J. S. C, Thesis, Univ. Leyden. C. Photochemistry of Cells 1943 Fan, C. S., Stauffer, J. F., and Umbreit, W. W., /. Gen. Physiol, 27, 15. 1946 Warburg, 0., and Liittgens, W., Biokhimiya (Russ.), 11, 303; Schwer- mctalle als Wirkungsgruppen von Fermenten, Berlin. 1946; Heavy Metal Prosthetic Groups and Enzxjme Action, Clarendon Press, Oxford, 1949. Aronoff, S., Plant Physiol, 21, 393. 1951 Clendenning, K. A., and Ehrmantraut, H. C., Arch. Biochem., 29, 387. Holt, A. S., Arnold, W., and Brooks, I. A., J. Gen. Physiol, 34, 627. 1952 Ehrmantraut, H. C, and Rabinowitch, E., Arch. Biochem. and Biophys., 38, 67. 1953 Brown, A. H., Personal communication. 1954 Clendenning, K. A., Personal communication. Chapter 36 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION* (ADDENDA TO CHAPTERS 3, 8 AND 9) A. IsoTOPic Carbon Tracer Studies* Abbreviations used in this chapter: GA, glyceric acid; PGA, phosphoglyceric acid; FA, pyruvic acid; PPA, phosphopyruvic acid; MA, maUc acid; and OOA, oxalacetic acid. TP, triose phosphate; DHAP, dihydroxyacetone phosphate; GMP, glucose mono- phosphate; FMP, fructose monophosphate; FDP, fructose diphosphate; RDP, ribu- lose diphosphate; and SMP, sedoheptulose monophosphate. Also DPN and TPN, diphospho- and triphosphopyridine nucleotides (coenzymes I and II, hydrogen carriers); coA, coenzyme A ("acetyl carrier"); LA, lipoic (or thioc- tic) acid (hydrogen carrier); and ADP and ATP, adenosine diphosphate and triphos- phate (high-energy phosphate acceptor and donor, respectively). The experiments and speculations described in this chapter are con- cerned with the primary fixation of carbon dioxide in photosynthesis and its subsequent reduction — topics that have been discussed in chapter 8 (sections 3 and 4) and chapter 9 (section 10) of Vol. I, respectively. Also included are observations that throw some light on the sequence in which different sugars, amino acids and other synthesized products appear in photosynthesis — a matter discussed in chapter 3. Some data will have to be included in this chapter concerning the effect of poisons on photosynthesis — a subject dealt with in chapter 12, and again in chapter 37D (section 2a). 1. Photosynthetic and Respiratory Fixation of C*02 It was hoped at first that the application of tracer carbon would rapidly lead to clarification of the hitherto mysterious chemical mechanism of car- bon dioxide reduction. The task proved to be more difficult than was an- ticipated for two reasons. In the first place, it has become apparent that, whatever the intermediates and primary products of carbon dioxide trans- formation in photosynthesis may be, they are very rapidly changed into a variety of other compounds; or, at least, the tracer carbon incorporated in them is rapidly redistributed through a multitude of metabolites. Within a few minutes, C(14) makes its appearance in compounds of differ- * Bibliography, page 1710. 1630 PHOTOSYNTHETIC AND RESPIRATORY FIXATION OF C*02 1631 ent types, including sugars, proteins and even fats. It seems that this redistribution of assimilated radiocarbon begins even before carbon dioxide has been reduced to the carbohydrate level, by side reactions of intermediate reduction products (such as amination of C3 or C4 acids) , or their involve- ment in respiratory reactions. Aronoff, Benson, Hassid and Calvin (1947) investigated the active constituents of barley seedlings exposed to light for one hour in C(14)-tagged carbon dioxide. Radiocarbon was found in all fractions, with the largest accumulation in sugars (Table 36.1). Table 36.1 Tracer Distribution in Barley Seedlings after One Hour of Photosynthesis IN Tagged Carbon Dioxide (after Aronoff, Benson, et al., 1947) Per cent of total C * Plants Plants with without Fraction roots roots Ether-extractable acids (fatty acids) 12 . 5 2.9 Lipids (including pigments) 6.1" 3.6* Amino acids 7.2 5.6 Acids 11.3 7.7 Sugars 25.8 35.0 Soluble proteins 0.7 6.9 Other water-soluble components 3.5 — Cellulose 2.8 3.6 Lignin, etc 8.3 9.4 Unaccounted 18.8 20.3 " 0.9% in carotenoids. " 0.8% in chlorins, 0.9% in phytol. Aronoff, Barker and Calvin (1947) subjected the sugar fraction in table 36.1 to fermentation (by yeast, or by Lactobacillus casei), and from the activity of the various products determined the relative amount of C* m the different positions in the C chain. The results are summarized in table 36.11; they indicate that even after a full hour of photosynthesis, the tagging of hexoses was still far from uniform, and C* was preferen- tially found in positions 3 and 4. Table 36.11 Tracer Distribution in Sugars Photosynthesized in One Hour by Barley Seedlings (Aronoff, Barker and Calvin, 1947) Material Product Fermented by C* in positions 1,6 2,5 3,4 Plants with roots Plants without roots. . Hydrolyzed sugar Yeast Crystalline glucose L. casei Nonhydrolyzed sugar " " Crystalline glucose " " 1632 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 Similar results were obtained by Calvin and Benson (1948). After one hour of photosynthesis, tagged glucose contained 61% of its C* in positions 3 and 4, 24% in posi- tions 2 and 5, and 15% in positions 1 and 6; after 2 hours of photosynthesis, the distri- bution was 37, 36, 27%— obviously approaching equilibrium. A similar spread of C(14) from position 1 (carboxyl) to positions 2 and 3 could be observed in alanine. Gibbs (1949) reported results quite different from those of Calvin and Benson. After 1 hour photosynthesis in C(14)-labelled CO2, he found glucose to be labelled prefer- entially in positions 1 and 6. All these early results appear to be in some contradiction to the more recent deg- radation experiments in Berkeley, where all preferential tagging (up to 90% C* in glu- cose in positions 3 and 4 after 30 seconds exposure, cf. table 36. VII) was found to yield to practically uniform tagging in a matter of minutes, not hours. In the first Russian experiments with C(14) as tracer, reported by Nez- govorova (1951, 1952^, long exposures to C*02 in light (of the order of one hour) were used, and the tracer was found largely in proteins. Doman, Kuzin, Mamul and Khudjakova (1952) went over to exposures of 1-2 seconds, following them by 0-300 seconds of photosynthesis in nontagged carbon dioxide. They used young leaves of 17 species, and obtained, by a ionophoretic method of fractionation, a large variety of results. For ex- ample, Phaseolus was found to fix C*, after 1 second exposure, only in anionic substances; with other plants, it was found also in cathionic sub- stances, with still others, in neutral substances. Nezgovorova (1952^) observed the dark uptake of C*02 by leaves, and found no ef- fect of preillumination. The insoluble tagged compounds synthesized by barley in 5-minutes exposure to C*02 are 95% polysaccharides; in Scenedesmus, they are 50% polysaccharides and 50% proteins, the latter containing mainly tagged alanine and aspartic acid (Calvin et al. 1951). This result may have some bearing on the observations of Smith et al. (Vol. I, page 36) that the product of photosynthesis in sunflower is almost 100% carbohydrates, and on the failure of other observers to obtain similar simple results with other plants, particularly with algae. The second complication is due to the fact that, as we now know, there occurs (in dark as well as in light), in addition to C* assimilation by photo- synthesis, also a C* uptake by exchange of C*02 with the carboxyl groups of certain respiration (and fermentation) intermediates. We mentioned in Vol. I (page 208) that, since 1936, many observations had revealed the capacity of bacteria (and of certain animal tissues) to assimilate carbon dioxide in the dark. This fixation often can be attributed to the reversibility of certain respiratory decarboxylations, such as: (36.1) COOHCH2COCOOH (oxalacetate) ;^=i CH3COCOOH + CO2 (pyruvate + carbon dioxide) PHOTOS YNTHETIC AND RESPIRATORY FIXATION OF C*02 1633 or of decarboxylations coupled with oxidation-reductions, such as (36.2) CH3COCOOH + A + H2O (pyruvate + oxidant) . CH3COOH + CO2 + AH2 (acetate + carbon dioxide + reductant) The oxidant in (36.2) is, in vivo, phosphopyridine nucleotide. Some irreversible reactions of the latter type become reversible if ac- companied by conversion of "low energy" into "high energy" phosphate. (For this, the oxidant and the reductant must be in the form of phosphate esters, and the ester of the reductant must release less energy on hydrolysis than that of the oxidant; this is the case, e. g., when the reductant contains a carbonyl and the oxidant a carboxyl group.) More recently, it has been found that the capacity for taking up radio- active carbon from carbon dioxide and distributing the tagged carbon atoms through a variety of metabolites is common to many animal and plant cells. It may perhaps be considered an inevitable concomitant of cellular respiration (and fermentation), since both processes involve revers- ible decarboxylations such as (36.1) or (36.2) and associated reactions, which, too, are either reversible in themselves, or can be reversed by cou- pling with degradation or high-energy phosphates. More specifically, all partial processes of glycolysis leading from sucrose to pyruvic acid, as well as the "Krebs cycle," by which pyruvic acid is decarboxylated and dehy- drogenated, are of this character. Therefore, what we observe as the net production of carbon dioxide in respiration (or fermentation) is the excess of reactions running in the direction of chain fission (and hydrogen trans- fer to oxidized pyridine nucleotides) over reactions by which the carbon chain is built up (and hydrogenated with hydrogen supplied by reduced pyridine nucleotides). The presence of C*02 in the atmosphere in which respiration or fermentation takes place must then lead to some C* atoms "creeping back," first into the intermediates of the decarboxylation cycle, and thence into those of glycolysis. Simultaneously, the C* atoms may also penetrate into compounds (such as certain amino acids), whose respira- tory breakdown is coupled with that of carbohydrates. Because of the existence of a C* uptake from C*02 which is unrelated to photosynthesis, it is not permissible to conclude that a certain compound is an intei-mediate (or an early product) of photosynthesis merely because it appears C* tagged soon after the exposure of the plant to C*02. Rather, specific evidence is needed to justify such a conclusion. In strong light, when photosynthesis far exceeds respiration, C* uptake determined by radioactive method can be compared mth the net CO2 uptake determined by manometry (or other analytical methods), and if the former does not ex- ceed markedly the latter, the hulk of the C*-tagged compounds must have been produced by photosynthesis. Even in this case, however, the tagged 1634 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 compounds which are present in relatively small amounts could be due to the reversal of respirative (or fermentative) decarboxylations, and have no relation to photosynthesis. The existence of these complications was not known when the first car- bon tracer studies of photosynthesis were made by Ruben and co-workers with the short-lived carbon isotope, C(ll). It is therefore difficult to de- cide which of their observations (described in Vol. I, pages 201 and 241) remain significant from the point of view of the mechanism of photosynthe- sis. Ruben, Kamen and co-workers found (1939, 1940) that in the dark, C"02 was incorporated, by an enzymatic reaction, into a water-soluble molecule, with a molecular weight of about 1000, forming a carboxyl group. After brief illumination, they found most of the active carbon in a similar (perhaps, the same) large molecule, but by now not, or not exclusively, in a carboxyl group. The active material was, however, still soluble in water and precipitable by barium. It was later observed, with some surprise, by Frenkel (1941) (also at Berkeley) that in the dark, the C* uptake was localized in the cytoplasm. After brief illumination, however, the radioactivity was found pre- dominantly in the chloroplasts. The amount of C(ll) taken up after several hours of exposure to C*02 in the dark was about 0.05 mole/liter of cell volume — roughlj^ equivalent to the amount of chloro- phyll present; however, the tagged product was not bound to chlorophyll. (We noted above that it could be extracted with water.) It seemed natural at that time to identify the C*02 acceptor compound, whose for- mation in the dark was indicated by these experiments, with the often postulated sub- strate of photochemical reduction (which we have variously symbolized by ICOal, ACO2 or RCOOH). However, this identification presented certain difficulties. One of them was the slowness with which C(ll) was taken up: about an hour was needed to "saturate" the cells with radiocarbon in the dark (c/. fig. 21, Vol. I); the same amount of carbon can be absorbed in about 20 seconds, either by photosynthesis in strong Hght, or by "pick-up" after intense photosynthesis in the dark (Vol. I, p. 206). Two explanations of the slowness of the observed C* uptake could be suggested. One is that this uptake is slow when it has to proceed through isotopic exchange: (36.3) C*02 + RCOOH , CO2 + RC*OOH but becomes /asi when it can occur by C*02 addition to free acceptor: (36.4) C*02 + RH , RC*OOH (However, on page 203, the "one hoxir uptake" itself was tentatively attributed to reaction with free acceptor — -in order to be able to explain the occurrence of additional CO2 uptake after evacuation.) During intense photosynthesis, in a medium which is low in carbon dioxide, the ac- ceptor may be present in the photostationary state predominantly in the decarboxylated state, RH. When illumination is stopped, CO2 rushes in, in a rapid gulp, as observed in "pick-up" experiments. In the experiments of Ruben and Kamen, on the other hand, the acceptor could have been present mainly in the carboxylated form, RCOOH, and C(ll) could enter it only by the slow exchange reaction (36.3). A second possible reason for the observed slowness of the C * uptake in the dark has become apparent after it had been established that C*02 can be taken up by mechanisms PHOTOSYNTHETIC AND RESPIRATORY FIXATION OF C*02 1635 unrelated to photosynthesis. The acceptor responsible for the slow uptake of C(ll) in the dark could be altogether different from that active in photosynthesis. Frenkel's observation that the C* uptake in the dark takes place in the cytoplasm, and not in chloroplasts, has been related on page 66 (Vol. I) to the incapacity of isolated chloroplasts to use carbon dioxide as hydrogen acceptor in light. This seems suggestive, but since the early products of carbon dioxide fixation in photosj'nthesis are water- soluble, it is possible for them to be primarily formed in the chloroplasts and yet to dif- fuse (or be eluted) into the cell sap cytoplasm fraction during the maceration and frac- tionation of the cell material. Clendenning and Gorham (1952) found that in multi- cellular algae {Nitella — the organism used by Frenkel — C/iara and Riccia), most of the (C14) taken up in brief periods of photosynthesis —down to 5 seconds— was found in the cell sap; in unicellular algae — Chlorella and Scenedesmus — a larger fraction could be lo- cated in insoluble lipoid and proteinaceous materials from the chloroplast fraction. The question of whether the enzymatic transformations of carbon dioxide in photosynthesis occur inside the chloroplasts, or outside these bodies (in which case they must be medi- ated by a reductant produced photochemically inside the chloroplasts and diffusing out of them), has not been definitely answered by these experiments. The observation of Arnon et. al. (1954), that intact chloroplasts take up C* in light, was described on pp. 1537 and 1615. Ruben and Kamen's observation that C( 11)02 is taken up by carbo.xylation, with the formation of a water-soluble carboxylic acid, has been confirmed by all the subse- quent investigations with C(14); it is true for photosynthetic as well as for respiratory C* fixation. On the other hand, the conclusion that this acid has a molecular weight of about 1000 appears to be inaccurate for both types of fixation, since the first product of photochemical fixation seems to be phosphoglyceric acid (molecular weight 187), while respiratory fixation is likely to lead to products such as pyruvic or oxalacetic acid, which, too, have low molecular weights. The possibility of "respiratory" C* fixation was not fully taken into ac- count also in the early stages of work with the long-lived isotope C(14) by Benson, Calvin and co-workers at Berkeley (1947) . It was first emphasized by Allen, Gest and Kamen (1947), and by Brown, Fager and Gaffron (1948), and subsequently recognized also by the Berkeley group. Allen, Gest and Kamen (1947) attempted to distinguish between the two types of C* absorption by recalling an early observation of Ruben, Kamen and Hassid with C(ll) (Vol. I, page 242). The latter had found that in Chlorella the cyanide inhibition of the C* uptake in the dark resembles that of photosynthesis, rather than that of respira- tion (for example 10"^ M KCN reduced the rate of both photosynthesis and dark C* fixation to 0.3% of the normal value, but left respiration almost unchanged). Allen, Gest and Kamen repeated these experiments, using C(14), with both Chlorella and Scenedesmus. (In the second species, respiration is more sensitive to cyanide than photosynthesis, cf. Vol. 1, page 305.) They found that, by the criterion of cyanide sen- sitivity, dark C* fixation in Scenedesmus appears to be related to photosynthesis rather than to respiration, while in Chlorella (in disagreement with earlier observation) the effect of cyanide on dark C* fixation was intermediate between those on photosynthesis and respiration. With Scenedesmus cells starved for 24 hours in darkness, the results were erratic, but with similarly starved Chlorella cells, the effect of cyanide on dark C* fixation was definitely similar to its effect on respiration and quite different from that on photosynthesis. Kamen concluded that one part of dark C* fixation is "photosynthetic" and another 1636 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 part is "respiratory" in origin, and that the latter part is particularly large in starved Chlorclla cells. Subsequently, CaKdn and co-workers achieved substantial progress by distinguishing clearly between three types of C*02 uptake; in the dark without preillumination, in the dark immediately after preillumination, and in light. (A further distinction is indicated by the work of Gaffron, Fager et al. — between preillumination in the absence of CO2, and that in the presence of CO2.) Most revealing are the results obtained by first illuminating cells in ordinary carbon dioxide with strong light long enough to establish a steady state of photosynthesis, then rapidly supplying them with tagged C*02, and killing them as quickly as possible after a very brief continuation of the light exposure. In some experiments of Calvin and co-workers, photo- synthesis in tagged C*02 lasted only 0.4 second. The results obtained at Berkeley in this study — which is being con- tinued— have been presented, between 1948 and 1954, in about twenty papers under the general title "Path of Carbon Dioxide in Photosynthesis," and in several reviews, e. g., by Benson and Calvin (1948, 1950) and Calvin (1952^-2), cf. also the reviews by Gaffron and Fager (1951), Brown and Frenkel (1953), and Holzer (1954). Brown and Frenkel paid particular attention to the question of whether all the Berkeley results were obtained in a steady state of photosynthesis, as was assumed in their interpretation. A source of uncertainty in this respect was the flushing of the vessel with air prior to the introduction of the tracer. The criterion of the steady state is the linearity of the C* uptake with time, which was observed in many, but not in all, of the Berkeley experiments. 2. C*02 Fixation in Darkness with and without Preillumination. A Surviving Reductant? Benson and Calvin (1947) began by resuming Ruben and Kamen's study of the C*02 uptake in darkness. They used two Chlorella suspen- sions; suspension I was kept in the dark for eight hours in 4% carbon di- oxide, while suspension II was strongly illuminated for one hour in CO2- free nitrogen. Both were then exposed to C*02 in the dark for 5 min. and the cells killed by acid (HCl -f CH3COOH). The purpose was to see whether preillumination of cells in the absence of carbon dioxide creates in them a chemical agent ("reducing power") able to cause subsequent up- take (which Calvin and Benson presumed must mean reduction), of carbon dioxide in the dark. The C* uptake was found to be five times greater in the preilluminated algae. The chemical distribution of radiocarbon also was different in the two samples: as much as 70% of total activity in sample I was found in succinic acid (about 85% of it in the two carboxyl C*02 FIXATION IN DARKNESS WITH AND WITHOUT PREILLUMINATION 1637 groups), 15% in amino acids, and 9% in "anionic substances" {i. e., sub- stances whose active component could be adsorbed on anion-exchange res- ins), while in (preilluminated) sample II, the largest fraction of total activ- ity (30%) was found in amino acids (mostly alanine), 25% in an unidenti- fied fraction (described as extractable from water by ether at pH 1), and 10% in anionic substances; only small amounts were found in succinic acid (6%), mahc acid (6%), and fumaric acid (1%); and 1.5% in sugars. c > O 01 w O UJ X TIME, min Fig. 36.1. Rate of dark fixation of radiocarbon by Scenedesmus. Curve A, one-day-old Scenedesmus cultures after one hour dark incubation in 4% CO2 in nitrogen. Curve B, same cells immediately after 10 min. preillumination (after Calvin and Benson 1947). Calvin and Benson (1947) gave the curves (reproduced in fig. 36.1) for the time course of the C* uptake in the dark without preliminary illumina- tion (curve A), and with 10-min. preillumination (curve B). In answer to suggestions that much if not all of the dark C*02 uptake previously reported by them, may have been due to the reversal of decarboxylations in the respiratory system (and thus bear no direct relation to photosynthesis), Calvin and Benson pointed out that figure 36.1 indicates the superposition of two processes: a slow uptake (curve A, and second part of curve B), which admittedly may be due to reversible reactions in the respiratory system, and afast uptake (initial part of curve B), which occurs only after preillumination and which (it was argued) must be related to photosyn- thesis. (This uptake could be considered as radiometric equivalent of the manometi'ically observed fast CO2 "pick-up.") 1638 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 This hypothesis is plausible; however, a period of intense illumination in the absence of external CO2 supply (the pretreatment used in these ex- periments) could lead to extensive decarboxylation not only of the carbon dioxide acceptor in photosynthesis but also of respiratory intermediates. Consequently, the fact that a "C02-gulp" was observed after such a period is not in itself convincing evidence that the C*, taken up in this gulp, enters the photosyiithetic reaction sequence. Calvin and co-workers saw an additional proof of their hypothesis in the distribution of tracer carbon among different compounds. This distribution was qualitatively similar, after the fast dark uptake by preilluminated cells, to that after a short period of photosynthesis in C*02, but quite different after the slow dark uptake without preillumination (c/. fig. 36.2). Tables 36. Ill and 36. IV show the fractionation of the tagged material obtained under different conditions. We note (table 36. Ill) thsitwithout pre- illumination a large part of total activity was in ether-soluble organic acids. (Only 5% of total C* was now found in succinic acid — as against 70% re- ported in 1947 — ^16% in malic acid, and 31% in other ether-soluble car- boxylic acids; 31% in amino acids, and only 16% in "anionic groups not extractable with ether.") In preilluminated cells, the proportion of tagged carbon present in ether-soluble fatty acids was much smaller. It decreased with the duration of preillumination, while the relative activity of amino acids and of ether- insoluble, anionically tagged substances increased. The distribution was similar to that obtained after brief photosynthesis in C*02 (table 36. IV). Tables 36. Ill and 36. IV are thus in agreement with the hypothesis Table 36.III C(14) Tracer Distribution in Chlorella pyrenoidosa'^ AFTER A Dark Fixation Time of Five Minutes (Calvin and Benson, 1947) No preillumina- tion Preillumination (total (total fixation, 10 rel. units) fixation, Fraction 1 rel. unit) 5 min. 60 min. 120 min. I. Carboxylic acids in ether ex- tract^ 52% 21% 14% 11% Including malic acid'= 16 11.5 7.4 — Including succinic acid'' 5.2 3.1 0.5 — II. Amino acids'' (adsorbed on cat- ion resin) 31 41 64 74 III. Anionic substances* (adsorbed on anion resin) 16 29 21 — IV. Sugars (nonionized compounds )-'^. . 0.45 1.0 0.96 1.0 " One-day-old cultures. '' Rapid, continuous 15-hr. extraction. "^ Separated by partition chromatography on silica gel column. '' Eluted from Duolite C-3 resin with 2.5 N HCl. " Eluted from Duolite A-3 resin with 1.5 N NaOH. / Effluate from both resins. C*02 FIXATION IN DARKNESS WITH AND WITHOUT PREILLUMINATION 1639 that the two kinetically different mechanisms of dark C* uptake lead to different tagged compounds. The slow mechanism, which is the only one operative in starved, not preilhiminated cells, produces predominantly tagged fatty acids (whether succinic, mahc, or others, is a secondary ques- tion) and also some tagged amino acids. The fast mechanism, active after preiUumination in absence of carbon dioxide, leads to large amounts of tagged anions of ether-insoluble organic acids (e. g., glyceric acid or other polyhydroxy acids which are much more hydrophilic than simple fatty acids, such as succinic, and the monohydroxy acids, such as malic acid). Table 36.1V C(14) Tracer Distribution in Chlorella and Scenedesmus after Exposure to 0*0^ IN Dark and in Light (Calvin and Benson, 1947) C* fixation in darkness C* fixation in light<= Chlorella'^ Scenedesmus b Chlorella" Scenedesmusb PreiUumination time 60 min. 10 min.-'^ — — Fixation time 1 min. 1 min. 30 sec. 30 sec. Total C" fixed (millions c.p.m.) 0.97 0.98 3.1 6.2 Proportion of C(14) in: I. Ether-extractable acids 13% 12% 2.5% 10% II. Cationic groups (amino acids). . 53% 39% 14% 11% III. A. Anionic groups, ammonia elutable'' 2.4% 4.2% 38% 44% III. B. Anionic groups, not am- monia elutable' 31% 42% 36% 27% IV. Nonionic compounds (sugars). . 0.3% 0.1% 4.5% 4.7% " Chlorella pyrenoidosa, one day old. '' Scenedesmus D-3, two day old. "= Cells, rapidly photosynthesizing, given HC*Oi' and shaken, then killed. '' Adsorbed on Duolite A-3, eluted with 1.5 A^ NH4OH. « Eluted with NaOH following ammonia elution. ^ "Time needed to ensure maximum C*02 uptake" (Calvin and Benson). These conclusions are confirmed by Figure 36.2, taken from a subsequent paper by Benson, Bassham et al. (1950). It shows the distribution of tagged compounds obtained in the dark (with and without preiUumination), and in brief photosynthesis, as revealed by paper chromatography and radioautography. We note the prevalence of C3 and C4 acids in the first radiogram (dark uptake without preiUumination), and the absence in it of tagged phosphate esters. The other two radiograms show the prevalence of phosphate esters of glyceric acid (PGA), of hexoses (HMP and HDP), and of pyruvic acid (PPA) ; they also have in common the appearance of tagged sucrose and of large amounts of tagged alanine. Malic acid and alanine contain considerable quantities of tracer carbon after all three tag- ging procedures. As mentioned above, Calvin and co-workers considered the similarity 1640 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 f -I s i IE >- a. (5 a z ♦ z z s 1 O a: z f < s o H IE < a < ^ O DC Ul 1 05 O 2 O o •z 3 o e V) A 1- ^a^i^' ttt tH ■Hh s a b. z jr ai kBEXk < ^, a: WmfP -J H< W^ < -J 3 ' §1 C*02 FIXATION IN DARKNESS WITH AND WITHOUT PREILLUMINATION 1641 O o ir z I J^ 73 &- c3 '^ T3 J3 «<-i •^ O G >> d O ■§ O r-i IC ■* ^^ W ^ v^-^ o CO CO y— N. o G ^^ ^ ;-. CJ a o 73 Q _c O HI a> rt o ^ ^ 1^ rt aj c -G .— o :^° '-13 a lo c3 -^ Ci X G "—I cd .2 ^~' '■+3 —J b G ^ •-^ Q> o S G o CO ^U 3 -a G S" o ■G cS u s-s & T3 M -u bC t3 S c3 =3 o -tJ « G .2 S 1 C o cSI o . J ^•*~' ■^ c3 -5 t-c v^ 03 «*- _: ^"^-^ o G _ •s ^ 02 £ a s c3 ^ o ^•— ' t4-. ^ o 03 d> u ■? N^ m 99% 62.1! ^■•^ ABC Fig. 36.11. C(14) fixation by Scenedesmus after 15 min. preillumination in the absence of CO2 and O2. Light-shaded areas, phosphoglyceric acid; dark-shaded areas, pyruvic acid; unshaded areas, other tagged compounds (after Gaffron et al. 1951). found in the nonsaponified part, in fatty acids (saturated as well as un- saturated), and in water-soluble products of saponification. These experiments showed that some C* enters lipoid products already in the first minute of exposure to C*02. In revision of their earlier results, Fager, Rosenberg and Gaffron (1950) 1652 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 Hi t^ o w m o V V ^ V V 92 z e u w ->C1 Si X o CM CO 1,1 trtl- u < I X I- 0. ^•9* V o m ? M o ^ 0 t- ', : a S I' O o3 o3 ^ ^ to g -O >- 4) 02 ..^ c o D< CD O — .2 -M .2 T3 a> » o .9 a a § a 50% of it in F, are located in compounds other than PGA and PA. This observation, too, remains in need of inter- pretation. Perhaps an explanation could be based on the difference be- tween cells filled with abundant intermediates of photosynthesis and cells bare of such intermediates. Observations of the type shown in fig. 36.22, revealing changes in the distribution of C* within the cells immediately after the end of illumination, are obviously relevant to the pick-up phenomena, whether observed analytically (chapters 8 and 33) or noted radiometrically. 5. Early Intermediates Other Than PGA; Paper Chromatography A most important step forward was made in 1948 by application, to the problem of carbon dioxide reduction intermediates, of a new chromato- graphic technique : paper chromatography. To combine this method (de- veloped by Consden, Gordon and Martin in England in 1944) with the radioactive tracer technique in the study of photosynthesis was first sug- gested by Fink and Fink (1947). Calvin and co-workers found this combination particularly suitable for rapid fractionation and identification 1656 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 of the numerous tagged compounds obtained in their experiments with C(14). The procedure consists in placing a drop of the cell extract (in aqueous alcohol or acid) in one corner of a sheet of filter paper, letting it 1 1 1 1 1 1 i • -c y HENYL ALANINE r J) FUMARIC - (^ ) SUCCINIC - / GLYGOLLIC - ^ ) MALIC - (^ J TYROSINE f J ISOCITRIC (^ y GLYCERIC ( ) ALANINE / j TARTARIC (^ j GLUTAMIC — ERUCTOSE (^ ^ / "^ GLYCINE (^^^ SERINE (^ ^ ASPARTIC C J GLUCOSE — P • PYRUVIC r J (^ ^SUCROSE - 1 1 ^ -^ TRIOSE-P C~^ 1 1 1 \^~^r Fig. 36.13. Scheme of paper chromatography of photosynthetic products (after Calvin 1950). The coordinates, taken from a large number of radiograms, were plotted by using serine and alanine as reference points. PGA, phosphoglyceric acids; HMP, hexose monophosphates; HDP, hexose diphosphates; A, B, C, unidentified sugar phosphates (for subsequent identification, see section 7); D, unidentified compound containing glucose and a glucose phosphate (for subse- quent identification, see section 10). dry and then allowing a solvent to pass slowly through the paper (which, for this purpose, is suspended vertically from a trough containing the solvent). The solvent moves the various constituents out of the original spot for different distances down the sheet. The sheet can then be dried, APPLICATION OF PAPER CHROMATOGRAPHY 1657 and a second fractionation made with another solvent, moving under right angle to the first one, thus producing a two-dimensional pattern. The re- sulting spots, usually invisible to the eye, can be "developed" by color reagents (such as ninhydrin for amino acids or molybdate for phosphoric acid); radioactive spots are easily detected by "radioautography" — for this purpose, the filter paper is pressed against an x-ray film, and the latter developed. Individual spots can be identified from the known solubilities of the compounds, whose presence appears likely, and the identification can be confirmed by coincidence checks with known compounds. If a spot con- tains sufficient material it can be cut out, eluted and investigated by the usual chemical methods. This technique was first applied by Stepka, Benson and Calvin (1948) to the C*-tagged amino acids from Chlorella and Scenedesmus. After 30-sec. photosynthesis in the presence of C*02, the material from Scenedesmus showed activity predominantly in aspartic acid, somewhat less in alanine. Radioauto- graphs revealed some activity also in asparagine, serine, alanine and phenylalanine. In similar material from Chlorella, there were about equal amounts of active aspartic acid and alanine. Calvin and Benson (1948), and Benson, Bassham, Calvin, Goodale, Haas and Stepka (1949) extended the chromatographic method to tagged products other than the amino acids. Figure 36.13 shows the location on the paper chromatogram of a num- ber of compounds which may be of interest in connection with the work on photosynthesis. Water-saturated phenol is the one (basic) solvent; a mixture of butanol with propionic acid and water the second (acidic) sol- vent. Cationic substances are moved toward the left, anionic toward the top of the figure. Lipids and lipophilic compounds are moved into the upper left corner, farthest from the origin ; while sugars, phosphate esters, and other neutral, hydrophilic compounds stay near the original spot, in the right bottom corner. A mixture containing acetic (instead of pro- pionic) acid was used when it was desirable to move phosphate esters more effectively. In these experiments, algae (or leaves) were killed by dropping them, after exposure to C*02, into hot 80% aqueous ethanol (instead of acid, as before). The extract was concentrated to 2 cc. and 0.01-0.2 cc. were ap- plied to a 1.5-cm. circle in the corner of the sheet. Amounts varying be- tween a few microgram (acids) and 1-2 mg. (sugars) could be separated in this w^ay. Radiograms were made from cells exposed to C*02 for between 5 and 90 sec. in light, after having been engaged in steady photosynthesis in 1-4% 1658 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 m g 4 f o i ^ i % Q. f O z ASPARTIC •" TRIOSE-P 9 i w A CO Q- O -J z: o Ul ^. y s ^ 2: «/) z u < o -4 X a. Lu s o Sen _i Wa- 5^ cT^io APPLICATION OF PAPER CHROMATOGRAPHY 1659 u o CQ O o CD m 05 u o -J < S 1 o _i o o Ji i " i i J on. o 1 0- 6 o q- o 8 • CL 5 UJ to o _J O o s< < UJ X 1- r> _j o UJ UJ CO 2 Q "J ^ fy* UJ 2" "^ — UJ o Z) t/5 < - d _J < cn 3 CO 50- m So ■ to O o 3 O o3 o CO CO 1660 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 inactive carbonate for 0.5 to 1 hr. During this exposure time, no activity entered any compounds insoluble in aqueous alcohol. In the carhoxylic acid field (upper right), a number of spots appeared which were identified as malic, succinic, glycolic and fumaric acids. In similar radiograms from dark -exposed (nonpreilluminated) cells ("respira- tory C* uptake"), iso-citric, succinic, fumaric, and malic acids were found. Each of these identifications was checked by "co-chromatographing" with the corresponding pure acids. ■LIPID SCENEDESMUS I HOUR P32 INORGANIC ^^ Fig. 36.15. Radiophosphorus-containing products of 1 hour photosynthesis in Scenedesmus (after Benson, Calvin et al. 1951). The shorter the exposure, the greater was the proportion of activity which stayed near the original spot (fig. 36.14). After an exposure of 15 sec. or 1 min., three or four dark areas could be distinguished there (fig. 36.14B,C), but after five-seconds exposure or less, only the uppermost of them was significant (fig. 36.14A). This spot was therefore considered as containing the earliest identifiable product of the fixation of carbon dioxide in light. From the general position of the spots, it could be surmised that the compounds in all three spots were phosphate esters; to test this hypoth- esis, radiograms were made of algae which had been allowed to photo- APPLICATION OF PAPER CHROMATOGRAPHY 1661 synthesize for an hour in normal carbon dioxide in the presence of radio- active phosphate. Active spots were found in the same region near the origin of the radiogram (fig. 36.15). A similar pattern could be produced also by using P (32) -labelled phosphorylated intermediates from the fer- mentation of glucose by yeast. By separating the several phosphate esters from the fermentation product (by selective elution from an anion exchange resin) and chromatographing them in one dimension, it was ascertained that the phosphoglyceric acid moves ahead of fructose-6-phosphate, and the latter moves ahead of fructose-l,6-diphosphate. The three above-men- tioned spots in C(14) radiograms from photosynthesizing cells were there- fore identified, from the bottom upward, as ''hexose diphosphate," "hexose monophosphate" and "phosphoglyceric acid." Since the third spot was the earliest to appear, and therefore com- manded the greatest interest, experiments were made to support its identi- fication as phosphoglyceric acid. P(32)-labelled material from algae which had been exposed for 1 hr. to ordinaiy CO2 in active phosphate was coprecipitated with synthetic barium phosphoglycerate, and the precipi- tate, after conversion to free acid, was cochromatographed with the ex- tract obtained from algae after 90 sec. photosynthesis in C*02. The perim- eters of the P* and C* activities in the region of the "third spot" were found to coincide, thus indicating that the C*-labelled compound must be identical with the P*-labelled compound (whose barium salt was proved before to be coprecipitable with barium phosphoglycerate). From products of 5 sec. exposure of Scenedesmus cells to C*02 (which, we recall, gave one strong radioactive spot only), phosphoglyceric acid was isolated and identified by the following procedure: 1-g. batches of cells, exposed for 5 sec. to C*02 in intense light, were added to 18 g. normal algae (to provide bulk for isolation). The algae were killed by acid (4CH3- COOH, glacial + 1 HCl, cone), all activity going into aqueous extract. (This method was again used, instead of killing with 80% alcohol, to re- move as much protein, cellulose and lipids as possible at the very beginning of the process.) Fractionation of the extract by precipitation at pH 7 and leaching out at pH 10, repeated precipitation with BaCl2 from 50% ethanol, and dissolu- tion in 0.05 A'^ HCl, led to a crystalline precipitate whose P content and specific rotation were close to those of barium-3-phosphogly cerate. Oxida- tion by periodate led to products (formaldehyde, formic acid, carbon dioxide) found also with glyceric acid. The somewhat lower rotation of the photosynthetic product was taken as indicating the presence of a small amount of 2-phosphoglyceric acid of higher specific activity than the main amount of 3-phosphoglyceric acid. On some of the radiograms, the PGA spot was in fact double; the "right half" could be changed 1GG2 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 into the "left half" by heating in 0.1 A^ HCl for 1 hr. It was suggested that the right half is 2-phosphoglyceric acid and the left one, 3-phospho- glyceric acid. The proportion of the 2-substituted acid was much higher in strong than in weak light. When the temperature during the C*02 ex- posure was lowered to 4° C, only the right half of the spot appeared. It thus seems that the 2-phosphoglyceric acid is the first formed, but less stable isomer. The crystalline product which thus had many properties of authentic phosphoglyceric acid weighed 10 mg. and accounted for one-third of total C* fixed in 5-sec. exposure. The last washings and supernatants from this preparation accounted for another 30%; when chromatographed, they, too, deposited their activities in the PGA spot. Easily hydrolyzable phosphate esters could be identified in the paper . chromatograms by spraying with molybdate to form molybdenum blue. In this way, glucose-1-phosphate, phosphopyruvic acid, and triose phos- phate (apparently, both 3-phosphoglyceraldehyde, and phosphodihydroxy- acetone) were identified in positions shown in figure 36. 13. Phosphopyruvic acid is one of the earliest of these compounds to appear in photosynthesis, as illustrated by figure 36.6a. According to Calvin et al. (1950), free glyceric acid often appears, together with malic acid and phosphopymvic acid, as the earliest companions of phosphoglyceric acid in the radiograms from briefly photosynthesizing cells. In one experiment, after 5-sec. photo- synthesis, 87% of the tracer was found in PGA, 10% in PPA and 3% in malic acid. These results agree with Gaffron and Pager's observations (Fig. 36.10), except that the latter listed PA rather than PPA as the main companion of PGA in the first seconds of photosynthesis. In all probabil- ity, both PGA and PPA may undergo hydrolysis in the extraction and sep- aration processes; free acids, GA and PA, will then be found instead of the originally present phosphate esters. The identification and order of appearance of amino acids was men- tioned above. It is significant that some of them — those with C3 and C4 chains— appear very early, even before tagged sugars or their phosphate esters. With these paper-chromatographic observations, the previous results, obtained with ordinary analytical methods and summarized in tables 36. Ill and 36.IV, were largely confirmed and greatly enlarged. The role of phosphoglyceric acid as the first compound which becomes tagged in photosynthesis was established much more firmly than before. The early appearance of phosphopyruvic acid and of the phosphates of trioses and hexoses was clarified. (In previous experiments, all these compounds must have been hidden in the second anionically tagged, "non ammonia- elu table," subfraction.) The early appearance of tagged amino acids was APPLICATION OF PAPER CHROMATOGRAPHY 1GG3 confirmed, and their specific nature established. Seldom has the applica- tion of a new method of analysis brought such sudden light into the dark- ness and permitted the identification of so many compounds present in a complex mixture only in microgram amounts. This work undoubtedly constitutes one of the greatest successes so far of the isotopic tracer tech- nique in chemical analysis ; but it could only be achieved by combination of this method with chromatography, an analytical method which, although almost fifty years old, has only recently been extended beyond its original narrow field of application to organic pigments. The distribution of tracer carbon within the carbon chain of PGA, hex- oses (cf. section 1), and of other early products also was reinvestigated. Typical results of these "degradation experiments" are shown in table 36. V. It shows the initial preferential (or exclusive) labelling of the car- boxyl group in PGA, malic acid and alanine, and gradual approach to equidistribution, already noted before. This approach to isotopic equilib- rium occurs in a closely parallel way in PGA, alanine and hexose. Table 36.V Distribution of Carbon 14 within the Carbon Chain in the Products of Photosynthesis (10,000 f. c.) (after Calvin et al. 1950) Barley Position 2 sec. 15 sec. 60 sec. Chlorella 5 sec. Phosphoglj^ceric acid COOH| CHOHJ I CH.OH 85 15 56 21 23 4-i 30 25 95 3 2 Alanine COOH CHNH, I CHs 67 30 30 48 44 <5 Sucrose C3,4 C2,5 Ci,6 52 25 24 37 34 32 Scenedesmus 30 sec. 87 7 6 Malic acid COOH 1 rest Scenedesmus 10 sec. 93.5 6.5 It will be noted that approximate equidistribution of C* in the sucrose molecule is reached, according to table 36. V, already after 60 sec. exposure. 1664 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 In the 1947 experiments (table 36.1), the distribution of C(14) in glucose was far from uniform even after a whole hour of photosynthesis; in 1948 experiments (also mentioned in section 1), uniform distribution was ap- proached, in glucose, only after about two hours. The reason for this dif- ference is not clear. Giljbs (1949, 1950) studied the distrilnition of C*(14) in glucose, fructose, dextrin, starch, alanine, and mahc acid, in materia! from sunflower leaves exposed to C*02 for 1.5, 2, and 4 minutes, with rather confusing and irrei)roducibIe results. In fructose, for example, the distribution was nonuniform after 1.5 and 2 minutes, but uniform after 4 minutes; in starch, on the other hand, the 3,4 positions were strongly favored even after 4 minutes; in malic acid, three quarters of total C* were found in the carboxyl both after 1.5 and after 4 minutes of exposure. \^'hen sunflower plants were exposed to C*02 in darkness, C* was found to >90%in the carbo.xyl groups in alanine and malic acid, and in 3,4 positions in sucrose, dextrin, and starch, after 16 or 27 hours of exposure. The rapid appearance of C* in the a and ^ positions in PGA is very significant. It proves — almost beyond doubt — that photosynthesis involves a cyclic process. An "acceptor" molecule. A, first takes up COo to form PGA; the newly added carbon is located entirely in the carboxyl group. The carboxyl-tagged PGA then luidergoes reduction and condensation, at some stage of which the products divide into two parts; one is transformed into permanent photosynthates (polysaccharides, proteins), the other is reconverted into the carbon dioxide acceptor, A. The latter now contains labelled carbon; it therefore gives, upon carboxylation, PGA molecules with the tracer not only in the carboxyl group (7, or 3-position), but also in the two other positions (a and jS, or 1 and 2). A cyclic process of this type was anticipated by biochemists, who saw the mechanism of photosynthesis, almost a priori, as a reversal of that of respiration, and therefore expected it to include a cycle in which hydrogen atoms (supplied by donors identical with, or analogous to, reduced co- enzymes I and II) and carbon dioxide molecules (supplied by appropriate carboxylases) are first "grafted" on a "stock" (a C3 or C4 compound); the C atoms in the product are then shuffled around (and H atoms thrown in) until a triose molecule is synthesized and the "stock" regenerated. Taking this type of mechanism for granted, the biochemists asked whether the "catabolic" cycle will prove to be simply the known anabolic cycle (Krebs cycle) run in reverse (which seemed to be the simplest hypothesis), or different from it. The above-described early findings with C(14) proved that the biochemists' main hunch was correct— a cyclic mechanism of the postulated type does exist in photosynthesis ; but it seems to be dif- ferent from any known respiratory cycle (e. g., it contains no tricarboxylic acids as intermediates). It is useful to realize that a cyclic mechanism was not an a priori neces- sity for photosynthesis (as it had not been an a priori necessity for respira- APPLICATION OF PAPER CHROMATOGRAPHY 1065 tion). Even leaving aside, as biochemically implausible, the once popular Baeyer's mechanism, which envisaged a straight reduction of Ci compounds (CO2 -^ HCOOH -^ CH2O) followed by condensation of formaldehyde to glucose, one could imagine other mechanisms in which Ci fragments were grafted onto a large carrier molecule, then reduced to the carbohydrate level, and the reduction product split off from the carrier without the latter having been directly involved in the reduction process. This, in fact, was the model used in most chemical and kinetic speculations concerning photosynthesis, such as the kinetic theories reviewed in chapters 27 and 28. The symbols ACOo or {CO2I were used there to designate a CO2 molecule "grafted" onto an acceptor molecule prior to being photochemically reduced; the acceptor, A, was assumed to be regenerated at some stage of the process, without change, and returned into the cycle. True, some evidence was found, even prior to the radiocarbon studies, that the CO2 acceptor itself is a product of photosynthesis. (For example, certain induction phenomena could be attributed, in chap. 33, sect. C3, to the disappearance of the CO2 acceptor in the dark and its regeneration in light) ; but even this did not prove that the acceptor molecule undergoes a continuous, cyclic transfor- mation in light. Radiocarbon studies first proved that the carbon skeleton of the acceptor molecule is assembled anew at the end of each cycle from carbon atoms derived, in part, from the carbon dioxide molecules it has taken up in the preceding turns of the cycle. (Without this throwing of the new and the old carbon atoms into a common reservoir, the labelling of the a and /3 carbons in PGA would not occur at all, or would occur much slower than it actually does.) One consequence of this finding is that in any future analysis of the kinetics of photosynthesis, one will have to consider the concentration of the acceptor, not as a constant (Ao in chapter 27), but as a function of the rate (and perhaps also of certain special conditions) of photosynthesis. A further complication will arise if the acceptor (or its carboxylation product) is produced also as intermediate (or by-product) of respiration (as seems to be the case with PGA; cf. chapter 37D, section 3). The proof of regeneration of the carbon dioxide acceptor by a rapid cyclic process can be considered as the second important result of the appli- cation of carbon tracer to the study of photosynthesis (the first one having been the identification of phosphoglyceric acid as the first intermediate). The study of the intramolecular distribution of labelled carbon was con- tinued by Bassham, Benson, and Calvin (1950). Degradation experiments with labelled phosphoglycerate showed that a and jS carbon atoms have approximately the same activity {e. g., in one experiment, after 60-sec. photosynthesis in C*02, 55% of all C* present in phosphoglyceric acid was located in the carboxyl, 25% in a, and 27% in 1666 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 j3 carbon). This indicates that at some point in the regenerative cycle a symmetric intermediate must be formed (e. g., oxaHc acid, succinic acid, or glycol) — a compound in which two equivalent groups have equal chance to contribute a or j3 atoms for the formation of phosphoglyceric acid. Calvin (1949) confirmed the previously mentioned observation that, while tagged C5 and Ce acids are absent among the products of brief photo- synthesis, they appear within minutes after photosynthesis is stopped. In these experiments, cells were allowed to photosynthesize in C*02, and the vessel was then swept by helium for 90 sec. before killing the cells. If, during the sweep, the cells were left in light, no tagged citric, isocitric or glutamic acid was found. These three acids appeared, however, if during the sweep the cells were left in darkness. It thus seemed that the inter- mediate products of photosynthesis were drawn, immediately after the cessation of illumination (but not before), into the respiration process, yielding C5 and Ce acids of the Krebs cycle (and related amino acids) (com- pare fig. 36.23). From observations of this type, and also from kinetic studies with C(14) by Weigl, Warrington and Calvin (1950), it was concluded that nor- mal respiration of the cells is largely inhibited during photosynthesis. However, this generalization is not supported by other experiments, e. g., those with isotopic oxygen. An alternative explanation of the tagging ex- periments by Steward and Thompson (1950) will be mentioned in Chapter 37D, (section 3), where the whole question of the relations between photo- synthesis and respiration will be reviewed. Another interesting result of the study by Weigl et at. — the strong iso- topic discrimination in photosynthesis between C(12) and C(14)^ — also will be discussed in chapter 37D, 6. The Role of Malic Acid : One or Two Carboxylations in Photosynthesis? Malic acid has been found, in all experiments of Calvin and co-workers (as well as those of Stutz, and Gibbs), to be a major early tagged product (c/. fig. 36.14); but its true role in photosynthesis is still uncertain. At first it appeared likely that malic acid is another "first product" of photosynthesis, possibly arising by reaction (36.1), followed by reduction of oxalacetic acid to malic acid ; or by direct reductive carboxylation of pyru- vic acid, as observed in chloroplast preparations by Vishniac and Ochoa, Tolmach, and Arnon {cf. chapter 35, equation 35.34), catalyzed by the so- called "malic enzyme." One could imagine, for example, that the forma- tion of PGA (by carboxylation of an unknown C2 acceptor) is followed by dehydration to pyruvic acid, and reductive carboxylation of the latter to malic acid. THE ROLE OF MALIC ACID 1667 However, subsequent experiments did not square with the assumption that malic acid is an intermediate in the main reaction sequence of photo- synthesis. Bassham, Benson and Calvin (1950) found that malonate inhibits, by as much as 70-97%, the formation of tagged malic acid in light, without inhibiting photosynthesis as a whole by more than 15-35%. Malonate had no effect either on the amount of tagged phosphoglyceric acid or on the distribution of tracer between the carboxyl and the a- and /3-carbon atoms in this compound. These experiments indicate that malic acid is not a link in the direct sequence of reactions leading from carbon dioxide via phosphoglyceric acid to sugars. Since, however, it is one of the first compounds to be labelled in light even when the labelling of fumaric acid and succinic acid is negligible, as well as in darkness (where the latter acids form the bulk of the tagged compounds), it appears that tagged malic acid is produced not only by the respiratory but also by the photosynthetic C*02 uptake mechanism. It was suggested that in the second case, malic acid is not an intermediate in the carboxylation cycle, but a side product, a "storage reservoir" of C*, isotopically equilibrated with a C4 product in the main reaction sequence (such as oxalacetic acid, OAA). The early ap- pearance of phosphopyruvic acid (PPA) suggested the scheme: _llQ +C*02 ^ main sequence to sugars (36.4A) PGA ^ PPA > OAA (enol ) 4-2[H] : ^ malic acid -2[H] This equilibration could conceivably be interrupted by malonate, while the main reaction sequence remains unaffected. Badin and Calvin (1950) chromatographed the products of C*02 fixation in Scenedesmus obtained by photosynthesis in very low light. The most interesting finding was that in low light the first identifiable C*- tagged product was malic acid (30% of total C* uptake, if extrapolated to zero time). Organic phosphates (including phosphoglyceric acid) ac- counted, in such low light, for less than 30%o of total C* (as against the 90% found in phosphoglyceric acid alone after 5 sec. photosynthesis in strong light). From these experiments, Calvin et al. concluded that two primary C*02 uptake reactions occur in photosynthesis: one leading directly to the for- mation of phosphoglyceric acid, the other, indirectly, to malic acid. The separate formation of these two products was supported by the observation that the percentages of total C* present in organic phosphates and malic acid extrapolated to finite values at time zero. Calvin and co-workers sug- gested that the primary uptake of C*02 in photosynthesis involves the two reactions: 1668 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 (36.5) Cj acceptor + C*02 > (phospho)gly eerie acid (36.6) Cs acceptor + C*02 > (oxalacetic acid) ^ ^ malic acid the first reaction being predominant in strong light, and the second one in weak light. Gaffron and co-workers (1951) disagreed with this hypothesis and sug- gested that (36.5) is the only carboxylation reaction directly related to Fig. 36.16. C*02 fixation in C2 and C3 compounds as function of time in Scenedesmus at 15° C. The cells (10 fi\.) were supplied with an amount of C*02 sufficient for only 4 min. of photosynthesis; this e.xplains the decUne of activity at longer times (after Benson et al. 1952). photosynthesis. They argued that the assumption of two different car- boxylations will require five different photochemically induced reduction steps (c/. section 12 below), which they considered theoretically implausible. The evidence for an independent carbo.xylation leading to malic acid they found inconclusive, even the low-light findings of Badin and Calvin (1950) (since in very weak light, dark catabolic processes occur at rates comparable with those of light-induced reactions). More recently (Bassham et al. 1954) this point of view was accepted also by Calvin and his group. This meant the abandonment of the argument that any intermediate whose tagging curve approaches the zero point on the time axis under a finite angle, must be tagged by the uptake of external C*02, without any sizable reservoirs of intermediates interposed between it and the medium. Malic and phosphoglyceric acids consistently showed this behavior (cf. figs. 36.16 and 36.18) ; but so did several other compounds, including some sugar phosphates {cf. fig. 36.17), which certainly could not be the direct products THE ROLE OF MALIC ACID 1669 of carboxylation. Bassham et al. (1954) discussed two mechanisms that could possibly account for the immediate appearance of tagged carbon atoms in compounds not connected directly with the external C*02 reser- voir. One of them is the possibility that some rapid enzymatic reactions 10 20 JO 40 Fig. 36.17. C*02 fixation in sedoheptulose phosphate, ribulose diphosphate, and mannose phosphate as function of time. 10 pi. Scenedesmus cells, 20° C. The separation of sedoheptulose and mannose is imperfect (both occurring in the same general monophosphate area) (after Benson et al. 1952). ASPARTIC ACID 10 to 30 40 SIC MINUTES Fig. 36.18. C*02 incorporation in some d compounds as function of time (10 All. Scenedesmus, 15° C). Insert shows the initial slope of the mahc acid tagging curve (after Benson et al. 1952). could occur during (or after) the supposedly instantaneous and complete stoppage of life processes by boiling alcohol, and could transfer some C* atoms from the primary into the "secondary" products. The second (rather vaguely described) possibihty is that of labelled C* "skipping" 1670 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 from enzyme to enzyme without undergoing full isotopic equilibration with the interenzymatie reservoirs. No suggestion was made as to whether the early tagging of malic acid could be ascribed to one of these two causes; but it was implied that some such mechanism, rather than the previously postulated more or less direct equilibration with the external tagged carlwn dioxide, must be responsible for it. The mechanism of formation and the function of C4 bodies in the reac- tion sequence of photosynthesis remains, at this writing, one of the uncer- tain parts in the interpretation of the radiocarbon experiments. 7. The C5 and C7 Sugars as Intermediates in Photosynthesis Benson, Bassham, Calvin, Hall, Hirsch, Kawaguchi, Lynch and Tolbert (1952) described the identification (first announced by Benson, Bassham and Calvin 1951, and Benson 1951) of two new photosynthetic intermedi- ates— a pentose and a heptose sugar or, more precisely, their phosphate esters. These compounds were found to appear in tracer quantities after a few seconds of steady-state photosynthesis in unicellular algae, higher plants, and purple bacteria (cf. figs. 36.17 and 36.21). Their identification resulted from a further development of the area of the paper chromatograms containing the phosphate esters — lower right corner in Figs. 36. 13 and 36. 15. This area contained, in addition to spots attributable to phosphoenolpyru- vic, phosphoglyceric, and phosphoglycolic acids, to triose (dihydroxyace- tone) phosphate, and to several hexose phosphates, also two previously unidentified major spots whose relative importance increased with de creasing exposure to C'*02, thus indicating that they were due to early in- termediates of photosynthesis. The sugar obtained by hydrolysis of one of the unknown phosphate esters was found to form reversibly an anhydride with the equilibrium constant characteristic of sedoheptulose. After the oxidation of this sugar by periodic acid, 15% of its tracer was found in formaldehyde, 55% in formic acid, and 28% in glycolic acid, in agreement with expectations for a C7 ketose. Periodic acid treatment of the poly- alcohol derived from the same sugar gave one molecule of tagged formalde- hyde per three molecules of tagged formic acid. Assuming equal labelling of all carbon atoms in the chain (a permissible assumption for carbohy- drates found after 5 min. of steady photosynthesis in saturating light, cf. table 36. V), the theoretical ratio is 1 : 2 for a hexitol, and 1 : 2.5 for a hepti- tol. The empirical value of 3 thus confirmed that the chain was longer than Ce. Co-chromatographing of the unknown sugar (or its an- hydride) with chemically pure sedoheptulose (or its anhydride) led to a defi- nite identification. The second unknown sugar was identified as a pentose by reduction. C5 AND Ct sugars AS INTERMEDIATES 1671 and more specifically, as ribulose, by co-chromatographing with the pure compound. Table 36. VI shows the relative amounts of the several tagged sugar phosphates found after 1 to 20 min. photosynthesis in C*02 in different organisms. Table 36.VI Relativj: Amounts of Different Phospiiorylated Sugars after Short Period of Steauy-State Photosynthesis in C*02 (after Benson et al. 1952) Duration of PS in Organism C*02 (min.) Rhodospir ilium ruhruni 20 Scenedesmus (1 day old). ... 5 Chlorella 1 Barley seedling leaves 1 Soy bean leaves 5 Percent C* in Glucose Fructose Sedo- heptulose Ribulose 50 20 15 15 40 10 16 34 40 14 40 6 53 16 17 13 39 24 36 1 In discussing these findings, Benson et al. noted that sedoheptulose ac- cumulation is known to occur during the photosynthesis of certain succu- lents. The same sugar (or its phosphate) was also identified in many other plants. Ribulose, on the other hand, is not known as a constituent of plants, although the corresponding aldose (ribose) is known to occur in them. Ribulose phosphate has been previously observed in bacteria, but not in photosynthesizing organisms. Bassham, Benson, Kay, Harris, Wilson and Calvin (1954) inquired into the distribution of labelled carbon in the ribulose and sedoheptulose skele- ton after brief periods of steady photosynthesis in 0*02," exposures as short as 0.4 sec. were used in this work. A suspension of Scenedesmus ohliquus was placed in a rectangular reservoir traversed by a stream of air charged with 4% CO2. The steadily photosynthesizing suspension was pumped out of this reservoir through a transparent tube. At a certain point in the tube, a solution of C*02 was injected into the suspension. The mixture was then discharged into boiling methanol. The time of exposure to C*02 in light (between injection and discharge) could be varied by chang- ing the rate of pumping; the illumination was about 40 klux from each side. Table 36.VII shows the distribution of C(14) in C3, Ce, C- and C5 chains formed after 5.4 see. or steady-state photosynthesis of Scenedesmus in C*02. Table 36. VII indicates clearly that the Ce chain is formed bj^ head-to- head condensation of two C3 chains. The mechanism of formation of the C7 and C5 sugars is less obvious; the suggested interpretation will be dis- cussed in section 12 below. 1672 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 Table 36.VII C(14) Distribution in Products Formed by 5.4 Sec. Exposure of Scenedesmus to C*02 DURING Steady-State Photosynthesis (after Bassham et al. 1954) Carbon no. C3 (glyceric acid) Ce (fructose) Ct (sedoheptu- lose) Cs (ribulose) 1 6 3 2 3 2 6 3 2 5 3 82 42 28 69 4 43 24 10 5 3 27 11 6 3 2 7 2 Sedoheptulose and ribulose have similar configurations (at the C-5 and C-6 atoms of the former and the C-3 and C-4 atoms of the latter). Cleavage of an aldoheptose into a "biose" (glycolaldehyde) and a pentose appears a likely reaction from recent chemical studies of this type of compound by Horecker (cf. below) ; in the case of sedoheptulose the product of this splitting will be ribose, which could isomerize to ribulose; a similar cleavage of ribulose must lead to a "biose" and a triose. Benson et al. therefore suggested that the function of sedoheptulose and ribulose in photosynthe- sis is to serve as sources of the C2 body (whose carboxylation leads to PGA) . The location on the chromatogram suggests that ribulose is formed as a diphosphate, and sedoheptulose as a monophosphate. This identification of the esters was further supported (Benson, 1952) by a determination of the ratio C(14):P(32) in the products of prolonged photosynthesis in doubly tagged medium. This ratio was 0.9 in the fraction suspected to be pentose diphosphate, as against 1.35 in the fraction containing PGA, 2.3 in that identified as glucose monophosphate, and 2.6 in that containing fructose and sedoheptulose. The theoretical values for the first three compounds (assuming complete isotopic equilibration with the medium) were 2.5, 3.0 and 6.0, respectively. Since all values were smaller than ex- pected, the saturation with C* must have been incomplete, but the order of the ratios clearly points to the first-named fraction as the one containing more phosphate residues than all the others. As mentioned above, when the C5 and C? intermediates were first found it was suggested that they are the source of the then postulated "C2 ac- ceptor" in photosynthesis. Subsequently, a simpler hypothesis was sug- gested— that the C5 sugar phosphate itself serves as the carbon dioxide ac- ceptor, giving rise to tw^o molecules of phosphoglycerate by hydrolytic splitting : (36.7) + HjO Cs + C, > 2 C3 Cs AND C7 SUGARS AS INTERMEDIATES 1673 It is easy to see that a pentose has the correct "reduction level" (L = 1.0, c/. p. 109) to give glyceric acid (L = 0.833) after dilution with one molecule CO2 {L = 0): 5/(5 + 1) = 0.833. The mechanism of reaction (36.7) involves a not implausible internal dismutation-conversion of two neigh- boring keto groups into a carboxyl and a hydroxyl group (c/. second arrow in equation 36.8) : (36.8) CH2O© 1 CHOH CHOH COOH :;ho© +CO1 I > CH2OH I +HjO _^ C=0 CH(0H)2 CH2O© C=0 CH2O© COOH I -^2 CHOH CH2O© where © is a phosphate residue, H2PO3. Reaction (36.8) recently was observed in vitro by Weissbach, Smyrniotis and Horecker (1954) with a leaf extract, and by Quayle, Fuller, Benson and Calvin (1954) with a Chlorella extract as catalyst, making its inclusion in the hypothetical reaction sequence of photosynthesis very plausible. In spinach extracts, reaction (36.8) is stimulated by TPN and ATP, sug- gesting that it may include a reversible oxidation-reduction stage. It will be noted that it is due to its occurrence as a diphosphate that ribulose can be split into two molecules of PGA (in reaction 36.8 the prod- ucts are one molecule of a-PGA and one molecule of /3-PGA) . As discussed in chapter 8 (section A4), carboxylation of RH to RCOOH is, under standard conditions, a reaction with a total energy close to zero, and a positive standard free energy of several kcal/mole (table 8. VIII) ; consequently, the equilibrium usually lies on the side of decarboxylation. The kinetics of CO2 fixation in photosynthesis, on the other hand, points to CO2 fixation reaction with an equilibrium far on the side of synthesis, even at 0.03% CO2 (c/. chapter 27). Even known reversible carboxylations, such as that of pyruvic to oxalacetic acid, do not satisfy this requirement. Bassham et at. (1954) attempted to estimate the free energy of reaction (36.8). Assuming pH 7, and concentrations of 5 X 10"* M for the pen- tose, 10-=^ M for carbon dioxide, and 1.9 X 10"^ M for glycerate (for justi- fication of these values, see section 9 below), they obtained AF = —7 kcal./mole, a value which would place the equilibrium far on the side of carboxylation. Using standard bond energies (table 9. II) one notes that the dismuta- tion reaction (second arrow in reaction 36.8) which is, in essence: 1674 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 2C=0 > C— O + O— C=0 should have a AH of —29 kcal. Empirically, the energies of this type of reactions ("Cannizzaro reaction") are not quite as large, but they are negative ; a coupling with dismutation could therefore reduce the carboxyla- tion energy markedly — e. g., bring it from the standard value of close to zero (table 8. VII) to as low as —20 kcal. This should shift the free energy of carboxylation (which, in table 8. VIII, differs by about 15 kcal from the total energy) to about —5 kcal, or close to the above estimate. This estimate of AH and AF is very crude, but it points to one impor- tant relationship. It has been repeatedly emphasized in this book that photosynthesis in- volves two reduction steps of different types: carboxyl to carbonyl, and carbonyl to hydroxyl. Because of the stabilization of carbon-oxygen bonds by accumulation at a single carbon atom, the first step requires consider- ably more energy than the second one. (According to table 9. IV, the nor- mal potentials of transitions of the first type are about -|-0.5 volt, those of the second type, about -fO.2 volt; the difference of 0.3 volt corresponds to a free energy of dismutation of about —2 X 0.3 X 23 = —14 kcal.) In schemes 36.IIIB and 36. V, light energy is used in one step only (conversion of PGA to TP) which is the more difficult of the two required reduction steps. By converting six carboxyl groups into six carbonyl groups — twice as many as are needed for the final carbohydrate synthesis — we acquire the possibility to dismute the three extra carbonyls, and to use the energy of their dismutation to drive forward an otherwise difficult reaction — fixa- tion of CO2 in a carboxyl. This is achieved in the coupled reaction (36.8) which can thus be considered as a chemical mechanism for indirect use of stored light energy to facilitate carboxylation. A further "assist" may be provided (as first suggested by Ruben, cf. p. 201) by a "high energy phos- phate" ; specifically, such a phosphate may be involved in the introduction of a second phosphate residue in ribulose diphosphate. However, the amount of energy stored in this way is likely to be considerably below the 10-12 kcal available in carboxyl phosphates. The assumption that sedoheptulose and ribulose phosphates are not involved in the synthesis of hexoses in photosynthesis is supported by the observation {cf. table 36. VII) that the distribution of tracer carbon in them bears no similarity to that in hexoses. (One does not see how a hexose with preferential C labelling in position 3,4 could be derived from a ribu- lose with preferential labelling in C atom 3, or a sedoheptulose with equal labelling of atoms 3, 4, 5.) If it were not for this argument, one would be tempted to consider the formation of the C5 and C7 sugars as related to the "alternative path" of respiration, first suggested by Warburg, and more recently established by Horecker {cf. Horecker 1951 and later), in which glucose phosphate is oxidized to phosphogluconate, decarboxylated in ribose, and the latter split into a triose and a C2 compound. THE C2-FRAGMENT 1675 8. The C2 Fragment The appearance of a C3 compound (PGA) as the first carboxylation product in photosynthesis naturally pointed to a C2 compound as carbon dioxide acceptor. If the carboxylation were not coupled with reduction, or phosphorylation (or both), the acceptor would have to be phosphogly- col: (36.9) n.POaCHOHCHaOH + CO, > H2PO3CHOHCHOHCOOH In the case of reductive carboxylation, the acceptor could be phospho- glycol aldehyde. Although C2 compounds have been found among the early tagged photo- synthesis products, none of them could be identified as the CO2 acceptor, and the hypothesis arose that this acceptor does not occur in the free state at all, its formation by splitting of a long-chain compound being coupled with carboxylation. The findings described in section 7 support this hypothesis. Nevertheless, the study of the early tagged C2 products remains of interest, because even in a reaction of the type (36.7), fleeting appearance of free C2 fragments remains possible. The actually found labelled C2 compounds may be, if not identical, at least related to an evasive C2 inter- mediate in the main reaction chain. Labelled phosphoglycolic acid, glycohc acid, and glycine are the main C2 compounds appearing on the chromatograms (cf. for example, fig. 36.14c). Glycolic acid has a reduction level (L = 0.75) too low to produce GA by carboxylation (which would require, according to equation 36.9, glycol with L = 1.50); however, glycohc acid could be a by-product of oxidation of the acceptor. Calvin and co-workers (1951) observed that if plants are illuminated in 0*02 until labelled C3 and C4 compounds appear, and illumination then continued in absence of CO2, the C*3 and C*4 compounds disappear, and labelled glycolic acid (CHoOH-COOH) and glycine (NH2CH2-COOH) increase on the radiograms. The two compounds may be related to the C2 acceptor, and accumulate whenever this acceptor is prevented from being converted to PGA by the absence of carbon dioxide. It is further of im- portance that glycolic acid is found to be labelled symmetrically in both carbon positions. This shows that it is derived from a symmetric precur- sor, such as glycol, OHCH2CH2OH. The appearance of glycolic acid was studied in more detail by Shou, Benson, Bassham and Calvin (1950). They confirmed that even after very short exposure the two carbons in glycolic acid are uniformly labelled (similarly to the a and j8 carbons in PGA). Shou et at. fed Scenedesmus C(14)-labelled glycolic acid at pH 2.8 (to suppress ionization) and found an 1676 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 appreciable assimilation. In the dark, under anaerobic conditions, the label was incorporated in glycine and serine. In light, tagged products similar to those obtained from C*02 were observed. The PGA obtained in this way was labelled uniformly in a and /3 positions, irrespective of the position of the label in glycolic acid. In section 12, we will refer to Calvin and Benson's speculations (1949) that the C2 acceptor in photosynthesis may be vinyl phosphate, CH= CHOPO3H2, a dehydration product of glycol phosphate (c/. equation 36.9). These speculations have been superseded by the above-mentioned hypothesis of direct carboxylation of a C5 sugar diphosphate to form two molecules of PGA. 9. Kinetic Studies Most conclusions presented so far in this chapter are based on qualita- tive kinetic information — the sequence in which different tagged com- pounds appear when photosynthesizing cells are offered tagged carbon di- oxide. On a few occasions, we have referred to more quantitative data. SLUCOSE MONOPHOSPHATE Fig. 36.19. C*02 incorporation in hexose phosphates as function of time of photosynthesis. 15° C, 10 /xl. Scenedesmus cells. "Unidentified glucose phos- phate" later identified as uridine diphosphate glucose + related compounds (after Benson et al. 1952). e. g., to the changes in the ratio of labels in phosphoglyceric and pyruvic acid with time {cf. fig. 36.10). More recently, the absolute and relative amounts of labelling on dif- ferent compounds as function of time have been studied more systematically by Calvin and co-workers (Benson, Kawaguchi, Hayes and Calvin 1952; Calvin and Massini 1952; Benson et al. 1954). KINETIC STUDIES 1677 Figures 36.16 to 36.20, taken from the first-named paper, show time curves of the tagging of several compounds in Scenedesmus cells exposed to C*02 during steady photosynthesis in saturating light, at 15-20° C. (The ordinates are the radioactivities eluted from the paper chromatogram — about }i of those measured by direct plating.) MINUTES Fig. 36.20. Course of C*02 incorporation in Scenedesmus in most important early tagged compounds presented on common time and activity scales (Scenedes- mus, 15° C, 10 ixl cells) (after Benson et al 1952). "reservoir" Figures 36.16 and 36.20 show clearly that no appreciable exists between C*02 and PGA. Figure 36.16 does not preclude a similar, direct C*02 incorporation in PPA. We have mentioned before arguments against such an assumption, presented particularly by the Chicago inves- tigators (Fig. 36.10), and suggesting that PPA is a secondary product, rapidly equilibrated isotopically with PGA. 1678 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 The malic acid tagging curve also appears, in fig. 36.18, to have a finite slope at ^ = 0, indicating the possibility of its formation by uptake of ex- ternal CO2 without sizable accumulation of the label in a "reservoir" (such as oxalacetic acid). However, here, as in the case of pyruvic acid, the con- clusion is not binding. Peculiar in the case of malic acid is the sharp in- crease in the slope of the curve about one minute after the beginning of exposure. The steeper slope is equal to that of PGA, (fig. 36.20) ; it can be taken as indicative of the formation of malic acid, at this stage, by car- boxylation of compounds derived from phosphoglyceric acid. Till ■ - I 1 GLUCOSE 60% 0— J 50% ^ ^ ^^^ 40% V" - 30% P 50 o < < 40- K O I- u. o 30 20 10- "1 r -1 1 1 r PGA o ^_^ — o TOTAL SUGAR PHOSPHATES 16 18 2 4 6 8 10 12 14 TIME (SECONDS) Fig. 36.25. Extrapolation of tagging in Scenedesnius to zero time (after Bassham et at. 1954). Benson (1952) described an independent determination of the photo- stationary concentration of various phosphate esters in steadily photo- synthesizing Scenedesmus. It consisted in ''saturating" the plants with radiophosphorus by 20 hours photosynthesis at 30 klux, in a nutrient solu- tion containing 2 mc. P(32), in 4% CO2, killing them with boiling ethanol in light, extracting with hot ethanol and water, chromatographing, and counting the several radioactive spots. Table 36. VIII showed the results in the last column. Considering the difference in conditions between the two experiments, the agreement with the data obtained by C(14) saturation is satisfying. Bassham, Benson et al. (1954) gave some additional kinetic data perti- KINETIC STUDIES 1683 nent to the problem of initial fixation. Figure 36.25 shows that extrapolat- ing the percentage of fixed C* to zero time, gives 75% for PGA and 17% for sugar phosphates, thus leaving 8% unaccounted for by compounds of these two types. This residue was distributed between malic acid (3%), free GA (2%) and phosphopyruvic acid (3%). In the sugar phosphate fraction, no single component shows a trend to become predominant at zero time (c/. fig. 36.21). 9 PGA O GLUCOSE PHOSPHATE © SEDOHEPTULOSE PHOSPHATE ® FRUCTOSE MONOPHOSPHATE • DIHYDROXYACETONE PHOSPHATE 8 10 12 TIME (SECONDS) Fig. 36.26. Tagging of several compounds in the first seconds of e.xposure to C*02 of steadily photosynthesizing Scenedesmus (after Bassham et al. 1954). This figure cor- responds to an enlargement of the left corner of Fig. 36.20. Using the estimates of photostationary concentrations given in table 36.VIII, the rate of increase in specific activity of the several intermediates could be calculated from curves of the type of those in figs. 36.21 and 36.26. (Figure 36.26 is an enlarged representation of the first seconds of tagging.) For the period between ^ = 2 and t = 10 sec, these rates are 0.3 for glucose monophosphate and 1.0 for PGA, with the values for FMP, DHAP, RDP and SMP falling between these two limits (after division by 2 for PGA and DHAP, and by 3 for SMP, to account for the 2 or 3 equally labelled C* atoms in these compounds). This illustrates well the rapidity with which the C(14) spreads over a multitude of cellular "reservoirs." 1684 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 In one set of experiments, the volumes of the several C* reservoirs during steady photosynthesis in 1% CO2 were compared with those after a sudden decrease of carbon dioxide pressure to 3 X 10~^ %. Figure 36.27 shows the result (which is opposite of that caused by darkening, cf. Fig. 36.22). The RDP reservoir increased sharply, reached a peak, and then settled to a new, higher steady level; the PGA reservoir, to the contrary, decreased, within 2 min. to a much lower steady level. The change in 20 O > UJ 10 UJ IT -1 1 1 r- 1% COj T (^ o r' PGA RIBULOSE Dl© AREA ^^00 } J =200" -100 SCENEDESMUS 6* C 0.003% CO2 300 TIME IN SECONDS Fig. 36.27. Effect of a sudden decline in CO2 concentration from 1 to 3 X 10~'% on the volume of tagged reservoirs in Scenedesmus (after Bassham et al. 1954). steady levels is in agreement with the concept that PGA is a product of carboxylation of RDP; but the transient peak of [RDP] requires interpre- tation (as do all the peaks and troughs observed in the transitional states). 10. The Sequence of Sugars In chapter 3 (p. 44) we discussed the evidence concerning the "first sugar" derived from chemical analysis of the products of photosynthesis. Some of this evidence favored the disaccharide, sucrose, as a precursor of free monosaccharides, glucose and fructose; other observations contra- dicted this hypothesis. I^owing now from tracer studies how brief must SEQUENCE OF SUGARS 1685 be the exposures to lead to reliable discrimination between primary and secondary products of photosynthesis, this earlier chemical evidence can- not be considered as very significant. The C(14) work, described in the preceding section, proved that the first products on the reduction level of carbohydrates (L = 1.0) are phos- phate esters. These include mono- and diphosphates of trioses, pentoses, hexoses and heptoses. The heptose and pentose phosphates seem to func- tion as intermediates in the regeneration of PGA, rather than in the forma- tion of the stable products of photosynthesis — sugars and proteins. The path to the latter leads from triose phosphates to hexose phosphates, and from these to sucrose; while free glucose and free fructose appear only later, as hydrolysis products of sucrose, or of their own phosphate esters. This sequence represents an approximate, but not exact, reversal of the common mechanism of glycolysis, which begins with free glucose and pro- ceeds via glucose diphosphate to fructose diphosphate, and thence to triose monophosphates. Following are the successive observations of Calvin and coworkers on the chemical mechanism of sugar transformations in photosynthesis. On their first paper chromatograms, Calvin and Benson (1949) noted that, of the three nonesterified sugars, sucrose was the first to appear; they suggested that it was formed directly from glucose monophosphate and fruc- tose monophosphates (which — as well as hexose diphosphate — regularly appeared earUer than the free sugars) ; and that free hexoses were formed afterwards by hydrolysis of sucrose. Of the two hexose monophosphates, the fructose-6-phosphate appeared to precede the glucose-1-phosphate; in agreement with this, sucrose at first contained more activity (after 30 sec, twice as much) in the fructose moiety than in the glucose moiety. Aronoff and Vernon (1950) made similar experiments with soybean leaves, and con- firmed most of the findings of Calvin and Benson ; they found, however, that glucose-1- phosphate appeared before the fructose-6-phosphate and suggested that the reactions leading to these relatively late products of photosynthesis do not need to proceed in exactly the same way in different plants. They also noted that tagged glyceric acid was more abundant than tagged phosphoglyceric acid after several minutes of photo- synthesis, but that PGA was predominant in the first seconds of photosynthesis. Bean and Hassid (c/. Hassid 1951) found that leaves (of barley, sugar beet or soy- bean) killed by being dropped into boiUng ethanol, after exposure to CO2, confirmed Calvin's conclusions: the tagging of sucrose precedes that of free hexoses. Extraction methods not leading to equally rapid inactivation of enzymes resulted in tagged hexoses being found before the tagged sucrose. It could be proved, however, that these free hexoses were the product of decomposition of phosphate esters. For example, if the cells were dropped into liquid nitrogen, ground in the cold, and extracted with ethanol at 20° C, the radioactivity was found predominantly in nonphosphorylated compounds- including glyceric acid and glucose; but if extraction was done with boiling ethanol, all activity was found in the phosphate esters. 1686 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 Similar experiments had been described earlier by Benson (1950). According to him, higher plants (barley) and purple bacteria (Rhodospirillum) contain a phosphatase which has considerable resistance to heat. Immediate killing of leaves with boiling alcohol produces no free hexoses or trioses and very little free glyceric acid ; but freezing in liquid air, grinding and extraction with boiling ethanol gives considerable amounts of free sugars and acids. Buchanan, Lynch, Benson, Bradley and Calvin (1953) called attention to a previously unidentified spot on the paper chromatogram, which ap- pears after about 30 sec. of steady photosynthesis in C*02, and is situated in the same general area as the known sugar phosphates. This material very easily produced glucose by hydrolysis. More extensive chromato- graphic study revealed that it also contained galactose and mannose. It was surmised that the spot may contain uridine diphosphates of these and other hexoses (compounds described by Leloir). This surmise was con- fi.rmed by co-chromatography with pure compounds and other chromato- graphic tests. The hexoses contained in esters of this type constitute a large fraction of labelled, nonpolymerized hexoses present after several minutes of steady photosynthesis in C*02. Adenine phosphate, adenosine- 5 '-phosphate and uridine-5 '-phosphate also were found in the hydrolysate from this radioactive spot. Benson, Kawaguchi, Hayes and Calvin (1952) gave fig. 36.21 for the course of tagging of several sugar phosphates, showing that fructose is labelled ahead of glucose (and mannose). Calvin and Massini (1952) suggested that the synthesis of sucrose occurs by the interaction of fructose phosphate with the above-described glucose phosphate-uridine complex (uridine diphosphoglucose). The first product could be sucrose mono- phosphate. By treating the material from the hexose monophosphate area of the chromatogram by a phosphatase free of invertase, Buchanan (1954; cf. Buchanan, Bassham et al. 1953) was able to demonstrate the actual presence of sucrose monophosphate, lending considerable support to the hypothesis. Calvin and Massini suggested that compounds of the type of uridine diphosphoglucose may serve as "glucose donors" also in the sub- sequent formation of polysaccharides, 11. Effect of Poisons and pH. on CO2 Fixation It was mentioned in section 6 that malonate was found to inhibit the tagging of malic acid, without much effect on the yield of sugar synthesis. Following are observations on the effect of some other poisons, according to Calvin et al. (1951) (based on the experiments of Stepka 1951). In the presence of 1.5 X 10~* M iodoacetamide, the total uptake of C* was reduced by 90%, but the formation of labelled sucrose was not af- fected at all. (At lower concentrations of the inhibitor, it was even in- EFFECT OF POISONS AND pU ON CO2 FIXATION 1687 creased.) In glycolysis, iodoacetamide is known to inhibit the reaction by which triose phosphate is oxidized to phosphopyruvate ; it thus could be ex- pected to block the synthesis of sugars if it occurred by the reversal of this reaction. The actually observed effect of iodoacetamide on the total C*02 fixation can be attributed to this source. However, the lack of effect on sucrose synthesis has then to be explained by the ad hoc assumption of another partial reaction beyond the triose stage which, too, is affected by iodoacetamide in such a way that the flow of intermediates through some channel by-passing sucrose is slowed down, and ten times more than the usual proportion of intermediates are converted into sucrose. The evidence concerning the effect of iodoacetamide on the respiration of green cells also was contradictory until lately, when Arnon (1952, 1953) and Holzer (cf. Holzer 1954) proved definitely that an iodoacetamide-sensi- tive respiration path does exist in green plants. This was taken as proof that their respiration, like that of animal tissues or yeast, proceeds, at least in large part, via the triose-pyruvate step, with triose dehydrogenase as catalyst, a pyridine nucleotide as hydrogen acceptor, and ADP as "energy acceptor." It seems, however, that in green cells this reaction is not DPN- specific, as usual, but can use either DPN or TPN. Because of the probable spatial separation of the sites of the main catabolic and anabolic processes, the proof of the presence of a triose de- hydrogenase in green cells is not a very strong argument for the participa- tion of this enzyme in photosynthesis. More convincing are the observa- tions on the inhibition of photosynthesis itself by iodoacetate (chapter 12, section 5, and chapter 37D, section 2). In conjunction with the data of Stepka et al. on the effect of iodoacetamide on the uptake of radiocarbon, these observations can be considered as lending support to the — anyhow plausible— assumption that the next step in photosynthesis after the for- mation of PGA is its reduction to triose by an iodoacetamide-sensitive hydrogenase. To a certain extent, these observations also make it plausi- ble that the hydrogen donor in this reduction is a pyridine nucleotide (DPNHa or TPNH2) and that consequently a high energy phosphate, such as ATP, must be supplied to make the reduction possible; however, these conclusions are by no means certain. A number of previous observations (cf. Vol. I, pages 301-311) led to the conclusion that cyanide is a specific poison for the carboxylation reac- tion in photosynthesis. Calvin and co-workers illuminated Scenedesmus for 30 min. in C02-free air to build up the CO2 acceptor, then added 3 X 10 -* M KCN, and one minute later exposed the cells to C*02. The tracer fixation was inhibited by 95%, but the compounds whose tagging was least inhibited were alanine, malic acid and phosphoglyceric acid— supposedly the immediate (or near-immediate) products of carboxylation ! 1688 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 The tagging of triose and hexose phosphates was inhibited much more strongly. It thus seems that cyanide establishes blocks not, or not only, in the carboxylation reaction, but also, or mainly, in the subsequent trans- formation of the carboxylated compounds. (This, of course, does not affect the conclusion that cyanide affects the participation of carbon dioxide in photosynthesis, and has little influence on the parts of the photosynthetic process which do not involve carbon dioxide — such as the primary photo- chemical process, and the reactions which lead to the liberation of oxygen from water.) 5 X 10-^ M hydroxylamine, added under similar conditions, produced a 75% inhibition of total C* uptake. The tagging of malic acid was least affected by this poison; the tagging of glutamic, succinic, fumaric and citric acids was even enhanced. Calvin interpreted this as an indica- tion that hydroxylamine removes a block that normally prevents, in light, the utilization of photosynthetic intermediates as material for respiration (c/. section 12), Gaffron, Fager and Rosenberg (1951) found cijanide to inhibit the post- illumination C*02 fixation strongly, whether it was added during the illumi- nation or after the light had been turned off. The inhibition affected par- ticularly the water-soluble fraction, indicating (in some contrast to the data of Calvin et al.) an inhibition of the PGA formation. Hydroxylamine, on the other hand, inhibited the postillumination fixation of C*02 only if given during the illumination period, but had no effect if given together with the tracer after the return to darkness. One-minute photosynthetic C* fixation in Scenedesmus was constant, according to Calvin et al. (1951), between pR 4 and 9; the rate dechned by about 50% at pR 2 and 10, and dropped abruptly to zero at pH 10.5. The rate-depressing effects of excess acidity (pH 2) or alkalinity (pH 10) did not increase with time; cells could be kept at these extreme pH values for 30 minutes, and the full rate of C* fixation restored upon return to pH 7. The main changes in C* distribution caused by variations in pH were an increase in the proportion of tracer found in malic acid (e. g., from 5% at pH 1.6 to 25% at pH 11.4) and a drop in its proportion in sucrose, from 7 to 0%. The absolute amounts of tagged malic acid and phosphopyruvic acid showed sharp maxima at pH 9. 12. Evolution of the CO2 Reduction Mechanism Between 1946, when the C(14) tracer was first used systematically for the study of photosynthesis, and mid-1954 (the time of the final revision of this chapter), our knowledge of the chemical mechanism of carbon dioxide EVOLUTION OF THE CO2 REDUCTION MECHANISM 1689 reduction has undergone rapid development. Chemical mechanisms based on almost pure speculation (such as Baeyer's formaldehyde theory), or on plausible analogy (such as Thimann's 1938 surmise that photosynthesis in- volves a reversal of glycolysis, c/. p. 183), which had been current before the new approach (and extensively used in the preceding parts of this monograph), have been replaced by schemes based primarily on experi- mental evidence, although still strongly influenced by analogies with other better known metabolic mechanisms. The Berkeley group (Benson, Cal- vin et al.) in particular, has proposed many such schemes. They were pri- marily heuristic hypotheses, repeatedly altered to fit the growing body of data. At this writing, some of the originally controversial questions have been settled, and several steps in the reaction sequence have been firmly established; others, while still speculative, have become at least highly plausible. As this section Avas revised in 1954, it seemed tempting to dis- card the chronological presentation of the several hypotheses adopted at its first writing in 1950, and use only the picture suggested m the most recent papers. However, retelling the gradual emergence of this picture may not be useless, even if somewhat confusing. It is an interesting record of gradual emergence of a landscape from the fog, through which at first only one or two disconnected landmarks were visible. Furthermore, this kind of presentation serves to underline the incompleteness and uncertainty of the current mechanism — instead of making it appear final and unalter- able. Since much of the study of the chemical mechanism of carbon dioxide fixation in photosynthesis has been guided, consciously or subconsciously, by what is known of the mechanism of carbon dioxide liberation in respira- tion, it is useful to begin by saying a few words about the latter process. In Volume I (p. 224, scheme 9. II) we reproduced an early version of the respiratory decarboxylation cycle. Since that chapter was first written, this cycle has been amended and enlarged, and has acquired a more or less definitive shape known as the "tricarboxylic acid cycle," the ''citric acid cycle," or— most commonly— the "Krebs cycle." It is reproduced in scheme 36.1. The main difference between this scheme and the earlier scheme 9. II is the insertion, between pyruvate and succinate, of a sequence of Ce and C5 tricarboxylic acids. Furthermore, the present cycle differs from the earlier one in the mechanism by which pyruvate is fed into the cycle. In- stead of a simple C3 + C3 condensation (pyruvate + pyruvate) starting a new turn of the wheel, as assumed in 9. II, we now postulate a C4 + C2 condensation (oxalacetate + "active" acetyl, i. e., acetyle + coenzyme A complex). Pyruvic acid now functions, in the steady state, only as source of the acetyl (being converted into it by oxidative decarboxylation, with 1690 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 CHjOH-CHOH-CHO (trio&e) C3 ^ CHa-CO-COOH (pyruvate) C3 CoA COOH-CHj-CO-COGH (oxaloacetate) " *" CHjCO-CoA (acetyl-CoA) OH I COOH-CH^-C-COOH CH2COOH 2H (citrate) COGH-CHrCHOH-COGH (malate) H2G CGOH-CH=CH-CGGH (fumarate) 2H H2O CGOH-CH=C-CGOH CHCGGH (aconitate) Ce H2O— ^ CGGH-CHGH-CH-CGOH I CHj-COOH (iso-citrate) CGGH-CH^-CH.-CGOH (succinate) C, CGOH-CO-CH-COOH I CHrCGGH (oxalosuccinate) Ca 2H C0GH-CH2-CH,-C0-CG0H (a-ketoglutarate) Ca Net reaction : Scheme 36.1. Krebs cycle. Before cycle CHeOs + H2O -* C0H4O2 + CO2 + 4[H] In cycle C2H4O2 + 2H2O -* 2CO2 + 8[H] Total C3H6O3 + 3H2O -* 3CO2 + 12 [H] lipoic acid as hydrogen acceptor) . The fact that pyruvate and oxalacetate are in a reversible carboxylation-decarboxylation equiHbriiim does not alter the fact that the Krebs cycle revolves upon oxalacetate rather than upon pyruvate as a pivot. It is the oxalacetate that takes up the acetyl group, merges with it in citrate, loses CO2 molecules and H atoms to decarboxyl- ases and dehydrogenases, and is regenerated at the end of the cycle. We now turn to the mechanism of carbon dioxide reduction. EVOLUTION OF THE CO2 REDUCTION MECHANISM 1691 Calvin and Benson assumed, when beginning their studies (1947), that carbon dioxide and a hydrogen donor ("reducing power") must be drawn into a cycle similar — but opposite in sense — to the known decarboxylation and dehydrogenation cycles in respiration. At first, when they made no distinction between "photosynthetic" and "respiratory" tracer uptake, the occurrence of tagged succinic, malic, and fumaric acids, and the absence of tagged C5 and Ce acids, led them to suggest that the carboxylation cycle in photosynthesis is the reversal of the short C3-C4 decarboxylation cycle shown in scheme 9. II (Vol. I, page 224) — rather than of the complete Krebs cycle, illustrated by scheme 36.1 above. The early appearance of the tracer in certain amino acids caused them to postulate side reactions of C3 and C4 acids in the cycle (such as pyruvic and oxalacetic acids, whose reduc- tive amination leads to alanine and aspartic acid, respectively). This scheme contained two primary carboxylation reactions — the Wood-Werk- man reaction : (36.10) C*02 + CH3COCOOH , C*00HCH2C0C00H and the Lipmann reaction: (36.11) C*02 + CH3COOH + 2[H1 , CH3C0C*00H + H^O (cf. equations 36.1 and 36.2). Later (1948, 1949), when succinic acid faded out of the picture as a likely intermediate of photosynthesis, and phosphogly eerie acid was found to be the main primary carboxylation product, the specific cycle postulated in 1947 had to be altered. The initial C(14) fixation in carboxyl group of phosphoglyceric acid, and its subsequent penetration into the two other positions (Table 36. V) was taken as evidence that phosphoglyceric acid is the pivot of the anabolic cycle — the role assumed in the catabolic cycle, as represented in scheme 9. II, by pyruvic acid. (It will be noted that enol pyruvic acid is a dehydration product of glyceric acid, and that the two acids have the same reduction level.) In the catabolic cycle, 9. II, one molecule of pyruvic acid accepts a second similar molecule and carries it through a series of reactions, as a result of which it is completely torn apart (all carbon being accepted by decarboxylases and later released as free CO2 gas, and all hydrogen transferred to dehydrogenases and thence, through a series of intermediate catalysts, to oxygen) . The other molecule of pyruvic acid is regenerated at the end of the cycle, so that the same series of reactions can be repeated. Calvin and co-workers thought (1948) that a similar cyclical mechanism, running in the reverse sense, can start with one molecule of phosphoglyceric acid and end with two such molecules, having assembled the second one from carbon dioxide (via one or several carboxylases) and from hydrogen atoms donated by water (via an unknown series of reactions, including at 1692 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 least one photochemical step) . Of the two phosphoglyceric acid molecules present at the end of the reaction cycle, one could be reduced to triose and thence dimerized to hexose (reversal of glycolysis), while the other could be returned into the cycle. The two reaction systems are presented in scheme 36.11, which omit phosphorylations and leave open the specific nature of intermediates in the two cycles. Calvin and co-workers did, however, speculate in some detail on the nature of these intermediates. The first fact relevant to these specu- lations was that after very brief photosynthesis in strong light, when >95% of total tracer taken up was contained in phosphoglyceric acid, practically all of it was located in the carboxyl group of this acid. This made it likely that PGA is a direct product of carboxylation, and since it is a C3 acid, leads to the assumption of a C2 carbon dioxide acceptor. Catabolic cycle }/2 hexose Anabolic cycle -> triose ^nexose <- triose -2[H] -HiO -> 2 pyruvate +3 H2O -3 CO2, -10[H] 1 pyruvate +2[H] 2 glycerate -2 H2O + 3 CO2, +10[H] -^ 1 glycerate Scheme 36.11. The catabolic cycle in respiration and the anabolic cycle in photosynthe- sis, according to Calvin, Benson and co-workers (1948). Retaining, at first, the remainder of the originally postulated cycle, Calvin and Benson (1949) suggested the interpolation between the last C4 acid in this cycle (succinic acid) and pyruvic acid, of the following reac- tion steps : OH CHsC/ acetyl phosphate + 2[H1 * CH3— C— H -H2O OPO3H2 phosphoacetaldehyde + CO2 CH2— CH— COOH (36.12) CH2=CHOP03H2 > \ \ + H2O OH OPO3H2 2-phosphoglyceric acid vinyl phosphate (C2-acceptor) -H^O CH2=C— COOH > I OPO3H2 2-pho-:pho-enol- P3^ruvic acid This sequence was to replace a reversal of Lipmann's reaction (reduc- tive carboxylation of acetate to pyruvate) in the original cycle; the other EVOLUTION OF THE CO2 REDUCTION MECHANISM 1693 carboxylation reaction — pyruvic to oxalacetic acid — was retained. In sup- port of the hypothesis that vinyl phosphate serves as carbon dioxide acceptor in photosynthesis, Calvin quoted the observation that some C(14) can be volatihzed from Chlorella material obtained after about 1 min. of photo- synthesis, in the form of acetaldehyde, by 10 min. hydrolysis with N HCl at 80°C.; the presence of a tagged vinyl ester offered a plausible explana- tion of this observation. Later (1950), Calvin and Benson l^ecame doubtful also of the correct- ness of the oxalacetate-malate-fumarate-succinate segment of the original cycle. The absence from the radiograms of active oxalacetic acid could be attributed to its instability, and the appearance of active aspartic acid (derivable from oxalacetic acid by reductive amination) could be consid- ered as indirect evidence of the occurrence of the latter. The scarcity of active succinic acid was, however, difficult to explain. In some experi- ments, glyceric acid, labelled in all three positions, as well as labelled hexose, have been obtained in complete absence of labelled succinic acid. Further- more it was necessary to account for the early appearance of radioactive glycolic acid (fig. 35.14) and glycine (probably derived from glyoxylic acid) . Malonate, which is known to inhibit the succinic acid-fumaric acid conversion, was found not to interfere with photosynthetic carbon dioxide reduction — not only with the production of active sugar but also with the appearance of active glycolic acid and glycine. (Only the formation of ac- tive malate was totally eliminated by malonate, cf. above, section 6.) This proved that a different path, avoiding the malate-succinate-fumarate series, must exist, leading — it was still assumed — from oxalacetate to acetyl phosphate or another C2 compound. Calvin and co-workers (1950) discussed the possibility that a C*i compound could serve as precursor of the C*2 acceptor (a hypothesis which is a priori implausible because of the poisonous nature of both formaldehyde and formic acid). Experimentally, only traces of labelled formaldehyde or formic acid were found. Assuming complete isotopic equilibration of these traces with the C*02 used, the quantities of C* found in Ci com- pounds corresponded to 2 X IQ-io mole H2CO, and 12 X IQ-'" mole HCOOH in 1 g. of wet cells. In all likelihood these small quantities were artifacts (cf. Vol. I, chapter lO.C). If the C2 acceptor is not formed via a Ci compound, the most likely mechanism of its formation is splitting in two of a C4 compound, as assumed in scheme 9.II. Calvin and co-workers proceeded to discuss the most likely mechanism of this sphtting. With succinic and fumaric acids eliminated as possible intermediates between oxalacetic acid and the C2 compound (their appearance in tagged form being now attributed to respiratory CO2 fixation), and with malic acid also excluded as main line intermediate by malonate inhibition experiments, 1694 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 a mechanism had to be invented by which oxalacetic acid could be con- verted into C2 fragments without being first reduced to malic acid. Glycolic or glyoxylic acid, for example, could be formed by hydrolysis of oxalacetic acid, either directly or via tartaric acid: (36.12a) COOHCH2COCOOH + H2O ( > COOHCHOHCHOHCOOH) oxalacetic acid tartaric acid > COOHCH2OH + COOHCHO glycolic acid glyoxylic acid However, no labelled tartaric acid has been observed in radiograms of products of short-time photosynthesis. Intermediate formation of dihydroxymaleic acid (to be split into two molecules of glyoxylic acid) or of diketo succinic acid (to be split into oxalic acid and glyoxylic acid) also were mentioned as possibilities by Calvin and co-workers. However, the formation of these C4 acids would mean oxida- tion of oxalacetic acid, while the cycle as a whole must be reductive. Another alternative discussed by Calvin et al. (1950) was the reduction of the dicarboxylic C4 acid to dialdehyde level before its splitting into C2 compounds. The cleavage of (diphospho) tartaric dialdehyde into (phos- pho)glycolaldehyde and (phospho)glyoxal is a plausible reaction, bearing resemblance to the splitting of fructose diphosphate by aldolase. The assumption of two primary carboxylations — (36.10) and (36.11) — was retained by Calvin and co-workers at that time for the reasons already explained in sections 6 and 9. The tagging of malic acid immediately at the beginning of the exposure to C*02 in light seemed to call for a more direct mechanism of its equilibration with C*02 in the medium than could be provided by secondary transformations of PGA. Reaction (36.10) of- fered itself as a possibility, even if the reduction of OOA to MA had now to be treated as a side reaction rather than as a step in the main reaction se- quence of photosynthesis. The experiments of Badin and Calvin (1950), mentioned in section 6, which indicated that in weak light, tagged malic acid appears even earlier than tagged PGA, were considered by Calvin et al. as supporting the hypothesis of a "second CO2 acceptor" — a C3 com- pound, as contrasted to the "first CO2 acceptor," which had to be assigned a C2 structure. It was suggested that in strong light the photochemically produced "C2" acceptor is the more abundant of the two ; while in darkness, this acceptor may disappear altogether. The "C3 acceptor," on the other hand, was as- sumed to be regenerated by glycolysis. At the beginning of illumination, C2 acceptor must first be built up to a steady level, which is higher the stronger the illumination. In weak light, the C*02 taken up by the C3 acceptor will be largely stored as tagged malic acid. In the steady state, EVOLUTION OF THE CO2 REDUCTION MECHANISM 1695 whether in weak or in strong light, both carboxylations must proceed at the same rate (more exactly, two CO2 molecules must be taken up by the C2 acceptor for each CO2 molecule taken up by one C3 acceptor, cf. scheme 36.IIIA). However, in very weak light, the establishment of this steady state may require a measurable time, during which the tagging of malic acid will outrun that of all other compounds, including PGA. Another possibility to be taken into consideration is that of tagged malic acid being involved in respiratory, catabolic processes, i. e., serving as a "bridge" be- tween photosynthesis and respiration. We will return to this important possibility when discussing the interaction of respiration and photosynthesis (chapter 37D, section 3), and its possible effect on quantum yield mea- urements (same chapter, section 3). Here, we will continue the evolution of the carbon dioxide reduction mechanism. It was reported in section 6 that the postulate of two different primary carboxylations, both essential for photosynthesis, was opposed by Gaffron, Fager and co-workers, who suggested that even Badin and Calvin's finding of a preferential tagging of malic acid in very weak light can be explained by transformations of the only primary tagged product, PGA (leading to a marked storage of the tracer in a C4 side product when the main sequence proceeds very slowly). The only way in which the C3 carboxylation could run steadily at a rate higher than one half of that of the C2 carboxylation, without causing malic acid to accumulate indefinitely, is for this acid to be drawn into cata- bolic reaction by which it is again decarboxylated. This would be in agree- ment with Calvin and Benson's conception of malic acid as a half-way "bridge" between photosynthesis and respiration. Fager, Rosenberg, and Gaffron (1951) suggested a reaction scheme making use of only one CO2 acceptor. The C2 acceptor was assumed to be regenerated from the final product — hexose sugar — by splitting of the latter into three C2 molecules (which must then be reduced to the level required for an acceptor able to produce glyceric acid by carboxylation) : ( + H2O) (36.13a) C2H4O3 + CO.. , C3H6O4 Carboxylation of d acceptor (glycol?) to PGA (36.13b) C3H6O4 + 2[II] > CsHeOa + H-O Reduction of PGA to triose (36.13c) C3H6O3 > 3^C6Hi206 Dimerization to hexose (36.13d) }4 C6H12O6 > C2H4O2 Dissociation of hexose to biose, glycolaldehyde (36.13e) C2H4O2 + 2[H] > C2H4O3 Reduction of glycolaldehyde to glycol (36.13) CO2 + 4[H] > Ve C^HuO, + H2O 1696 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 The relation of the Chicago model (36.13) to that preferred in Berkeley in 1950 is best illustrated by schemes 36.IIIA and B. Ce < 2 C3 Ce < 3 Ce < 6 Cj (A) 4C3 (B) 2C6 +6C0j -^GC, 4C2 <- ^Q 2C4 Scheme 36. III. Transformation of the carbon chain in the carhoxylation cycles. (A) Calvin, Bassham et al. (1950). (B) Fager, Rosenberg and Gaffron (1950). Calvin and co-workers sustained until 1953 the belief in the "second carhoxylation" because of the kinetic evidence detailed in section 9. While the question of the C3 acceptor (and, more generally, of the role of C4 com- pounds in the reaction sequence of photosynthesis), thus remained in dis- pute, a new important experimental finding was added by the Berkeley group — the identification of sedoheptulose and ribulose phosphates as early tagged products in photosynthesis (described in section 7 above). These observations, and the indications that the C7 and C5 compounds may serve as precursors of the C2 acceptor (rather than as intermediates in the formation of the final products of photosynthesis), led Calvin and co- workers to a new scheme for the transformation of the carbon chain in photosynthesis. This mechanism is shown in scheme 36. IV. In this scheme, the unknown acceptor C2 is regenerated in two successive reactions — one half of it in the reductive splitting of a heptose, producing a pentose, and the other half in a similar splitting of the pentose. The heptose is supposed to be produced by reductive condensation of a triose and a C4 compound (oxalacetate?). The malate still appears as a by-product (and a possible "bridge" to the respiratory system). The essential difference between scheme 36. IV and Calvin's earlier schemes is the replacement of the direct split C4 ^- 2 C2 by a roundabout mechanism C4 + C3 ^ C7 -^ C2 + C5 -* 2 C2 + C3, with a C3 (triose) molecule acting as a catalyst and being regenerated at the end. When Bassham et al. (1954) finally decided — as mentioned in section 6 — to follow Gaff ron et al. in postulating only one carhoxylation, scheme 36.IV had to be changed, and evolved into scheme 36. V. Eliminating one car- hoxylation permitted reduction of the number of hydrogenations from four to one — the reduction of glyceric acid to glyceraldehyde. The rest of the EVOLUTION OF THE CO2 REDUCTION MECHANISM 1697 (pentose) (heptose) Scheme 36.IV. CO2 reduction mechanism with regeneration of C2 ac- ceptor via Ct and C5 sugars (two carboxylations, Ci and C2, four reductions, Ri, R2, R3, R4) (Calvin et al. 1951). reactions in scheme 36.V are disproportionations or condensations of sugars (more precisely, sugar phosphates) without changes in the reduction level : C3 + C4 -* C7; C3 + C7 -^ 2 Cs; Ce + C3 -^ C4 + C5. Malic acid must be derived in scheme 36. V from an as yet unidentified C4 sugar (tetrose). The C2-acceptor formation had been eliminated in passing from scheme 36.IV to scheme 36.V in favor of a direct carboxylation of pentose coupled with splitting into two C3 fragments. (An unstable intermediate Ce com- pound— on the reduction level of a uronic acid — could be postulated here.) The simplification of the scheme to a single carboxylation and a single hydrogenation makes it fundamentally similar to Gaffron, Fager and Rosenberg's scheme 36.IIIB, with the important elaboration of the mecha- nism by which % of the primarily formed triose is converted back into phosphoglyceric acid. 1698 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 + 12H +3C02\C 3C (pentose) Scheme 36.V. Mechanism of carbon dioxide reduction (single carbox- ylation C, single reduction R, all other reactions are disproportionations of sugars) (Bassham et al. 1954). The sugar transformations postulated in this scheme have been made plausible by recent investigations of the chemistry of pentoses and heptoses by Horecker and co-workers (c/., for example, Horecker 1952). 13. The Lipoic Acid Hypothesis Calvin and Barltrop (1952) (c/. Calvin 1952^) suggested that the inhibi- tion of respiration in light — or at least, the failure of fresh photosynthates to be involved in respiration as long as illumination continues (which they deduced from the lack of tagging in tricarboxylic acid in light) — can be ex- plained by the assumption that a catalyst, needed for the respiration cycle, is involved also in photosynthesis, and is kept in the reduced state as long LIPOIC ACID HYPOTHESIS 1699 as photosynthesis is going on, thus preventing it from catalyzing the oxida- tion process. (Another way to state the same fact is to say that the com- petition between the photochemical and the thermal supply of H atoms to the oxidizing catalyst is won by the former in practically every case.) Specifically, Calvin (19522) suggested that this catalyst is lipoic acid (LA), which is needed, in accord with scheme 36.1, as hydrogen acceptor for the conversion of pyruvate into "active acetyl ' ' (acety 1-CoA) . This suggestion formed part of a hypothesis ascribing to lipoid acid a key role in photosyn- thesis (Calvin 1952^; Calvin and Barltrop 1952; Barltrop, Hayes and Calvin 1954). This compound, whose presence in green plants and blue- green algae has been demonstrated by Cayle, Holt, and Punnett (1953), contains a ring of three carbon and two sulfur atoms. Calvin et al. sug- gested that this ring is under such a strain that it can be opened, and lipoic acid thus converted into a biradical (a dithiyl) by an energy quantum of ~40 kcal (i. e., by a single photon of red light), although the standard value for the energy of a S— S bond is (according to Pauling's tables) > 60 kcal. Calvin and Barltrop (1952) argued that the shift of the absorption peak in the series: straight-chain disulfide - 4,8-thioctic acid - 5,8-thioctic acid - 6,8-thioctic (lipoic) acid, from 250 to 330 mju, indicates that the dissocia- tion energy of the ground state is decreased, from the open chain to the five-membered ring, by someting like 10,000 cm.-^, or ~20 kcal. How- ever, this estimate is based on the assumption of an approximately un- changed energy of the excited state — while it seems more likely that the latter changes even more strongly than the former. From stereochemical considerations, Barltrop, Hayes, and Calvin (1954) estimated a strain of about 10 kcal/mole in the 5-membered ring. Postulating that the strain in the five-membered ring is sufficient to reduce the S— S dissociation energy to below 40 kcal, Barltrop and Calvin suggested that the photochemical formation of the dithiyl biradical: -S— -S— is followed by a dark reaction with water: (36.14, <4 > ^_ /— S— /— SH (36.15) < + H2O V_s_ " \— SOH The addition product was assumed to dismute, liberating O2 (perhaps with H2O2 as intermediate) : (36.16) 2 < > < + < I + H2O + V2O2 \— SOH \— SH \— S 1700 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 (Arguments in favor of such a dismutation reaction of a sulfenic acid were presented by Barltrop et al. 1954.) The result of the reaction (36.16) is the utiHzation of two quanta for the oxidation of one molecule of water, and reduction of one molecule of lipoic acid. The normal redox potential of the couple : -S / /— SH (36.17) was assumed by Calvin et al. (1954) to be close to that of the pyridine nucleotides {i. e., about +0.3 volt), so that the standard free energy of reaction (36.18): /— S /— SH (36.18) < I + H2O > < +1^02 ^— S \— SH —which is the sum of (36.14), (36.15) and (36.16)— became, at pH 7: (36.19) ^F = 2 X 23.0 (0.8 + 0.3) = 50 . 6 kcal/mole (where 0.8 volt is the potential of the oxygen electrode at pW 7). It was suggested that the dithiol : ,— SH -SH reduces DPN (or TPN) in the dark, and the reduction of PGA to triose is then achieved by DPNH2 (or TPNH2) with the assistance of ATP, as re- peatedly suggested before. The high energy phosphate, of course, has to be also produced by light, e. g., by an "energy dismutation" of the type repeatedly discussed before (c/. also section B below) . According to Lehninger, three high energy phosphates can be produced by autoxidation of one molecule of DPNH2. The supply of the one ATP molecule needed to reduce PGA would thus increase the quantum require- ment from 2 to 2V3 per two H atoms transferred; if this were all the energy needed for the synthesis of a triose, and if no additional energy (e. g., in the form of more ATP) were required to convert triose to hexose, the overall quantum requirement of photosynthesis would be 5V3- Calvin et al. sug- gested, however, that one additional ATP molecule may be needed to phosphorylate ribulose monophosphate to ribulose diphosphate; this would raise the over-all quantum requirement to 6.0. We will return to these estimates in chapter 37D (section 4e). Here we must point out that the theory of lipoic acid as the key catalyst in photosyn- thesis is at this writing unsupported by evidence. Calvin et al. (1954) found that under special conditions the rate of the Hill reaction (with quinone as oxidant) can be accelerated by lipoic acid. Barltrop, Hayes and Calvin (1954) made interesting photochemical experiments on sensi- TRACER STUDIES OF SPECIAL FORMS 1701 tized photoxidation of model compounds, such as trimethylene disulfide, but as yet the findings do not seem to bear much relation to the suggested function of lipoic acid in photosynthesis. One argument against Calvin's hypothesis that lipoic acid is the "quan- tum acceptor" in photosynthesis is as follows: If it is assumed that the S — S bond in lipoic acid is looser by 20 kcal (or more) tlian the standard S— S l)ond, then the reducing power of the couple (3G.17) should be correspond- ingly weaker (since there is no reason why S — H bonds in the dithiol : -SH -SH should not have the normal strength). Assuming approximate parallelism of free energies and total energies of reduction for the different disulfides, the redox potential of the couple (36.17) should be V2 X 20:23 or about 0.45 volt less positive than that of a similar "standard" system (such as cystine/cysteine). The normal potential of the latter pair is about +0.35 volt (close to that of pyridine nucleotides); that of the couple (36.17) should then be negative, and thus quite insufficient to reduce TPN or DPN. 14. Tracer Studies of Special Forms of Photosynthesis The observation of Tobert and Zill (1954) on the COa fixation and tracer distri- bution in squeezed-out material from the giant Chara and Nitella cells were described in chapter 35 (p. 1536). In the same chapter we briefly reported also the findings of Arnon, Bell and Whatley (1954, cf. p. 1615) of the fixation of C*02and ATP '^ formation by whole chloroplasts, and the apparently competitive character of these two processes. Some carbon tracer measurements have been made with hydrogen adapted algae Gaffron, Fager and Rosenberg (1951) "stabilized" adapted Scenedesnms cells with phthio- col {cf. chapter 6) to be able to observe photoreduction in relatively strong light. They noted that under these conditions, the scattering of C* over the three fraction A, B, C {cf. section 4) was almost as rapid as in true photosynthesis. The tagged primary products thus could be converted into various metabolites, including fats and proteins under strictly anaerobic conditions, i. e., without any help of respiratory energy. Badin and Calvin (1950) made similar experiments, but did not inhibit "de-adaptation"; they had therefore to work at very low hght intensities and use very long exposures; the same kind of tagged compounds was found after photoreduction as after prolonged photosynthesis in weak light. Comparatively Uttle C* was found, by Badin and Calvin (1950), to be fixed in the oxyhydrogen-carbon dioxide reaction of hydrogen adapted Scenedesmus {cf. chapter 6, section 3); apparently none of the tracer passed into the insoluble fraction (polysaccha- rides, proteins). Benson (1950) stated that purple bacteria {Rhodospirillum rubrum) produced a greater variety of tagged compounds in 5 min. photoreduction in H2 -1- C*02 than did barley leaves in equal period of photosjmthesis. Unicellular algae were found to stand midway between higher plants and bacteria in respect to the complexity of tagged prod- ucts. Benson also noted that Rhodospirillum produced no sucrose, but formed a poly- saccharide, "probably starch." 1702 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 This may be the place to mention the study, by means of the C(14) tracer, of the de-acidification of succulents in Ught— a phenomenon described in chapter 10 (section D2). The two alternative interpretations suggested there, were: (1) "photosynthesis" with malic (or citric) acid as "substitute oxidant" (replacing CO2 as hydrogen acceptor), and (2) photoxidation of these acids to carbon dioxide, followed by normal photosynthe- sis. Tracer experiments could perhaps permit a choice between these two hypotheses. Thurlow and Bonner (1948) and Varner and Burrell (1950) found that, in darkness, acidification in tagged C*02 produced, in Bryophyllum calycinum, malic acid tagged preferentially in the carboxyl group, indicating its probable formation by the Wood- Werkman reaction (i. e., carboxylation of pyruvic to oxalacetic acid followed by reduction of the latter to mahc acid. It will be recalled that the same mechanism has been discussed — but more recently abandoned — by Calvin and co-workers also as explanation of the tagging of malic acid in photosynthesis.) In fight, the hexoses which are formed in the succulent when malic acid disappears, were tagged preferentially in the 3,4 position (as in ordinary photosynthesis, cf. table 36. VII). This indicates that the C* dicarboxyfic acid is either photoxidized completely to CO2, or, at least, decarboxylated to a C3 acid (with the remaining carboxyl still tagged) before it is used for the synthesis of sugars by the reduction of the C3 acid to triose and head-on condensation of two triose molecules to a hexose (leading to the accumulation of labelled carbon in the two middle atoms of the Ce chain). If mafic acid were to take direct part in Ce synthesis, e. g., in the way postulated by Benson, Calvin et al. in 1949-1950 {i. e., spfitting of the C4 chain into two C2 fragments, and their carboxylation), this would lead to a C3 acid with labelled carbon in the /3- position, rather than in the carbo.xyl; reduction of the product to triose and condensation of the latter to hexose would then place the labels on carbon atoms 2 and 3 in the Ce chain. Stutz (1950) exposed Bryophyllum calycinum to labelled carbon dioxide for several hours in light and chromatographed the synthesized organic acids on a column ; activity was found in succinic, oxafic, malic, citric, and iso-citric acid. After 12 hours in light and 12 hours in darkness, almost 90% of C* was found in mafic acid, with only 58% in iso-citric acid and 4% in citric acid (although the absolute amount of iso-citric acid present was larger than that of malic acid). Similar preferential labelling of mafic acid was noted after V2, 2 or 4 hours in fight without a subsequent dark period. When Cali- cynum leaves, exposed to C*02for 15 min. in light, were left in darkness (without C*02) for 4 hours (to observe the shift in the relative concentration of the different acids, cf. p. 269), the specific activity of all acids was found to increase (obviously at the cost of some nonacid C* reservoir formed in light), despite a strong increase in the absolute quantity of all acids. (Mafic acid increased, in the dark period, by a factor of 5 in abso- lute amount and a factor of 2 in specific activity, citric acid by a factor of 3 in quantity and 6 in specific activity, iso-citric acid by a factor of 2.8 in quantity and 2.4 in specific activity.) Similar experiments were carried out with tobacco leaves (which also store malic and citric acid, cf. p. 264). There, too, C* was fixed predominantly in malic acid. It was shown that 100 mg. quantities of heavily labelled mafic and citric acid can be pre- pared by growing Bryophyllum or tobacco seedlings in C*( 14)02. B. Photosynthesis and Phosphate Metabolism* As described in part A, most reactions in the reduction of carbon dioxide in photosynthesis involve phosphate esters rather than free organic com- * Bibliography, page 1712. PHOTOSYNTHESIS AND PHOSPHATE METABOLISM 1703 pounds. Phosphoglyceric acid, phosphopyruvic acid, phosphoglycolic acid, triose phosphates, pentose phosphate, hexose phosphates, heptose phosphate and sucrose phosphate have all been found among the early tagged products of photosynthesis. (For a review of phosphates identified in tracer experiments with C(14) and P(32), see Buchanan et al. 1952.) Some of the observed phosphate esters contain one, some two phos- phoric acid residues; none of them — except phosphoenolpyruvate, whose function in photosynthesis is not clear^ — are true "high energy phos- phates." If it is true, however, that the main reduction step in photosyn- thesis is (as postulated in the "one-carboxylation, one-reduction" mecha- nism in section A, 12) the reduction of PGacid to PGaldehyde by reduced pyridine nucleotide, then the cooperation of a high-energy phosphate is indispensable; i. e., PGA has to be phosphorylated to DPGA (with one H2PO3 residue attached to the carboxyl group, forming a "high energy" ester) before it can be reduced. It has been suggested that the needed high-energy phosphate (ATP) is produced by partial reoxidation of an intermediate, such as TPNII2. If this is correct (and as of this writing it appears a plausible hypothesis), then the beginning of photosynthesis in a cell should lead to an increase in the concentration of high energy phos- phate (and a corresponding consumption of orthophosphate, or of a "low- energy" phosphate ester). The steady progress of photosynthesis beyond the reduction level of PGA will require a certain steady concentration, [ATP], which must be higher the higher the light intensity (at least up to saturation). In the dark, this concentration will decline again as the cell uses up its energy reserves (c/. Lynen 1941, 1942). In chapter 9, section 5, we reported some experiments by Emerson, Stauffer and Umbreit (1944) indicating an effect of photosynthesis on the phosphate household of Chlorella. (A more direct, but quantitatively not very convincing evidence of storage of energy in high-energy phosphates was obtained by Vogler and Umbreit (1943) with chemosynthetic bacteria, c/. p. 114.) More recently, more significant evidence has been supplied. Wassink, Tjia, and Wintermans (1949) observed shifts in the phosphate content of a medium containing the purple bacterium Chromatium D upon transfer from light to darkness. The uptake of inorganic phosphate oc- curred in light in N2, H2, and — to a smaller extent — in N2 -f- CO2 ; shift to darkness caused a small release of phosphate if the gas phase remained unchanged, and a marked release if No + CO2 was substituted for H2. These results were interpreted as supporting the hypothesis of Vogler and Umbreit that, in the energy-storing period (oxidation of sulfur in chemo- synthetic bacteria, illumination in Ho in photosynthetic bacteria), high- energy phosphate bonds are built up, while in the energy-utilizing period (CO2 supply in the absence of O2 in chemosynthetic bacteria, CO2 supply 1704 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 in darkness in photosynthetic bacteria) the energy of these bonds is used up for the fixation and reduction of carbon dioxide. Wassink, Wintermans and Tjia (19510 made similar experiments with Chlorella. In addition to measuring the inorganic phosphate in the me- dium, the amount of phosphate in the cells extractable with trichloroacetic acid (TCA) also was determined. In the absence of carbon dioxide, phos- phate was taken up from the medium in light, and up to 30% of the TCA- soluble phosphorus in cells was converted into TCA-insoluble form. The shifts were much smaller in the presence of carbon dioxide. Wassink, Wintermans and Tjia (1951*) found that glucose had the same effect on the phosphate transformation in Chlorella as carbon dioxide. Wintermans and Tjia (1951) described the liberation of the largest part of the phosphate, made insoluble in TCA by illumination in the absence of carbon dioxide, by hydrolysis in 1 N HCl. The (much smaller) amount of TCA-insoluble phosphate, synthesized in the presence of carbon dioxide, is divided about equally between a labile (i. e., HCl hydrolyzable) fraction and a fraction that is stable in 1 A^ HCl at 100° C. Kandler (1950) also made experiments on the "phosphate level" of Chlorella pijrenoidosa in darkness and in light. He determined the diffusi- ble "inorganic" phosphate and the "TCA-soluble" organic phosphate. The former declined by 20% in the first V2 min. of exposure to light, rose to a peak at 1 min., and, after a second shallow minimum at about 3 min., set- tled to a constant level, about 10% below that in darkness. Upon switch- ing the light off, the level of TCA-soluble phosphate rose in about 2 min., to a maximum about 30% above the steady value in light, then declined again and assumed a constant level about 10% above that in light. These "transients" show remarkable similarity to those observed in gas exchange and chlorophyll fluorescence (chapter 33) and C* incorporation in certain intermediates (this chapter, section 9). Kandler interpreted these variations in phosphorus content as evidence of the participation in photosynthesis of high-energy phosphate (ATP) and discussed two alternative mechanisms for its formation and utilization in the photochemical process (c/. below). Simonis and Grube (1952) and Grube (1953) took up the study, first attempted by Aronoff and Calvin (1948), of the incorporation of P(32) tracer in different fractions of green cells in darkness and in light. Aronoff and Calvin could find no uptake of P in the TCA-soluble fraction of Chlorella in light (which would be a sign of increased concentration of or- ganic phosphates, such as ATP). Kamen, Gest and Spiegelmann (cf. Gest and Kamen 1948, Kamen and Spiegelmann 1948) observed, in light, mainly a change in the TCA-insoluble P* fraction; the effect was de- creased by HCN poisoning. Simonis and Grube resumed this study using PHOTOSYNTHESIS AND PHOSPHATE METABOLISM 1705 Elodea densa leaves instead of unicellular algae. They suggested that leaf pieces can be freed from adsorbed radiophosphorus, by rapid rinsing before killing, much more effectively than unicellular algae. After killing in 10% TCA at o° C. the leaA'es were extracted three times with TCA, and the P* activity determined in the total extract; the inorganic orthophosphate (including perhaps some phosphate from very easily hydrolyzaV)le organic esters) was precipitated by magnesium, with inactive orthophosphate as carrier, and its activity measured; so was the P* activity of the remaining TCA-soluble organic material and of the TCA-insoluble residue. T 20 T 1 1 r 40 60 80 100 LIOMT INTENSITY (VOLTS) 120 -I 140 Fig. 36.28. The effect of light on the steady ATP level in Chtorella (in the presence of CO2, after 2-3 hours of anaerobic incubation). Measured by the fiiefly extract luminescence method (after Strehler 1952). The experiments indicated a decrease in inorganic orthophosphate and an increase in TCA-soluble organic phosphate in light; the effect was stronger in the presence than in the absence of carbon dioxide (but was definitely present also in the CO2 free system). The change could be ob- served after 1-10 min. of illumination. An increase in TCA-insoluble^P* (observed before by Kamen and Spiegelmann) was noted only after about 1 hour. The P* uptake in TCA-soluble fraction increased at first (and that in an organic phosphate decreased) with light intensity, but reached a constant level at about 2.5 klux. Strehler and Totter (1952) and Strehler (1953) (cf. also Strehler 1952) applied to the problem of phosphate metabolism in photosynthesis the sen- sitive techni(iue (discovered by McElroy) based on the stimulation of chemiluminescence of firefly extracts by ATP. They estimated that in the steady state of photosynthesis in Chlorella about one molecule ATP is added to the cellular pool for every six molecules of liberated oxygen. If one assumes that one (or two) ATP molecules are degraded to ADP to make the liberation of one molecule of oxygen possible, the observed net 1706 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 increase in ATP indicates an excess of 7.5% (or 15%) of the photochemi- cally produced over the simultaneously consumed ATP. Figure 36.28 illustrates the relation of the ATP concentration to light intensity. Contrary to expectation, this concentration has a maximum at a certain relatively low light intensity; at the higher light intensities it settles to a steady level, which is still markedly higher than in darkness. A further peculiarity is illustrated by fig. 36.29: the ATP \'alue increases in the dark immediately after the cessation of illumination. This concen- tration shows a characteristic transient fluctuation also at the beginning of the illumination period (fig. 36.30) which should be considered in conjunc- tion with the induction curves of gas exchange and fluorescence, described in chapter 33, and with the transients in the concentration of tagged com- pounds, observed by the Berkeley group (figs. 36.22 and 36.23). ■ 10 I MINUTES Fig. 36.29. Changes in ATP level in Chlorella duiing and after 10 min. il- lumination in the absence of CO2 (after Strehler 1952). The concentration of AtP in Chlorella can be enhanced also })V the ad- mission of oxygen to anaerobically incubated cells (i. e., by stimulation of respiration). The photochemical enhancement is much less sensitive to a lowering of temperature than the respiratory enhancement, which seems to indicate a direct relation of ATP formation to the primary photochemical act, rather than to the enzymatic stages of photosynthesis. Strehler interpreted these results as supporting the hypothesis that for- mation and utilization of ATP is a part of photosynthesis. Rather than thinking of ATP as merely an "assistant" in the reduction of RCOOH to RCHO, he suggested that direct photochemical hydrogen transfer from water occurs only to an acceptor with a potential not highei- than 0.0 volt (cf. chapter 35, section B4 for evidence which could be adduced in support of this postulate), and that the further enhancement of reducing power ("energy dismutation") is brought about by reoxidation of some of the reduced products and storage of their oxidation energy as phosphate bond energy. It may be useful to juxtapose here the different variants of the theory of PHOTOSYNTHESIS AND PHOSPHATE METABOLISM 1707 phosphate bond energy storage in photosynthesis. The "extreme" variant (suggested, e.g., in the paper by Emerson, Stauffer and Umbreit 1944) is that all light energy utiHzed in photosynthesis is first converted to phos- phate bond energy; our objections, based on the undesirabiHty of sphtting 43 kcal energy quanta into 10 kcal portions before accumulating them again (chapter 9, p. 228), were directed against this extreme theory. An "in- termediate" suggestion is that one (luantum of light is used to lift hydrogen from water (Eq ^ 0.8 volt) to an intermediate redox catalyst with a poten- tial around 0.0 \olt, and that high energy phosphates (created by partial 90^ 60 120 leo TIME OF ILLUMINATION (SEC) Fig. 36.30. Transient changes in ATP eoncentration (E) at the begin- ning of illumination of Chlorella, compared with those of fluorescence (A), chemiluminescence (B), and C02-uptake (D) (after Strehler 1952). reoxidation of this intermediate) take over from there (Kandler 1950, Strehler 1952). Finally, there is the "modest" suggestion (Ruben 1943, Calvin et al. 1954), which assigns to phosphate energy only the bridging of the final gap between Eo — 0.3 volt (pyridine nucleotides) and the energy needed to reduce a carboxyl to a carbonyl {Eo ^ 0.5 volt). All the above-discussed hypotheses suggested partial reoxidation of intermediates as source of ATP energy. Burk and Warburg postulated in- stead (c/. chapter 37D, section 3), partial reoxidation of the final products of photosynthesis (carbohydrates) . To complete the review of alternatives, the oxidant in the back reaction need not l^e molecular oxygen— as postu- lated by most of the above-enumerated authors— but could be an inter- mediate oxidation product ("oxygen precursor") formed by the photo- chemical process. (This assumption becomes particularly plausible if one wants to apply the same general concept to photosynthesizing bacteria and 1708 CHEMICAL PATH OF CARBON DIOXIDE KEDUCTIOX CHAP. 36 "hydrogen adapted" algae, which do not produce free oxygen, but ne\er- theless reduce CO2 to the carbohydrate level.) By encompassing all these alternatives, one is brought to the most general picture — that of "energy dismutation" (chapter 7, section 6 and chapter 9, section 7) by back reac- tions between intermediate or final oxidation products and intermediate or final reduction products of photosynthesis, with the one specific sugges- tion that the energy of these back reactions is temporarily stored as phos- phate bond energy. Energy storage in phosphate bonds is, however, not the only conceivable mechanism of energy dismutation; and though the available experimental indications of the participation of ATP in photo- synthesis are suggestive, they are not yet conclusive. Table 36.IX Relative Size of Phosphate Reservoirs in the Ste.\dy State of Photosynthesis IN Scenedesmus (after Goodman et al. 1953) Compound Phosphoglj'cerate Hexose monophosphate. . Uridine diphosphoglucose . Sugar diphosphates ..... ADP ATP 1 hr. light 1 + hr. lifiht 1.5 iiiin dark 1 hr dark 1 T hr. dark 15 niin. light 33 AO 29 39 7 6 7 / 14 11 7 10 19 14 17 9 9 10 7 10 18 30 28 25 Goodman, Bradley, and Calvin (1953) initiated a systematic quantita- tive study of the incorporation of P(32) tracer in tlifferent compounds by dark and photochemical metabolism of Scenedesmus obliquus, using paper chromatography to separate and identify the labelled compounds. After about 30 sec. exposure to radiophosphate in the dark, 72% of incorporated P* were found in ATP and 9% in ADP, with 11% in sugar diphosphates, 5% in the hexose monophosphate area, and 3% in PGA. After 30 sec. in light, the P* distribution was much more uniform (39% in sugar diphos- phates, 19% in hexose monophosphates, 179c i'l PGA, 15% in ADP, and 10% in ATP). This means that even 30 sec. is too long a time, in strong light, to identify the port of entry of P*. The early P* labelling of PGA in light suggests a close relation between photosynthesis and the organic binding of mineral phosphate (which is indicated also by the above-de- scribed analytical experiments). The considerable labelling of ADP is worth noting, because the terminal phosphate in ADP is not usually con- sidered as contributing to metabolic transformations. By longer exposures to P*04~, enough labelled compounds could be NITRATE METABOLISM AND PHOT(3SYXTHESIS 1709 obtained to permit more detailed fractionation of the tagged compounds. The P* distribution between glucose-()-phosphate, fructose-6-phosphate, and mannose-6(sedoheptulose)-phosphate corresponded to the equilibrium concentrations of these three esters, after exposures of from 1 to 25 min. in dark or in light. Still longer experiments (1 hour exposure), permitted a first estimation of the volumes of the ^•arious phosphate reservoirs in the cell (on the assumption that after this time, a steady state has been reached in respect to these volumes, and all reservoirs have been uniformly la- belled). Table 36.IX shows the relative reservoir volumes computed on this basis. It is interesting to note the smaller proportion of P contained in ATP in light (compensated by a greater proportion in sugar phosphates). Frenkel (1954) described the formation of ATP from ADP + ortho- phosphate in light by sonically disintegrated RhodospiriUum rubrum. Arnon et al. made similar observations on whole chloroplasts. At first (Arnon, Bell and What ley 1954) oxygen was reported to interfere with "photosynthetic phosphorylation"; but later (Arnon, Whatley and Bell 1954) it was reported to occur also under aerolMc conditions. Mg + + ions, the vitamins C (ascorbic acid) and K, and certain other compounds, were found to act as "co-factors," stimulating the ATP formation. C. Nitrate Metabolism and Photosynthesis* This chapter would be the place to discuss also the relation of photo- synthesis to the nitrate metabolism of plants. It, too, is affected by illu- mination, and affects photosynthesis, e. g., by changing the photosynthetic ratio ACO2/AO2 from about 1.0 in solutions containing no nitrogen or only ammonia nitrogen, to much lower values in solutions containing nitrate, in consequence of partial substitution of HN0,3 for CO2 as ultimate hydro- gen acceptor. Early experiments on the photochemical reduction of nitrate by ChloreUa by Warburg and Negelein have been described in chapter 19, section Bl. It was stated there that "unfortunately, the sub- ject has received no further attention since 1920." Since this was written, some new studies have appeared in this field. Space limitations prevent us from entering here into their description; we can only refer to the series of papers by Burstrom (1942-1945), Myers and co-workers (1948, 1949), Davis (1952) and Kessler (1953), and to the discussion by Kandler (1950). The radiocarbon tracer experiments, described in part A of this chapter, have indicated how early the reduction of carbon dioxide in photosynthesis can branch out into various side reactions, including transaminations leading to simple aminoacids (and perhaps from there to proteins) before the reduction stage of the carbohydrates has been reached. At some stage * Bibliography, page 1713. V 1710 CHEMICAL PATH OF CARBON DIOXIDE REDUCTION CHAP. 36 (or stages) of the reduction mechanism of photosynthesis, HNOs may be- come substituted for CO2 as hydrogen acceptor whenever photosynthesis is carried out in a nitrate-containing medium. Various possibiHties of this coupHng can be envisaged, such as nitrate serving as Hill oxidant in com- petition with carbon dioxide (as oxygen competes with quinone according to Mehler, c/. chapter 35, section B4). Another possibility is that of nitrate substituting for oxygen as oxidant in the often postulated partial reversal of the primary photochemical process. Bibliography to Chapter 36 Chemical Path of Carbon Dioxide Reduction A. Isotopic Tracer Studies 1939 Ruben, S., Hassid, W. Z., and Kamen, M. D., /. Am. Chem. Soc, 61, 661. 1940 Ruben, S., Kamen, M. D., anrl Hassid, W. Z., ibid., 62, 3443. Ruben, S., Kamen, M. D., and Perry, L. H., ibid., 62, 3450. Ruben, S., and Kamen, M. D., ibid., 62, 3451. 1941 Frenkel, A. W., Plant Physiol, 16, 654. 1947 Benson, A. A., and Calvin, M., Science, 105, 648. Aronoff, S., Barker, H. A., and Calvin, M., J. Biol. Chem., 169, 459. Aronoff, S., Benson, A. A., Hassid, W. Z., and Calvin M., Science, 105, 664. Allen, M. B., Gest, H., and Kamen, M. D., Arch. Biochem.. 14, 335. Wassink, E. C, Antonie van Leeuwenhoek, 12, 281. Calvin, M., and Benson, A. A., Science, 107, 476. Fink, R. M., and Fink, K., Science, 107, 253. 1948 Stepka, W., Benson, A. A., and Calvin, M., Science, 108, 304. Brown, A. H., Fager, E. W., and Gaffron, H.. Arch. Biochem., 19, 407. Benson, A. A., and Calvin, M., Proc. Cold Spring Harbor Symp. Quant. Biol, 13, 6. Thurlow, J., and Bonner, J., Arch. Biochem., 19, 509. Kamen, M. D., in PliotosyntJiesis in Plants. Iowa State College Press, Ames, la., 1949, p. 365. Brown, A. H., Fager, E. W. and Gaffron, H., ibid., p. 403. Fager, E. W., ibid., p. 423. Benson, A. A., Calvin, M., Haas, V. A., Aronoff, S., Hall, A. G., Bas- sham, J. A., and Weigl, J. W., ibid., p. 381. Weigl, J. W., and Calvin, M., .7. Chem. Phys.. 17, 210. Calvin, M., and Benson, A. A., Science, 109, 140. 1949 Gibbs, M., /. Biol. Chem., 179, 499. 1950 Benson, A. A., Bassham, J. A., Calvin, ^M., Goodale, T. C, Haas, V, A., and Stepka, W., /. Am. Chem. Soc, 72, 1710. Benson, A. A., and Calvin, M., J. Exptl. Botany, 1, 63. Bassham, J. A., Benson, A. A., and Calvin, M., J. Biol. Chem., 185, 781. Calvin, M., Bassham, J. A., and Benson, A. A., Federation Proc, 9, 524. Badin, E. F., and Calvin, M., /. Am. Chem. Soc, 72, 5266. BIBLIOGRAPHY TO CHAPTER 36 1711 Shou, L., Benson, A. A., Bassham, J. A., and Calvin, M., Physiol. Plan- tarum, 3, 487. Stutz, R. E., CO2 Assimilation in Biological Systems, Brookhaven Confer- ence Report, Assoc. Univ. Inc., Upton, N. Y., pp. 77-96. Benson, A. A., ibid., pp. 119-138. Gibbs, M., ibid., pp. 139-145. Fager, E. W., Rosenberg, J. L., and Gaffron, H., Federation Proc, 9, .525. Eager, E. W., and Rosenberg, J. L., Science, 112, 617. Benson, A. A., and Calvin, M., An7i. Rev. Plant Physiol., 1, 25. Clendenning, K. A., Arch. Biochem., 27, 75. Aronoff, S., and Vernon, L., ibid., 27, 239. Varner, J. E., and Burrell, R. C, ibid., 25, 280. Steward, F. C, and Thompson, J. F., Nature, 166, .593. 1951 Calvin, IVI., Bassham, .J. A., Benson, A. A., Lynch, V. H., Ouellet, C, Schou, L., Stepka, W., and Tolbert, N. E., Symposia Soc. Exptl. Biol., 5, 284. Gaffron, H., Fager, E. W., and Rosenberg, J. L., ibid., 5, 262. Gaffron, H., and Eager, E. W., Ann. Rev. Plant Physiol, 2, 87. Hassid, W. Z., in Phosphorus Metabolism. Vol. I, John Hopkins Univer- sity Press, Baltimore, pp. 11-66. Horecker, B. L., ibid., pp. 117-144. Nezgovorova, L. A., Compt. rend. (DoMady) acad. sci. USSR, 79, 537. Benson, A. A., ./. Am. Chem.. Soc, 73, 2971. Benson, A. A., Bassham, J. A., and Calvin, M., ibid., 73, 2970. Weigl, J. W., Warrington, P. M., and Calvin, M., ibid., 73, .5058. Utter, M. F., and Wood, H, G., in Advances in Enzymology. Vol. 12, In- terscience, New York-London, p. 41. Stepka, W., Thesis, ITniv. of California. 1952 Benson, A. A., Bassham, J. A., Calvin, M., Hall, A. G., Hirsch, H. E., Kawaguchi, S., Lynch, V. H., and Tolbert, N. E., /. BioJ. Chem., 196, 703. Benson, A. A., Kawaguchi, S., Hayes, P. AL, and Calvin, AI., ./. Am. Chem. Soc, 74, 4477. Calvin, M., Harvey Lectures, Ser. 46, pp. 218-251. Calvin, M., Harrison Hoive Lecture, UCRL, Report iVo. 2040. Calvin, M., and Massini, P., Experientia, VIII 12, 445. Calvin, M., Bas.sham, J. A., Benson, A. A., and ]\Ia.ssini, P., Ann. Rev. Phys. Chem., 3, 2\5. Calvin, AL, and Barltrop, J. A., /. Am. Chem. Soc, 74, 61.53. Benson, A. A., Z. Elektrochem., 56, 848. Clendenning, K. A., and Gorham, P. R., Arch. Biochem. and Biophys.. 37, 56. Fager, E. W., and Rosenberg, J. L.. Arch. Biochem. and Biophys., 37, 1. Horecker, B. L., /. Cellular Comp. Physiol., 41, Supplement 1, pp. 137- 164. Arnon, D. I., Science, 116, 635. Nezgovorova, L. A., Compt. rend. (Doklady) acad. .j. Fig. 37A.11. Cross section through Mougeotia chloroplast, fixed for 45 min. in 1% OSO4, showing disintegration into band-shaped laminae (after Steinmann 1952^). The much larger, also thin, non-folded structures, usually round or o\'al, but occasionally streaky or fibrous (Vatter 1952, and Thomas et al. 1952, cf. fig. 37A.9), were interpreted by Frey-Wyssling as dried-out residues of lipide drops. Most of these large myelin "pancakes" lie free on the film, but some seem to protrude from grana, like ham slices from a sandwich. This supports Frey-Wyssling's suggestion that some "myelin" material origi- ELECTRON MICROSCOPY 1 / '2'^ nates in the grana. However, estimates of the total amount of Hpide in grana and in the stroma (cf. below) convinced Frey-Wyssling that this can- not be true of all myelin; rather, a hirge part of it must come from the stroma. The above-mentioned small granules, embedded in the "myelin" masses, or scattered on the supporting film, are, according to Frey- Wyssling, globular protein molecules, originally associated with the lipides in the lipoproteinaceous stroma. The grana and the discs, into which they disintegrate, appeared as the most striking — and potentially significant — structures in the photosyn- thetic apparatus. It was therefore important to establish whether they occur in all photosynthesizing organisms. Fig. 'MAA2. Irregularly shaped patches obtamed by sonic disinte- gration of himinar chloroplasts of Spirogyra, fixed for 15 min. in 1% Os04 {cf. Figure 37A.13) (after Steinmann 1952'). Thomas (1952) published electron micrographs .showing the presence of grana in higher plants {Spinacia oleracea), green algae {Chlorella vulgaris), blue-green algae {Synochoccus spp.), red algae (Porphyridium cruenium), diatoms {Niizchia), purple bacteria {Chromatium D, Rhodo spirillum ruh- rum), and green sulfur bacteria. The diameter of the grana reproduced in his paper ranges from 1.3 m {Nitzschia), through 0.7 m (Porphyridium), 0.6 M {Chlorella, Syncechococcus), and 0.4 m {Spinacia}, to about 0.15 ^ in bacteria. Another set of electron micrographs of material obtained from the same organisms shows thinner, round bodies of the same diameter. These are interpreted by Thomas as single protein discs from disintegrated grana ; however, on some photographs (in particular, those of purple and green bacteria), the distinction in thickness between the "grana" and the "discs" is not very great, at least in the reproductions. Colorless bacteria showed no grana, but small grana were found in the colorless alga, Proiotheca zopfii. 1726 CHLOROPLASTS, CHROMOPLASTS AND CHROMATOPLASM CHAP. 37A These electron microscope observations were supported by ultracentri- fuge studies (to be described in section 3), also indicating the presence of grana in blue-green algae and purple bacteria. However, just when one may have become inclined to postulate that the grana are universal (and therefore probably indispensable) elements of the photosynthetic appara- tus, complications appeared. Steinmann (1952') found in the large, spiral band-shaped chloroplasts of the algae Spirogyra and Mougeofia, fixed in 1% Os04, a lamellar struc- ture extending through the whole chloroplast, without evidence of grana. Fig. 37A.13. A circular lamella ("disc") from the granum of a tulip chlorplast, fixed 25 min. in 1% OSO4 and disintegrated by sonic oscillations (after Steinmann 19522). Note granulated surface. Compare with Fig. 37A.12. Fig. 37A.14. Cross section of u tulip chluroplast fixed for 15 niin. in 1% OsO^, showing laminated grana imbedded in granulated stroma (after Steinmann 19522). The lamellar structure is clearly A^isible on longitudinal sections of such fixed Spirogyra chloroplasts (fig. 37A.10). Mougeotia chloroplasts easily dis- integrated into long, narrow bands (fig. 37A.11). By sonic vibrations, these bands could be further disrupted into irregular-shaped patches (fig. 37A.12), the thinnest of which were about 7 m/x thick, i. e., of the same thickness as the "discs" in the grana. Albertson and Ley on (1954) reproduced a very interesting electron ELECTRON MICROSCOPY 1727 micrograph of a fixed and sectioned Chlorella cell, showing that the ciip- shaped chloroplast consists of 4-6 dark shells, each about 35 m,x thick, separated by lighter layers of about the same thickness. Each dark shell seems, in its turn, to consist of several (usually four) thinner shells each about 5 m/x thick. No grana are visible in the picture. On the other hand, stacks of disc-shaped lamellae (of the type of those shown in fig. 37A.8) were obtained, in the same study by Steinmann, from ca. 5000 i ca.lOOA Fig. 37 A. 15. Idealized cross section of a granum disc, consisting of two monolayers of macromolecules (after Frey-Wyssling and Steinmann 1953J. As observed under the electron microscope, the discs are only 50-70 A. thick because of shrinkage. Aspidistra chloroplasts; they contained up to 40 thin discs of uniform diam- eter. Figures 37A.18a and 37 A. 18b show the cross section of lamellae in fixed and sectioned chloroplasts of a higher plant according to Palade (1953). Steinmann drew attention to the occurrence of lamelhir structure in visual rods, and suggested that this similarity may have something to do with the common purpose of the two cellular structures— utilization of light energy. In another note, Steinmann (1952^) showed that single thin round protein discs can be obtained from granular chloroplasts {e. g., those of tulip, or Aspidistra) by sonic disintegration (fig. 37 A. 13). Electron micro- graphs of sections through fixed tulip chloroplasts clearly showed that such discs are the structural elements of intact grana. Incidentally, these elec- tromicrographs— c/. fig. 37A.14— also confirmed the layered arrangement of the grana in each chloroplast, first noted by Heitz on ultraviolet micro- graphs (cf. fig. 39c in Vol. I). No carrier lamellae which support the grana in a layer, accordnig to Strugger (1951), are visible on fig. 37A.13. Leyon (1954') noted, how- ever, that in Aspidistra elatior, the lamellae within the grana extend from it into the surrounding stroma, the grana being merely regions of accentu- ated lamination. Frey-Wyssling and Steinmann (1953) suggested that the discs (thick- ness, 7-10 mM) obtained by disintegration of grana from the chloroplasts of 172S CHLOROPLASTS, CHROMOPLASTS AND CHROMATOPLASM CHAP. 37A Aspidistra, consist of only two layers of macromolecules, as represented schematically in fig. 37A.15. (The same structure can be suggested also for the bands or shells of approximately the same thickness, observed in Mougeotia and Chlorella, respectively'. ) This conclusion was drawn from the observation of the transformation of the grana, suspended in 1 M sucrose solution, caused by dilution with water. Under the phase contrast microscope the grana could be seen swell- ing, flaking off, and growing into long strands; the fluorescence microscope Fig. o7.\..ItJ. Cluuiitf ol culluiJsL'tl \y.ig^, ubUauL'tl Ironi a graiium of Aspidrntia by treatment in distilled water (after Frey-Wyssling and Steinmann 1953). Fixed in 1% Os04, chromium shadowed. It is suggested that this is a "monej^ roll" (Fig. 37A.8) in which each protein "coin" has swelled up by penetration of water between two mono- laj-ers (Fig. 37A.15), precipitation membrane has formed on the outside, and the bags have collapsed. showed that the latter still carried the chlorophyll. Under the electron microscope the strands appeared to consist of round, collapsed, and folded bags, about 1 n in diameter (fig. 37 A. 16). These bags were interpreted as the protein "discs," similar to those shown in fig. 37A.8, which had swollen into bubbles in water and then collapsed. The growth and col- lapse of the grana bags was considered by Frey-Wj^ssling and Steinmann as an indication that the grana had originally consisted of two layers of macromolecules (as shown in fig. 37A.15) which have been pushed apart by the osmotic pressure of water penetrating into the space between them. The fine granulation, observed on the best electron micrographs of the grana surface, suggests that these macromolecules are spherical; they could not 3'et be measured exactly, but certainly are under 0.0 1 fx in diameter, and must therefore have a molecular weight of <400,000. The "laminated" chloroplast structure again appeared in Wolken and Palade's (1952, 1953) investigation of two flagellates — protozoans which ELECTRON MICROSCOPY 1729 become autotrophically active when exposed to light, but switch to hetero- trophic metabolism in darkness. The organisms used were Euglena gracilis, which, in the active state, measures about 70 X 20 ^ and contains 8-12 chloroplasts, and Poteriochromonas stipitata, a much smaller, round cell, with usually only two chloroplasts. The cells were fi.xed with 1% OSO4 at pH 7-8, swelling V)eing prevented i)y the addition of sucrose. The preparations were embedded in plastic and sectioned to 0.1 n thickness, or thinner. The chloroplasts of active Fig. 37A.17. Sectiun of an Aspi({i)7. Nagai suggested that this selective silver deposition is the result of migration and adsorption of silver atoms on the chloroplast surface, rather than evidence of the presence of a reducing agent directly in the latter. In a second paper, Nagai (1951) noted that several plant substances — particularly ascorbic acid and di- oxyphenylalanine (to a lesser extent, also tannings and flavone derivatives) — can reduce silver nitrate under the conditions of the Molisch test. Later (1952) Nagai observed the Molisch reaction in many marine and fresh water algae, and concluded that the only appropriate reducing compound found in all of them is ascorbic acid. Algae which showed a negative Molisch test also gave no evidence of ascorbic acid in chromatographic separation. In sorrel and similar "oxalate plants" the Molisch test gave no clear positive results, despite the presence of ascorbate, but this could be attributed to the interference of oxalic acid. The effect of light was considered by Nagai in still another paper (19522). He found that the Molisch test was negative in etiolated seedlings, but became positive after 2 hours exposure to light, simultaneously with the visible formation of chlorophyll. Nagai suggested that the photochemical effect consists in attracting to the chloroplast the silver formed elsewhere by a non-photochemical reaction of ascorbic acid with silver nitrate, and consequent formation of localized black deposits instead of a diffuse, plas- matic precipitate. This seems a much less plausible explanation than sensitized photo- chemical reduction of silver ions (or photochemical reduction of chlorophyll by ascor- bic acid, followed by re-oxidation of reduced chlorophyll by silver ions). Hagene and Goas had noted (1945) that the disintegration of chloroplast struc- ture (e. g., by heat) is accompanied by a cessation of fluorescence, and inquired (1949) whether the capacity of chloroplasts for the reduction of silver nitrate shows a paral- lelism with their fluorescence. They found this to be true, but not under all condi- tions: fluorescence is preserved, e. g., after freezing chloroplasts in liquid air, while the reducing capacity disappears. No general parallelism exists between the capacity for silver nitrate reduction and for photosynthesis. It is not clear whether in these experi- ments the photochemical component of the silver nitrate reduction was considered rather than the — perhaps quite unrelated — thermochemical reduction. Metzner (1952') observed the distribution of silver deposits in sections from the lower surface of leaves of Agapanthus umbellatus and cotyledons of Impatiens parviflora. Blackening was found to occur in the cytoplasm as well as in plastides and nuclei. Within the chloroplasts, the reaction occurred preferentially in the grana and at the surface (perhaps at a membrane). The chloroplast reaction becomes predominant at pH 3, that in the cytoplasm at pH 7-8. Both reactions occurred in the dark; light ac- celerated the reduction. The action spectrum of AgNOs reduction by (initially) living Agapanthus cells showed peaks in the red and in the violet, indicating sensitization by chlorophyll and suggesting that direct photochemical decomposition of unstable silver salts (precipitated in the cells) does not contribute much to the blackening. In cells killed by several minutes immersion into silver nitrate solution in the dark, the action spectrum (for silver deposition in a following light period) was quite different, showing no peaks in the chlorophyll bands — except for the reaction in the grana. Cells killed by lead acetate still reduced AgNO,,, but only in the plastides; cells killed by brief boiling were completely inactive. Metzner (1952^) tried to sensitize the photochemical silver nitrate reduction in chloroplasts by vital staining with Rhodamin B, but the action spectrum of stained cells stiU showed the "green gap" characteristic of sensitization by chlorophyll. Thomas, Post and Vertregt (1954) first used the electron microscope to study the distribution of silver in chloroplasts and their fragments. They 1738 CHLOROPLASTS, CHROMOPLASTS AND CHROMATOPLASM CHAP. 37A showed that suspensions of chloroplast fragments (consisting predomi- nantly of grana), suspended in a silver nitrate solution, liberate silver in light. Whereas, according to Metzner, the photochemical reduction of AgNOs in "living" chloroplasts is best observed at pH 3, the same reac- tion in "grana suspensions" was fastest in neutral solution. The rate in- creased with light intensity slower than linearly, but without definite indi- cation of saturation, which would indicate the participation of an enzyme — 200 ISO 100- 50- • Christiansen-Weichert o -• ■■ *lcm6°/oCuS0t m ■■ ■*R62*B620 a - ■ - • -fRGS 4 •- ■■*RG5 a •' .+RGe 750 700 650 600 550 500 400 450 mfj Fig. 37A.19. Three sections of the action spectrum of Molisch reaction (Ag precipitation from AgNOs in acid solution) in a suspension of chloroplasts from Hibiscus rosa sinensis, obtained with different sets of filters (after Thomas, Post and Vertregt 1954). Abscissae: relative rates, measured by increase of opacity with time (rate in Na light = 100). Light filters as indicated in figure. in agreement with Metzner's observations that it can occur also in cells killed by silver nitrate; Thomas et al. found that 10 minutes boiling of the suspension also did not stop the reaction. Experiments with colloidal suspensions of pure chlorophyll in silver nitrate solution indicated that chlorophyll in itself has the capacity to re- duce AgNOs in light, without the aid of other cellular components; the chlorophyll-free alga, Prototheca zopfii, showed no increase in the rate of AgNOs reduction in light. These observations suggest that, even while ascorbic acid appears essential for the silver nitrate reduction by chloro- plasts in the dark, it may not be recjuired for the photochemical reduction. The latter may occur at the cost of chlorophyll itself (although it will then have to stop soon), or at the cost of other cellular hydrogen donors. Re- duction at the cost of water (Hill reaction with Ag+ ion as oxidant) seems unlikely because of the apparently non-enzymatic kinetics, and also be- ELECTROX MICROSCOPY 1739 Fig. 37A.20. Uncoated Hibiscus chloroplast granum (marked 6), and grana first smoothly coated (marked 8), and then deformed (marked 9) by silver deposits from AgNOa (after Thomas, T'ost and Vertregt 1954). Micrograph marked 7 shows a heap of grana overgrown with silver deposit. 1740 CHLOROPLASTS, CHROMOPLASTS AND CHROMATOPLASM CHAP. 37 A cause of the continuation of the reaction in killed cells. Addition of ascor- bic acid to grana suspensions considerably increases the yield — perhaps by reducing again the chlorophyll oxidized by Ag+ ions. Thomas et al. also measured the action spectrum of AgNOs reduction by grana suspension, and found it to be quite similar to this absorption spec- trum, and essentially determined by chlorophyll (fig. 37 A. 19). Electron diffraction study of the grana after the Molisch reaction showed a strong scattering pattern, indicating the presence of crystalline silver deposits (no diffraction rings were visible in the scattering pattern of grana not covered with silver). Electron-microscopic observations of grana after the Molisch reaction showed uncoated (fig. 37A.20 (top)) grana as well as grana in different stages of silver coating (fig. 37 A. 20 (middle)) ; after prolonged reaction, silver de- posits grew and hid the grana (fig. 37A.20 (bottom)). When grana were disintegrated into single discs by ultrasonic treatment, and the latter exposed to silver nitrate in light, they, too, were overgrown with silver deposits. 3. Ultracentrifuge Study Pardee, Schachman and Stanier (1952) disintegrated purple bacteria by grinding, sonic waves or sudden release of pressure, and subjected the products to fractional ultracentrifugation. They found that all pigments — the bacteriochlorophylls as well as the carotenoids — came down in a single fraction, consisting of large particles, and called the latter "chromato- phores." It seems that these particles, though smaller than the grana of the higher plants, may be functionally similar to them ; we will therefore refer to them as "grana." Particles of the same size could not be found in non-photosynthetic bac- teria, or in photosynthetic bacteria grown in dark and containing no pig- ment. Their presence thus seems to be associated with that of the photo- synthetic pigments — conceivably, they are held together by the strong res- onance attraction that exists between pigment molecules. From the sedimentation constant of the colored fraction of bacteria, Schachman et al. calculated — assuming spherical shape — a particle diam- eter of 0.04 n. Electron micrographs of the same fraction showed the pres- ence of flattened ellipsoids, or low cylinders, with a diameter of 0.11 m- The authors suggested that these particles originally had been spherical, but be- came flattened in the preparation of samples for electron microscopy. On this assumption, a value of 0.06 /x was calculated for the diameter of the original spheres — in fair agreement with the value derived from sedimenta- tion experiments. However, agreement can be achieved also by assuming that the })acterial grana have the approximately cylindrical shape observed OPTICAL EVIDENCE OF CHLOROPLAST STRUCTURE 1741 under the electron microscope, but that their thickness is such that, in sedimenting edgewise, they have a cross-section similar to that of a 0.04 n sphere. For a diameter of 0. 1 1 m, this thickness is 0.05 m- These dimensions are small, but not so small as to make implausible the assumption of func- tional identity of the bacterial "chromatophores" with the grana of the higher plants and algae. More recently grana of approximately this size have been in fact found in bacteria by electron microscopy (Thomas 1952; cf. section 2 above). Schachman et al. estimated that a single bacterial cell contains about 5,000 colored particles; but this estimate involved several rather uncertain premises. In a suspension prepared by grinding blue-green algae, fractional ultra- centrifugation also showed the presence of a single chlorophyll-bearing fraction. The colored particles were even larger than in bacteria. Electron microscopy indicates the presence in blue-green algae of grana of the same, or even somewhat larger, size than in higher plants — about 0.8 n across (Vatter 1952, Thomas 1952). The large-particle fraction of blue-green algae contained all chlorophyll, but phycocyanin was found mainly in a slow-sedimenting fraction. However, the occurrence of phycobilin-sensi- tized chlorophyll fluorescence in algae {cf. Chapter 24), argues against spatial separation of the two pigments in the living chromatoplasm, and for this separation having been effected in the grinding of the cells. This is sup- ported by the observations of McClendon (1952, 1954), cf. section 6 below. 4. Optical Evidence of Chloroplast Structure In Volume 1 (p. 365), measurements of double refraction and dichroism of chloroplasts were described. Frey-Wyssling's (1938) interpretation of these observations given there, was that the "positive" double refraction of chloroplasts in the natural state is a "morphic" birefringence, caused by lamellar structure, while the "negative" double refraction, found after im- bibition of chloroplasts with glycerol, is an "intrinsic" birefringence caused by the presence in the chloroplasts of an orderly array of long rod-shaped hydrocarbon chains, such as exist in lipides and phospholipides. However, Frey-Wyssling noted that the intrinsic double refraction of chloroplasts is weak, and suggested an imperfect alignment. Frey-Wyssling and Steinmann (1948), in a more precise study of the double refraction of Mougeotia chloroplasts, used different fixatives, and imbibed the chloroplasts with varying amounts of different liquids. They obtained in this way a set of hyperbolic curves showing double refraction as a function of refractive index. For all fixating solutions, the hyperbolas had a peak at n=1.58 (which is close to the refractive index of protein). 1742 CHLOROPLASTS, CHROMOPLASTS AND CHROMATOPLASM CHAP. 37A With osmic acid as fixative, the peak corresponded to a weak positive double refraction; with Zenker's sohition, to no double refraction at all (fig. 37A.21). The authors suggested that osmic acid fixes both proteins and Hpides, while Zenker's fluid fixes proteins only. If the negative morphic "2 ^

>■ <0 1500 do not penetrate into the plastides of these algae and thus do not cause the phycobilins to dissolve and ooze out into the medium. (Their separation from chlorophyll is accompanied by a sudden increase in fluorescence yield, noted in Chapter 32.) The Hill reaction of the plastids of Griffithsia, Antithamnion and Coral- Una is quite strong in Carbowax 4000, but drops greatly in buffer solution. Bibliography to Chapter 37A Morphology and Composition of Choroplasts 1936 Heitz, E., Ber. deut. botan. Ges., 54, 362; Planta, 26, 134. Hubert, B., Rec. trav. botan. neerland., 32, 323. 1937 Geitler, L., Planta, 26, 463. 1938 Frey-Wyssling, A , Submikroskopische Morphologie des Protoplasmas und seiner Derivate. Springer, Berlin. English edition: Submicroscopic Morphology of Protoplasm and Its Derivatives, Elsevier, New York, 1948. 1939 Hanson, E. A., Rec. trav. botan. neerland., 36, 183. 1940 Kausche, G. A., and Ruska, H., Naturwiss., 28, 303. Nezgovorov, L., Compt. rend. (Doklady) Acad. Sci. USSR, 29, 60. 1941 Liebich, H , Z. Boianik, 37, 129. Smirnov, A. I., and Pshennova, K. V., Biokhimiya, 6, 29. Nezgovorov, L., Compt. rend. (Doklady) Acad. Sci. USSR, 30, 258. 1943 Jungers, V., and Doutreligne, Soeur J., La Cellxde, 49, 409. Timm, E., Z. Botanik, 38, 1. 1944 Roberts, E. A., Am. J. Botany, 29, 10. Matuzamo, H., and Hir-ano, Z., Acta Phytochem. Japan, 14, 131. 1945 Hagene, M., and Goas, L., Compt. rend. soc. bioL, 139, 159. Day, R., and Franklin, J., Science, 104, 363. 1946 Bonner, J., and Wildmann, S. G., Arch. Biochem., 10, 497. Warburg, 0., and Llittgens, W., Biokhimiya, 11, 303. Aberg, B., Ann. Agr. Coll. Sweden, 13, 241. Gollub, M., and Vennesland, B., /. Biol. Chem. 169, 233. 1947 Granick, S., and Porter, K. R., Am. J. Botany, 34, 545. Algera, L., Beijer, J. J., van Iterson, W., Karstens, W. K. H., and Thung, T. H., Biochim. et Biophys. Acta, 1, 517. Wildmann, S. G., and Bonner, J., Arch. Biochem., 14, 381. Bradfield, J. R. G., Nature, 159, 467. Vecher, A. S., Biokhimiya, 12, 156. Boichenko, E. A., ibid., 12, 153. Vennesland, B., Ceithaml, J., and Gollub, INI., ibid., 171, 445. Stoll, A., and Wiedemann, E., Schweiz. Medizin. Wochenschr., 77, 664. 1948 Bracco, M., and Euler, H. v., Kem. Arbeten (Univ. Stockholm), New Series No. 10, March 1948. Arnon, D. I., Plant Physiol, 24, 1. Boichenko, E. A., Biokhimiya, 13, 88. 1756 CHLOROPLASTS, CHROMOPLASTS AND CHROMATOPLASM CHAP. 37A Euler, H. v. and Hahn, L., Ark. Kem. Miner. Geol, 25B, No. 1. Sisakyan, N. M., and Kobjakova, A. M., Biokhimiya 13, 88. Wassink, E. C, Enzymologia, 12, 362. Euler, H. v., Deut. vied. Wochschr., 73, 265. Frey-Wyssling, A., and Steinmann, E., Biochim. et Biophys. Acta, 2, 254. Frej^-Wyssling, A., Suhmicroscopic Morphology of Protoplasm and Its Derivatives, Elsevier, New York. Rezende-Pinto, M. C. de, Protug. Acta Biol. (A)2, 111. 1949 Frey-Wyssling, A., Faraday Soc. Discussions, No. 5, p. 130. Frey-Wyssling, A., and Miihlethaler, K., Vierteljahrsschrift naturforsch. Ges. Zurich, 94, No. 3. Rezende-Pinto, M. C. de, Portug. Acta Biol, A(2), 367. Hagene, M., and Goas, L., Compt. rend. soc. biol. 143, 147. Arnon, D. I., Nature, 162, 341. Vennesland, B., Gollub, M. C, and Speck, J. F., /. Biol. Clietn., 178, 301. Steemann-Nielsen, E., and Kristiansen, J., Physiol, plantarum, 2, 325. Euler, H. v., and Hahn, L., Ark. Kem. Miner. Geol., 26A, No. 11. Rosenberg, A. Y., and Ducet, G., Compt. rend., 229, 331. Conn, E. E., Vennesland, B., and Kraemer, L. M., Arch. Biochem... 23, 179. Sisakyan, N. M., and Shamova, K. G., Compt. rend. (Doklady) Acad. Sci. USSR, 67, 377. Sisakyan, N. M., and Kobjakova, A. M., ibid., 67, 703. Sisakyan, N. M., and Filipovich, 1. 1., ibid., 67, 517. Godnev, T. I., Kalishevich, S. V., and Zakharich, G. F., ibid., 66, 957. o Aberg, B., Physiol. Plantarum, 2, 164. Vennesland, B., /. Biol. Chem., 178, 591. Ceithaml, J., and Vennesland, B., ibid., 178, 133. 1950 Strugger, S. Naturwiss., 37, 166. Nagai, S., J. Inst. Poly tech. Osaka City Univ., 1, 33. Sisakyan, N. M., Bezinger, E. N., and Kuvaeva, E. B., Compt. rend. (Doklady) Acad. Sci. USSR, 74, 385. Osipova, 0. P., and Timofeeva, I. V., ibid., 74, 979. Holt, A. S., Oak Ridge Natl. Lab. Rept. ORNL 752. Osipova, 0. P., and Timofeeva, I. V., Compt. rend. (Doklady) Acad. Sci. USSR, 80, 449. Waygood, E. R., and Clendenning, K. A., Can. J. Research, C28, 673. Conn, E. E., and Vennesland, B., in "Brookhaven Conference Report," Assoc. Univ., Upton, N. Y., pp. 64-76. Laties, G. G., Arch. Biochem., 27, 404. 1951 Strugger, S., Ber. deut. botan. Ges., 64, 69. Nagai, S., /. Inst. Polytech. Osaka City Univ., 2, 7. Arnon, D. T., Nature, 162, 341. Hill, R., and Scarisbrick, R., New Phytol., 50, 98. Davenport, H. E., and Hill, R., Proc. Roy. Soc. London, B139, 327. Hill, R., Symp. Soc. Exptl. Biol, 5, 222. Whatley, F. R., Ordin, L., and Arnon, D. I., Plant Physiol, 26, 414. BIBLIOGRAPHY TO CHAPTER 37A 1757 Sisakyan, N. M., "Enzymatic Activity of Protoplasma Structures" (in Russian) Moscow, Acad. Sci. USSR. Bhagvat, K, and Hill, R., New Phijtol, 50, 112. Whatley, F. R., Ordin, L., and Arnon, D. I., Plant Physiol, 26, 414. Arnon, D. I., Nature, 167, 1008. 1952 Pardee, A. B., Schachman, H. K., and Stanier, R. Y., Nature, 169, 282. Takashima, S., Nature, 169, 182. Thomas, J. B., Bustraan, M., and Paris, C. H., Biochim. et Biophys. Ada, 8, 90. Vatter, A. (personal communication). Dlivel, D., and Mevius, W., Jr., Naturwiss., 39, 23. Schmidt, H., Protoplasma, 41, 336. Rezende-Pinto, M. C. de, Protoplasma, 41, 336. Rezende-Pinto, M. C. de, Portug. Acta Biol, (A)3, 281. Thomas, J. B., Proc. Roy. Acad. Amsterdam, C55, 207. Bustraan, M., Goedheer, J. C, and Thomas, J. B., Biochim. et Biophys. Acta, 8, 477; 9,499. -Metzner, H., Protoplasma, 41, 129. Metzner, H., Nachr. Akad. Wiss. Gottingen, Math, physik. Kl lie. Biol physiol chem. Aht., No. 1. Steinmann, E., Exptl. Cell Research, 3, 367. Steinmann, E., Experentia, VIII, 8, 300. Weier, T. E., and Stocking, C. R., Botan. Rev., 18, 14-75. Wolken, J. J., and Palade, G. E., Nature, 170, 114. Nagai, S., and Ogata, E., J. Inst. Polytech. Osaka City Univ., 3, 37. McClendon, J. H., and Blinks, L. R., Nature, 170, 577. Weier, E., and Stocking, C. R., Botan. Rev., 18, 14-75. Van Fleet, D. S., Botan. Rev., 18, 354. Godnev, T. B., Shlyk, A. A., and Tretyak, N. K, Compt. rend. (Doklady) Acad. Sci. USSR, 87, 493. Clendenning, K. A., Waygood, E. R., and Weinberger, P., Can. J. Botany, 30, 395. McClendon, J. H., Am. J. Botariy, 39, 275. Metzner, H., Naturwiss., 39, 64, Biol Zentr., 71, 17. Weinberger, P., and Clendenning, K. A., Can. J. Botany. 30, 755. Holzer, H., and Holzer, E., Chem. Ber., 85, 655. Sisakyan, N. M., and Chernyak, M. S., CojnpL rend. (Doklady) Acad. Sci. USSR, 87, 469. Calvin, M., and Lynch, V., Nature, 169, 455. 1953 Wolken, J. J., and Palade, G. E., Ann. N. Y. Acad. Sci., 56, 873. Leyon, H., Exptl. Cell Research, 4, 371. Leyon, H., ibid., 5, 520. Frey-Wyssling, A., Fortschritte der Botanik, 14, 66. Frey-Wyssling, A., and Steinmann, E., Vierteljahresshu. naturforsch. Ges. Zurich, 98, 20. Heitz, E., and Maly, R., Z. Naturforsch., 8b, 243. 1758 CHLOROPLASTS, CHROMOPLASTS AND CHROMATOPLASM CHAP. 37A Rabinowitch, E., Ann. Rev. Plant Physiol., 3, 229-264. McClendon, J. H., Am. J. Botany, 40, 260. Krasnovsky, A. A., and Kosobutskaya, L. M., Compt. rend. (Doklady) Acad. Sci. USSR, 91, 343. Hill, R., Northcote, D. H., and Davenport, H. E., Nature, 172, 947. Algal Culture, from Laboratory to Pilot Plant, J. S. Burlew, Editor, Car- negie Institution Publication No. 600, Washington, 1953; in particular Fogg, G. E., and Collyer, D. M., "The Documentation of Lipides by Algal," pp. 177-181; Alilner, H. W., "The Chemical Composition of Algal," pp. 285-302. Rodrigo, F. A., Biochim. et Biovhys. Acta, 10, 342. 1954 Wolken, J. J., /. Gen. Physiol, 37, 111-120. Thomas, J. B., Post, L. C, and Vertregt, N., Biochem. etBiophys. Acta, 13, 20. Chiba, Y., Arch. Biochem. and Biophys., 54, 83. Leyon, H., Exptl. Cell Research, 7, 265. Albertson, P. A., and Lej-on, H., ibid., 7, 288. Leyon, H., and Wettstein, D. v., Z. Naturforsch., 9b, 472. Thomas, J. C, Rapp. Comm. Sme Congr. Botanique, Paris, Section 11, p. 28. Holzer, H., Angew. Chemie, 66, 65. Weissbach, A., Smyrniotis, P. Z., and Horecker, B. L., J. Am. Chem. Soc, 76, 3611. Quayle, J. R., Fuller, R. C, Benson, A. A., and Calvin, M., ibid., 76, 3610. McClendon, J. H., Plant Physiol. 29, 448. Arnon, D. I., Rosenberg, L. L., and WTiatley, F. R., Nature, 173, 1132. Daly, J. M., and Brown, A. H., Arch. Biochem. Biophys., 52, 380. B. Chemistry of Chloroplast Pigments* (Excluding Photochemistry) (Addenda to Chapters 15 and 16) No important progress has been achieved, since the pubUcation of Volume I in 1945, in the synthesis of chlorophyll or in further elucidation of its structure. Important observations have been made, however, concern- ing the biogenesis of chlorophyll and its oxidation-reduction reactions. 1. Biosynthesis of Chlorophyll ; The Protochlorophyll (Addendum to Chapter 15, section B2) In Vol. I (p. 405), after the description of protochlorophyll and its possible function as chlorophyll precursor in plants, the remark was made that "the whole problem of chlorophyll development in seedlings is in need of renewed study." Since then, the subject has been taken up by Smith and co-workers at the Stanford Laboratory of the Carnegie Institution. Through quantita- tive determination of protochlorophyll, chlorophyll a and chlorophyll h in etiolated seedlings after exposure to light, they proved that up to 90% of the protochlorophyll, accumulated in the dark, are quantitatively con- verted, within a minute or less of moderately strong illumination, into chlorophyll a. After this initial period, the formation of chlorophyll a continues (and that of chlorophyll h gets under way) at a much slower rate, to reach saturation in a day or two; during this second stage, the rate is governed by thermal reactions, involving other precursors besides proto- chlorophyll— e.g., those containing no magnesium in ether-soluble form. Whether this slow synthesis goes through protochlorophyll as intermediate, or by-passes this compound (as was assumed in Lubimenko's scheme, illus- trated on p. 405), remains an open question. The presence of small amounts of protochlorophyll in fully green barley plants, noted by Koski and Smith (1948^), could be explained by protochlorophyll's being a side product, as well as by its being a necessary intermediate in chlorophyll synthesis. Following is a brief summary of the studies of Smith and co-workers. Smith (1947) determined the changes in the total magnesium content, ether-soluble magnesium, and chlorophyll magnesium, caused by exposure to light of etiolated barley seedHng, in attached as well as in excised leaves. The results are summarized in Table 37B.I. * Bibliography, page 1790. 17.59 1760 CHEMISTRY OF CHLOROPLAST PIGMENTS CHAP. 37B Table 37B.I Synthesis of Organic Magnesium Compounds by Etiolated Barley Seedlings in Light (after Smith 1947) Weight per cent of dry matter before and after exposure Total Mg Total ether-sol. Mg Chi Mg Leaves Before After Before After After Attached(71hr. exposure).... 0.16 0.22 5.2X10-" 2 X 10"^ 1.8 X 10"* Detached (46 hr. exposure) .. . 0.19 0.216 5.4X10-' 1.2X10-2 l.llxiO-« The fastest increase in ether-soluble magnesium was noted in the first two hours of exposure to light; during this time, the total formation of ether-soluble magnesium compounds ran considerably ahead of that of chlorophyll, indicating the accumulation of other magnesium-containing organic molecules^perhaps chlorophyll precursors (c/. section 2 below). After a day or two, chlorophyll formation caught up with that of total organic magnesium compounds; in attached leaves, in particular, over 90% of the total ether-extractable magnesium was found, at that time, to be present in chlorophyll. Leaves left in the dark as controls formed no more ether-soluble mag- nesium compounds after germination, indicating that light is needed not only for the formation of chlorophyll, but also for the accumulation of the other magnesium-containing organic compounds. Smith (1949) gave time curves for the formation of total ether-soluble magnesium and chlorophyll magnesium in etiolated barley seedlings at three different temperatures. At 0° C, the maximum amount of ether-soluble magnesium compounds formed was only 6% of that formed at 19° C. ; the synthesis reached satura- tion in a few hours, after which the concentration of ether-soluble mag- nesium began to decline. The chlorophyll synthesis fared even worse — the amount s.yntheeized never exceeded 15% of total ether-soluble mag- nesium, and started to decline already after two hours of exposure. At 7° C, a fast initial formation of a small amount of chlorophyll was noticeable in the first hour (without much change in the total ether-solu- ble magnesium) ; a delay ensued in the next few hours, after which the formation of ether-soluble magnesium compounds got under way and con- tinued steadily even after 40 hours' exposure. The formation of chlorophyll followed, and gradually caught up with, that of the total ether-soluble magnesium compounds; after 40 hours, 70% of total ether-soluble mag- nesium was in chlorophyll. At 19° C, the chlorophyll formation, after a slight initial lag, caught up in a few hours with the total formation of ether-soluble magnesium com- pounds (in agreement with the results shown in Table 37B.I for attached BIOSYNTHESIS OF CHLOROPHYLL) THE PROTOCHLOROPHYLL 1761 leaves). The concentration of organic magnesium reached saturation after about 40 hours' exposure, after which the amount of chlorophyll began to decline slightly, probably because of photoxidation. These experiments, particularly that at 7° C, clearly pointed to a two- stage process of chlorophyll formation — a fast photochemical reaction and a relatively slow thermal reaction, brought almost to a standstill at 0° C. The formation of ether-soluble phosphorus compounds was found to follow a course parallel to that of the formation of ether-soluble magnesium compounds. The isolation of sizable quantities of pure protochlorophyll by Koski and Smith (1948^), and the absolute determination of its absorption co- efficients (c/. fig. 21.8 in Vol. 11,1, page 618) made it possible to correlate quantitatively the formation of chlorophyll with the disappearance of protochlorophyll. Smith (1948) made the first such comparison by determining the amount of chlorophyll formed in the first two hours of illumination at 0° C. and the amount of protochlorophyll present at the beginning of the e.xposure. He found an approximate proportionality between the two magnitudes, with the proportionality constant <1. In fact, the amount of synthesized chlorophyll corresponded, under these conditions, to little more than one-half of the available protochlorophyll. A more precise study was carried out by Koski (1950), by restricting the measurements to the first few minutes of illumination. This allowed working at room temperature, without the slow thermal reactions inter- fering significantly with the fast photochemical transformation. Corn seedlings were used. They were germinated for 15 days in the dark and then exposed to fluorescent light (150 foot-candles) at room tem- perature. Protochlorophyll, chlorophyll a and chlorophyll h were deter- mined spectrophotometrically at different times after the beginning of illumination, using the following specific absorption coefficients: X a,p (Chi. a) a.p (Chi. 6) a (protochlorophyll)" 663 m/j. 95 5.1 0.19 644 m^ 15.4 57.5 1.15 624 iQfi 13.9 9.9 39.9 " Contrary to the observations of Seybold and Egle on pumpkin seeds (Vol. I, p. 404 and Vol. II. 1, p. 619), no evidence of the presence of a protociilorophyll b was obtained by Smith and Koski. Fig. 37. Bl shows the rapid formation of chlorophyll a in the first minute of exposure, and an exactly equivalent decrease in protochlorophyll con- tent. After the first minute, the amount of chlorophyll formed begins to trail somewhat, but not significantly, behind the amount of proto- 1762 CHEMISTEY OF CHLOROPLAST PIGMENTS CHAP. 37b chlorophyll lost, so that the total pigment concentration drops very slightly. After about one hour of illumination, chlorophyll b appears, and < 5£ 0.0004 6 8 10 12 EXPOSURE TIME, min. Fig. 37B.1. Quantity of chlorophyll formed (O) and of protochlorophyll used up (•) in etiolated corn seedlings (after Koski 1950). _ 2 3 4 ^ EXPOSURE TIME, hr. Fig. 37B.2. Formation of chlorophyll a (O) and b (•) at 12° C. in etiolated corn seedlings (after Koski 1950). Original protochlorophyll content of the seedings was 0.003 mg./g. fresh weight. BIOSYNTHESIS OF CHLOROPHYLL; THE PROTOCHLOROPHYLL 1763 the amount of chlorophyll a begms to rise above the almost stationary- level that was established after the first few minutes of illumination (fig. 37. B2). This slow, but steady, increase continues for hours; it must be determined by the rate of replenishment of the pool of chlorophyll pre- cursors by slow thermal reactions. The initial fast pigment formation un- doubtedly is the result of photochemical conversion of pre-existent proto- chlorophyll into chlorophyll. Only the a component is formed in this way. 400 700 500 600 WAVE LENGTH, m/i Fig. 37B.3. Action Bpeetrum of transformation of protochlorophyll to chlorophyll a in etiolated corn leaves: (O) normal, ( + ) albino (Koski, French and Smith 1951). The sharp absorption band in the violet, noted by Koski and Smith (1948) in the spectrum of protochlorophyll, has an analogue in the action spectrum of chlorophyll formation, first noted by Frank (1946) in experi- ments on oat seedlings in filtered light. The finding of this band invali- dated a previously used argument against identification of the precursor which produces chlorophyll in light, with protochlorophyll. Frank's study left the question of carotenoid contribution to the action spectrum peak in the blue-violet unsettled; an "auto-photocatalysis" by chlorophyll (c/. Vol. I, p. 430) seemed to be indicated, but could not be definitely proved. The action spectrum of chlorophyll formation was re- investigated by 1764 CHEMISTRY OF CHLOROPLAST PIGMENTS CHAP. 37B Koski, French and Smith (1951) using a powerful monochromator with normal and albino corn. The albino mutant {cf. Koski and Smith 1951) formed, in the dark, more protochlorophyll than the normal strain ; it con- verted it into chlorophyll in hght, but this chlorophyll was rapidly de- stroyed by further illumination, indicating that albinism was due to lack of protection of chlorophyll against photoxidation. Other, "virescent," mutants, described by Smith and Koski 1948, produced only very little protochlorophyll in the dark, but kept forming chlorophyll slowly upon prolonged illumination, and once formed, did not lose it easily. The albino seedlings had the great experimental advantage of contain- ing almost no carotenoids. The action spectra for chlorophyll synthesis of the two strains are shown in fig. 37. B3. The ordinate is inversely proportional to the number of quanta needed to transform 20% of the total available protochlorophyll into chlorophyll; the peak at 650 m^ corresponds to this transformation being completed in 18.4 sec, in a flux of 80 erg/(cm.2 X sec). The absorp- tion peaks are situated, in both cases, at 445 and 650 m^. The ratio of the ordinates in the two peaks is 1.89 in albino and 0.66 in normal plants, clearly indicating the relative (or complete) inefficiency of the carotenoids. (Frank, 1946, had found a ratio of 1.47. This result may indicate either a lower content, or a higher efficiency of carotenoids in oat seedlings. A more trivial explanation also is possible — a smearing-out of the action spectrum, caused by the use of light filters.) The red action peak in fig. 37. B3 is shifted by 21 mju, and the violet peak, by 11 m/z, toward the longer waves, compared to the absorption peaks of protochlorophyll in methanol solution; both figures correspond to a shift of about 500 cm.~^ on the wave-number scale. It seems plausible that the absorption bands of protochlorophyll in vivo are shifted by that amount toward the red from their position in vitro. A transmission mini- mum was in fact noted in the spectrum of squash seeds at 650 m/z {cf. Chapter 37C, section 2e). When plants in which one-half of protochlorophyll had been converted to chlorophyll, were illuminated with a monochromatic band centered at 680 m/x (a wave length only slightly absorbed by protochlorophyll, but strongly absorbed by chlorophyll a in vivo) , very little additional chlorophyll formation was observed, indicating the practical absence of "auto-photo- catalysis." Smith (1951, 1954) used Franck and Pringsheim's phosphorescence quenching method (Vol. 11,1, p. 851) to determine whether any oxygen was produced during the photochemical reduction of protochlorophyll to chloro- phyll. The conversion was not inhibited by an atmosphere of pure hydro- gen (<10~" per cent O2) needed for the application of this method. The BIOSYNTHESIS OF CHLOROPHYLL; THE PROTOCHLOROPHYLL 1765 observed oxygen liberation was <2.5% of that calculated on the assump- tion that the reduction occurs at the expense of Avater (protochlorophyll + H2O ->- Chi + O2). Under similar conditions, in leaves which contained some chlorophyll, oxygen evolution from photosynthesis could be easily observed; this shows that the lack of oxygen production in etiolated leaves was not due to immediate re-utilization of this gas by the cells. The con- clusion is thus justified that the photochemical conversion of protochloro- phyll into chlorophyll is not coupled with the oxidation of water. (It therefore cannot represent the oxygen-liberating step in photosynthesis, as has been occasional^ suggested.) Smith and Benitez (1954) inquired into the temperature dependence of the initial protochlorophyll -»► chlorophyll conversion in light, over a wide range, from -175° to -^55° C. They noted that heating above 40° C. progressively destroyed the capacity for conversion and suggested that this indicates the denaturation of a protein (to which protochlorophyll can be assumed to be attached in vivo, and which may be essential for the transformation; dissolved protochlorophyll is not converted to chlorophyll by illumination). Cooling below 40° C. slowed down the photochemical conversion. This, and the fact that the reaction followed a bimolecular law (strictly at 589, 579 and 546 m/i ; less strictly at 436 m^u) indicated interaction between an excited and a normal protochlorophyll molecule. (Proportionality of the rate Avith the first power of light intensity excludes a reaction between two excited molecules.) Perhaps the reaction is a dismutation with one protochlorophyll molecule being reduced and one oxidized, but no oxidation product has yet been observed. The temperature curve can be followed down to —80° C, at which temperature the conversion is still appreciable in both rate and total amount. It becomes unobservable at —195° C. Tha\ving after freezing destroys the capacity for conversion; slow passage through the freezing zone damages it. In order to measure the rate at —10° or —20° C, the leaves must be snap-freezed at —80° C. and then warmed up. Some experiments were made by Smith and co-workers on the rate of protochlorophyll formation by barley seedlings in the dark. A sigmoid curve was obtained, almost no protochlorophyll being formed in the first three days of germination, an accelerated formation occurring between the third and the sixth day, and the synthesis slowing down after the sixth day. The development of carotenoids followed a similar course. Godnev and Terentjeva (1953) found that the conversion of protochloro- phyll to chlorophyll can be achieved, in etiolated corn seedhngs, instead of by exposure to light, also by infiltration with an extract from pine seed- lings (conifers, like algae, can form chlorophyll in darkness). 1766 CHEMISTRY OF CHLOROPLAST PIGMENTS CHAP. 37B Krasnovsky and Kosobutskaya (1953) observed that when etiolated leaves (or plastid fragments from them) were exposed to light the forma- tion of chlorophyll occurred in two stages. In the first 2-3 hours, the ab- sorption peak was at 670 m^i, and most of the chlorophyll formed was easily bleachable by strong light. Later, the band peak gradually shifted to 678 m^, and most of chlorophyll became photostable. As mentioned above, the development of chlorophyll b in etiolated barley occurs, according to Smith, only in the second, slow stage that follows the initial rapid conversion of preformed protochlorophyll into chlorophyll a. It is possible, but not certain, that the synthesis of the h component in- volves a ''protochlorophyll 6." Such a compound does not accumulate in etiolated seedhngs; but according to Seybold and Egle (c/. pp. 404 and 619) it is present in pumpkin seeds. It is very unlikely that chlorophyll b is formed by oxidation of chlorophyll a. The delayed formation of chlorophyll b was noted also by Goodwin and Owens (1947) in oat seedlings. The appearance of chlorophyll b is not essentia] for the beginning of photosynthesis, according to Smith (1954). Some x-ray mutants of barley were found to be permanently deficient in chlorophyll b (Highkin 1950), but capable of photosynthesis. In some mutants of corn, chlorophyll b was formed only after a delay of from 4 to 14 days, depending on temperature (Schwartz 1949). Certain x-ray mutants of Chlorella, obtained by Granick (1951), pro- duced chlorophyll only in light (similarly to the higher plants, but unlike normal algae). Granick suggested that the enzymatic system capable of reducing protochlorophyll to chlorophyll in the dark was absent in these mutants. The effect of narcotics and of colchicine on the synthesis of chlorophyll in etiolated wheat seedlings was studied by Brebion (1948, 1950). Ether, acetone, benzene and phenol vapors were found to delay the greening in high concentrations and to stimulate it in low concentrations (with phenol, as low as 1 X 10"^%). Among new observations of the influence of different external factors on the development of chlorophyll, we can mention the specific effect of streptomycin, first noted by von Euler, Bracco and Heller (1948). Seeds germinating on paper wetted with streptomycin solution developed com- pletely colorless coleoptiles and first leaves. Streptomycin had no effect on leaves already containing chlorophyll. Provasoli and co-workers (1948, 1951) found that streptomycin causes Euglena gracilis cells to lose their capacity for forming chloroplasts ; this deficiency was inherited by their offspring. Bogorad (1950) noted that the cotyledons of a pine (Pinus Jeffryi), which normally can form chlorophyll in the dark, will do it only in light if BIOSYNTHESIS OF CHLOROPHYLL; THE PROTOCHLOROPHYLL 1767 the seeds had been germinated in the presence of streptomycin. This antibiotic seems to prevent the formation of the enzyme that converts protochlorophyll into chlorophyll. It must be kept in mind, however, that this interpretation (whether applied to x-ray mutants, or to the action of streptomycin) takes it for granted that all chlorophyll formation occurs via protochlorophyll — a question which we have left open. The relation between the appearance of chlorophyll and the beginning of photosynthesis was mentioned in Chapter 32 (section 3). Additional ob- servations have been since reported by Blaauw-Jansen, Komen and Thomas (1950), and Smith (1954). Blaauw Jansen et al. found that etiolated oat leaves had only a slight capacity for photosynthesis when first illuminated. This capacity increased, upon exposure to light, faster than the chlorophyll content; and it was noted that it grew parallel with the increase in the [h]:[a] ratio, until the latter attained its normal value. Smith (1954) found, by the phosphorescence-quenching method, that even after 85% of the protochlorophyll present in etiolated barley leaves had been converted into chlorophyll a by illumination in pure hydrogen, the leaves were still incapable of photosynthetic oxygen production. If such leaves (containing only chlorophyll derived from the original reservoir of protochlorophyll) were exposed to air in the dark, they acquired a shght capacity for photosynthesis. This capacity was greatly increased by brief illumination — the increase being out of all proportion with the accumula- tion of new chlorophyll. It thus seems that the chlorophyll formed from protochlorophyll in the first minute or two of exposure of etiolated leaves to light is photosynthetically inactive. This inactivity should be due to difference, either in chemical structure, or in the composition of the pigment complex, or in the arrangement of the pigment molecules. (For example, activity may require the formation of monomolecular layers of the pigment on protein discs, as suggested in Chapter 37A.) Alternatively, the delay in the acquisition of photosynthetic capacity could be caused by deficiency of a catalytic component other than chlorophyll. (This component must be rapidly formed in the cycle: brief anaerobic irradiation — exposure to air in the dark — brief aerobic irradiation; and only more slowly in continuous light.) As mentioned above, Krasnovsky and Kosobutskaya (1953) found that the first chlorophyll formed has an absorption peak at 670 m/i while the main mass of the pigment, formed later, has a peak at 678 m^. They suggested that the (more easily photoxidized) first batch ("Chi 670") is the only "photoactive" part of the pigment; while the subsequently formed bulk ("Chi 678"), is in an inactive, polymeric form, and contributes to photosynthesis only by energy transfer to the "active" form. The J 1768 CHEMISTRY OF CHLOROPLAST PIGMENTS CHAP. 37B above-described observations by Smith and Koski suggest the reverse hypothesis — that only "Chi 678" is photosynthetically active. (The capacity for photoxidation may well be antiparallel, rather than parallel, to photosynthetic activity.) In Chapter 37C (section 6b) we will quote another argument, pointing to the same conclusion: Resonance energy transfer always is directed toward the pigment with the lower excited state (while Krasnovsky's hypothesis calls for energy transfer in the opposite direction). The 670 mn absorption band may belong to as yet "unor- ganized" (but already protein-bound) chlorophyll, while the 678 m^i band may be that of chlorophyll arranged in monomolecular layers; the shift may then be due to the electrostatic interaction of the pigment mol- ecules in a regular array, to be discussed in section 3 of Chapter 37C. Yocum (1946) observed the inhibition of chlorophyll formation in excised, etiolated bean leaves by poisons. Carbon monoxide reduced both respiration and chlorophyll for- mation in red light, in approximately the same ratio; strong blue light reversed the inhibition in both cases. Cyanide (10~^ .1/) reduced the formation of protochlorophyll in the dark, but did not affect the conversion of protochlorophyll to chlorophyll. The hiodecomposition of chlorophyll was discussed by Noack (1943), who attributed the destruction of the green pigment in autumn to the action of hydrogen peroxide, not decomposed by the, now inactivated, catalase, on water-soluble chlorophyllides formed by the action of chloroph3'llase. 2. Biogenesis of Chlorophyll ; Earlier Precursors If one wants to speculate on the mechanism of chlorophyll synthesis in the living cell beyond the experimentally established photochemical con- version of protochlorophyll to chlorophyll, one must take recourse to in- direct evidence. Granick (1948,^-^ 1950,^'2 1951, 1954) obtained such evidence from ex- periments in which Chlorella cells were exposed to x-rays, producing a vari- ety of mutants. Several of these — ^'iable in glucose solution but incapable of photosynthesis — were found to contain no chlorophyll: instead, some of them carried pigments of the porphin type not normally encountered in algae. One mutant, in particular, dark brown in color, contained globules of pivtoporphyrin 9. In another mutant, which, when grown on a solid nutrient medium, developed an orange-brown color, small amounts of the magnesium derivative of the same protoporphyrin could be identified (in addition to the protoporphyrin itself). A third, yellow mutant yielded magnesium-vinyl pheoporphyrin 05, i. e., protochlorophyll without the phytol chain. Granick postulated that the appearance of these pigments signifies BIOGENESIS OF CHLOROPHYLL; EARLIER PRECURSORS 1769 that the development of chlorophyll has been interrupted at an intermediate stage l)y lack of a specific enzyme (caused by damage to a gene). He de- rived from these observations the scheme of the l)iosynthesis of chloro- phyll, which is represented in the right side of fig. 37B.4. Protoporphyrin 9 is the porphyrin most closely related to hemoglobin (in fact, it is the heme minus iron). According to Granick, this porphyrin is the common precursor of both the red blood pigment and the green plant pigment. Granick speculated further on the probable mechanism of synthesis of protoporphyrin 9 from ultimate, small building blocks — assumed to be glycine and acetate. The first synthesized pyrrole derivative he assumed to be (I) in fig. 37B.4. Uroporphyrin (II) was assumed to be the first tetrapyrrole formed from this monopyrrole derivative, and coproporphyrin (III), the intermediate between uroporphyrin and protoporphyrin 9. Additional results were reported by Bogorad and Granick (1952). They found new Chlorella mutants, containing porphyrins with 2, 3, 4, 5 and 8 carboxyl groups. One of them, the monovinyl hydroxy dicarboxyl- porphyrin, was suggested as a likely immediate precursor of protoporphyrin ; its own precursor could be hematoporphyrin, also found in these mutants. The role of glycine in supplying nitrogen for the synthesis of porphin derivatives was first demonstrated by Shemin and Rittenberg in 1940, by experiments on human erythrocytes with N^^ tracer. The (indirect) acetate origin of the carbon atoms (other than those supplied by the a- carbons of glycine), also was supported by isotope tracer studies of Rit- tenberg and co-workers. Salomon, Altman and Rosa (1950) showed that the a-carbons of acetic acid and of glycine are used in the synthesis of chlo- rophyll in the plant cell; this lends support to Granick's hypothesis that the formation of chlorophyll follows — up to a certain point — the same path as the synthesis of heme. Granick suggested, as a further hypothesis, that biosynthesis, as it occurs in plants now, repeats the path of evolution. In other words, he postulated that organisms synthesizing uroporphyrin, coproporphyrin, protoporphyrin and, finally, protochlorophyll and chlorophyll, have evolved in this order. He assumed that, at the time when any one of these compounds represented the final product of synthesis, it was used for some metabolic process by the organism; this particular process fell into disuse as the synthesis advanced another step. More specifically, Granick suggested that all porphin pigments presently encountered in cells have been used, at some stage of evolution, as photochemical sensi- tizers. He saw a confirmation of this view in the capacity of "vestigial" pigments, such as chlorophyll c, to sensitize photosynthesis. (He had found chlorophyll c to be a porphin, rather than a chlorin, cf. section 5 1770 CHEMISTRY OF CHLOROPLAST PIGMENTS CHAP. 37b COOH-CH.-C .c=:n -'^■^ H Glycine + acetate I n steps Hypothetical pyrrole COOH 4 I 4 HOOC CH, 3 I I 3 H.C CH, 2 I ; I 2 c-rC 1 (I) II rl- 1 HCy C-COOH / N ' H 71 steps Uroporphyrin type III (hypothetical first tetrapyrrole) COOH I CH, COOH CH2 a CH2 I H I C r^; C, Bacteriochlorophyll CH. CH H Me MeCH=-CH, HC. Mg CH (VIII) Chlor 6 -e- V"' ioOCH, COOCooH,, Chlorophyll a t Mg vinyl pheoporphynn a^ phytyl ester (yiT\ or protoclilorophyll i H C \ C "C-CHj-CHo-COOH H ffiH (i (II) COOH-C-C. Ic; ic[ ^C-CH,-COOH I H I C T C C.Hi / ,C-N C I COOH C I COOH 4 ste|)S (?) Chlor f Coproporphyrin type III CCOH CH (VI) C CH2 \H I C-C=0 CH. I I COOCH, COOH Mg vinyl pheoporphyrin a^ 4-5 *teps Mg protoporphyrin (V) H3C HjC CHo-CH,-COOH (III) CH=CHj .CH (IV) HN — C n steps COOH COOH Phycoerythrin ) Phycocyanin / Fig. 37B.4. Scheme of chlorophyll biosynthesis fafter Granick 1951). ISOMERS AND SOLVATES OF CHLOROPHYLL 1771 below, and considered it a side product of evolution of Mg vinyl pheo- porphyriii as, as indicated in fig. 37B.4.) Granick saw in the same light the role in photosynthesis of the phycobilins, which he considered another survival from an early stage of evolution. This approach differs, in its general tendency, from the theoi-ies that consider organic evolution as having been accompanied by a loss of syn- thesizing abilities rather than by their gradual aquisition. Also, Granick sees the role of accessory pigments in somewhat different light than the — presently most plausible — hypothesis that these pigments serve as sup- pliers of energy, by resonance, to the one pigment adapted for the photo- catalytic function proper — chlorophyll a {cf. Chapters 30 and 32) . 3. Isomers and Solvates of Chlorophyll (Addendum to Chapter 15, section Bl) In \'ol. I, page 403, we mentioned that Strain and Manning (1942) had observed the reversible conversion of chlorophylls a and h into slightly different forms, which they called a' and b'. (It will be recalled that four such forms: d, iso-f/, d' and iso-r/' were observed with chlorophyll d.) Strain (1949, 1953) gave additional details on these compounds. When chlorophyll is extracted from leaves, hydrolytic and oxidative reactions are apt to occur, leading to a multiplicity of colored bands in the chromatographic column. The relative rates of these reactions depend on the plant species used, and on incidental factors such as the method of mincing (for example, chopping causes less oxidation than grinding or crushing). Light is to be avoided during preparation and storage of the minced leaf material. Rapid extraction of chopped mallow or barley leaves by grinding in methanol or acetone, to which some petroleum ether had been added, gives, according to Strain, only chlorophylls a and b. If, however, chopped leaves are placed into boiling water foi- 1-2 minutes before extraction, a' and b' are also found in the extract. (Plants with acid sap, such as Opuntia, Pelargonium, and Bryophyllum, rapidly form pheophytin when heated in air.) Chopped leaves, dried in air for 24 hours, yield small amounts of a' and 6' ; heating to 100° C. in air for 15 minutes after drying increases the quantity of these isomers considerably. No new pigments are produced, in either mallow or barley leaves, by freezing and thawing, or by standing with saturated ammonium sulfate for 20 hours and wasliing (by dialysis into distilled watei') for 20 hours. Isomerizations similar to those occurring with chlorophylls a and b can be produced also with the chlorophyllides, e. g., by heating their pro- panol solutions, or permitting them to stand for several days at room tern- 1772 CHEMISTRY OF CHLOROPLAST PIGMENTS CHAP. 37B perature. The isomeric chlorophyllides are less adsorbable than the original chlorophyllides a and h. Separated on the column and redissolved in propanol, they can be (at least partially) reconverted to a and h by heating. The isomerization equilibrium of the ethyl and methyl chlorophyllides seems to correspond, at room temperature, to [a]: [a'] = [6]: [6'] = 3:1. Isomerization is not affected by the presence of air or dimethyl- aniline (which, respectively, bring about and prevent allomerization). It is accelerated by alkali: 0.2% KOH produces isomeric eciuilibration in 1-5 minutes, followed by a slower oxidation. In discussing the possible mechanism of isomerization. Strain pointed out that its absence after allomerization points to a role of the C(10) atom, and suggested that the isomers differ only by the spatial arrangement of the two groups (COOCH3 and H) attached to C(10); they could then be con- verted into each other via their common enol, in which the H-atom is trans- ferred from C(10) to C(9) to form a hydroxyl group there. This could ac- count for the catalytic effect of alkalis on the isomerization. Freed and Sancier (1951) noticed a variability of the spectrum of chloro- phyll 6 preparations — particularly of the small peak at 481.5 ni/u. Subsequently, Freed, Sancier and Sporer (1954) isolated from such preparations a fraction which they called chlorophyll b" . It had a spec- trum very similar to that of h, but gave no phase test. Freed suggested that 6, 6' and h" may be the three isomers anticipated on page 444, differing only in the routing of the conjugated bond system. On the other hand, Freed's b" appears similar to Holt's "fraction 3" of allomerized chloro- phyll a (cf. page 1775). The main subject of the studies of Freed and Sancier (1951, 1952, 1953, 1954) was the transformation of the chlorophylls and their derivatives at low temperatures. At first (1951) the observed reversible spectroscopic changes were interpreted as evidence of the formation of new isomers. Because of the analogy between these spectral changes and those caused by various solvents, it was suggested that the solvent effect, too, may be caused by isomerization. After the experiments of Livingston and co-workers, Evstigneev, and others {cf. Chapter 21, section Bl; Chapter 23, sections 4-6, and Chapter 37C, section 2) had revealed that chlorophyll forms complexes with water and other solvents which differ in their absorption spectrum and capacity for fluorescence. Freed and Sancier (1954) reinter- preted the temperature changes of the absorption spectra as also resulting from the formation and dissociation of solvates. However, the two types of transformation — isomerization and solvation — may be coupled; e. g., association with a protophilic solvent is likely to favor enolization. It is a matter of arbitrary choice whether to discuss the solvent effects and temperature effects on the chlorophyll spectrum in the present chapter OXTDATIOX, ALLOMERIZATIOX AND REDUCTION OF riTLOROPHYLL 1773 (as isomerization or soh'ation phenomena), or in Chapter 37C under the heading of "Solvent Effects on Absorption Spectra." Since the latter ar- rangement had been adopted in Chapter 20 we will adhere to it; the description of Freed's low temperature studies will be therefore found in Chapter 37C, section 2a. Only the part of these experiments related to the mechanism of certain irreversible reactions of chlorophyll (with alkalis and amines} will be described in section 4 below. 4. Oxidation, AUomerization and Reduction of Chlorophyll (Addendum to Chapter 16, section B3) In Chapter 10 (p. 459) the so-called aUomcrization of chlorophyll was discussed — a transformation that occurs upon standing in alcoholic solution, involves the uptake of oxygen, and can be brought about also by addition of oxidants such as quinone. It was interpreted as oxidation of the "lone" hydrogen atom at C(10). The allomerized chlorophyll differs from the intact compound by its incapacity to give the "brown phase" in alkalysis, and l)y its spectrum. A significant feature of this spectrum (fig. 2 1.4 A) is the apparently complete absence of the short-wave satellite of the main blue-violet band, that is present, with varying intensity, in the spectra of all chlorophylls, chlorophyllides, pheophorbides, etc. {cf. figs. 21.1 and 21.26). This satellite could be due to a vibration (of the order of 750 cm.~^), or to the doublet structure of the excited electronic term (perhaps, corresponding to charge oscillations in two mutually per- pendicular directions in the porphin plane, cf. Chapter 37C, section 1) ; or to a tautomeric form (as suggested tentatively on p. 027, Vol. 11,1). What- ever the correct interpretation of the satellite band may be, its disaijpear- ance upon allomerization must be significant. It has long been known that upon extraction from leaves and standing in solution, homogeneous chlorophyll preparations very soon form several fractions which are adsorbed in separate bands on the chromatographic column. Some study was devoted to these changes by Strain (1949), who placed mallow or barley leaves in methanol for over 24 hours, and studied the products formed in the presence and in the absence of air. Prolonged treatment with methanol turned chopped mallow leaves yel- low; crystals of the methyl chlorophyllides a and h, and a small amount of acidic chlorophyllides {i. e., free chlorophyllins) a and b (extractable with 0.01 N KOH), were formed in them. In addition, new green and yellow-green pigments were produced, the spectra of which were similar to those of the chlorophylls a and h, but which gave no positive phase test 1774 CHEMISTRY OF CHLOROPLAST PIGMENTS CHAP. 37B and could thus be presumed to be oxidation products. They were not formed in the absence of air, or in dried leaves moistened with water and then immersed into methanol (even in the presence of air). Formation of the methyl chlorophyllides as well as of the colored oxidation products could be prevented by preliminary exposure of leaves to boiling water or steam, or to a temperature of 100° C. in a sealed tube — presumably because heat destroyed l)oth the chlorophyllase and the enzymes catalyz- ing oxidation. Sometimes pheophytins also were formed during the treatment of leaves with methanol; this could be prevented by the addition of a base, such as dimethylaniline. The formation of the chlorophyllides was faster when a smaller volume of methanol was used, or aqueous methanol (50-75%) was employed instead of pure solvent. In barley leaves, no chlorophyllase action occurred upon standing in methanol ; if air was absent, they yielded mainly the original chlorophylls a and b ; in air, the two above-mentioned, nonacidic oxidation products. The oxidation enzymes resisted freezing but not dehydration. They were most active in young seedlings. Strain analyzed products formed by standing in methanol also in leaves of seventeen other species, and in isolated chloroplasts of Swiss chard. Some gave mainly unaltered chlorophylls, and others chlorophyllides, pheophytins or oxidized chlorophylls. A 10-20 hour treatment with acetone or methyl acetate caused chloro- phyll in freshly chopped, or dried and moistened, mallow leaves to be con- verted largely to chlorophyllin. Heating of dried leaves to 100° for 10 minutes did not prevent this conversion, but it could be avoided by a 1-2 minute immersion into boiling water. When chopped mallow leaves were permitted to stand in acetone in air (but not in vacuum) for 24 hours, very strongly adsorbable pigments — presumably oxidation products of the chlorophyllins — were formed. Barley leaves treated with acetone in air yielded primarily the same oxidation products, which were also observed in treatment with alcohols. The green pigments formed from chlorophyll a by enzymatic oxidation were found to differ from those formed by allomerization in solution. Allomerized chlorophyll a is blue-green; absorption peaks lie {cf. fig. 21.4A) at 650 and 420 m/x (in methanol); on a sugar column, it forms a band well above chlorophyll a. Xo trace of this band is visible in the chro- matograph of enzymatic oxidation products from chopped barley leaves left standing in methanol. On the other hand, the most strongly adsorbed of the several minor green products formed in allomerization of chloro- phyll a in methanol, appeared to be identical with the main enzymatic oxidation product formed in leaves. Its band maximum lies at 662 m/x OXIDATION, ALLOMERIZATION AND REDUCTION OF CHLOROPHYLL 1775 in petroleum ether (blue-green solution) and at 667 m// in methanol (green solution). AUomerization of chlorophyll a' in methanol yielded the same series of pigments as that of chlorophyll a. Chlorophyll h also produced, l)y allomerization in methanol, several products, with the principal one having absorption bands shifted toward the shorter waves (to 631 and about 442 m/x in petroleum ether, 636 and about 458 mju in methanol ) . Experiments with etiolated barley plants, immersed into chlorophyll solution in methanol, showed that the chlorophyll entering the tissues is rapidly deposited there in the form of green ''crystals," which upon re- dissolving give no brown phase and therefore must be considered as oxi- dation product. Once plants had liecome green, the assortment of chlorophyll pigments in the leaves remains remarkably constant— in darkness as well as in light, in air, oxygen, carbon dioxide or hydrogen. In the disappearance of chlorophyll in ripening fruit or autumn leaves, no colored transformation products of the natural chlorophylls could be observed. Holt and Jacobs (1954) also found that allomerized chlorophyll (or chlorophyllide) a, formed by standing in air in methanolic solution, can be separated by chromatography into several components. In addition to the main fraction, with the spectrum of the tyj)e illustrated by Fig. 21. 4A ("fraction 2"), there was one more readily absorbed fraction ("fraction 3") with a spectrum practically identical Avith that of "native" chlorophyll a, and a smaller, less readily absorbed "fraction 1," with a red peak further toward the longer waves (c/. Fig. 37.C.3). All three fractions gave no "brown jjhase" with alkali. All of them could be reduced in light by ascorbic acid ("Krasnovsky reaction," cj. Chapter 35). All three fractions could be reversibly decolorized (oxidized?) by ferric chloride ("Rabino- witch- Weiss reaction," cJ. Chapter 16, section B3, and this section, further below). Infrared spectroscopy (c/. Chapter 37C, section 2e) indicated that fraction 3 alone contains the C=0 group in position 9. If negative phase test is attributed to incapacity for enolization of this group, this incapacity must be attributed, in fraction 3, to an oxidative substitution of the H-atom in position 10, e. g., by a methoxy group: V III V III llO 9 I +CH3OH I I ^^ ^ (aJA.l) HC— C— C= > CH3OC— C— C= + HoO I II +'/'^' I II o o In fractions 2, allomerization must involve a more drastic change — probably disruption of ring V, with the formation of a lactone (as suggested by Fischer and Pfeiffer, 1944): 1776 CHEMlSTllY OF CHLOROPLAST PIGMENTS CHAP. 37B V III III llO 9 I +CH3OH I I (37A.2) HC— C— C= > CH3OC— O— O— C— C= + H,0 I II +1V2O2 I II o o The difference between the fractions 2 and the (small) fraction 1 re- mains to be interpreted. The chlorophyll reaction with ferric chloride, described by Rabino- vvitch and Weiss and interpreted by them as oxidation (Vol. I, p. 464), also was studied by Strain (1949). He noted that extraction of the yellow product with water and petroleum ether led to regeneration of chlorophyll a, and that addition of dimethylaniline prevented the formation of the yel- low product. He concluded from these observations that the reaction in ciuestion is not an oxidation. However, the interpretation of Rabinowitch and Weiss was based on the assumption of an equilibrium between the re- duced and the oxidized form — and in this form their hypothesis is not contradicted by Strain's observations. Strain confirmed that the decoloration reaction with ferric chloride occurred also in allomerized chlorophyll a, and in the above-described green product of enzymatic oxidation in chopped leaves. Objections against the interpretation of the reaction of chlorophyll with ferric salt as reversible oxidation were raised also by Ashkinazi, Glikman, and Dain (1950). They noted that bleaching of the red al)Sorption band of chlorophyll in methanol can be caused not only by Fe+^ salts, but also by the salts of Al+^ and Sn+" (while Cu^^ and Zn+2 caused an enhancement of the red band, and K+, Ca+'-, Pb+'' and Mn+2 had little or no effect). The authors suggested that the bleaching effects of Al+^, Sn+- and Fe+^ are caused primarily by the acidity of these salts, which leads to pheophytiniza- tion of chlorophyll (and conseciuent weakening of the redbandandincreasein absorption in the green). The ions Fe+^, Cu+- and Zn+- (but 7iot Sn+- and Al+^) react with pheophytin, replacing hydrogen and forming metal complexes; in these complexes, the red band is restored to the same (or even greater) intensity it had in chlorophyll. Ashkinazi et al. suggested that a similar complex, but with a much weaker red band, is formed also by pheophytin with Fe+^ ions; the Rabinowitch- Weiss "reversible oxida- tion" was thus reinterpreted by them as conversion of Mg-pheophytin (chlorophyll) into olive-green pheophytin, followed by formation of greenish- yellow ferripheophytin complex, or a bright green ferropheophytin com- plex. Ashkinazi and Dain (1951) prepared a pheophytin-ferrous iron com- plex by heating ferrous acetate with pheophytin (a -|- h) solution in acetic acid, extracting with chloroform, evaporating to dryness, washing out the iron salt and dissolving in ethanol. Exposure to air caused oxidation of the OXIDATION, ALLOMERIZATIOX AND REDUCTION OF CHLOROPHYLL 1777 complex, with a band shift from 645 to 610 m^u; in strong light, the process was reversed and the band returned to 645 Tn/j,. Reactions of this type had been observed also by Rabinowitch and Weiss; however, these reactions are quite different from the instantaneous and completely reversible transformation to which the hypothesis of re- versible oxidation was applied. Ashkinazi et al. loft chlorophyll and ferric chloride to react in the dark for 24 hours before observing the spectral change; according to Rabinowitch and Weiss, reversibility is lost after a few minutes. It is quite likely that a chlorophyll solution left standing for several hours with ferric chloride will be converted to pheophytin. This conversion is essentially irreversible (except via the Grignard reaction). Green pheophytin complexes of different divalent ions, Zn + +, Cu + "'", etc., are well known; they can be formed by direct substitution, but not in- stantaneously; and their absorption spectra differ markedly between them- selves, and from that of the Mg + + derivative (chlorophyll). Watson (1953) confirmed the finding of Rabinowitch and Weiss that the decoloration of chlorophyll a in methanol by ferric chloride is spectro- scopically fully reversible by immediate addition of reducing salts, such as CU2CI2, and therefore must be attributed to regeneration of the original pigment rather than to the formation of new green compounds (such as the metal complexes of pheophytin, suggested by Ashkinazi et al.). On the other hand, the much slower restoration of the green color upon stand- ing, or upon the addition of non-reducing salts, such as NaCl, leads to "allomerized" chlorophyll, with its distinct spectrum; regeneration by hydrociuinone results in a still different, unknown green compound. In- terpretation of the reversible reaction wdth FeCls as oxidation is supported, according to Watson, by analogy with the transitory bleaching of chloro- phyll observed upon addition of bromine or iodine, and by the formation of allomerized {i. e., oxidized) chlorophyll upon standing of the decolorized solution. The acceleration of the transformation by non-oxidizing salts may be similar to the known acceleration of allomerization by LaCl2; dissolved oxygen is needed in the latter case as oxidant. Dunicz, Thomas, Van Pee and Livingston (1951) built a flow system to study spectroscopically the intermediates of chlorophyll reaction with alcoholic alkali {"phase test"). They found that the brown intermediate [an ether, ionized either at C(10) or at the enol at C(9)] is formed without the participation of oxygen, and is converted into a green chlorine by reac- tion with oxygen. In the absence of oxygen, a slow irreversible reaction ensues. Weller (1954) measured the absorption spectrum of the phase test intermediate with much better precision than Dunicz et al. Its bands were located (in pyridine) at 683, 524, -186, 428 and 375 m/x for chlorophyll a, and at 630, 558, 560, 505 and 444 mu, for chlorophyll h (main bands itali- 1778 CHEMISTRY OF CHLOROPLAST PIGMENTS CHAP. 37b 3000 4000 5000 WAVE LENGTH A 6000 7000 Fig. 37B.5. Absorption spectra of the "brown phase." Chlorophyll a in 10% mono- isopropyl amine, 10% isopropyl benzene and 80% 1:1 propane-pro pene: ( ) 190° K., ( ) 170° K., ( ) 160° K. (after Freed and Sancier 1953). Curve (...) has been derived by Dunicz el al. (1951) for the transient intermediate of chlorophyll a at 300° K cized). Weller suggested that the brown intermediate is a triplet (diradi- cal) form of the anion formed by enolization and acid dissociation in posi- tion 9. This diradical is stabilized by resonance between several struc- tures, with free valencies in different positions in the — normally conjugated — ring system. The striking similarity of the absorption spectra of the phase test intermediate, the anaerobically bleached chlorophyll (in steady or flashing light), the reversibly reduced chlorophyll and the reversibly oxidized chlorophyll (in rigid solvent) can be understood if all these prod- ucts have a similar radical or biradical structure, with interrupted all-round conjugation. OXIDATION, ALLOMERIZATION AND REDUCTION OF CHLOROPHYLL 1779 The brown intermediate of the "phase test" was studied by Freed and and Sancier (1953) by low-temperature spectroscopy. They found that if chlorophyll a (or b, or h') is dissolved at low temperature in a base (isopro- pylamine), a brown (or red) solution is formed, which becomes green upon warming and again brown (or red) upon (immediate) cooling. The reac- tion becomes irreversible if the solution in pure amine is allowed to stand after warming; but if the amine is diluted (e. g., 10% amine in 45% pro- pane -I- 45% propene; or 10% amine in 40% propane + 40% propene + 10% isopropyl benzene), the reversibility is preserved. The absorption spectrum of the brown (or red) solution is very similar to that given by Dunicz et al. (1951) for the "brown phase" (Fig. 37B.5). The green color of the solution in amine-containing solvent can be "frozen in" by sudden cooling from room temperature to —190° C; upon warming up, the "brown phase" reappears in a certain interval of temperatures. It is thus confirmed that the brown product exists in equilibrium mth green chloro- phyll, and that its formation (by enolization?) requires an activation en- ergy. The suggested identification of the brown (or red) low-temperature products of the reaction ^\^th amines, with the similarly-colored intermedi- ates of the phase test, is supported by the observation that no such products are obtained with allomerized chlorophyll. No red compound could be ob- tained from chlorophyll b in the presence of diisopropylamine, indicating that basic reaction alone is not sufficient for its formation. The reversible chemical reduction of chlorophyll to a leuco compound by zinc and organic acid, claimed by Timiriazev, and Kuhn and Winter- stein, but found to result in irreversible changes by Albers, Knorr, and Rothemund {cj. Vol. I, p. 457), was again studied by Kosobutskaya and Krasnovsky (1950). They used chlorophyll (o + b, a, b), pheophytin, the complexes of pheophytin with zinc and copper, and magnesium phthalo- cyanide. The addition, to 3 ml. of 10^^ M pigment solution in pyridine, of 0.3 g. zinc powder in 0.0() ml. glacial acetic acid, under vacuum, caused brownish discoloration of all pigments; phthalocyanin became entirely colorless. Re-admission of air restored the green color in every case; however, only with two compounds — Mg phthalocyanin and Zn pheo- phytin— were both the red and the blue-violet peak restored exactly to their original positions. In compounds of the a-series, the bands of the final product lay at 661 and 431 m^— the positions they have in Zn pheophytin. With chlorophyll b, the bands were at 656 and 470 m^u before, and at 644 and 459 mn after the "cycle." When very concentrated solutions of chlorophyll b were used, a weak band in the region 510 530 m/x was noticeable in the reduced state (cf. Chapter 35 for the presence of this band in photochemicaUy reduced chlorophyll). With the nonfluorescent Cu pheophytin, the "cycle" leads to the an- 1780 CHEMISTRY OF CHLOEOPLAST PIGMENTS CHAP. 37B pearance of fluorescence; the shape of the final absorption spectrum could be explained by assuming the presence of a mixture of Zn pheophytin with unchanged Cu pheophytin. In all cases, the "cycle" led to the destruction of a considerable fraction of the pigment (50% in the case of chlorophyll a, 80% in that of phthalo- cyanin). The order of increasing velocities of reduction was phthalocyanin, chlorophyll a, chlorophyll 6, Zn pheophytin, pheophytiti. Reoxidation required 1.5-2 hours at room temperature with chlorophyll a, and up to 10 hours with phthalocyanin. The general conclusion was that Timiriazev's reaction is irreversible because it leads to the replacement of magnesium by zinc, and not because it causes hydrogenation of the vinyl group in ring I (as tentatively suggested in Vol. I, p. 457). Godnev (1939) applied Timiriazev's method to protochlorophyll, and found similar results — decoloration with zinc and acetic acid, restora- tion of green color upon admission of air. Marked differences between the absorption spectra before and after the experiment were noted by Godnev and Kalishevish (1944). The experiments of Linschitz and co-workers on 'photochemical oxida- tion, and of Krasnovsky and co-workers on photochemical reduction of chlorophyll and its analogues, and the reversals of these reactions in the dark, have been described in chaptei- 35. The abnormal position of chlorophyll a baud in piperidine was noted on p. 642. More recent studies (Weigl and Livingston 1952; Holt, unpub- lished) showed that this band shift is caused by an irreversible reaction of the pigment ^\dth this strongly basic solvent. Similar reactions occur with other strong bases, e. g., parabenzoyl amine (cf. Chapter 37C, section 2a). A peculiar behavior was noted by Freed and Saucier (1954) also when chlorophyll b was dissolved in a solvent containing a secondary amine (e. g., diisopropylamine). At — 188° C. this solution showed a sharp double band (430 and 480 m/x) in place of the single band (at 450 m/x) which characterizes chlorophyll b solutions at room temperature; in other solvents only a broad absorption at 450-480 m^ appeared at low tempera- tures, indicating the formation of solvates. Upon warming of the solution of diisopropylamine, an irreversible transformation of chlorophyll b took place. Weller and Li^'ingston (1954) pointed out that a clea^'age of i-ing V by amines was observed in 1936 by Fischer and Gobel in methyl pheo- phorbide, and that the product was identified by them as chlorin-6-acet- amide. The bond between the C-atoms 9 and 10 is broken in this reaction, the H-atom being added in position 10 and the amide residue in position 9. OXIDATTOX, ALLOMERIZATIOX AXD REDUCTIOX OF CHLOROPHYLL 1781 /- (37B.3) HC — C I II O Weller and Livingston measured the rate constants of this reaction spectroscopically with seven different amines. They ranged, in the case of chlorophyll a at 20° C, from log /.' = — 1.88 sec.~^ with piperidine, to log k' = —5.38 sec.~' for phenylhydrazine, and < — 8.0 sec.~' for aniline. The order of rate constants parallels (for both chlorophylls a and /;) the order of basicity of the amines (as measured by p/v^ in water). Allomerized chlorophyll a in isobut3damine showed no reaction even after 3 days. Temperature effect on the rate of reaction of Chi a with isobutylamine indi- cated a very small activation energy (<() kcal) ; the (negative) entropy of activation must therefore be large (to account for the relatively low abso- lute rate at room temperature) . In weekly basic amines the reaction is so slow that the reversible for- mation of an amine -|- pigment complex can be observed, as described by Krasnovsky and Brin (cf. Chapter 37C, section 2a). The phase test reaction with alkali differs from the reaction of chloro- phyll with amines in several respects : (7 ) It goes via a brown intermediate, which is formed at room temperature within less than 1 sec, and disap- pears within 20-30 sec. (in the system chlorophvll a in pyridine -f meth- anolic KOH); {2) the energy of activation is >20 kcal, and (3) the pres- ence of at least a small amount of alcohol is needed for the reaction. The allomerization process has characteristics similar to those of the phase test. In analogy with the accepted mechanism of aminolysis of esters, Weller and Livingston suggested the following mechanism of cleavage of ring \ , in its keto form, by amines: (37B.4) V HC -C II o + RNHo Transient — addition > product V 10 9 HC- CNHo+R O ~* ilO 9 H2C CNHR O The basicity of the amine is determined by the charge accumulation on the 1782 CHEMISTRY OF CHLOROPLAST PIGMENTS CHAP. 37B nitrogen, and the same factor also determines the tendency of the amine to attach itself to the electrophilic carbon atom (9). According to Holt (1954) the brown "phase test intermediate" can be obtained instantaneously by the action of sodium methanolate on chlorophyll or other solvents (and is then comparatively long-lived at room temperature). Weigl and Livingston (1952) could observe no isotopic exchange of hydrogen and deuterium between watei' and chlorophyll a (or pheophylin a) in neutral organic solvents (ether, dioxane, acetone, benzene); .similar negative results, ohtaincil by Xorris, Ruben and Allen with tritium, were described on p. .557. 5. Crystallization and Stability of Chlorophyll and Bacteriochlorophyll (Addendum to Chapter 16, section A5) Despite occasional reporting of microcrystalline chlorophyll prepara- tions, it seems that all pre\dously recommended methods lead to amor- phous, more or less wax-like products. In particular, the procedure by which Willstatter and StoU had obtained a product described as "micro- crystalline," was found by Hanson (cf. p. 448) to yield a preparation with- out a sharp x-ray diffraction pattern (Fig. 37B.7) — which is decisive proof of the absence of a regular molecular arrangement. Jacobs, Vatter and Holt (1953, 1954) were able to obtain crystalline chlorophylls a and b, as well as crystalline bacteriochlorophyll, by using as guide the shift of the absorption band upon crystallization, first noted with alkyl chlorophyllides (cf. Chapter 37C, section 3). The several meth- ods they found to produce crystalline chlorophyll preparations had in com- mon the presence of water during the precipitation, although this presence alone is not sufficient for the purpose. Thus, chlorophyll precipitated from acetonic solution by dilution with water remains amorphous unless a small amount (> 100 p. p.m.) of Ca + + ions is present. (Calcium is not incorporated in the precipitate; its effect must be an electric one, destroy- ing the stability of the amorphous colloid.) Other and better methods to prepare crystalline chlorophyll are: (1) adding water to a chlorophyll solution in ether, and slowly evaporating the ether in vacuum; (2) adding pentane to water-saturated ethereal solution of chlorophyll and removing the ether by repeated washing with water; and (3) adding hexane to the ether- water solution and removing ether by evaporation. In contrast to the chlorophylls (and bacteriochlorophyll), the pheophytins can be crys- tallized by evaporation of their ethereal solutions also in apparently com- plete absence of water. The difficulty of crystallizing chlorophyll is thus caused by the presence, in the same molecule, of the long hydrophobic phytol chain, and the polar magnesium atom. Fig. 37B.6 is an electron micrograph of crystalline chlorophyll a. It shows that the crystallization extends mainly in two dimensions; the ma- CRYSTALLIZATION AND STABILITY 1783 terial forms sheets only a few molecules thick, which have the tendency of rolling into cylinders upon drying. The extreme thinness of the crystals explains why their diffraction patterns (Fig. 37B.7) are less sharp than those of crystalline ethyl chlorophyllide, in which the crystals (although, they, too, have the tendency to grow mainly in two dimensions) reach much A IJUL w^sstiiaa C IJU B IJLX D IjJL Fig. 37B.6. Electron micrograph of crystalline chlorophyll a (after Jacobs, V'atter and Holt 1954). greater thickness. Fig. 37B.8 shows microcrystals of ethyl chlorophyllide a of different size. Fig. 37B.9 is a model of their crystal structure, as sug- gested by Hanson (cf. Chapter 16, p. 448). The difference in the x-ray diffraction patterns indicates that Hanson's chlorophyllide model does not apply to chlorophyll, but no definite struc- ture can yet be proposed for the latter. The diffraction pattern of methyl 1784 CHEMISTRY OF CHLOKOPLAST PIGMEXTS CHAP. 37b bacteriochlorophyllide shows no evidence of a three-fold screw axis, sug- gesting that the successive layers may be oriented in parallel (or rotated by 90°, instead of by 120°, as in chlorophylHde; compare the shapes of the conjugated ring systems in Fig. 37C.1!). The diffraction pattern of the CHLOROPHYLL 14.8 7.6 CHLOH)PHniL b 7.6 4 MCTERIOCHLOBOPHYLL CHLOROPHYLL a (Hanson) Fig. 37B.7. X-ray diffraction patterns of microcrystals of chlorophyll a and b and of bacteriochlorophyll (after Jacobs, Vatter and Holt 1954). o o pheophorbides shows a spacing of 16 A, instead of 12.8 A as in the chloro- phyllides, suggesting that the molecules in each layer may be standing up vertically, instead of being inclined by 55° (16 = 12.8/sin 55°) (Jacobs 1954). The absorption spectrum of crystalline chlorophyll will be described in Chapter 37C, together with that of the crystalline chlorophyllides (cf. Figs. 37C.16to 18). CRYSTALLIZATION AND STABILITY 1785 In contrast to the often reported instability of amorphous chlorophyll preparations, crystalline powders of chlorophyll and bacteriochlorophyll were found unchanged after six months storage at room temperature, at »*«?^ .<* Pig. 37B.8. Microcrystals of the chlorophyllides (after Jacobs and Holt 1954). least as judged by the criterion of a high ratio between the absorption coef- ficients in the two main peaks and in the trough between them. (These spectroscopic "purity ratios," suggested by Zscheile, are particularly sen- sitive to loss of magnesium.) 1786 CHEMISTRY OF CHLOROPLAST PIGMENTS CHAP. 37B J^ Fig. 37B.9. Model of the crystal struetiire of chlorophyllides (after Hanson] 6. Nature of Chlorophyll c (Addendum to Chapter 15, section B3) The chemical nature of chlorophyll c was investigated for the first time by Granick (1949). The absorption spectrum of this faintly green pigment {cf. insert in Fig. 21.5C, p. 615) is quite similar to that of proto- chlorophyll (Fig. 21.8). By analogy, it can be surmised that chlorophyll c is a porphyrin and not a chlorin. Granick's tests indicated that chloro- phyll c contahis magnesium (bound more strongly than in chlorophyll) but no phytol. Positive phase test indicates the presence of the cyclopen- tanone ring. 7. Chemistry of Bacteriochlorophyll and Bacterioviridin (Addendum to Chapter 16, section A3) Holt and Jacobs (1954') found that bacteriochlorophyll is not as un- stable as it has been described. A green oxidation product, which is easily BACTERIOCHLOROPHYLL AND BACTERTOVIRIDIN 1787 formed during the extraction, can be separated from the bulk of the blue pigment by chromatography; after bacteriochlorophyll has been so puri- fied it proves (luite stable, not only in crystalline form (c/. section 5 above), but also in ethereal solution. The green oxidation product, obtained in the preparation of bacterio- chlorophyll, probably has the oxidation level of dihydroporphin, and may be identical with the green compound obtained by Schneider (c/. p. 445) by oxidation of bacteriochlorophyll with ferric salts, iodine, quinone and other mild oxidants, and by Krasnovsky and Voynovskaya (1951) by the action of quinone on a bacteriochlorophyll solution in toluene. The latter ob- servers reported that the oxidation can be reversed by ascorbic acid— an interesting suggestion which needs confirmation. The oxidation product has a chlorophyll-type spectrum, with the main absorption bands at 432 and 677 mu (as measured by Holt and Jacobs in ether) ; it fluoresces with. red hght. Fischer (p. 445) suggested that it may differ from bacterio- chlorophyll only by the absence of two H atoms in ring II, and also that the "bacterioviridin" of green sulfur bacteria may be identical with it. Seybold and Hirsch (1954) reported that the azure-blue, non-fluorescent (more correctly: infrared fluorescent!) bacteriochlorophyll (which they called "bacteriochloro- phyll a") is, in vitro, "extraordinarily unstable," and is converted "in a few minutes" into a green "bacteriochlorophyll b" (obviously the above-mentioned green oxidation product!). The question of the chemical identity of hacteriovindin remains to be settled. Barer and Butt (1954) found, in pigment extract from green bac- teria, two main absorption bands, at 664 and 434 m^, respectively. The first one is markedly different from the red band of the oxidation product of bacteriochlorophyll (677 m/x) ; both bands are almost coincident with the peaks of chlorophyll a. Barer asked whether ''bacterioviridin" could not be identical with chlorophyll a; but noted differences in the position of the minor bands and in the chromatographic behavior, which argued against this identity (cf. also Katz and Wassink's curves in Fig. 21.7!). Seybold and Hirsch's absorption curve of an extract from Microchloris showed three peaks, including, in addition to Barer's "red" and "violet" peaks, also a peak in the far red, at 770 m/x, just where the bacteriochlorophyll band is usually found. Seybold and Hirsch interpreted this as evidence that green bacteria contain, in vivo, the same "bac- teriochlorophyll a" as the purple ones, but that this pigment is rapidly converted (oxi- dized) into "bacteriochlorophyll 6" ( = bacterioviridin) upon extraction. An alternative explanation is, however, that Sej'bold and Hirsch's culture of Microchloris, when used for pigment extraction, was contaminated with purple bacteria. (The absorption curve given by them for live Microchloris cells shows a weak but unmistakable band at 850 m^, typical of purple bacteria ! ) We will see in Chapter 37C (section 6c) that the absorption spectra of live green (and purple) bacteria, given by Seybold and Hirsch, also disagree with those found by 1788 CHEMISTRY OF rilLOROPLAST PIGMENTS CHAP. 37b other observers, and suggest contamination of the green cells with purple ones, and of tlic puii)l(' foils with green ones. 8. Molecular Structure and Properties of Phycobilins (Addendum to Chapter 17, part B) Lemberg and Legge (1949) reviewed the data on bile pigments, in- eluding the phycobilins. Structural formulae of phycoerythroljilin and phycocyanobilin, suggested by Seidel (1935), are represented below in a form which emphasizes their similarity to porphyrins and chlorins. COOHCH.CH HaC C2H5 (I) Seidel's mesobiliviolin ( = phycocyanobiliii ?) COOHCH.CH H3C C.H. (II) Seidel 's mesobilirhodiu (, = phycoerythrobihn?) COOHCH,CH CH.CHoCOOH H3C C2H5 (III) Lemberg and Legge's suggested formula for phycoerythrobilin The two formulae, I and II, differ only by the position of one imino hydrogen, and consequent alteration in the path of the (seven-membered) conjugated double bond system. The question was raised in Vol. I (p. 443) whether, in the case of porphin derivatives, arrangements of this type would be stable isomers, tautomers or mesomers. In the case of bilan derivatives, mesomery or tautomery is less likely, because of the open- ness of the whole structure. Their isomers could therefore be stable; the considerable spectroscopic difference between the two pigments makes it unlikely that they are isomers containing conjugated double bond systems of the same length, as implied in the formulae (I) and (II). Lemberg and PHYCOBILINS 1789 Legge suggested, for phycoerythrobilin, the structure III, with only five conjugated double bonds. This seems, however, too short for a compound with a strong absorption band in the green. (Lemberg and Legge said that the spectroscopic properties of erythrobihn are similar to those of mesobilin b — a compound with a structure of type III — but mesobilin h is yellow and its first aljsorption band lies at 452 m/i in dioxane, and at 450 m/x in HCl + alcohol, while those of the red-violet erythrobilin lie at about 530 and 560 ni/x, respectively.) (Only one band, at 498 m/x, was listed in our table on p. 665; but Lemberg and Legge give, for erythro- bilin in 5% HCl, two bands — at 560 and 495 m/x, respectively.) Lemberg and Legge refer to one difficulty, pointed out on p. 666 (Vol. II, 1): the implausibly high absolute values (3 X 10^) calculated for the molar extinction coefficients of the phycolnlins. They suggest that the estimate of 8 chromophores (ii/ = 536) per chromoproteid molecule {M = about 200,000) used in this calculation (about 2% phycobilin by weight !) has been too low, and that the number of chromophores per chromoproteid molecule may be as high as 16; the previously calculated molar extinction coef- ficients will then have to be halved. The phycocyanin of Oscillatoria was hydrolyzed at 100° C. in 6 N HCl, and subjected to fractionation by paper chromatography by Wassink and Ragetli (1952). Sixteen amino acids were found, and 13 of them identi- fied, one of the "unknowns" was a major component. Non-occurrence of arginine is the only notable difference of the phycocyanin protein from the proteins of Chlorella, or leaf chloroplasts, or hemoglobin. Swingle and Tiselius (1951) developed a chromatographic method for the separation of phycochromoproteins. Its application led to the dis- covery of several new pigments of this class. Koch (1953) found a new phycoenjthrin (in addition to the two or three varieties encountered in Rhodophyceae and Cyanophijccae) in one of the Bangiales, a rather primiti\'e red algal group; it is characterized by the absence of the 495 myu absorption peak (c/. Fig. 21.39). A new phycocyanin was isolated chromatographically, from l)oth blue- green and red algae, by Haxo, O'hEocha and Strout (1954), and tentatively named "P-phycocyanin." It has a single absorption peak at 650 m/x, as contrasted to the "R-phycocyanin" from red algae (which has peaks at 555 and 617 m^t), and the "C-phycocyanin" from blue-green algae (which has a single peak at 615 m/tx, cf. Fig. 21.40). BUnks (1954) suggested that "R-phycocyanin" may be a complex of C-phycocyanin (X^ax. 617 m/x) with phycoerythrin (A^ax. -555 m^u). The evidence is electrophoretic and chromatographic ; in both types of experi- ments, a mixture of the two pigments is observed to behave, under certain conditions, as a single, anionic entity. 1790 CHEMISTRY OF CHLOKOPLAST PIGMENTS CHAP. 37B Haxo, O'hEocha and Strout (1954) found typical "P-phycoerythrin," with peaks at 497, 537 and 566 mn, in Rhodymenia palmata and Polyneura latissima, and typical "C-phycoerythrin," with a single peak at 560 m/x, in Phormidium persicinum, Ph. fragile, and Nostoc. On the other hand, Ph. ectocarpi yielded a phycoerythrin with only two peaks, at 542 and 56() m/x. In Bangiales, such as Prophyra tenera or P. perforata, they found a phyco-> erythrin with peaks at 497 and 566 mn, but with no definite middle peak. Phycoerythrin from another of the Bangiales, Porphyridium cruentum, had only one broad band, at 545 mu, and three bands in the ultraviolet, at 368, 306 and 276 m/t, the second of which is not present in R-phycoery- thrin. Krasno^'sky, Evstigneev, Brin and Gavrilova (1952) found that phy- coerythrin can be extracted from CaUilhamnion rybosum more conveniently than from Ceramium. They purified the extracted pigment by chromatog- raphy on a tricalcium phosphate column. Developing with 0.15 M di- sodium phosphate permitted complete separation from phycocyanin. The product, extracted into 0.15 M Na2HP04, showed two protein fractions in the ultracentrifuge, Avith M = 300,000 and M ^ 50,000, respectively. The phycoerythrin solution was found to be photochemically stable against oxidation by air, as well as against reduction by ascorbic acid. It therefore did not sensitize the reduction of riboflavin or safranin by ascorbic acid. (These results, belonging in Chapter 35, were not mentioned there.) Pho- toxidation was accelerated by dioxane or pyridine, and seemed to in\-olve a partial separation of the chromophore from the protein (drop of absorption at 565 and 540 m/x, preservation of the 495 m/x peak). This observation may be interesting from the point of view of a search for better methods of separation of erythrobilin from its associated protein — a most important hurdle in the study of these pigments. The methods in use at present — hydrolysis with hot hydrochloric acid or alkaline hy- drolysis (Bannister 1954) — destroy a large part of the chromophore while prying it off the protein. Bibliography to Chapter 37B Chemistry of Chloroplast Pigments 1939 Godnev, T. N., I'chenyje Zapiski Beloniss. Univ. (Chem. Ser.), 1939, No. 1, 15. 1943 Noack, K., Biochem. Z., 316, 166. 1944 Fischer, H., and Pfeiffei'. H., Ann. Chem. Ju.^lus Liehigs, 555, 94. Godnev, T. N., and Kalishevish, S. V., Trudy Inst. Fisiol, Rmtmij,, 1944, No. 2, 160. BIBLIOGRAPHY TO CHAPTER 37B 1791 1946 Frank, S. R., /. Gen. Physiol., 29, 157. Yocum, C. S., Paper at Boston Meeting of Am. Botan. Soc, December, 1946. 1947 Goodwin, R. H., and Owens, 0. H., Plant PhijsioL, 22, 197. Smith, J. H. C, J. Am. Chem. Soc, 69, 1492. 1948 Koski, V. M., and Smitli, J. H. C, ibid., 70, 3558. Smith, J. H. C, and Koski, V. M., Carnegie Inst. Wash. Yearbook. 47, 95. Granick, S., J. Biol. Chem., 172, 717; 175, 333. Smith, J. H. C, Arch. Biochem.. 19, 449. Euler, H. von, Bracco, M., and Heller, L., Compt. re7ul., 227, 16. Provasoli, L., Hutner, S. H., and Schatz, A., Proc. Soc. E.rptl. Biol. Med., 69, 219. Brebion, G., Galliga Biolog Acta, 1, 24, 124. 1949 Smith, J. H. C., "Processes Accompanying Chlorophyll Formation," in Photosynthesis in Plants, Iowa State College Press, Ames, Iowa, 1949, pp. 209-217. Strain, H. M., "Functions and Properties of the Chloroplast Pigments," ibid., pp. 133-178; and private communication. Schwartz, D., Botan. Gaz., HI, 123. Lemberg, R., and Legge, J. W., Hematin Compounds and Bile Pigments, Interscience, New York-London. Granick, S., /. Biol. Chem., 179, 505. 1950 Granick, S., ibid., 183, 713. Blaauw-Jansen, G., Komen, J. G., and Thomas J., B., Biochim. et Biophis Acta, 5, 179. Koski, V. M., Arch. Biochem., 29, 339. Highkin, H. R., Plant Physiol., 25, 294. Kosobutskaya, L. M., and Krasnovsky, A. A., Compt. rend. (Doklady) Acad. Sci. USSR, 74, 103. Bogorad, L., Abstr. Botan. Soc. of America, Sept. 1950. Granick, S., Harvey Lectures, 44 (1948-49), C. C Thomas, Springfield, Illinois, 1950, pp. 220-244. Salomon, K., Altman, K. I., and Rosa, R. D., Federation Proc, 9, 222. Ashkinazi, M. S., Glikman, T. S., and Dain, Al. Y., Compt. rend. (Doklady) Acad. Sci. USSR, 73, 743. 1951 Koski, V. M., French, C. S., and Smith, J. H. C, Arch. Biochem., 31, 1. Smith, J. H. C, Carnegie Inst., Dept. Plant Biol. A^in. Repts., 50, 123. Granick, S., Ann. Rev. Plant Physiol., 2, 115-144. Ashkinazi, M. S., and Dain, M. Y., Compt. rend. (Doklady) Acad. Sci. USSR, 80, 385. Provasoli, L., Hutner, S. M. and Pintner, 1. J., Cold Spring Harbor Symp. Quant. Biol., 16, 113. Freed, S., and Sancier, K. M., Science, 114, 275. Krasnovsky. A. A., and Voynovskaya, K. K., Compt. rend. (Doklady) Acad. Sd. USSR, 81, 879. 1792 CHEMISTRY OF CHLOROPLAST PIGMExNTS CHAP. 37B Dunicz, B., Thomas, T., Van Pee, M., and Livingston, R., /. .4m. Chem. Soc, 73, 3388. Swingle, S. M., and Tiselius, A., Biocheni. J. London, 48, 171. Koski, V. M., and Smith, J. H. C, Arch. Biochem. and Biophys., 34, 189. 1952 Bogorad, L., and Granick, S., Am. Soc. Plant Physiol. Abstr., Cornell Univ. Meeting, Sept. 1952. Weigl, J. W., and Livingston, R., J. Am. Chem. Soc, 74, 3452. Weigl, J. W., and Livingston, R., ibid., 74, 4160. Freed, S. and Sancier, K. M., Science, 116, 175. Wassink, E. C, and Ragetli, W. J., Proc. Roy. Acad. Amsterdam (6), 55, 462. Krasnovsky, A. A., Evstigneev, V. B., Brin, G. P., and Gavrilova, V. A., Compt. rend. (Doklady) acad. sci. USSR, 82, 947. 1953 Jacobs, E. E., Vatter, A. E., and Holt, A. S., /. Chem. Phys., 21, 2246. Godnev, T. N., and Terentjeva, AT. V., Compt. rend. {Doklady) acad. sci. USSR, 88, 725. Krasnovsky, A. A., and Kosobutskaya, L. M., ibid., 91, 343. Koch, W., Arch. MikrobioL, 18, 232. Freed, S., and Sancier, K. M., Science, 117, 655. Strain, H. H., ibid., 117, 654. Watson, W. F., J. Am. Chem. Soc, 75, 2522. 1954 Smith, J. H. C., and Benitez, A., Plaiit Physiol., 29, 1 35. Smith, J. H. C., ibid., 29, 143. Freed, S. and Sancier, K. M., /. Am. Chem. Soc, 76, 198. Weller, A., and Livingston, R., ibid., 76, 1575. Haxo, F., O'hEocha, C., and Strout, P. M., Rapports Commnn., Section 17, 8th Congr. Botany, Paris, Jul}' 1954. Blinks, L. R., "The Role of Accessory- Pigments in Photosynthesis," in Symposia Soc. Gen. Microbiol., 4, 224, Cambridge LTniv. Press, Cam- bridge. Granick, S., "Metabolism of Heme and Chlorophyll," in Chemical Path- ivays of Metabolism, Vol. II, Chapter 16, Academic Press, N. Y. Bannister, T. T., Arch. Biochem. and Biophys., 49, 222. Jacobs, E. E., Vatter, A. E., and Holt, A. S., Arch. Biochem. and Biophys, 53, 228. Weller, A. and Livingston, R., /. Ayn. Chem. Soc, 76, 1575. Weller, A., ibid., 76, 5819. Freed, S., Sancier, K. M., and Sporer, A. H., ibid., 76, 6006. Holt, A. S. and Jacobs, E. E., unpublished. Jacobs, E. E., unpubhshed. Holt, A. S., unpublished. C. Spectroscopy and Fluorescence of Pigments* (Addenda to Chapters 21-24) 1. Theory of Chlorophyll Spectrum (Addendum to Chapter 21, Section A4) In Vol. II, Part 1 (pp. 619-635), a purely empirical discussion of regular- ities in the spectra of porphin derivatives was given. Since then, several papers have appeared dealing with the theoretical analysis of the spectra of large conjugated bond systems in general, and of porphin and its deriva- tives in particular. Simpson (1949), Kuhn (1949), and Nakajima and Kon (1952) used the free-electron model; Piatt (1950'''-) and Longuet-Higgins, Rector and Piatt (1950), the one-electron LCAO model (LCAO = linear combination of atomic orbitals). They arrived, in general, at similar results concerning the sequence of low excited states of porphin and tetrahydroporphin. (These two molecules were treated, in preference to dihydroporphin, be- cause of their simpler symmetry — D4h in porphin and Doi, in tetrahydro- porphin.) The basis of all theoretical considerations in this field is the well-founded postulate that ring systems such as porphin, containing a closed sequence of conjugated double bonds, are, similarly to benzene, naphthalene, etc., rigid planar structures. Their chromophoric properties are due to the excitation of the so called " x-electrons" in the conjugated ring system — electrons which can be considered as belonging to this system as a whole, rather than to an individual atom. In the more nearly circular porphin structure (fig. 37C.1A), the electric dipole oscillations corresponding to the combination of the ground state with the low excited states, while confined to the plane of the ring system, are not further restricted ("polarized") in respect to any special direction in this plane. An electron in a round potential box therefore provides an appropriate first approximation for the analysis of the term system of porphin and its derivatives, such as proto- chlorophyll — although the side chains, particularly such containing double bonds conjugated with the main ring system, may strongly affect the close- ness of the approximation. In the more elongated tetrahydroporphin * Bibliography, page 1882. 1793 1794 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37c structure, Avbich can be treated, in first approximation, as a rectangular — rather than circular— potential box (fig. 37C.1C), oscillations || and -L to the long axis of symmetry should have different frequencies and intensities; the corresponding absorption and emission bands should therefore consist of two components, differing in wave length and strength, polarized in two mutually perpendicular directions. ■=&. PORPHIN ( PROTOCHLOROPHYLL) B. DIHYDROPORPHIN (CHLOROPHYLL) RECT- ANGULAR BOX APPROXI- MATION TETRAHYDROPORPHIN TBACTERIOCHLOROPHYLL) Fig. 37C.1. Porphin ("round field"), dihydroporphin and tetra- hydroporphin ("long field") conjugated ring systems. Table 37C.I summarizes the term estimates, made by the two methods, for the two lowest excited terms of porphin, and the four lowest excited terms of tetrahydroporphin, and their attempted correlation with the empirical absorption bands. All these bands arise by transfer of the electron in the filled shell which has the highest angular momentum, 4, into the lowest empty shell. The combination of the angular momentum, 5, of this excited electron, with the angular momentum, 4, of its partner left behind after excitation, produces THEORY OF CHLOROPHYLL SPECTRUM 1795 two states, with the angular momenta of 1 and 9, respectively. The second state lies below the first one; its excitation gives rise to the red band, while transition to the state with angular momentum 1 must be responsible for the blue- violet band. Each of the two main levels is split in two, when the T)^^^ symmetry of porphin is destroyed by substitutions, or partial hydrogenation of the double bonds. Table 37C.I Calculated and Observed Energy Levels of Porphin and Tetrahydroporphin (after Platt 1950) Estimated level in cm."' (center of gravity, singlet + triplet) Polariza- tion Free electron model LCAO model Observed singlet (relative transition (band to Uncorr. Corr. L^ncorr. Corr. peak in ether) Transition type long axis) for N atoms for for N atoms N atoms for N atoms i>, cm ' X. niM PORPHIN A,u -* Eg A, u ->- Eg None None 9300 20200 15600 9000 20200 21300 14500 16500 16100 23200 621" 420* TETRAHYDROPORPHIN Bi „ — »■ Bsg ± 8900 13300 8700 13200 13000 772"" Au ^ Bsg 16200 16400 16100 10300 17400 575' Blu—*- B-ig ± 24100 2;)700 22200 21000 (25300) (395^'^ ) Blu — *■ Bog 23000 24100 23000 27200 27800 360" " 613 ni/ii in porphin in dioxane, cf. fig. 21.9; 621 mn in protochlorophyll in ether, cf. fig. 21.8. ^ "Soret band," 430 ni/i in porphin in dioxane, rf. fig. 21.9; same wave length in protochlorophyll in ether, cf. fig. 21.8. " 772 lii/j. in baeteriochlorophyll in ether, 771 m^i in methanol (Weigl 1952). ■^ 575 niju in baeteriochlorophyll in ether, 609 ni/u in methanol (\Veigl 1952): in Piatt's table a value of 613 ni/u was used. '" A weak band at 513 niyu, noted by Weigl in impure baeteriochlorophyll prepara- tions, belongs to spirilloxanthin (Holt and .Jacolis 1954-). ■'' A "satellite" of the 360 m/j. band, cf. section 2, below. " "Soret band" of baeteriochlorophyll; the different value (417 m^) in Piatt's original table was based on French's curve (fig. 21.6) that did not extend below 400 m/i. The spectrum of dihydroporphin (chlorin) which has only one short two-fold axis of symmetry, cf. fig. 37C.1, has not yet been analyzed; but we can expect it to resemble that of tetrahydroporphin more than that of porphin, because the circular symmetry of the conjugated system is destroyed by the hydrogenation of a single pyrrole ring. It can therefore be expected that the absorption bands of dihydroporphin derivatives — similarly to those of tetrahydroporphin derivatives — will be split into components polarized in the direction of the symmetry axis of the con- 1796 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37c jugat.ed bond system and normal to this direction, respectively. One can expect, however, to find this split less wide than in tetrahydroporphin. The spectrum of chlorophyll (and other dihydroporphin derivatives) shows, instead of the two widely separated long-wave bands of bacterio- chlorophyll (located, in ether, at 772 and 575 m/x, respectively), only one main band (at 660 mn in chlorophyll a in ether, cf. fig. 21.1). However, the relatively weak "orange" band (located at 613 m/x in ethereal solution), 20000r 16000 - 12000- 8000 - 4000 - 613 m^ 16,300 cm"' a„= 1.7x10* 660 m^ 15,200 cm"' am = 9x10* 575 m^ 17,400 cm "' «„ = 2.0x10* 772 m^ 13,000 cm."' om-- 9.6x10* PROTOCHLOROPHYLL CHLOROPHYLL BACTERIOCHLOROPHYLL Fig. 37C\2. Low frequenry hand.s of protochlorophyll, chlorophyll a, and bacterio- chlorophyll (in ether). The symbols || and _L refer to the long axis in baoteriochloro- phyll and the corresponding direction in chlorophyll (which is not a sj-nimetr}' axis). which was interpreted in table 21. VI as the first vibrational sub-band of the red band, could be due partially, or even mainly, to an independent elec- tronic transition. (This was suggested, as an alternative, on p. 631 in Part 1 of Vol. II.) The bands at 660 and 613 m^ would then correspond to the long-wave band pair of bacteriochlorophyll, but with narrower sepa- ration. This hypothesis is illustrated in fig. 37C.2. An alternative is to assume that one component of the doublet is "prohibited" in dihydro- porphin. Stupp and Kuhn (1952) concluded, from fluorescence polarization measurements {cf. below section 4b) that the weak "green" band near 530 m/i (527.5 m^u in table 21. VI), rather than the "orange" band at 613 m/z, must be attributed to a separate electronic transition with an electrical momentum normal to that of the main red band. If this conclusion is accepted, values of the constants of the " || band" of chlorophyll a, given in figure 37C.2, must be changed to 530 mn, 18,900 cm.~S and am = 0.4 X lO"", re- spectively. The separation between the || and the ± component would thus be in- creased to 3700 cm."' — approaching that assumed for bacteriochlorophyll (4400 cm.~')- However, the experimental results of Stupp and Kuhn are in need of confirmation {cf. section 4b). THEORY OF CHLOROPHYLL SPECTRUM 1797 The attribution of the long-wave (red) band in chlorophyll, as well as in bacteriochlorophyll, to a vibration parallel to the symmetry axis, and of the short-wave (orange) component to a vibration perpendicular to this axis, is plausible because the molecular structure changes, in the series porphin-dihydroporphin-tetrahydroporphin, much less strongly in the direction normal to this axis than parallel to it {cf. fig. 37C.1). Of the two bands interpreted above as components of the red doublet, the short- wave (orange) band changes comparatively little, in position or intensity, in the transition from porphin to tetrahydroporphin {i. e., from proto- chlorophjdl to bacteriochlorophyll) ; while the long-wave (red) band is strongly shifted and enhanced by this transition (fig. 37C.2). This difference becomes less pronounced if the constants of the "|| band" in chloro- phyll are changed, as suggested above, to account for the polarization experiments of Stupp and Kuhn. The blue-violet band is a doublet, in chlorophyll as well as in most of its derivatives. (AUomerized chlorophyll appears to be an exception, cf. fig. 21 .4A.) The relative intensities of the main "blue" peak and of its "violet" satellite change strongly from preparation to preparation {cf. p. 607). Whenever the violet satellite appears particularly prominent, suspicion arises that pheophytin is present as an impurity; however, careful chro- matographic purification never eliminates the satellite band entirely, but only reduces its prominence. The separation and relative intensity of the violet satellite band depends considerably on the solvent {cf. fig. 21.26). Its interpretation is at present uncertain. A vibrational sub-band could occur on the short-wave side of the blue peak. Isomerism (or tautomerism) could account for a doublet structure. A final possibility is that the two bands correspond to the two theoretically expected electronic transitions. In the last case the blue band of chlorophyll a (at 429 m^ in ether) and its violet satellite (at 410 m^u in ether) are to be considered as components of an electronic doublet, the first one polarized |!, and the second one J_ to the short axis of the molecule. They are then analogous to the two ultra- violet bacteriochlorophyll bands (395 and 360 mn, respectively) ; but the "satellite" is located at the short-wave side of the main component in chlorophyll, and on the long-wave side of it in bacteriochlorophyll. According to Stupp and Kuhn (1952), the polarization of chlorophyll fluorescence, excited by absorption in the blue-violet region, can be interpreted by a superposition of two bands, with electric momenta in two mutually perpendicular directions {cf. fig. 37C.21). The wide error margin of the calculations, which is obvious from table 37C.I, the arbitrariness in the selection of weak bands interpreted as inde- pendent electronic transitions, and the uncertain isomeric homogeneity 1798 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C of chlorophyll and bacteriochlorophyll preparations, combine to make the suggested interpretation of the spectra uncertain. It provides, however, a better starting point for further investigation than the purely empirical suggestion, made on p. 622, that the addition of a pair of hydrogen atoms to the porphin system generates a new low electronic state, and pushes all the old levels upward. According to the present hypothesis, the homologous bands are 621 m^u in protochlorophyll, 660 m^ in chlorophyll, and 775 mju in bacteriochlorophyll; in other words, the shift with increasing hydro- generation is toward the longer waves. In Part 1 of Vol. II it was argued — by analogy with polyene chains, or with the benzene-naphthalene-anthracene series — that addition of new links to a conjugated double bond system (which occurs in the series tetra- hydroporphin-dihydroporphin-porphin) should lower, rather than raise, the first excited level (z. e., produce a "red" rather than "blue" shift of the first absorption band). Longuet-Higgins and Piatt (1950) pointed out, however, that a shift of the absorption band toward shorter waves with increasing number of conjugated bonds has been noted also in other ring structures, in which this addition broadened, rather than elongated the conjugated system. To make the theoretical treatment of the spectra of poi-phin pigments more precise and reliable, both better calculations and better experimen- tation are needed. In particular, spectroscopic measurements on oriented molecules (in flowing solutions, monolayers, or single crystals) could be helpful, by providing direct evidence of the polarization of the several bands in respect to the symmetry axes of the molecule. Further develop- ment of the above-mentioned studies of fluorescence polarization also is desirable. The life-time of the lowest excited state of chlorophyll a and h in solu- tion, first estimated by Prins (p. 633), was re-estimated by Livingston (personal communication) for ethereal solutions and by Jacobs (1952, 1954) for acetonic solution. They used a quantum-mechanical equation given by Lewis and Kasha (1945), instead of the classical relation used by Prins, to calculate the "oscillator strength," /, and the mean life time of excitation, r, from the integral under the absorption band. The results of the two calculations do not agree well: Livingston Jacobs Chi a: T = 1.8 X lO^Ssec. 1.27 X 10-8 .=!ec. (/ = 0.38) Chi />.• T = 4.4 X lO^^sec. 1.62 X 10-«sec. (/ = 0.28) Peirin (1929) and Stupp and Kuhii (1952) estimated the life time of chlorophyll a excitation in solution from the relative polarization of fluorescence in two solvents, one of these (castor oil) so viscous as to practically suppress Brownian rotational movement. The calculated (actual) excitation time was r^ = 3 X 10"^ sec. (Perrin) and r^ = 1.5 X MEASUREMENTS OF ABSORPTION SPECTRA OF PIGMENTS 1799 10-« sec. (Stupp and Kuhn); the natural life time must be r = t^/^ sec, where ,p, the quantum yield of fluorescence, is about 0.25 {rf. table 87CTV). 2. New Measurements of Absorption Spectra of Pigments in Solution (Addendum to Chapter 21, Sections Al, 2) (o) Chlorophyll, Chlorophyllides and Pheophorbides. Holt and Jacobs (19540 made absolute extinction measurements on the ethyl chlorophyllides a and b in ethyl ether, in the visible and the ultra- violet (figs. 37C.3 and 4), from 220 to 780 niM- 380 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 ruyU Fig. 37C.3. Absorption spectrum of ethyl chlorophyllide a in ether (after Holt and Jacobs 1954). Dots represent Zscheile's data for chlorophyll a. Similar measurements were made also with the two ethyl pheophorbides (figs. 37C.5 and 6). According to these studies, the molar absorption curves of chlorophyll a and etliyl chlorophyllide a in ether are almost identical (within ±3% in respect to intensity, and ±0.5 m^ in respect to the position of the band peaks between 250 and 700 m^, cf. fig. 37C.3). In the case of the b com- ponent, the main peaks of the chlorophyllide appear shifted towards the shorter waves by 1.5-2 mn from their position in chlorophyll, without noticeable change in intensity. This corrects our earlier statement (p. 626), which was based on Stern's absorption curves (fig. 21.16), that the absorption peaks of porphin derivatives become sharper when a short- chain alcohol is substituted for phytol. Weigl and Livingston (1952) confirmed the finding of Katz and Wassink (Vol. TI, Part 1, p. 642) that in piperidine, the absorption peak of chloro- 1800 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C I O K z 380 400 420 440 460 480 500 520 540 560 580 600 620 640 660 Fig. 37C.4. Absorption .spectrum of ethyl chlorophyllide h in etlier (after Holt and Jacobs 1954). "I — ' — 1 — I — I — 1 — I — I — I — ' — I — ' — r _L ■ I I I 1 I I 1 L. "T — ' — r 1 1 r- 220 240 260 280 300 320 340 360 380 380 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 700 Fig. 37C.5. Absorption spectrum of ethyl pheophorbide a in ether (after Holt and Jacobs 1954). phyll a is located at much shorter waves than in any other solvent. They found it at 642.5 m/i; it was 35% lower than in ether, while the blue peak was 65% higher. Evaporating and re-dissolving in ether did not bring the band back to its normal position, indicating an irreversible chemical MEASUREMENTS OF ABSORPTION SPECTRA OF PIGMENTS 1801 180 380 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680 Fig. 37C.6. Absorption spectrum of ethyl pheophorbide h in ether (after Holt and Jacobs 1954). change. A similar spectral shift is caused by dissolving chloroi)liyll in parabcnzoylamine (Livingston, Watson and McArdle 1949). The observations of Freed and Sancier (1951), concerning the irre- versible transformation of spectra of chlorophyll solutions in secondary amines, belong to the same category. All these results were mentioned in chapter 37B (section 4) as evidence of irreversible aminolysis of the cyclopentanone ring in chlorophyll, which proceeds faster the more basic the amine. Freed and Sancier (1951-54) studied the absorption spectra of the chlorophylls and their derivatives at low temperatures. As stated in chapter 37B (section 4), they first attributed the observed changes to a reversible isomerization of the pigment; later (1954) it was suggested that they are caused by a reversible formation of solvates. (That the two phenomena may be coupled was also mentioned in chapter 37B.) Freed and Sancier (1951) measured the spectra of the chlorophylls a, b, and b' in a mixture of w-propyl ether (20 p.), propane (60 p.) and propene (60 p.), from —198° to —63° C; and in a mixture of n-propyl ether (20 p.) and n-hexane (80 p.) from —63° C. to room temperature. Figs. 37C.7 and 8 show the observed transformations. They indicate 1802 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C 7000 6000 t ioocA Wavelength Fig. 37C.8. Changes with tem- perature of the absorption spectrum of the mixture of isomers of chlorophyll h in solution (Freed and Sancier 1951). Wovelength Fig. 37C.7. Visible absorption spec- tra of solutions of chlorophyll a at 230° K (dashed line) and at 75° K (solid line) (Freed and Sancier 1951). Con- centrations are different at the two temperatures. the replacement of "room temperature bands" by "low temperature bands," generally situated at somewhat longer wave lengths (c/. par- ticularly fig. 37C.8). The changes shown in figs. 37C.7 and 8 are completely reversible. The same spectra could be obtained by approaching a certain temperature from below or from above; they are therefore to be considered as belonging to equilibrium mixtures of different solvates, formed practically \\dthout any activation energy. The chlorophylls a and h are half-converted into their low-temperature, solvated forms at —93° C, chlorophyll h' , at — 43°C. Linschitz and co-workers (1952) noted that spectral changes similar to those described by Freed and Sancier for the chlorophylls occurred also upon cooling of ethyl chlorophyllide solutions (a or h) in EPA (ether -\- pentane + ethanol, 8:5:3) to -193° C. The rigidity of the solvent over a part of this temperature range did not interfere with the completion of the changes. As noted also by Freed and Sancier, allomerization does not prevent the transformation, which thus cannot depend on the keto-enol isomerism in the cyclopentanon ring. Strain (1952) suggested that the effects observed by Freed and Sancier could be due to the formation of colloidal (amorphous or crystalline) particles. Freed and Sancier (1952) argued against this interpretation, pointing out that what they had observed was substitution of new bands for the original ones, and not a gradual shift of the latter. However, we MEASUREMENTS OF ABSORPTION SPECTRA OF PIGMENTS 1803 mil see below in section 3, fig. 37C. 16-18, that the formation of crystalline particles also produces new bands, which then shift with the growth of the particles. A more convincing argument against Strain's suggestion is that phase changes are not likely to be reversible or lead to an equilibrium. In the same paper, Freed and Sancier (1952) described the low-tempera- ture transformation of the spectrum of chlorophyll b' in diisopropyl amine (10%) + propane (45%) -\- propene (45%); the red band, located at 645 m/x at -43° C, was replaced, at -103° C, by one at 662 mn. (The transformations in the blue-violet region indicate that in this system, irreversible aminolysis is superimposed on solvation.) Freed and Sancier (1954) based the above-mentioned reinterpretation of the temperature effect on its similarity with the effects caused by changes in solvent. Cooling of chlorophyll b solution in (not specially dried) hydrocarbons enhanced the long-wave band components, at- tributable to complexes with water, at the cost of the short-wave com- ponents, attributable to water-free pigment. This, however, is only a partial explanation, since a similar trans- formation was found also in dry (nonflu orescent) solutions. It had to be attributed there to thermal excitation of a vibration {AH = 1.4 ± 0.2 kcal/m.). A similar AH value (1.32 kcal m., corresponding to 460 cm.-^ was derived from the vndth of the band split in the red; the split of the blue band gave AH = 1.42 kcal/m. In wet solvent (n-propyl benzene), on the other hand, the temperature dependence indicated a AH of only 2.5 kcal/m.; this lower value suggested that, in this case, vibrational excitation was coupled with the dissociation of a hydrate. Thermally excitable vibrational states were noted, in addition to chlorophyll a, also in chlorophyll b, and possibly in bacteriochlorophyll, but not in the pheo- phytins, allomerized chlorophylls, or metal porphyrins. By vaiying the concentration of a polar admixture (water, propyl ether, pyridine, diiso- propylamine) in a nonpolar solvent, spectroscopic evidence of the forma- tion of both mono and disolvates could be obtained. With pyridine as admixture, and chlorophyll b as solute, an equilibrium constant K = [ChlPy2]/([ChlPy][Py]) = 10 was calculated (at 2° C). Solvations of this type occur also with allomerized chlorophyll, the pheophytins, and metal porphyrins. With isopropylamine as admixture, a second type of solvation can be observed, leading to bro-wn compounds (phase test intermediates, cf. chapter 37B, section 4). If this phenomenon is attributed to interaction of the amine with the enol group in position 9, the first type of solvation could be interpreted as involving the polar group in the center of the molecule (magnesium-nitrogen bonds in chlorophyll, imino groups in pheophytine, metal-nitrogen bonds in metal porphyrins). 1804 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C 13.0 FRACTION*ls FRACTION •3^ FRACTION *2 380 420 460 500 540 580 620 660 700 WAVELENGTH M;1 Fig. 37C.9. Absorption spectra of three fractions of allomerized chlorophyll in ether (after Holt and Jacobs 1954). Red peaks arbitrarily matched in height. Livingston, Pariser, Thompson and Weller (1953) measured the ab- sorption spectrum of pheophytin a in methanol in the presence of acids or bases. In acid solution, a reversible conversion of the "neutral" into an "acidic" form was revealed by spectroscopy; in basic solutions, an irreversible conversion into a "basic" form took place. In dilute solutions of strong bases, or in solutions containing weak bases (e. g., aliphatic amines), the conversion into the basic form was gradual. Neutralization of the base changed the spectrum of the basic form, but did not convert it back into that of the neutral (or acidic) form, Neuberger and Scott (1952) interpreted the spectral changes, obtained in several porphyrins by pH variation, as evidence of transformation of the free base, P, first into a monovalent cation, PH+ (often around pH 7) and then into a divalent cation, PH2 "*""*■ (often at about pH 4). This would mean that porphjTins are stronger bases than pyri- dine; the authors attributed this increased proton affinity to additional resonating struc- tures that become possible when one or two H + ions are added to the pyrrole nitrogens. Scott (1952) suggested that a similar addition of H+ ions to ring nitrogens accounts for the effect of acidity on tetraphenylporphin spectrum (cf. fig. 21.14), with the peculiarity that the addition of the second H ■•" ion to the ring causes, in this particular case, the ap- pearance of a "chlorin type" spectrum. (The spectrum of the intermediate, low- acidity form seems to be similar to that of the monocations of other porphyrins.) Holt and Jacobs (1953) also measured the absorption curve of allo- merized chlorophyll a in ether, and found it very similar to that given by Livingston for the same material in methanol (fig. 21.4A). The main peaks are at 420 m^ (Soret band) and 650 m/i, with minor peaks at 610, 567 and 520 mju. The most notable feature of this spectrum is the absence of the "satellite" band on the short-wave side of the Soret band. Closer investigation by Holt and Jacobs (1953) confirmed the findings MEASUREMENTS OF ABSORPTION SPECTRA OF PIGMENTS 1805 of Strain (c/. chapter 37B, section 4), that "allomerized chlorophyll" is not a single compound, but can be separated chromatographically into at least three fractions. Fig. 37C.9 shows the visible absorption spectra, in ether, of the three fractions of allomerized ethyl chlorophyllide a. The most abundant one of them (No. 2) has the spectrum shown in fig. 21.4A; another (No. 3) a spectrum very similar to that of chlorophyllide itself; the third one (No. 1) an absorption band further to the red, near 680 mju. Evstigneev and Gavrilova (1953) made the first quantitative study of the absorption spectrum of photochemically reduced chlorophijUs a and b. 1.0 - 1 tog^- r % 1 7 ' 0.5 " 4l -• —'' ^V ' / A f\ I > 1 1 1 600 500 400 Fig. 37C.10. Absorption spectra of reduced chlorophyll a (after Evstigneev and Gavrilova 1953). Curve 1: before reduction. Curve 2: after reduction. Curves 3, 4: same after 30 and 160 min. of reoxidation in darkness with- out air. Curves 5, 6: same after 100 additional min. and 1 day in air. Cui"ve 7: absorption curve of reduced pheophytin a (larger scale). 700 600 500 400 XiTi m)x. Fig. 37C.11. Absorption spectra of reduced chlorophyll b (after Evstigneev and Gavrilova 1953). Curve 1: before reduction. Curve 2: after reduction. Curves 3 and 4: different stages of re- oxidation, 25 min. in darkness without air and same plus 60 min. in air. Curve 5: shift of red peak after 24 hr. As described in chapter 35 (part A), chlorophyll can be reduced by illumi- nation in the presence of ascorbic acid in basic medium ("Krasnovsky reaction"). The product has been desciibed as "pink," with an ab- sorption band at about 530 m^ (fig. 35.3); but its rapid re-oxidation after the cessation of illumination had prevented a quantitative study of the spectrum. Evstigneev and Gavrilova found that if the reaction is carried out in toluene (instead of pyridine), and with phenyl hydrazine serving as both the reductant and the basic ingredient, the back reaction is slowed down so much that a comparatively cojicentrated chlorophyll solution can be completely reduced, and the absorption spectrum of the product can be measured before re-oxidation occurs. Figs. 37C.10 and 11 show the results for the chlorophylls a and 6, respectively. The different behavior 1806 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C of the bands at 518 and 585 m/z (in a), and at 565 and 635 m^ (in &), in- dicates that they belong to two different forms of reduced chlorophyll; Evstigneev and Gavrilova suggested that the "long- wave" bands belong to nondissociated semiquinones (derived from the chlorophylls by one-step reduction, cf. Krasnovsky's hypothesis on p. 1503), and the ''short-wave" bands to their ions. T — ' — I — ' — I — ' — I — ' — 1 — ■ — I — I — r 220 260 300 340 380 420 460 500 540 580 620 660 700 740 780 820 TT\/jL Fig. 37C.12. Absorption spectrum of methyl bacteriochlorophyllide in ether (after Holt and Jacobs 1954). (6) Bacteriochlorophyll and Derivatives New measurements have been made in the absorption spectrum of bacteriochlorophyll, by Manten (1948), Krasnovsky and Vojnovskaja (1951), Weigl (1952) and Holt and Jacobs (1954^). Fig. 37C.12 shows the absorption spectrum of methyl bacteriochlorophyllide in ether. The main red peak is located, according to Weigl, at 772 m^u in ether, 771 mju in acetone and methanol, and 782 mju in benzene; and according to Holt and Jacobs, at 769 m^ in ether, all in satisfactory agreement with fig. 21.25. Weigl noted a striking effect of polar solvents on the "orange" band: it is found at 575 mn (577 mii according to Holt and Jacobs) in ether, 580 mn in acetone, and 581 mju in benzene, but is shifted to 609 mn in methanol. (French, as well as Manten, reported a similar position — 605 m/i.) Weigl converted bacteriochlorophyll to bacteriopheophytin, MEASUREMENTS OF ABSORPTION SPECTRA OF PIGMENTS 1807 calculated the concentration of the latter from French's absorption data (c/. fig. 21.21) and derived in this way a value for the maximum (decadic) absorption coefficient (^max. = 9.6 X 10^ 1. mole~^ cm.~0 in the peak of infrared band of bacteriochlorophyll. Absolute measurement by Holt and Jacobs confirmed this calculation. (They found a:,nax. = 9-5 X 10* 1. mole~^ cm.~'.) The position of the peak of the main short-wave band of bacteriochloro- phyll, which we assume to be analogous to the^Soret band" of chlorophyll, could be given previously only as <400 mju (c/. fig. 21.6). It has since been determined by Manten (1948) as 362 m/x (in methanol), by Weigl (1953) as 358 m/x (in ether or acetone), 362.5 mn (in benzene), or 364 m^u (in methanol) and by Holt and Jacobs (1954^) as 356 m^u (in ether). A satellite band appears at 390-400 m/x (392 m/x in ether). The variable relative intensity of this long-wave satellite reminds one of the behavior of the short-wave satellite of the main blue-violet band in chlorophyll (cf. Part 1 of Vol. H, p. 646). According to Weigl (1953) the absorption peaks of baderiopheophytin lie, in ether, at 750 mju (amax. = 6.3 X 10* 1. mole"^ cm.~0> 680 mju, 620 m/x, 525 m/n, 384.5 m^ and 357 m/x. Barer and Butt (1954) confirmed the earlier observations (cf. pp. 618- 619) that the red absorption band of haderioviridin, extracted from green sulfur bacteria, lies at 664 m/x — quite close to that of chlorophyll a; they found the same to be true of the main violet band, at 434 m/x. Neverthe- less, on the basis of chromatographic behavior and minor spectroscopic characteristics, Barer and Butt agreed with the suggestion (p. 619) that bacterioviridin is not identical with chlorophyll a. Larsen (1953) gave an absorption curve of an acetonic extract from Chlorobium thiosulfato- philum, showing two sharp main peaks, at 435 and 660 m/x, respectively; and minor bands at 770 (very weak), 625, 496, 467, 412, 392 (shoulder), and 338 m/x; those at 496 and 467 probably indicate the presence of carotenoids. Again, the main bands almost coincide with those of chloro- phyll a, but the minor bands have different positions. Whether the much wider shift of the red band toward the longer waves, found in green bacteria (as compared to algae), is due to a difference in the chemical structure of the two pigments, or to a different state of aggregation (or binding to a different protein-lipoid complex), remains to be established. The shift is similar to that found for bacteriochlorophyll in purple bacteria. The divergent results of Seybold and Hirsch (1954) already were noted in chapter 37.B. These observers found, in extracts from what they called "purple spirillae," as well as in those from green bacteria (Microchloris), absorption bands at both 770 and 660 m/x, and contended that green and purple cells carry one and the same pigment, 1808 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C which they called "bacteriochlorophyll a" (characterized by the 770 niju band); they suggested that upon extraction this pigment is very rapidly converted into "bacterio- chlorophyll b" (= bacterioviridin), characterized by a band at 660 mju. The finding with purple bacteria could be related to the observations of Schneider; Holt and Jacobs; and others, who also noted the oxidation of bacteriochlorophyll in vitro to a green product (cf. chapter 37B); but since the absorption curve of live purple bacteria, given by Seybold and Hirsch, indicated the presence of green cells (cf. below, section 6(c)), this contamination may have been the main reason also for their results in vitro. Their finding with green bacteria similarly suggests a contamination of green cultures by purple cells. (c) Protochlorophyll and Other Porphin Derivatives The absorption spectrum of protochlorophyll, reproduced in Vol. II, Part 1 (fig. 21.8), from Rudolph (1933) and Koski and Smith (1948), was re-determined by Krasnovsky and Vojnovskaja (1949). Absorption peaks were found at G23, 571, 533 and 433 m/x in ether, and at 633, 588, 550 and 453 m/x in pyridine {cf. table 21. IV). The transmission minimum of proto- chlorophyll in coats of winter squash seeds was observed at 645-G50 m/x (cf. p. 705). Krasnovsky, Kosobutkaya and Voynovskaya (1953) noted that in etiolated leaves the corresponding transmission minimum was located at 635 m/x, and considered this as evidence that protochlorophyll can be present, in vivo, in two forms — an "active" form, "Pchl 635," capable of conversion into chlorophyll in light, and an "inactive" (polymeric ?) storage form, "Pchl 645." (Protochlorophyll in seed coats is not con- verted to chlorophyll by illumination.) This suggestion is analogous to Krasnovsky's hypothesis of the existence of two states of chlorophyll and bacteriochlorophyll in vivo (cf. below, section 66). In extension of the observations of Livingston and co-workers, and of Evstigneev et al., on the effect of complexing (with water or organic bases) upon the absorption spectrum of chlorophyll, Livingston and Weil (1952) Table 37C.IA Complexing Constants (Ki) Calculated from Absorption Changes (after Livingston and Weil 1952) Pigment Mg-Chlorophyll Activator Mg-porphyrin Zn-porphyrin ( = Chi a) Zn-Chlorophyll Aniline 41 — 46 — Benzyl alcohol 1 . 120 — 2 . 900 — Quinoline 8.300 1800 13.300" 15.200 Heptylamine 110.000 — 160.000 — " Similar value obtained from fluorescence measurements. MEASUREMENTS OF ABSORPTION SPECTRA OF PIGMENTS 1809 measured the absorption spectra of miscellaneous porphin derivatives, (Mg-, Zn- and Ca-complexed mesoporphyrin dimethyl ester; Mg-complexed tetraphenylchlorin, and Zn- or Ca-substituted chlorophylls) in pure, nonpolar solvents, and in the presence of alcohols or organic bases. The association constants in table 37C.IA (to be compared with constants calculated from fluorescence and listed on p. 768 in table 23.IIIA!) were calculated from absorption measurements. Imino compounds (in which the metal is replaced by two hydrogen atoms), such as pheophytin, showed no tendency at all for complexing with bases. This, and the unchanged complexing tendency of allomerized chlorophyll, supports the hypothesis of Evstigneev et al., attributing com- plexing to the central metal atom, in preference to Livingston's initial hy- pothesis of complexing through enol formation in ring V. In section 2a above, we have noted Freed and Sancier's evidence for the occurrence of both types of association with bases, the association in the cyclopentanon ring leading to "discolored" compounds (brown in the case of chlorophyll a). (d) Spectra of Accessory Pigments A few new measurements of the absorption spectra of plant carotenoids, particularly of those of purple bacteria, have been added to those sum- marized in chapter 21 (part C). Polgar, van Niel and Zechmeister (1944) gave absorption data for spirilloxanthin from Rhodospirillnm ruhrum (peaks at 540.5, 503.5 and 473 m/x in dioxane; 548.5, 510 and 479 m)u in benzene; 571.5, 532 and 495 van in carbon disulfide). Comparison with the absorption spectrum of Karrer and Solmssen's rhodoviolascin from Rhodovibrio (table 21. IX, p. 659, and fig. 37C.13) indicates the probable identity of the two pigments. Two methoxyl groups have been identified in both of them, and their most likely formula is therefore that suggested by Karrer, C4oH5o(OCH3)2. Natural spirilloxanthin is the all-^rans isomer; partial conversion to cis form produces a new peak at 495 myu. Some new absorption measurements on phycohilins also must be mentioned. According to Lemberg and Legge (1949), the phycoerythrin from Rhodophyceae ("R-phycoerythrin") has three absorption bands, while that from Cyanophyceae (*'C-phycoerythrin") has only one. However, fig. 21.39 shows considerable variations in the relative intensities of the different bands in phycoerythrin from different red algae, and it is not impossible that these variations may occasionally lead to spectnim with only one prominent peak. From this point of view, the absorption spec- trum of phycoerythrin from Porphyridium cruentum, measured by Koch (1953), is of interest — it shows only one main peak at 535 m^t, but a satelUte 1810 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C at 565 mix and a shoulder at 500 m^ also are noticeable. Blinks (1954) suggested (cf. chapter 37B, section 8) that this shows the presence, in Bangiales, of a third type of phycoerythrin ("B-phycoerythrin"); alter- natively, the same chromophore can be associated with different proteins (or other molecules), or present in different forms (as an adsorbed mono- layer, a colloid, or a solution). The situation is quite analogous to that in purple bacteria, where bacteriochlorophyll exhibits, in Thiorhodaceae, three infrared peaks of varying relative intensity and in A thiorhodaceae only one major (and one minor) peak (cf. chapter 22, p. 702). toge ''3A 2.9 2A 1.9 250 300 350 AOO 450 500 550 m-Xinmfi Fig. 37C.1.3. Absorption spectrum of rhodoviolasin (after Karrer and Jucker 1948). e(= 2.30a) = loge (h/I)/cd. For phycocyanin, Lemberg and Legge (1949) distinguished two varie- ties— "phycocyanin R" (from Rhodophyceae) , with two absorption bands {e. g., at 614 and 551 m^u), and "phycocyanin C" (from Cyanophyceae) with a single band (e. g., at 615 mju). Blinks (1954) suggested {cf. chapter 37B, section 8) that "phycocyanin R" may be a complex of phycocyanin C with phycoerythrin (with the 551 m/x band due to the latter). Haxo, O'hEocha and Strout (1954) extracted a new variety of phyco- cyanin from several red and one blue-green alga; it had an absorption peak at 645-650 m/x (halfway between those of phycocyanin C and chloro- phyll). A review of the absorption and fluorescence spectra of chlorophyll, carotenoids and phycobilins, with all curves replotted in a unified way, has been prepared by French and Young (1953). MEASUREMENTS OF ABSORPTION SPECTRA OF PIGMENTS 1811 (e) Infrared Spectra of Chlorophtjll and Its Derivatives The early infrared absorption curves, reproduced in fig. 21.4, obtained with solid deposits of unknown density, can be now replaced by better data, obtained with solutions of known concentration by Weigl and Livingston and by Holt and Jacobs. Weigl and Livingston (1953) published absorption curves for chloro- phyll a, chlorophyll h, pheophytin a, bacteriochlorophyll, and allomerized chlorophyll a. They gave a list of bands (a) common to all five compounds, (b) common to several compounds and (c) unique to single compounds, and suggested (in addition to several rocking and bending frequencies) the following identifications of bond-stretching vibration frequencies: V (cm."') (in ecu or dry films) 3400 (pheophytin a) 2862-2956 (all compounds) 1740 (all compounds) 1700 (all compounds) 1660 (all phytol-containing compounds) 1610 (all but bacteriochlorophyll) 1380 (all phytol-containing compounds) Assignment N— H C— H 0 C \ OR (ester) C— 0 c— C c— C (ketone) (phytol) (in ring) C— CH 3 No O — H band was noted in any of the five compounds, including chloro- phyll a (where it could be produced by enolization), and allomerized chlorophyll (where it would be present if oxidation in position 10 were I 1 to lead to an HC(10)OH, rather than to an HC(10)OCH3 group). More i I remarkably, no "aldehyde" C=0 band could be noted in chlorophyll b. Holt and Jacobs (1954) obtained the infrared spectra of the same compounds, as well as of several other chlorophyll derivatives, dissolved in chloroform, carbon tetrachloride and pyridine. Some of the results are represented in fig. 37D.14. Most interesting are the changes in the C=0 and 0 — H bands, indicating transformations in the cyclopentanone ring (enolization, chelation, allomerization), which characterize this ring as the most reactive center in the molecule. These transformations are sensitive to changes in solvent (for example, a basic solvent stabilizes the enol group, and prevents its chelation with the adjoining ester group); they are characteristically affected by the presence of the magnesium atom in the molecule (which favors enolization) : 1812 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C H R O ^c— c— c R O Mg I, ^ C— CuT . „ A o ^C— c— C R O -- C— Ci'o Mg Ketone (stabilized by the absence of Mg) 1^ ^-b-C dH...NC5H5 '^H— O Enol Chelated enol (stabilized by base) In pheophytins, no enolization (with or without chelation) is noticeable in CCI4 sohition. In contrast to Livingston and Weigl's observation, Holt and Jacobs' curves show the expected extra "aldehyde" C=0 band in compounds of the h series. Fig. 37C.15 shows the effect of allomerization (by exposure to air in methanol) on the infrared absorption spectrum of ethyl chlorophyllide a in chloroform. In the curve set (a), most striking is the practical dis- appearance of the (ketone) C=0 band at 1680 cm.~^ in the main allo- merized fraction, No. 2. This is in agreement with Fischer and Pfeiffer's hypothesis of lactone formation (as indicated under the top curve in set (b)) ; this transformation increases the frequency, since CO bonds generally are stabilized by accumulation at one C atom (cf. p. 215). The band at 1720 cm.~"^ can therefore be attributed to the lactone (the ester band at 1740 cm.~^ showing only as a shoulder). The two bands are similarly merged also in pyridine. In CCI4, the ketone C=0 band seems to re- appear (at 1675 mfi), suggesting a dismutation equilibrium between the I I I lactone, CH3— 0— C(10)— 0— C(9), and the ketoester, 0=C(10)-f- I II 1 0 I CH3O — C(9) (analogous to the equilibrium suggested by Fischer for the II O I I corresponding alcohol, HO— C(10)— O— C(9); cf. Fischer and Stern 1940, I II O p. 88). In CCI4 or CHCI3, none of the allomerized forms shows an OH-band, confirming Fischer's view that allomerization ordinarily leads to an OCH3 (rather than an OH) group in position 10 (except when no alcohol is pres- ent, as in the allomerization of methyl pheophorbide by dilute KOH in pyridine). MEASUREMENTS OF ABSORPTION SPECTRA OF PIGMENTS 1813 (a) (b) 17 16 15 FREQUENCY CM"' Fig. 37C.14. Infrared absorption spectra: (a) of chlorophyll a and pheophytin a, in ecu and CsHsN; (b) of chloropliyll b, in CCI4 and CHC'U, and of pheophytin b in ecu (after Holt and Jacobs 1954). Note (1) additional "aldehyde" C=0 band in b- compound; (2) bands of "chelated" C=0 in chlorophylls a and 6 in CCU* (absent in CsHjN and CHCI3!); (3) absence of "chelated" C=0 band in plieophytins; (4) presence of OH band— indicating enolization— in chlorophyll a and pheophytin a in pyridine, and its practical absence in CCU; (5) presence of OH band in chlorophyll b in CCU, and its absence in pheophytin b in the same solvent. * * Experiments with fresh chlorophyll b prei)arations showed no evidence of chelation in ecu. Also, the OH— band seen before in chlorophyll b in CCU was weakened by prolonged evacuation, indicating that it may be due to traces of water. 1814 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C The OH group re-appears in fraction No. 2 in pyridine. This is re- markable since, with the bond between Cg and Cio assumed to be broken 90 eo 70 60 50 40 30 KX) S90 |eo t70 «60 2 50 -40 -90 BO 70 60 50 40 30 20 10 jm PYRIDINE VS PYRIDINE '° \ /I /pyridine O-HP 36 32 28 24 20 19 18 16 15 14 13 12 rREOUENCY'CM"lx lO'^ (a) 17 « 15 frequfncy;Cm"'xio"' (b) Fig. 37C.15. Infrared absorption spectra of allomerized ethyl chlorophyllide a. (a) Intact chlorophyllide, and three fractions of its allomerization product, all in CHClj. Note differences in the C=0 band region (1600-1800 cm.-i). (6) Main allomerization product ("fraction No. 2") in CHCU, CCI4 and C5H5N (after Holt and Jacobs 1954). Note evidence of a lactone structure in CCI4 (and probably also in CHCI3), and of an OH group in pyridine. in the allomerized state, where can the H-atom needed for enolization come from? Shall perhaps this result be taken as an argument for re- SPECTRA OF CRYSTALLINE AND COLLOIDAL PIGMENTS 1815 turning to Fischer's original concept of ring III (rather than ring IV) as site of the two extra H-atoms? (This would mean interchanging the "short" and the "long" axes in the chlorin and bacteriochlorin systems!) Fraction No. 3 may be simply the C(10)-methoxy derivative, without lactone formation. In the optical spectra of several preparations of this fraction, considerable variations (from 1.33 to 1.59) in the intensity ratios of the blue and the red peak were noted; but all of them gave essentially the same infrared spectrum. 3. Spectra of Crystalline and Colloidal Chlorophyll Derivatives (Addenda to Chapter 21, Section B2) The scanty data on the absorption spectra of solid chlorophyll and chlorophyll derivatives (summarized in Part 1 of Vol. II, p. 649) have been considerably augmented by a study by Jacobs et al. {cf. Jacobs 1952, Kromhout 1952, Rabinowitch, Jacobs, Holt and Kromhout 1952, Jacobs, Vatter and Holt 1953, 1954, Jacobs, Holt and Rabinowitch 1954, Jacobs and Holt, 1954, Jacobs, Holt, Kromhout, Rabinowitch and Vatter 1954) of the properties of chlorophylls, alkyl chlorophyllides, pheophorbides, and bacteriochlorophyllide, in the form of microcrystals, colloids and mono- layers. The solid preparations were obtained by diluting acetone solutions of the pigments with water (occasionally, gum arable, or water-soluble cellulose, was added to slow down crystal growth). Immediately upon dilution (within 0.1 sec.) the red transmission minimum of the suspension was observed to move (in compounds of type a) from 660 to 670 m^t; fluorescence disappeared at the same time. When chlorophyll a was used (in Ca++-free solvent) the band remained at 670 m/x indefinitely, and no crystal formation could be observed at all. With chlorophyllide, on the other hand, the transmission minimum continued to move further towards the infrared; after a few seconds (or minutes), depending on temperature, viscosity and concentration, it reached a final position in the region of 735-745 m^. Kromhout (1952) {cf. Rabinowitch, Jacobs, Holt and Kromhout 1952) was able to follow this transformation of the red band by means of a rapid-action, rotating mirror spectrophotometer, in which the visible spectrum was scanned within 0.01 sec. at 0.1 sec. intervals by a synchronized photomultiplier-cathode ray oscilloscope- photographic camera system. Later, similar sequences of spectra could be obtained also by ordinary spectroscopy, by conducting the crystallization at 0° C. (Jacobs, Holt, et al., 1954). Fig. 37C.16 shows, in graphs a and b, a sequence of trans- mission spectra of growing chlorophyllide a microcrystals. Graph c 1816 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C >- CO z UJ o < o \- Q. o TaTi I I I I — I — I — I — I — I — I — I — r— 1 — I — I — r 400 500 600 700 WAVE LENGTH (M^) 00 400 — -r I X~-t—l—t-+--t--i-rt--t--t"f I 1-^J l_J L. 500 600 700 WAVE LENGTH (M^) 800 «■ ' I i ' TT- 620 700 800 WAVE LENGTH (M//) Fig. 37C. 16(a), (b), (c). Absorption spectra of ethyl chlorophyllide a crystals of different size (after Jacobs et al. 1954). Curve 1, solution; curves 2-10, small micro- crystals at consecutive stages of growth; (c), large microcrystals. Dashed lines: scattering correction. SPECTRA OF CRYSTALLINE AND COLLOIDAL PIGMENTS 1817 represents the transmission spectrum of still larger microcrystals, obtained by using a more concentrated pigment solution. Electron microphotographs of some of these crystalline preparations were reproduced in fig. 37B.8. In the case of the 6-compound, only relatively large microcrystals could be obtained. The spectrum of the microcrystals was measured in the collimated trans- mitted beam, in aqueous suspension. The presence of a Tyndall cone indi- cated that all suspensions produced marked scattering. For crystals with dimensions small compared to the wave length of light, this scattering is not too disturbing, as shown by the dotted line in fig. 37C.166. Although the z O < o p Ql O 1.00 .50 - .00 1 r- — I 1 — "T 1 •" — 1 « r 1 /'> /> . \ \ 5/ \ ■ ' \ \ \ \ \ \ \ \ \ 1 1 _i. ^ \ \ 1 /v 1 u_ 350 450 550 650 750 850 WAVE LENGTH (M^a) 950 Fig. 37C.17. Spectrum of methyl bacteriochloroplij^llide in solution (dashed line) and in large microcrystals (solid line). (Jacobs ei al. 1953.) scattering is predominantly of the "selective" type, with peaks on the red side of the transmission minima, it is not strong enough to shift the peaks ; nor does it affect significantly the shape of the bands. For crystals Avith linear dimensions >0.5/x, (c/. figs. 37B.8and37C.16c), the shape of the transmission curves is more strongly influenced by scat- tering. The experimental scattering curve of a suspension of such crystals is shown by the dashed line in fig. 37C.16c. It has a sharp peak at 765 m/x- — considerably on the long-wave side of the transmission minimum at 745 m^ — and a minor peak at 715 m/i. The scattering dechnes only slowly in the infrared; this must be largely responsible for the similar behavior of the transmission curves of the larger microcrystals. With such crystals, correcting the transmission curve for scattering— to trans- form it into true absorption curve — does change the position of the peak significantly, shifting it by about 5 m^ and making its slope on the long- wave side much steeper. Fig. 37C.17 shows the effect of crystallization on the spectrum of methyl bacteriochlorophyllide. As described in chapter 37B, section 5, similar spectroscopic shifts 1818 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C were observed with ehlorophjdl itself, and showed the way to reUable preparation of crystalHne chlorophylls a and b, and of crystalline bacterio- chlorophyll (figs. 37B.6 and 7). Fig. 37C.18 gives a comparison of the spectra of chlorophyll a in acetonic solution and in microcrystalline form (the latter corrected for scattering!). Monolayers of chlorophyll, chlorophyllide and bacteriochlorophyllide were obtained by Jacobs (1952, cf. Jacobs, Holt and Rabinowitch 1954) by spreading on water a drop of a solution of these pigments in petroleum ether. (The solubility of these pigments in pure petroleum ether is very 1.00 — 1 — 1 — 1— 1 1 1 r- -T 1 1 1—1 1 1 1 '-1 1 — r— > 1- z o ' 1 1 1 ■ ' 1 / 1 l\ /I 1 , I < o 1- a. o .50 .00 4C V 1 1 vj , JurTr-i=, /I / 1 / 1 / 1 / I 1 1 1 1 1 1 M 1 1 . l^ )0 500 600 700 WAVE LENGTH (Mju) Fig. 37C.18. Absorption spectrum of chlorophyll a micro- crystals, corrected for scattering (after Jacobs and Holt 1954). Dashed line, solution; solid line, crystals. small; but it can be made adequate by first dissolving the pigment in a small amount of pyridine, and then adding a large amount of petroleum ether.) The absorption spectra were obtained by picking up the monolayers on glass plates. Even a single monolayer absorbs enough to permit the locali- zation of the main absorption peaks; a stack of 5-10 glass plates, each carrying a single monolayer, can be used to obtain a complete spectrum. (Attempts to collect several monolayers on a single plate were not suc- cessful.) Figs. 37C.19 and 20 show the absorption spectra of monolayers of chlorophyll a, and ethyl chlorophyllide a. It will be noted that in chloro- phyllide (and bacteriochlorophyllide) monolayers, the red band is shifted toward the longer waves almost — but not quite — as far as in "large" microcrystals (cf. table 37C.II). With chlorophyll itself, two kinds of monolayers were observed, as illus- trated in fig. 37C.20. One kind — obtained from solutions containing SPECTRA OF CRYSTALLINE AND COLLOIDAL PIGMENTS 1819 >10-« mole/1. Ca + +— has a very high optical density, and a red band shifted as far as 735 m^. The red band of monolayers obtained from .000 400 500 600 700 WAVE LENGTH (M/i) Fig. .37C.19. Absorption spectrum of ethyl chlorophyllide a monolayer on water (after Jacobs, Holt and Rabinowitch 1954). Dashed line, solution; solid line, monolayer. Ordinate: optical density of a single monolayer. —I r 1 1 1 ' ' a' 0.020 - /M / >- (- / 1 i S 0.015 - \ / — o A^ I 1 _i < o ►r 0.010 /\\ A 1 1 - o I \ / \ / 1 0.005 nnnn 1 , ^^--^r^ ^-^\j\ 400 450 500 550 600 650 WAVELENGTH IN MU 700 750 Fig. 37C.20. Absorption spectra of chlorophyll a monolayers (after Jacobs e< al. 1953). Dots, "compressed gas" or "colloidal" type monolayer (prepared without Ca + + ions); circles, "crystalline" monolayer (prepared in the presence of Ca + + ions). Ordinate, optical density of a single monolayer; arrows, location of absorption peaks m solution. Ca++-free solutions is located at 670 m/x; their optical density in this band is much lower {cf. table 37C.III). According to Hanson (cf. Vol. I, p. 448), chlorophyllide crystals consist 1820 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C Table 37C.II Red Absorption Band of Chlorophyll Derivatives in Different States of Ag- gregation (after Jacobs) Position ' of red peak (X in m^i) Shift in (obs 1000 cm.-i lerved) Pigment Soln. in acetone Gas (extra- pol.) Microcrystals Small Largea Mono- layer Gas to large crystal" Gas to mono- layer EtChl. a... . 660 648 718 740 730 1.9 1.7 EtChl. 6... . 645 632 — 710 — • 1.7 — EtPheo. a.. . 665 652 690 715 — 1.3 — EtPheo. b.. . 655 642 680 690 — 1.1 — MeBact . 760 740 840 860 845 1.9 1.5 CoUoid Crystal Monolayer Chlorophyll a. . 660 648 670 735 675, 735 1.8 0.43,1.8 Chlorophyll b.. , 645 632 655 695 — 1.4 — Bacteriochloro- phyll.... . 760 740 - ( >860) — (>1.9) — Corrected for scattering. Table 37C.III Optical Density of Single Monolayers (at Red Peak) Pigment Et chlorophyllide a Me bacteriochlorophyllide . . . . Chlorophyll a (Ca"*""^ present). (noCa ++) log (lo/I) gas Shift monolayer 1000 cm.-i 0.019 ± 0.03 1.69 0.019 1.64 0.026 1.81 0.011 0.43 of layers built and arranged as shown in fig. 49 (Vol. I). Fig. 37B.9 rep- resents a model of this structure. Monolayers of chlorophyllide on water have, according to Hanson, the same structure as the layers in the crystal, with the plane of the porphin ring inclined by 55° to the water-air interface^ — perhaps, because of the affinity of the magnesium atom for water. The surface density is one molecule per 0.69 mju^. In colloidal chlorophyll rnonolayers, each molecule occupies 1.06 m/x^. The additional 0.37 m^u^ (compared to the surface requirement of ethyl chlorophyllide) is needed to accommodate the phytol chain. More pre- cisely, the presence of phytol prevents the orderly arrangement charac- teristic of chlorophyllide monolayers, and leaves the pigment molecules more or less randomly oriented — except that their symmetry axes seem to prefer the orientation parallel to the water surface also in this "com- pressed gas type" monolayer; this is indicated by the fact that, as shown SPECTRA OF CRYSTALLINE AND COLLOIDAL PIGMENTS 1821 in table 37C.III, the absolute optical density of the chlorophyll monolayers of this type (in the peak of the red band), is smaller than that of the chloro- phyllide monolayers, in the ratio of 11 : 19 = 0.58 — which is not much less than the ratio of molecular surface densities (70: 106 = 0.6G). The optical density of the chlorophyll monolayers of crystalline type is about 2.4 times that of colloidal ones, and 1.4 times that of ethyl chloro- phyllide monolayers. It seems probable (but needs experimental con- firmation) that its molecular surface density is correspondingly higher. The "red shifts" of the absorption band — which may be considered as in- dices of the closeness of the chromophore packing — follow the same trend (table 37C.II). That it should be possible to rearrange a monolayer of the type shown in fig. 37B.9, into a 30% denser pattern would be startling enough even if we were dealing with chlorophyllide itself; it is even less likely if we are, at the same time, substituting, for chlorophyllide, chloro- phyll, whose phytol "tail" must interfere with close packing. One could suggest that an increase in optical density could be achieved without a proportional increase in molecular density. For this, the vibra- tion planes of the optical electrons should be re-oriented so as to enhance the absorption of light passing through the monolayer normal to its plane. However, in Hanson's model of the monolayer, the short axis of the por- phin ring is oriented parallel to the water surface; and, according to Piatt (table 37C.I), this is the plane of vibration of the main red absorption band. If these assignments are correct, one does not see how any re-orientation of molecules could enhance the optical density of the monolayer in the red band. A more plausible interpretation of the high density surface layer of chlorophyll is that it is a two-molecular layer. If this is so, the shift of the red absorption band to 735 m/x may be due either to an increased inter- action within each layer, or to interaction between the adjacent pigment molecules in the two layers. The blue-violet absorption band also shifts toward the longer waves in crystals and monolayers, but its behavior is more complex, because of its double structure. According to Kromhout (1952) the observed changes in this band can be explained by taking into account the different polarizations of the two components (c/. table 37C.I) and the preferential growth of the microcrystals in one plane: when crystal particles grow mainly in one plane, transmitted light consists increasingly of rays that have passed through the crystals under the right angle to this plane. The strong scattering of light by the larger crystals — already mentioned in connection with the red band — adds complication to the shape of the blue-violet doublet. According to fig. 37C.16c, the selective scattering of large microcrystals has its peak at 485 m/x, on the long-wave side of the 1822 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C 450-470 m/i peak; it thus increases the apparent absorption in the blue component of the doublet much more strongly than in the violet one. The curves in Fig. 37C.16a-c were drawn so as to equalize the heights of the red peaks. Estimates indicate that crystaUization causes no large change in the area under the red band. The area under the blue-violet band, on the other hand, seems to be substantially reduced by crystallization. (In the case of ethyl pheophorbide b, the optical density in the peak of the blue band is seven times that in the peak of the red band in solution, but about equal to it in small microcrystals!) In part, at least, the explanation may be the same as suggested above for mono- layers: the light transmitted through thin plate-shaped crystals consists mainly of beams that have passed through the crystals normally to their planes; the oscillation dipole of the red band lies in this plane, while that of the "Soret band" is mainly (or entirely) normal to it. Probably the formation of a high-density solid chlorophyll phase accounts for the observations of Strain (1952). He dissolved chlorophyll (a, a', b, or b') in petroleum ether + 5% methanol; upon extraction of methanol with water, the chlorophyll solution in petroleum ether became "colloidal" and the red band moved to 710 m/* (in a and a') and to 690 m/j, (in b and b'). A theoretical interpretation of the band shift in crystals and mono- layers can be obtained (cf. Jacobs, Holt et al. 1954) by considering the interaction of an orderly array of resonating "virtual" dipoles (which are associated, in quantum theory, with the transition between two stationary states, and determine the "oscillator strength" of the transition). Each molecule in the array has the same average probability of being excited by an incident light wave. Heller and Marcus (1951) showed that the excitation energy of an infinite isotropic lattice of such virtual dipoles differs from that of an individual dipole by a term equal to the "classical" interaction energy of a system of actual oscillating dipoles of the same average magnitude, having a phase difference of e'""^ ^ (where k is the wave — > number vector of the incident wave, and R the distance vector between two lattice points). Jacobs applied Heller and Marcus' equations to a finite isotropic crystal, and obtained an expression for Er , the energy of the excited state as a function of the radius R of the crystal. Because of the mutual cancellation of the effects of dipoles with "unfavorable" phase differences, this function grows only very slowly, until the crystal dimensions reach the order of magnitude of a wave length, after which the increase becomes more rapid; in other words, the contribution to the interaction energy of spherical shells closest to the center is smaller than that of the shells \vith R > 100 m/n. The sigmoid-shaped 7?'^' = f{R) curve reaches saturation at (37C.1) E''' = Eo + ^E"' = Eo- {ATuVmi) SPECTRA OF CRYSTALLINE AND COLLOIDAL PIGMENTS 1823 where Eo is the energy of an isolated molecule, and AE^, the change in energy caused by an infinite cubic lattice with a lattice constant Ro and a virtual dipole n in each lattice point. For a monomolecular layer traversed by light perpendicular to its plane, the phase difference is zero for all molecules in the plane, and the inter- action energy is therefore a rapidly converging function of the radius of the monolayer. For a circular, isotropic array of dipoles, ^vith a radius R, the interaction energy in the excited state is: (37C.2) AE„ = Rl \ r) The saturation energy oXR = oo is: (37C.2A) AE = TTfJ. VRl The value of A£'„ is equal to V4 of that of A£'„ given in (37C.1) for an infinite cubic crystal with the same lattice constant. The dipole moments of the transitions can be estimated from the in- tensity of absorption {i. e., the total area under the absorption band), and the lattice constant, from x-ray diffraction studies. Using these data, Jacobs obtained table 37C.IIIA for the crystals whose structure is known or can be surmised (no such surmise is as yet possible for chlorophyll, cf. chapter 37B, section 5). The table shows that the "limiting" shifts, found in three-dimensional crystals of the two chlorophyllides and their pheophorbides, are close to the theoretical estimates. Table 37C.IIIA Red absorption peak in crystals. Red absorption peak in isolated molecules (extrapo- lated as in fig. 21.25), m/i Molecular density in crystals (molecules per cc.) X 10-21 Oscillator strength of "red" transition, / Band s (cm hift A.- Crystal Theoret- ical (for 2-dimen- sional crystals) X 10-3 Experi- mental (table 37C.II) X 10-' Ethyl chlorophyllide a . 740 648 1.1 0.38 1.9 1.9 Ethyl chlorophyllide b . 710 632 1.1 0.28 1.4 1.7 Methyl bacteriochloro- phyllide 860 740 1.1 0.79 4.4 1.9 Ethyl pheophorbide a . . 715 652 0.89 0.31 1.2 1.3 Ethyl pheophorbide h . . 690 642 0.89 0.27 1.1 1.1 The experimentally observed dependence of the band shift on crystal size, illustrated in fig. 37C.16, indicates rapid rise to saturation at 72 10-'^ sec. (the period of an intermolecular vibiation). This permits •■ 0, the values of

(after Forster and Livingston 1952) Pigment Solvent Exciting X, m/i

■ b "cross-quenching" all grow with increasing concentration. An approximate kinetic analysis led to the following tentative conclusions. (1) Energy transfer from Chlb to Chla becomes significant above 2 X 10^"* mole/1. — an order of magnitude earlier than self-quenching of either a or b z < o z < o z o o UJ o !^i^ct"°"''^o :.> Chl * _L u. 650 700 750 WAVE LENGTH, m/i Fig. 37C.23. Fluorescence spectra of chlorophyll a + b in ether (1.2 X 10 ~' M) excited by X = 429 (70% absorbed by chlorophyll a) and by X = 453 ni/i (5% absorbed by chlorophyll a). Corrected for self-absorption. Fluorescence in (quanta emitted)/- (quanta absorbed), multiplied by arbitrary factor. The peak at 670 ni/i is due to chloro- phyll a, the hump at 650 m/x to chlorophyll b. Vertical arrow shows large contribution of chlorophyll a to fluorescence at 670 n\n excited by X 453 m^, where the absorption of chlorophyll a is small; it indicates energy transfer from chlorophyll b to chlorophyll a (after Duysens 1952). (this may be attributed to stronger overlapping of the absorption band of a with the fluorescence band of 6, as compared to the overlapping of the ab- sorption band of each pigment and its own fluorescence band). (2) The quenching of chlorophyll b fluorescence by chlorophyll a is highly efficient; that of a fluorescence by b is negligible. (3) The quenching of 6 by a increases with concentration faster than the sensitization of a. (It appears as if the first effect may be proportional to the square, the second one to the first power, of concentration.) Above 10~^ mole/1., quenching begins to reduce markedly the yield of sensitization, and the latter drops to zero above 10 "2 mole/1. Duysens (1952) determined the ratio of the intensities of fluorescence excited in an ethereal solution, 1.2 X 10^^ il/ in both chlorophyll a and FLUORESCENCE OF CHLOROPHYLL in vitVO 1837 chlorophyll h, by X 429 and 453 m/x, respectively. Chlorophyll a al)sorbs fourteen times more quanta at 429 than at 453 m^; nevertheless, the in- tensity ratio was only 1.7. Since the spectrum of fluorescence excited at 453 m/x indicates that only a minor part of it was due to the emission by chlorophyll h (fig. 37C.23), the intensity measurements indicate that the energy absorbed by chlorophyll h is made available for fluorescence of chlorophyll a; the efficiency of this transfer could be estimated as 40-50%. At lower concentrations (5 X 10^^ M, and 1.2 X lO"" M in each component), the effectiveness of the transfer declined to 30-40% and 20%, respectively. (/) Fluorescence of Colloidal Chlorophyll Solutions Table 37C.IIIB shows that Krasnovsky and Brin (1948), while con- firming the nonfluorescence of most colloids and adsorbates of chlorophyll, found some fluorescence in adsorbates on paraffin, palmitic acid and magnesia, and strong fluorescence in colloidal solutions containing deter- gents, whether prepared by dilution of alcoholic chlorophyll solutions with aqueous detergents (2b), or by similar dilution of alcohofic solutions con- taining alcohol-soluble proteins (3e), or by treatment of chloroplast or grana suspensions wdth aqueous detergents (5d) (c/. pp. 775-777 for earlier observations of weak fluorescence in some colloidal chlorophyll prepara- tions). The non-fluorescence of microcrystalline chlorophyll suspensions was noted above in section 3. (g) Fluorescence of Bacteriochlorophyll and Protochlorophyll The presence of two fluorescence bands, at 695 and 810 m/i, respectively, in the fluorescence spectrum of bacteriochlorophyll, was noted in Part 1 of Vol. II (p. 748, fig. 23.4). An interpretation of this phenomenon — if it were real— could be sought in the theory of the tetraporphin spectrum {cf. table 37C.I), according to which the "orange" absorption band of bac- teriochlorophyll corresponds to a dipole oscillation parallel to the long axis of the conjugated double bond system, while the red absorption band corresponds to an oscillation perpendicular to this axis. Internal conver- sion of the "parallel" into the "perpendicular" oscillation could be difficult; the orange band of bacteriochlorophyll (in contrast to the blue-violet bands of both chlorophyll and bacteriochlorophyll) could then have a fluorescence band of its own associated ^\^th it. However, it seems more likely that the two-band fluorescence spectrum in fig. 23.4 was obtained not with pure bacteriochlorophyll, but with a mixture of bacteriochlorophyll and a green oxidation product (described 1838 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C in chapter 37B, section 6). Purified bacteriochlorophyll has no visible fluorescence, while the green oxidation product fluoresces red. French (1954) made similar observations: A two-band fluorescence spectrum was found in a "crude" bacteriochlorophyll preparation; after chromatography a green fraction was separated which gave only one fluorescence band, at 687 mn. However, the main blue fraction still showed, in French's experiments, the two original bands. Their relative intensity depended on the wave length of the exciting light, confirming the surmise that they belonged to two different compounds, in other words, that the blue fraction still was a mixture rather than a pure pigment. French (1954) gave also a spectral curve for the fluorescence of proto- chlorophyll in acetone. It showed a main peak at 630 m/i and a secondary peak at 685 m^i. (h) Fluorescence of Phycobilins Bannister (1954) found that the quantum yield of fluorescence of the purest phycocyanin preparations (i. e., preparations freed as much as possible from adventitious proteins), excited with monochromatic ultra- violet light, was constant (within 10%) from 250 to 400 m/z, including the region around 275 mju, where a considerable fraction of the absorbed light must be assigned to the protein moiety of the chromoprotein molecule, more specifically, to tryptophane and tyrosine residues. (Crude estimates indicated the protein share of the absorption in this region to be of the order of 50%.) This indicates an efficient excitation energy transfer from the protein to the chromophore — a type of energy transfer suggested before as explanation of CO-removal from hemoproteins by light absorbed in their protein moiety. In the latter case, however, an alternative to transfer of electronic excitation is conceivable: the conversion of elec- tronic excitation into vibrations, and accumulation of enough vibrational energy in the iron-carbon monoxide bond to dissociate it; no such alterna- tive explanation seems to be possible in the case of protein-sensitized fluorescence of phycobilins. 5. Chemiluminescence of Chlorophyll in vitro and in vivo The only evidence of chemiluminescence of chlorophyll, described in Part 1 of Vol. II, was the light emitted, upon heating, by solutions of chlorophyll in tetralin (p. 751). Since then, two studies have been de- veloped in this field. One deals with a system in vitro, the other with photosythesizing cells. Linschitz and co-workers (1952) found that chlorophyll (and other metalloporphyrin dyes) chemiluminesce when reacting with organic per- (a) H2 H2 R C 0 OH + DH. -* R C— 0— + DH (b) H.> H R C 0 + DH -> DH.* + RC— 0 (c) DH2* -* DHa + hf CHEMILUMINESCENCE OF CHLOROPHYLL til vHvO AND in VIVO 1839 oxides (at / > 100° C). The kinetics of luminescence decay (studied with Zn-tetraphenyl porphin and tetralin hydroperoxide) indicated that the over- all process includes: (1) a second-order reaction of peroxide, P, with the dye, DH2; (2) an approximately second-order catalytic decomposition of the peroxide; and (3) a slow first-order, noncatalyzed decomposition of the peroxide. Between 20 and 60 peroxide molecules are decomposed for each permanently destroyed dye molecule. The decay of chemilumines- cence follows the second-order law, / ^ [PlfDHo]. The following free- radical reaction mechanism (of the type first proposed by J. Weiss for the chemiluminescence of luminol) is suggested for the main catalytic reaction, and the emission of light : H, + H2O H.> H (37C.6) H2 R — C — 0 — and DH — are free radicals, and their reaction, (b), should lib- erate enough energy to produce an electronically excited dye molecule, DH2*. It will be noted that reading reactions (c), (b), and (a) backwards, one gets the over-all reaction : H2 +DH2 H2 (37C.7) RC=0 + H,0 > R— CO— OH + hi' i. e., a dye-sensitized photochemical peroxide formation. This illustrates the inverse relation between photosensitization and chemiluminescence. The chemiluminescence of photosynthesizing cells was discovered by Strehler and Arnold (1951), using sensitive methods of fight detection. It can be observed in higher plants as well as in green and red algae. In its low yield, it resembles the weak luminescences which Audubert and co- workers have found to accompany many common chemical reactions. An important difference is that the emission occurs in the visible, not in the ultraviolet as in Audubert's experiments. Its spectrum could be deter- mined only rather crudely, but appeared to be identical with that of chloro- phyll fluorescence. The emission could be followed for several minutes after the cessation of illumination; about 0.1 sec. after the exciting light had been turned off, the intensity of the afterglow was about 0.1% of the intensity fluorescence has had during the illumination (in which less than saturating light had been used). The intensity of emission decays in the dark, following an approximately bimolecular law at 6.5° C. and a lower order law at higher temperatures. That the emission is due to chemiluminescence (and is not a fluores- 1840 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C cence or phosphorescence of long duration) is indicated by the many relations between the rate of photosynthesis and the intensity of the emission. Both show saturation in the same region of light intensities; both have a temperature maximum at about 35° C. ; inhibitors of photo- synthesis, including azide, cyanide, hydroxylamine, dinitrophenol and ultraviolet light, stop both photosynthesis and the "afterglow." 2 S \ iti)nr> v't Nunit cells, sue )a and cil irently in (and ofte ated solui have also ium has i 'urves of sperroato( tlie tech; Fig. 37C.28. Removal of scattering in a bacterial suspension by the use of sus- pension medium with properly adjusted refractive index (Barer 1955). One particular point, which needs clarification, is to what extent scattering can explain the high apparent absorption by cells in the green and in the far red. We will see below (section 6 A) that, according to Duysens (1954), ordinary Rayleigh scattering can account for both the shallowness of the "green trough" and the existence of a "red tail" in the transmission curve; but that the observations of Latimer (1954) indicate the occurrence of a strong selective scattering on the long-wave side of the main absorption bands, which is bound to contribute to the diminution of transmission, par- ticularly in the far red. Sharp selective scattering bands are associated not only with the main absorption bands of chlorophyll, but also with those of the carotenoids, although the latter are merely ripples in the absorption curve of live Chlorella cells. On p. 709, we noted that the dips in the green and in the red were as shallow in Noddack's "pure absorption" curve as in Emerson's "absorption + scattering" curve LIGHT ABSORPTION BY PIGMENTS in VWO 1847 of Chlorella. In Noddack's measurements, however, scattering was not eliminated (as it was supposed to be in Barer's experiments), Ijut merely excluded from measurement; therefore, the possibility of the scattering increasing significantly the true absorption in the minima (by the so-called "detour factor," cf. chapter 22) could not be ruled out. Barer noted that the position of the main red peak varied, in different green algal suspension, between 675 and 680 myu, and that of the main violet peak, between 435 and 440 m/i. New absorption bands Avere foinid by Barer, in Chlorella, at 385 and 340 m.n, probably corresponding to the solution bands of chlorophyll a at 380 and 330 m^, respectively (c/. fig. 21.3). In Closterium, an absorp- tion band was noted at 620 myu, of which only a slight indication can be found in the Chlorella spectrum ; it was suggested that it is the equivalent of the 612.5 m^c band of chlorophyll a in solution (fig. 21.1). Closterium spectrum showed no evidence of the 645 m/x band, which would indicate the presence of chlorophyll h. The absorption peaks of green algae at 480 and 490 m/x, presumably' due to caro- tenoids, were noted by Barer to become progressively more pronounced in the transition from yellow, through yellow-green, to green Chlorella strains, indicating a parallel increase in the contents of chlorophyll and the carotenoids (unless the 480-490 m/x band is attributed — at least in part — to chlorophyll b\). (h) Two Forms of Chlorophyll in vivof Krasnovsky and Brin (1948) suggested that chlorophyll is present in leaves in two different forms. The less abundant form, with a band at 665-670 mn, they supposed to be "non aggregated," but bound to a lipide, or lipoprotein. It was also supposed to be fluorescent and photochemically active. The main amount of chlorophyll, responsible for the band at 677- 678 mil, was supposed to be in an "aggregated" (colloidal), nonfluorescent and photochemically inactive form — a kind of "chlorophyll reserve." This hypothesis is very similar to that propounded earlier by Seybold and Egle (c/. p. 818) . The argument used on pp. 746 and 819 against the latter 's assumption that two different forms of chlorophyll account for the main absorption band and the fluorescence band in vivo — the approximately equal shifts of both bands from their positions in solution — was considered by Krasnovsky and Brin to be unconvincing because, in their opinion, self-absorption of fluorescence could have caused an apparent shift of the fluorescence band toward the longer waves. (The fluorescence band, corresponding to the absorption band at 665-670 m/x of the "nonaggregated" photoactive chlorophyll, could be located under the absorption peak of the "aggregated" form at 677-678 m/x, and distorted by absorption so as to show an apparent peak at 680 m/x. French's fig. 37C.45 shows this type of shift.) 1848 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C No convincing spectroscopic evidence of a doublet structure of the red absorption band in mature plants has been published. Duysens (un- published) has suggested, however, that the greater width of the absorption band of chlorophyll in vivo (compared to that of the same pigment in vitro), may be interpreted as due to the superposition of two bands, with peaks close enough to merge into a single band. The situation in green plants would then be similar to that in purple bacteria, where two (or three) infrared bands, present in vivo, are replaced by a single band after the destruction of the pigment complex. An alternative explanation of the broadening could be based, however, on enhanced coupling of electronic excitation Avith vibrations in the pigment-lipoid-protein complex (c/. the explanation of the greater width of the bands in chlorophyll crystals in section 3 above). To sum up, from the point of view of absorption and fluorescence spectra, the hypotheses of Seybold and Egle, and of Krasnovsky and Brin, must be considered as unproved, even if not impossible. The second pastulate of Krasnovsky and Brin — that the bulk of chloro- phyll is photochemically inactive — is quite implausible. The high quan- tum yield of photosynthesis seems incompatible with the assumption that the bulk of chlorophyll in the cell is photochemically inactive. The fluores- cence experiments of Duysens, and of French and Young (chap. 24, sec. 7; this chapter, sec. 7), and action spectra of photosynthesis of green, brown and red algae, give convincing evidence that light quanta absorbed by all forms of chlorophyll are transferred, with high efficiency, to chloro- phyll a, and utihzed there for photosynthesis. Krasnovsky 's ''chlorophyll reserve," with an absorption band overlapping the fluorescence band of "active" chlorophyll would be a "sink" into which all excitation energy would disappear without the possibility of it ever being used for photo- synthesis (assuming this "reserve" itself is photochemically inert, as suggested by Krasnovsky). In a more recent publication, Krasnovsky, Kosobutskaya and Voynovskaya (1953) suggested that the "inactive, polymerized" Chi 678 contributes to photosynthesis by resonance energy transfer to the "active, monomeric" Chi 670. However, energy transfer in this direction should be negligible compared to that in the opposite direction, from Chi 670 to Chi 678, because of the relative position of the absorption and fluo- rescence bands of the two forms. We will see in section (c) below that Krasnovsky's analogous explanation of the several absorption peaks of bacteriochlorophyll in vivo runs into even greater difficulties. The assumption of photochemical activity of "chlorophyll 670" and photochemical inertness of "chlorophyll 678" was based on the observation of Krasnovsky and Kosobutskaya (described in part B of this chapter, section 1) that, when plastid fragments from etiolated leaves, or these LIGHT ABSORPTION BY PIGMENTS in VIVO 1849 leaves themselves, are exposed to light, the protochlorophyll band (at 635 him) is first replaced by a band at 670 m/x, and only upon longer ex- posure of the leaves is shifted to 678 mju; simultaneously with the latter transformation, the sensitivity of chlorophyll for photoxidation declines. These changes may well be associated with the formation of grana or lamellae, which can be considered as a special type of chlorophyll aggrega- tion, or of chlorophyll attachment to proteins; and in any case, increased resistance to photoxidation does not necessarily mean general loss of photochemical activity, as postulated by Krasnovsky and co-workers. We see in Krasnovsky's experiments no reason to abandon the — ad- mittedly speculative^ — picture, developed in section A of this chapter, according to which chlorophyll is present, in green cells, in monolayers interlarded between proteidic and lipoidic layers. We consider it most likely that all of it contributes uniformly to fluorescence, and to photo- synthesis as well. Krasnovsky's observations on the shift of the absorption peak in the process of cliloiophyll formation can perhaps be explained as indicating the transformation of "unorganized" chlorophyll-protein complexes into "organized" structures (such as cohesive monomolecular layers), the additional band displacement being due to pigment-pigment interaction, as discussed above in section 3. The spectroscopic difference between the "active" protochlorophyll of etiolated leaves (subject to photochemical conversion to chlorophyll and having a peak at 635 mn, cf. Krasnovsky, Kosobutskaya and Voynovskaya 1953), and "inactive" protochlorophyll in pumpkin and squash seeds (mth a peak at 645-650 m^) can perhaps be explained in a similar way. The situation may be different in those red and blue algae in which a large part of chlorophyll appears inactive, and only the part intimately associated with the phycobilins contributes to photosynthesis and fluo- rescence {cf. chapter 30, section 6 and chapter 32, section 6). It would be interesting to obtain evidence of a difference between the absorption spectra of these two fractions. (c) Absorption Spectra of Purple and Green Bacteria As mentioned in section (a) above. Barer (1953) suspended cells in protein solutions to minimize scattering. The results were particularly satisfactory with bacteria {cf. fig. 37C.28). Barer's absorption curves of the purple bacteria Rhodopseudomonas spheroides, Rhodo spirillum {rubrum, palustris, and capsulatus), reproduced in fig. 37C.29, show (compared to figures 21.30A, 21.30B, 22.26, 22.27 and 22.36) not only a considerable extension of the spectral range, but also the elimination of much of the 1850 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C scattering (indicated by the higher ratios of the optical densities in the absorption peaks and the troughs between them). Except for increased sharpness, the infrared absorption bands of the purple bacteria, measured by Barer, are similar to those observed earlier by the Dutch investigators; they, too, show two or three peaks of varying relative intensity. Barer suggested that the 585-590 mju absorption band (cf. figs. 22.27 and 37C.29) is correlated with the band in bacterio- chlorophyll solutions located, depending on the solvent, between 575 and 610 m/i {cf. section 2(&) of this chapter). On p. 704, it was suggested that 370 410 450490 530 570 610 750 790 830 870 910 Fig. 37C.29. Absorption spectra of Rhodopseudo- monas spher aides (A) in the absence of iron, (B) in the presence of iron. Absorption spectrum of Rhodospirillum rubrum (C) (Barer 1955). this solution band has no counterpart in vivo; Barer's interpretation implies that such a counterpart does exist, but that, in contrast to the two other main bands, it is not significantly shifted towards the longer waves. For comparison, the main near-ultraviolet band of bacteriochlorophyll is visible, on Barer's absorption curves of purple bacteria, at 375 m^, indicat- ing a shift by 1500 cm.~^ from its position in ethereal solution (at 356 mju, cf. section 2(b) above); while the long-wave band is shifted by 500-2000 cm.~^ (assuming that all three peaks, at about 800, 850 and 890 m/z, corre- spond to the one solution band at 770 m/x). Similarly to French's two differently colored vaiieties of Streptococcus varians, mentioned on p. 707, Barer obtained, from the same inoculum, two strains of Rhodo- pseudomonas spheroides, with the spectra shown in fig. 37C.30. The red variety is characterised by peaks at 550 (weak), 512, 477, and 450 m/x; the green one, by peaks at 490, 460, 425 and 405 m^. Duysens (1952) found a difference between the absorption LIGHT ABSORPTION BY PIGMENTS 171 VIVO 1851 RHODOPSEUDOMONAS SPHEROIDES 5s^ 3bO 370 380 390 4004I0420 430 440440 460 470 460 490 SCO ilO S20 53C 540 iiO 560 570 iSO iio bOO 2 4 RHODOPSEUDOMONAS SPHEROIDES . , , . . »-*^r ■ , I I I — I — I — I — I I 1 1 1 1_ Vi 1 J— -J no 740 750 7b0 770 780 790 800 8IO »20 830 840 850 860 870 880 890 900910 920930 Fig. 37C.30. Absorption spectra of a red and a green strain of Rhodopseudomow (Barer 1955). as 1852 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C spectra of "young" and "old" cultures of Rhodospirillum rubrum, the former ones con- taining more rhodopin, the latter ones more spirilloxanthin (which is probably identical with fhodoviolascin, cf. this chapter 37B, section 2(d) and is characterized bj^ a band at 550 mn). Whether all these differences are based on analogous changes in the assort- ment of the carotenoids remains to be seen. Krasnovsky, Voynovskaya and Kosobutskaya (1952) suggested a new interpretation also of the three infrared absorption bands of bacterio- chlorophyll in vivo. As mentioned above, they attributed them to three different states of aggregation of bacteriochlorophyll (rather than to three different protein complexes, as suggested by Wassink and Katz, and Duysens); they pointed out that at least two, and perhaps all three, of these peaks can be observed also in colloids or solid films of chromato- graphed pigment containing no proteins or other foreign molecules {cf. section 3 above). This is not an implausible alternative to Wassink's interpretation, since, apart from spectroscopic data, no other evidence for the existence of three different bacteriochlorophyll-protein complexes in bacteria is pres- ently available. More difficult to reconcile with experimental data is, however, the additional hypothesis of Krasnovsky that the "aggregated" forms of bacteriochlorophyll, ^vith absorption bands at 800, 850 and 890 m/x, which account for the bulk of the pigment in vivo, are "storage forms," nonfluorescent and inactive, while the monomeric form, with a (hypo- thetical) absorption peak at 780-790 van, is the only "active" one. Krasnovsky, Kosobutskaya and Voynovskaya (1953) noted a prefer- ential bleaching and extraction of "Bchl 890" (compared to "Bchl 850" and "Bchl 800"), and suggested that "Bchl 890" is the one of the three "polymeric" forms most easily decomposed into the "monomeric" form. However, they acknowledged that no absorption band of the "monomer" could be observed in the expected location (780-790 m/z). To these difficulties of Krasnovsky's hypothesis, one can add the physical im- possibility of energy migration from the Bchl forms 890, 850 or 800, to the hypothetical, active "Bchl 780"; and the experimental proof by Duysens of energy migration in the opposite direction — towards Bchl 890. The fluorescence of the latter indicates that it serves as the final energy acceptor, and probably also as the actual photocatalyst in bacterial photosynthesis. We conclude that Krasnovsky's hypothesis is even less plausible in the case of purple bacteria than we found it to be in that of green plants. Krasnovsky and co-workers made some observations on the effect of external factors on the three absorption peaks of colloidal dispersions of bacteriochlorophyll-bearing material from purple bacteria, which origi- LIGHT ABSORPTION BY PIGMENTS lU VIVO 1853 11 ally had the three bands (800, 850, 890 m^) characteristic of live ^iAzor/io- daceae. Warming to 50-80° C. shifted the main absorption to 800 m/x; further heating to 100° C. brought it to 770 m/i (in analogy to the shift of the chloro- phyll band in boiled leaves) . Wetting with methanol, acetone, pyridine or dioxane causes a gradual transformation of the spectrum into the single-peak spectrum of bacterio- chlorophyll in solution, pyridine being the fastest acting agent. 500 JOO wavelength in mid Fig. 37C.31. Absorption spectrum of Chromatium (after Duysens 1952). Absorption at 950-570 my. is due to bacteriochlorophyll, that at 430-570 mix mainly to carotenoids. The first region is analyzed into three bands, attributed to the three molecular species, "B 890," "B 850" and "B 800." Outside the infrared peak, the spectra of these species are similar, with a maximum at 590 van. Action of acids weakens the 800 and 890 m^t bands, and enhances that at 850 m^i, indicating conversion to bacteriopheophytin, whose "aggre- gated form" has a band at 850 m^. No band shift occurs when acid- treated bacteria are heated — in accordance with the behavior of bacterio- pheophytin in solid films (c/. section 3 above). Duysens (1952) gave fig. 37C.31 for the absorption spectrum of Chro- matium and the contribution to it of the several carotenoids and the three bacteriochlorophyll-protein complexes, postulated by Wassink (p. 703). The latter have band peaks at 800, 850 and 890 mju, respectively. Only the "890 m/x type" (with absorption peak at 873 m^t) is present in Rhodo spirillum, rubrum. Upon heating autolysates from Chromatium, the absorption band at 890 m/i disappears; the other two infrared peaks are shifted slightly to- ward shorter waves and enhanced in intensity, thus compensating for the loss of absorption at the longer waves (fig. 37C.32). This change is plau- 1854 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C sibly explained by assuming that the "B 890" complex is converted, by heating, into the "B 850" and the "B 800" complexes. In Rhodospirillum, heating causes the 873 mix band to drop considerably in intensity, and a new band to arise at 780 m^ (fig. 37.33). Duysens suggested that this new band may indicate the formation of colloidal bacteriochlorophyll. Bsgo RHODOSPIRIL L UM RUB RUM absorption spectra _, -healed autolysate non heated auloiysate Fig. 37C.32. Absorption and fluores- cence spectra of autolysates of Chroma- tiurn (after Duysens 1952). Constant concentration. The long-wave shoulder disappears in heated autolysate, while the absorption at 850 mju, and particularly at 800 mju, is enhanced. Fluorescence peak of non-heated autolysate indicates fluo- rescence of B 890, that of heated autolysate, fluorescence of B 850. As long as B 890 is present, B 850 transfers excitation energy to it and does not itself fluoresce. Fig. 37C.33. Absorption spectra of autolj^sate of Rhodospirillum rubrum (after Duysens 1952). After heating, the main absorption maximum of the autolysate decreases and new absorption arises at about 780 m/x, indicating that bacterio- chlorophyll had passed into a new state. (Location of the band and absence of fluorescence suggest colloidal state.) The 590 m^t peak has a height proportional to the total amount of bacteriochlorophyll, as if all three complexes contributed equally to it — a striking fact, because one would expect complexing to influence also the higher excited electronic levels of the molecule, and not only the lowest one. Fig. 37C.31 illustrates the predominant role of carotenoids in the ab- sorption spectrum of Chromatium between 400 and 550 mju. Larsen (1954) gave an absorption curve of a suspension of the green sulfur bacteria, of which only the long-wave part had been described LIGHT ABSORPTION BY PIGMENTS in vivO 1855 previously by Katz and Wassink {cf. p. 704). He found, in Chlorobium thiosulfatophilum, two main absorption peaks at 747 and 457 mju; an ultraviolet band at 338 m/x, and shoulders, indicating hidden bands, at 423, 515 and 810 m/x (where Katz and Wassink observed a second strong infrared band). One enrichment culture showed a band peak at 732 miJL, instead of 747 mju. Barer (1955) also measured the absorption spectrum of ChloroUum. As shown by fig. 37C.34, he, too, found the two main bands at 745 and 455 van, respectively. No band at 810 m/x was visible. 350 370 390 410 430 450 470 490 510 530 550 570 590 610 630650 670 690 710 730750 770790 810 X(nn^) Fig. 37C.34. Absorption spectrum of green bacteria (after Barer 1955). The blue-violet (Soret) band of green bacteria is, in both position and shape, very similar to that of chlorophyll a in green cells; but the position of the red band is quite different. Perhaps this indicates that chlorophyll a is present in these bacteria in a special form (Wassink, 1954, pointed out that 745 m^ is the approximate location of the absorption band of chlorophyll a in crystals!); more Hkely, however, "bacterioviridin" is a compound different from chlorophyll a (although belonging to the same type, as suggested by Fischer and Stern, cf. p. 445), and that its absorption band undergoes — in comimon with that of bacteriochlorophyll — a much wider red shift in the transition from solution to the living cell than does the absorption band of chlorophyll in vivo. E.xperiments give even less support for the identification of bacterioviridin with bacteriochlorophyll, suggested by Seybold and Hirsch (1954) on the basis of absorption curves showing both a 750 m^i and an 850 van peak in green bacteria. One can only surmise that the cultures used by these investigators, had been contaminated with purple bacteria. Similarly, Seybold and Hirsch's finding of an absorption band at 1856 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C 750 m/i, in a culture of "Purpurspirillen" (and of a band at 660-670 m/i in extracts from this culture) points, according to Barer, to a contamination of purple bacteria cultures with green bacteria. (d) Phycohilin Spectra in vivo. New absorption curves of blue-green and red algae, with the allotment of absorption at selected wave lengths to the several pigments, were given by Duysens (1952); they will be reproduced in section 7 together with the action spectra of fluorescence (figs. 37C.51 and 53). (e) Changes in Absorption Spectrum during Photosynthesis An old problem is whether chlorophyll (or other cell constituents) undergo a chemical change during photosynthesis that could be detected C/ 03 02 RHODOSPIRILLUM RUBRUM .«-\ g I (3 9 B690 \ I \ ■absorption spectrum \ I \ « spectrum of the charge "of absorption (enlarged)-.^ •< &< (3 wave length in mfi __l l_ 08- OC 002 OM- 0.01 Q2 -001 CHROMATIUM ^ fx cbsorptijn ? \ « / t / 'J spectrum B890 A Baso \ \ P 1 1 . 1 Baoo 1 f spectrum of the J change of absorption ■ / (enlarged, cf ordinate sea faj 1 y wai/e length in mfi 1 y^ 1 - OOIS -0010 - 0.005 -0005 900 850 aoo 150 Fig. 37C.35. Absorption spectrum, and spectrum of the reversible change of ab- sorption by irradiation of Rhodospirillum rubriim, in tap water (left); and of colloidal extract from Chroviatimn (right) (after Duysens 1952). The ordinates of solid curves are on a scale ten times (left) or fifty times (right) smaller than those of the dashed curves. Decrease in absorption is plotted upwards, increase downwards. The main absorption maximum of B 890 is indicated by arrow. spectroscopically. Of the two previously known phenomena belonging to this field, one — the slow change in transmission of light by photo- synthesizing leaves (p. 680) — has been attributed partly to chloroplast realignment, and partly to scattering by newly formed starch grains. The LIGHT ABSORPTION BY PIGMENTS m VIVO 1857 other — variation in intensity of chlorophyll fluorescence associated with variations in the rate of photosynthesis {cf. chapter 24, section 4, and chapter 33, part B)— has been explained by changes in the composition of compounds associated ^\dth chlorophyll in the "pigment complex" (oxidation, reduction, formation and deposition of "narcotics"). Bell (1952) found the transmission of leaves to increase rapidly, by 3-5% (in the region 350-670 m/x) upon exposure to strong light [intensity up to 185 kerg/(cm.2 sec.)], in the absence of carbon dioxide; the trans- mission decreased again (by about 0.6%) when carbon dioxide was admitted. The increase in transparency was ascribed by Bell to reversible chlorophyll RHODOSPIRILLUM RUBRUM Od 0.03 002 ^ V , absorption spectrum of 8% fraction absorption spectrum of changed fraction 008 006 OO'i' 002 —9 RHODOSPIRHLUM RUBRUM b A _.absopption spectrum I' If T / \ of 20% fraction 6 1 1 1 1 1 j \\ ij if ll ij 11 \h absorption spectrum \\ qfchan ged fraction 1 1 1 J 8 uvave length 1 in mf2 1 L 1 — 850 ROO 900 850 800 750 Fig. 37C.36. Absorption sj>octra of bacteriochlorophyll in vivo during illumination, calculated from fig. 37C.35 by assuming that 8% (left) or 20% (right) of bacteriochloro- phyll (B 890) is transformed by u-radiation (after Duysens 1952). bleaching through transfer into a long-lived (> 0.02 sec.) excited state; he suggested that the back reaction is accelerated when the stored energy can be utihzed for photosynthesis {i. e., when carbon dioxide is present). No evidence of a change in spectroscopic composition (of the trans- mitted or fluorescent light) has been sought in all these studies. Duysens (1952) found, however, that the absorption spectrum of bacteriochloro- phyll changes when the bacteria are photochemically active. By a com- pensation method, permitting measurement of very small, sudden changes in transmission, Duysens found that illumination of a suspension of purple bacteria by a 500 w. lamp, [about 30 kerg/(cm.2 sec.)], causes an immedi- ate change in the absorption spectrum, which is rapidly reversed in the dark. Figiu'e 37C.35 (left and right) illustrates these changes for a Rhodo- spirillum suspension and in Chromatium extract. In Rhodospirillum 1858 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C (which contains "B 890" only) ilhimination causes a decrease in absorp- tion above 805 m/x, and an increase below 805 mju. In the first region, the changes are strongest at 872 and 812 uifx, in the second one, at 790 m^u. A similar picture was obtained with the colloidal extract from Chromatium (which has three absorption peaks). The maximum changes in fig. 37C.35 are of the order of 3 to 5%. These effects were strongest in the absence of hydrogen donors, i. e., when photosynthesis was prevented; as suggested by Bell (cf. above), spectral changes may be more pronounced under these conditions, because f 0.001 0 001 _ 0 002 - CYTOCHROME C OXIDIZED- REDUCED. i-Plt-i O Q ^^^65^ RHODOSPIRILLUM RUBRUM IRRADIATED- DARK WAVE LENGTH (m^) 600 550 500 4 50 •400 Fig. 37C.37. Changes in absorption spectrum in the green-blue-violet region in illuminated bacteria (after Duysens 1954^. the usual path by which the photochemically changed pigment returns into its normal state, is closed. It will be noted that the first maximum of the bleaching effect coincides with the absorption peak of "B 890." If the whole change is attributed to conversion of "B 890" into a different pigment, at least 8% of "B 890" must be changed. (Fig. 37C.36a and h show the spectrum of the changed pigment, calculated with two arbitrarily chosen percentage changes — 8 and 20% ; one sees that, if one would assume < 8% conversion, the cal- culated absorption coefficient at 820 mju would become negative.) Prob- ably, much more than 8% of "B 890" is changed, in light, into a pigment (or pigment complex) with a spectrum only shghtly different from that of "B 890" itself (as assumed in fig. 37C.246). Later, Duysens (1952, 1954^) noted that important reversible changes occur, in irradiated purple bacteria, also in the region 400-570 mju. These are represented in fig. 37C.37, for Rhodo spirillum ruhrum suspended in peptone solution (or in 0.03 M sodium acetate + phosphate buffer, pH 6.8), under anaerobic conditions and in relatively low light. The change is complete within a time of the order of one second. LIGHT ABSORPTION BY PIGMENTS 171 VIVO 1859 In stronger light, another spectral change becomes superimposed on the one represented in fig. 37C.37, probably correlated with the changes in the infrared represented in fig. 37C.35. While the latter reveals trans- formations in the pigment complex, the difference spectrum in fig. 37C.37 clearly indicates— as shown by the dashed line— the oxidation of a cyto- chrome-type pigment. A cytochrome similar to cytochrome c had been extracted from Rhodo spirillum ruhrum by Vernon (1953) and Elsden, Kamen and Vernon (1953). Duysens (1954^) made similar studies with Chlorella. No significant effect could be noted in the long-wave region of chlorophyll absorption; characteristic changes were observed, however, in the violet, blue, and green SSO 500 iSO Fig. 37C.38. Changes in absorption spectrum in illu- minated Chlorella (after Duysens 19542). (cf. fig. 37C.38); the peak at 420 m/x was tentatively attributed to the oxidation, in light, of cytochrome /—a compound found in chloroplast material by Hill and co-workers (cf. chapter 35, section B4(/)). The two other peaks — a "negative" one at 475 m/x and a "positive" one at 515 mix — could not be associated with any known pigment. These two peaks were found also in an illuminated leaf, an algal thallus, and a grass blade. (No measurements were made on these objects at 420 m/x.) The temptation is great to associate the 515 m/x band with bands in sunilar positions, which appear in reversibly reduced and reversibly oxi- dized chlorophyll in vitro, as well as in chlorophyll transferred, by intense illumination, into the metastable triplet state. We commented elsewhere in this chapter on the spectroscopic similarity of these products, and quoted Weller's suggestion that this similarity indicates that all of them are radi- cals (or biradicals) with interrupted all-round conjugation in the porphin system. There is, however, one difficulty in the way of attributing the 515 m/u band in illuminated Chlorella cells to chlorophyll-derived radicals: 1860 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C The apparent absence of an equivalent loss of absorption in the main chlorophyll bands (cf., however, page 1989). Witt (1954) applied to photosynthesizing leaves and algae a flash illu- mination technique, similar to that used by Livingston and Ryan (cf . chap- ter 35, part A) on chlorophyll solutions. The apparatus he used permitted measurement of absorption changes of <1%, produced by flash illumina- tion. Witt found a new absorption band which arose, in light, at 515 m/x; it disappeared, in the dark, within about 10"^ sec. The absorption at 475 m^ was reversibly weakened during the flash; but no mention was made + 1 ■4- -2-J^ 550 500 450 400 Fig. 37C.39. Changes in absorption spectrum of illuminated Porphyri- dium compared to the difference between the spectra of oxidized and reduced cytochrome / (after Duysens 19543). of a similar change at 670-680 m^, so that an uncertainty remains as to whether the above-quoted objection against attributing the 515 m^u band to a chlorophyll-derived radical remains valid in the case of Witt's experi- ments. Witt himself suggested such an attribution, and identified the decay period of the 515 m/i band with the " Emerson- Arnold period" in photosynthesis. However, it was pointed out in the discussion of that period in chapters 32 and 34, that the catalytic agent requiring 10"^ sec. for recovery must be present in a much smaller concentration than chloro- phyll to account for the observed saturation yield in flashing light and saturation rate in steady light. Witt found a light saturation of the photochemical processes generating the 515 mjLt band at flash intensities of the same order of magnitude as those known to cause flash saturation of photosynthesis. We return now to those absorption changes which, Duysens suggested, were due to reversible photoxidation in the cytochrome system. Lundegardh (1954) observed such changes by a flow method, in LIGHT ABSORPTION BY PIGMENTS ITl VIVO 1861 which a pre-ilkiminated Chlorella suspension was conducted through the cuvette of a spectrophotometer. Only the narrow range 540-570 m/x was examined — a range in which the absorption spectrum of Chlorella, pretreated with ascorbic acid, shows the a-bands of the reduced cyto- chromes b, c, and /. (The amounts of c and / are similar, the concentration of c being much higher in Chlorella than in the leaves of the higher plants.) Upon illumination, the 556 m^ band of reduced cytochrome / is found to be weakened, indicating oxidation, while those of the cytochrome b and c are unchanged; that of b may be even slightly enhanced. Duysens T 1 1 \ 400 350 Fig. 37C.40. Changes in the ultraviolet absorption spectrum of illu- minated Chlorella and Porphyridium (after Duysens 1954''). fig. 37C.38 did not show any effects above 550 m/x, but his subsequent experiments (1954^) revealed a small dip at 555 mju, in agreement with Lundeg&rdh's findings. Duysens (1954'') applied his stationary crossed-beam method also to a suspension of the red algae Porphyridium cruentum. The difference spectrum is shown in fig. 37C.39. Bands at 420 and 555 m^, character- istic of the oxidation of cytochrome / (or of another very similar pigment) appear more clearly on this picture, because the relatively stronger spectral effects, exhibited by Chlorella between these two bands are absent in the red alga. Duysens (1954^) also noted a reversible increase, during the illumina- tion of both Chlorella and Porphyridium, of the absorption in the near ultraviolet, with the indication of a peak at about 350 m/z (fig. 37C.40). This effect was tentatively interpreted as evidence of the reduction in light of a pyridine nucleotide; if this interpretation is correct, about one molecule of pyridine nucleotide must be reduced, in the photostationary state, for about 100 chlorophyll molecules present in the cell. 1862 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C After addition to a chloroplast suspension, of 10~^ m./I. DPN, a re- duction of 0.05% of the added nucleotide could be observed in the photo- stationary state. This very small shift in the oxidation-reduction equi- librium, whose discovery was made possible by the extreme sensitivity of the compensation apparatus, explains why no reduction of DPN could be observed directly in Hill reaction experiments (chapter 35, section B4(/)), and why DPNH2-mediated reductions of organic substrates could be obtained, by means of illuminated chloroplasts, only with very low yields: the rate of the back reaction (reoxidation of DPNH2) is so high as to make a significant accumulation of reduced DPN in the stationary state of the illuminated system impossible. As to the significance of the observations of Duysens, and of Lunde- gardh, concerning photostationary oxidation of cytochrome-like com- pounds in illuminated photosynthesizing cells, they can be interpreted either as evidence of direct participation of these compounds in the photo- chemical hydrogen transfer from water to carbon dioxide (or, rather, to an organic compound into which CO2 had been incorporated, such as PGA), or as evidence of their participation in oxidative processes (back reactions), coupled with the reduction process. The first hypothesis (Hill, Lunde- gardh) suggests photochemical transfer of electrons from reduced cyto- chrome to the organic acceptor (perhaps via DPN or TPN). The transfer of hydrogen (or electrons) from H2O to the oxidized cytochrome would then require another photochemical reaction. To account for the observed shift, the relative probability of the two photochemical reactions would have to be such as to establish a photostationary state with most of the cytochrome in the oxidized state. The quantum requirement of the hydrogen transfer reaction as a whole would be (at least) 8, since two quanta ^vill be needed to transfer each of the four required H atoms (or electrons), first from water to the cytochrome, and then from the cyto- chrome to the final acceptor. The other hypothesis (preferred by Duysens) is that the photochemical hydrogen transfer from H2O to TPN (or to another compound of a similar reduction potential), occurs directly, i. e., by a single quantum, but that one part of the reduced photoproduct reacts back \vith the oxidized photo- product, with a cytochrome as final or intermediate H acceptor, and the reoxidation energy may become available to assist in a further reduction step (as first suggested in the "energy dismutation" hypothesis in Vol. 1, chapters 7, p. 164 and 9, p. 233). If the reoxidation energy is stored as phosphate bond energy, the ATP produced in this way may, for example, enable reduced pyridine nucleotide to reduce PGA to a triose (as repeatedly suggested before, e. g., in Chapter 36, p. 1717). LIGHT ABSORPTION BY FIGMENTS in VWO 1863 (/) Calculation of True Absorption Spectra from Transmission and Fluores- cence Spectra of Suspensions Duysens (1952) suggested two methods to determine the true absorp- tion of light in small colored particles, such as Chlorella cells. The first method (Duysens and Huiskamp, 1953) is based on comparison of the intensities of fluorescence emitted "forward" and "backward" (in relation to the incident beam). If the particles are idealized as tiny plane-parallel vessels (thickness d) filled with pigment solution and illuminated normally to their "front wall," the intensity of fluorescence of wave length X/ excited by monochromatic light of wave length Xj, emitted through the "front wall" into a solid angle o-, is: (37C.8) Fl = ^ f^^ he'^^'^'fih, Xi)e'"^'' dx while that of the fluorescence escaping through the "rear wall" of the par- ticle (into an equal solid angle) is : (37C.9) Fx, = ^ /;/or"^^^/(V, XOr-^^^'^-^^ dx where f(\f\i) is the yield of fluorescence of wave length X/, excited by Xi, while a^f and ax.- are the (natural) absorption coefficients for these two wave lengths of the pigment solution inside the particle. According to equations (37C.8) and (37C.9), the ratio of the "forward" and the "backward" fluorescence emissions is a function of the products ax^d and a^id. If a second relationship between these two products is known, a^i and ax/ can be calculated from relative measurements of fluorescence intensity in the two directions. In applying this method to a suspension of Chlorella cells, Duysens and Huiskamp (1953) used the relationship: (37C.9A) (1 - exp(-«Xid))/(l - exp(-ax/rf)) = DpM/Dp.\f in which i)p,x.- and Dp^/ are the optical densities of the cell suspension at the two wave lengths. (This relationship follows from equation 37C.1G, derived below.) The wave length of the incident light (Xi) was 420 m/x, that of the measured fluorescence (X/), 680 m/x. The suspension was diluted so strongly that the transmission of the vessel was > 80%. Since the transmission of each single cell is about 40%, practical absence of mutual shading of the cells was assured. (This is a necessary condition for the application of the above equations, derived for individual particles, to the fluorescence emission of the suspension as a whole.) The calculation was further improved by assuming spherical rather than plane-parallel particles. An absorption of 64% was calculated in this way for a single Chlorella cell, at 680 mju. 1864 SPECTROSCOPY AND FLUORESCENCE OF PIGMENTS CHAP. 37C The second method of calculation, also devised by Duysens and Huis- kamp (1953), is based on the apparent depression of absorption coefficients by accumulation of colored molecules in small particles (as compared to the absorption caused by the same number of molecules in true molecular solution; c/. chapter 22, p. 714). The following derivation was made for the idealized case of cubic particles with one edge parallel to the incident beam. For a given wavelength, the optical density of a single particle (subscript p) is: (37C.10) d^ = In {l/T)p = In (///o)p = aCr>d where a is the molar absorption coefficient of the pigment, Cp its molar concentration (within the particles), and d the edge of the (cubic) particle. If the area density of the particles in suspension is A'' (per cm. 2), dis- persing the pigment uniformly would produce a solution (subscript S) with a molar area concentration c^Nd'^ (per cm. 2) and an optical density: (37C.il) Ds = In {h/I)s = aCj>Nd^ Assuming the molecular absorption coefficient to be the same in (37C.10) and (37C.11), we have: (37C.12) Ds = In {l/T\Nd'' Since d"^ is the cross section of a single particle, the average number p of particles the light beam crossing the suspension traverses is Nd^, and we can therefore wi'iie, instead of (37C.12): (37C.13) Ds = p\n{l/T\ When the molecules are bunched together in particles, the average optical density of the suspension can be calculated by using Poisson's formula for the probability of a beam encountering a certain number of particles, k, on 1 cm. of its path through the suspension. This probability is: (37C.14) Pk = e-^p^/k\ where p is, as above, the average number of particles encountered. The beam that traverses k particles is weakened by the factor Tp/ the average transmission of the suspension is therefore obtained by summation of: over all values of k from 0 to oo . This gives : (37C.15) f = e-pd-^p) or an average optical density of suspension (subscript P) : (37C.16) Dp = p(l - Tp) as compared with the optical density (37C.13) of the same pigment in LIGHT ABSORPTION BY PIGMENTS in VIVO 1865 molecular dispersion. The optical density is thus reduced, in consequence of particle formation, in the ratio: (^^^■'^'^^ Ds " In il/Tfp which is dependent only on Tp, the transmission of a single particle, and not dependent on the concentration of particles. According to this reason- ing, the "flattening" effect of particle formation on the absorption band should be independent of concentration. In other words, the "sieve effect" persists as "bunching effect" even when the concentration of parti- cles is so high that no beam can traverse the cell without passing through several particles (c/. below) . Relationship (37C.17) can be derived even more simply by applying Beer's law first to the pigment "solution" in the particle, and then to the "solution" of particles (considered as giant molecules). If the absorption coefficient of a single molecule (its "cross section for photon capture," to use the language of corpuscular physics) is a, and the number of mole- cules per unit area of the particle is n, we have, for the transmission of a single particle : (37C.18) Tp = (I/h)p = e-n-y and for the absorption : (37C.19) ^p = 1 - Tp = 1 - e-n^ = 2 where 2 can be considered as the photon capture cross section of the particle as a whole. Applying Beer's law a second time, to a suspension contain- ing A^ particles per unit area, we obtain : (37C.20) {I/h)p = e-N^ = e-m-e-'n = e-^d-^p) (37C.21) Dp = In (//7o)p = A^(1 - Tp) If the same total number of molecules, nN, is distributed at random over the same area. Beer's law gives : (37C.22) Ds = In (h/Ds = nN