mmmm mimm |r .... • '!«w: i' »■ ,M!(');i-;i liti.T.'lii; : Dl II Q Dl ID m E3 m 3^^^^^c3^^^^^^^^SE I Marine Biological Laboratory Library \^ J Woods Hole, Mass I I 1 ID D ra J Presented by (1 H I IE Interscience Publishers / c^.O i^y A meeting was held at (!atliiil)urg, Tennessee, October 25-29, 1955, to discuss some new observations and hypotheses in photosyn- thesis. The present volume owes its existence to the wish of the National Research Council, which sponsored the meeting, and of the National Science Foundation, which supported it financially, to have a record of the proceedings, permitting a greater number of scientists to benefit from the meeting than could be present at Gatlinburg. A book of this kind, if it contained only the papers read at the meeting, would offer only the type of information usually supplied by the scientific journals, in which these papers probably would have been published; its only advantage would have been the collection of a number of related articles under one cover. The meeting was held, however, primarilj^ to provide opportunity for extensive discussion; to convey its spirit, the editors decided to reprint more or less verba- tim those parts of the discussion which suggested unsolved problems, or pointed out where different investigators disagree in the interpre- tation of recent observations. The material now contained in the book includes about one-third of the total discussion taken down by the stenotypist. This material has been edited from the point of view of grammar and clarity; for the rest, the remarks have been left unchanged, except where it is expressly stated that the text has been altered or new material added in proof. The reader will find that occasionally, the discussion following a paper returns to subjects al- ready dealt with in connection with earlier papers. Nearly always it is obvious to what problem the riuestions and answers refer, and we did not try to transpose them. Transposition of discussion remarks has been made, however, in a few cases, where a whole paper was transferred in the book, into a group other than that in which it had been presented at the meeting. Experience gained at an earlier meeting had taught us that dis- cussions of this kind cannot be profitably extended beyond a period of about five days, and that, on the other hand, this period is not suf- ficient to cover thoroughly all facets of the problem of photosynthesis. The Subcommittee on Photobiology of the National Research Council decided therefore to omit from the agenda, among others, two big Vi PREFACE topics in photosynthesis: the chemistry of the reduction of carbon dioxide, and the (inestion of the smallest number of quanta capable of accomplishing this reduction. At the time of the meeting, the first of these topics had been treated in several reviews, which sug- gested general agreement concerning the experimental results and their interpretation. The second topic, too, was in a static condition — that of a disagreement which had persisted for twenty years. The program was organized mainly around the problem of the primary photochemical process in photosynthesis about which we know yet next to nothing — with the inclusion of those of the sec- ondary reactions which seemed most likely to furnish clues to the mechanism of the primary process. The editors wish to acknowledge the help of Dr. John Brugger in editing and transposing the discussion. Together ^^ith Miss Dolores Heffernan and Miss Marlene Roeder, he also took care of a great part of proofreading and indexing. It is a pleasant duty to express sin- cere thanks to all of them. June 1957 H- G. CONTRIBUTORS Edwin W. Abrahamson, Stale University of Forestry, Syracuse University, Syra- cuse, Xew York (Formerly at Department of Chemistry, Brookhaven National Laboratory, Upton, Long Lfland, A'ew York) Ingrid Ahrne, Slot tsg rand, Uppsala, Stceden (Formerly at Department of Plant Biology, Carnegie Institution of Washington, Stanford, California) F. L. Allen, Arthur D. Little, Inc., Cambridge, Massachusetts (Formerly at Research Institutes, University of Chicago, Chicai,o, Illinois) M. B. Allen, Laboratory of Plant Physiology, Department of Soils and Plant Nu- trition, University of California, Berkeley, California William Arnold, Biology Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee Daniel I. Arnon, Laboratory of Plant Physiology, Department of Soils and Plant Nutrition, University of California, Berkeley, California S. Aronoff, The Institute for A fomic Research and Departments of Botany and Chem- istry, Iowa State College, A mes, Iowa William E. Arthur, Western Electric Co., Chicago, Illinois (Formerly at Research Institutes, University of Chicago, Chicago, Illinois) Margareta Baltscheffskj^, Johnson Foundation for Medical Physics, University of Pennsylvania, Philadelphia, Pennsylvania Thomas T. Bannister, Department of Botany, University of Illinois, Urbana, Illinois J. A. Bassham, Radiation Laboratory, University of California, Berkeley, Cali- fornia Ralph S. Becker, Department of Chemistry, University of Houston, Houston, Texas Allen Benitez, Department of Plant Biology, Carnegie Institution of Washington, Stanford, California L. R. Blinks, Hopkins Marine Station of Stanford University, Pacific Grove, Cali- fornia Lawrence Bogorad, Department of Botany, University of Chicago, Chicago, Illinois F. S. Brackett, Laboratory of Physical Biology, National Institute of Arthritis and Metabolic Diseases, National Institutes of Health, Bethesda, Maryland Allan H. Brown, Department of Botany, University of Minnesota, Minneapolis, Minnesota T. E. Brown, Charles F. Kettering Foundation, Yellow Springs, Ohio John E. Brugger, Research Institutes, University of Chicago, Chicago, Illinois J. A. Businger, University of Wisconsin, Madison, Wisconsin (Formerly at Labora- tory for Plant Physiological Research, Agricidtural University, Wageningen, Netherlands ) Warren L. Butler, Agricidtural Experiment Station, USD A, BellsviUe, Maryland (Formerly at Research Institutes, University of Chicax^o, Chicago, Illinois) J. B. Capindale, Laboratory of Plant Physiology, Department of Soils and Plant Nutrition, University of California, Berkeley, California Ruth V. Chalmers, Department of Botany, University of Illinois, Urbana, Illinois vii Vm CONTRIBUTORS liritton Chance. Johnson Research Foundation for Medical I'hysics, University of Pennsylvania, Philadelphia, Pennsylvania K. A. Clendenning, Division of Marine Biology, Scripps Institution of Oceanog- raphy, La Jolla, California {Formerly at Charles F. Kettering Foundation, Yelloxo Springs, Ohio) J. W. Coleman, Photosynthesis Laboratory, University of Illinois, Urbana, Illinois It. G. Crickard, Department of the Army, Ordnance Corps, Diamond Ordnance Fuze Laboratories, Washington, D. C. (Formerly at Laboratory of Physical Biology, National Institute of Arthritis and Metabolic Diseases, National In- stitutes of Health, Bethesda, Maryland) L. N. M. Duysens, Biophysical Laboratory, Nieuwsteeq, Slate University, Leiden, Netherlands (Formerly at Department of Physics, University of Utrecht, Utrecht, Netherlands) Robert Emerson, Department of Botany, University of Illinois, Urbana, Illinois A. Gene Ferrari, Western Electric Co., Chicago, Illinois (Formerly at Research In- stitutes, University of Chicago, Chicago, Illinois) James Franck, Research Institutes, University of Chicago, Chicago, Illinois C. S. French, Department of Plant Biology, Carnegie Institution of Washington, Stanford, California Albert W. Frenkel, Department of Botany, University of Minnesota, Minneapolis. Minnesota Hans Gaffron, Depanment of Biochemistry and Research Institutes, University of Chicago, Chicago, Illinois Arthur T. Giese, Department of Plant Biology, Carnegie Institution of Washington,. Stanford, California S. Granick, Rockefeller Institute for Medical Research, New York City, New York Helen M. Habermann, Research Institutes, University of Chicago, Chicago, Illinois (Formerly at Department of Botany, University of Minnesota, Minneapolis, Minnesota) Daniel D. Hendley, Department of Microbiology, University of Sheffield, Sheffield, England (Formerly at Department of Biochemistry and Research Institutes, University of Chicago, Chicago, Illinois) Toyoyasu Hirokawa, The Tokugaioa Institute for Biological Research and Depart- ment of Botany, Faculty of Science, University of Tokyo, Tokyo, .Japan A. Stanley Holt, Division of Applied Biology, National Research Laboratories, Ottawa, Canada (Formerly at Photosynthesis Laboratory, University of Illinois, Urbana, Illinois) Leonard Horwitz, University of California, Berkeley, California (Formerly at Re- search Institutes, University of Chicago, Chicago, Illinois) Martin D. Kamen, Department of Biochemistry, Brandeis University, Waltham, Massachusetts (Formerly at Edward Mallinckrodt Institute of Radiology, Wash- ington University Medical School, St. Louis, Missouri) Erich Kessler, Botanisches Institut, Universitdt Marburg, Marburg/Lahn, Germany (Formerly at Research Institutes, University of Chicago, Chicago, Illinois) B. Kok, Laboratory for Plant Physiological Research, Agricultural University, Wageningen, Netherlands CONTRIBITTORS ix Albert R. Krall, RIAS, Inc., Baltimore, Maryland {Formerly at Biology Division, Oak Ridge Notional Laboratori/, Oak Ridfjc, Tennessee) Donald W. Kupke, Depart.mrni of Biochemistry, School of Medicine, University of Virginia, Charlottevillc, Virginia {Formerly at Deportment of Plant Biology, Carnegie Institution of Washington, Stanford, California) Paiil Latimer, Department of Physics, Vanderhilt University, Nashville, Tennessee {Formerly at Photosynthesis Laboratory, University of Illinois, Urbana, Illinois) Henry Linschitz, Department of Chemistry, Brandeis University, Waltham, Massa- chusetts {Formerly at Department of Chemistry, Syracuse University, Syracuse, New York) Robert Livingston, School of Chemistry, Institute of Technology, University of Min- nesota, Minneapolis, Minnesota) Josef E. Loeffler, Hochschule filr Bodenkultur {Chemie), Vienna, Austria {Formerly at Department of Plant Biology, Carnegie Institution of Washington, Stanford, California) Rufus Lumry, School of Chemistry, Institute of Technology, University of Minne- sota, Minneapolis, Minnesota V. H. Lynch, Department of Plant Biology, Carnegie Institution of Washington, Stanford, California Shigetoh Miyachi, The Tokugawa Institute for Biological Research and The In- stitute of Applied Microbiology, University of Tokyo, Tokyo, Japan Jack Myers, Department of Zoology, University of Texas, Austin, Texas Jack W. Newton, Edward Mallinckrodt Institute of Radiology, Washington Uni- versity Medical School, St. Louis, Missouri John M. Olson, Johnson Research Foundation for Medical Physics, University of Pennsylvania, Philadelphia, Pennsylvania R. A. Olson, Laboratory of Physical Biology, National Institute of Arthritis and Metabolic Diseases, National Institutes of Health, Bethesda, Maryland Gerald Oster, Polytechnic Institute of Brooklyn, Brooklyn, New York A. Pirson, Botanisches Institut, Universitdt Marburg, Marburg /Lahn, Germany Eugene I. Rabinowitch, Photosynthesis Laboratory, University of Illinois, Urbana, Illinois George H. Reazin, Jr., Research Department, Joseph E, Seagram & Sons, Inc., Louisville, Kentu/;ky J. L. Rosenberg, Department of Chemistry, University of Pittsburgh, Pittsburgh, Pennsylvania Lawson L. Rosenberg, Laboratory of Plant Physiology, Department of Soils and Plant Nutrition, University of California, Berkeley, California K. Shibata, Tokugawa Institute for Biological Research, Toshitnaku, Tokyo, Japan {Formerly at Radiation Laboratory, University of California, Berkeley, Cali- fornia) James H. C. Smith, Department of Plant Biology, Carnegie Institution of Washing- ton, Stanford, California Lncile Smith, Johnson Foundation for Medical Physics, University of Pennsylvania, Philadelphia, Pennsylvania (Formerly at MoUeno Institute, University of Cam- bridge, Cambridge, England) X CONTRIBUTORS John D. Spikes, Department of Experimental Biology, University of Utah, Salt Lake City, Utah C. J. P. Spruit, Landhouwhoqe School, Wageningen, Netherlands Bernard L. Strehlor, National Heart Institute, National Institutes of Health, Bclh- esda, Maryland {Formerly at Department of Biochemistry and Research In- stitutes, University of Chicago, Chicago, Illinois) S. Takashima, School of Chemistry, Institute of Technology, University of Minne- sota, Minneapolis, Minnesota Hiroshi Tamiya, The Tokugawa Institute for Biological Research and The Institute of Applied Microbiology, University of Tokyo, Tokyo, Japan N. E. Tolbert, Biology Division, Oak Ridge National Laboratory, Oak Ridge, Ten- nessee Hemming I. Virgin, Botanical Laboratory, University of Lund, Lund, Sweden {Formerly at Department of Plant Biology, Carnegie Institution of Washington, Stanford, California ) Wolf Vishniac, Department of Microbiology, Yale University, New Haven, Connecti- cut E. E. Walldov, Charles F. Kettering Foundation, Yellow Springs, Ohio E. C. Wassink, Laboratory of Plant Physiological Research, Agricidtural University, Wageningen, Netherlands F. R. Whatley, Laboratory of Plant Physiology, Department of Soils and Plant Nutrition, University of California, Berkeley, California C. P. Whittingham, The Botany School, Cambridge University, Cambridge, England H. T. Witt, Physikalisch-chemisches Institut, Universitat Marburg, Marburg /Lahn, Germany L. P. Zill, Research Institute for Advanced Studies, Inc., Baltimore, Maryland {Formerly at Biology Division, Oak Ridge National Laboraiory, Oak Ridge, Tennessee) CONTENTS PART I: Absorption, Fluorescence, Luminescence, and Photochem- istry of Pigments in Vitro The Photochemistry of ChlorophvU in Vitro Robert Livingston 3 The Electronic Spectra of Chlorophylls and Related Molecules Ralph S. Becker 13 Fluorescence Spectra of Protochlorophyll, Chlorophylls c and d, and Their Pheophytins C. S. French, James H. C. Smith, and Hemming I. Virgin 17 General Remarks on Chlorophyll-Sensitized Photochemical Reactions in Vitro James Fra nck 19 Reversible Spectral Changes in Chlorophyll Solutions, Following Flash Illumination Henry Linschitz and Edwin W. Abrahamson 31 Infrared Spectroscopy of Chlorophyll and Derivatives A. Stanley Holt 37 Dark- and Light-Activated Chemiluminescence of Chlorophyll in Vitro A. Gene Ferrari, Bernard L. Strehler, and William E. Arthur 45 Photoreduction of Synthetic Dyes Gerald Oster 50 PART II: Absorption, Scattering, Fluorescence, Luminescence, and Primary Photochemical Process in Vivo Methods for Measurement and Analysis of Changes in Light Absorp- tion Occurring upon Illumination of Photosynthesizing Organisms L. N. M. DuYSENS 59 Reversible Bleaching of Chlorophyll in Vivo J. W. Coleman, A. Stanley Holt, and Eugene I. Rabinowitch 68 Reaction Patterns in the Primary Process of Photosynthesis H. T. Witt 75 Spectroscopy of Flash-Illuminated Chloroplasta J. L. Rosenberg, S. Takashima, and Rufus Lumry 85 Absorption Spectrum Changes in Chlorella and the Primary Process Bernard L. Strehler and V. H. Lynch 89 -^ Selective Scattering of Light by Pigment-Containing Plant Cells Paul Latimer and Eugene I. Rabinowitch 100 The Absolute Quantum Yields of Fluorescence of Photosynthetically Active Pigments Paul Latimer, Thomas T. Bannister, and Eugene I. Rabino- witch 1*^' XI 75442 Xll CONTENTS fluorescence Yield of Chlorophyll in ChloreUa as a Function of Light Intensity John E. Brugger 113 Introductory Remarks on the Luminescence of Photosynthetic Organ- isms Bernard L. Strehler 118 Decay of the Delaj'ed Light Emission in ChloreUa William Arnold 128 Some Observations on the Chemiluminescence of Algae John E. Brugger 134 A Theory of the Photochemical Part of Photosynthesis James Franck 142 PART III: The Possible Role of Cytochromes Hematin Compounds in the Metabolism of Photosynthetic Tissues Martin D. Kamen 149 Investigations in the Photosynthetic Mechanism of Purple Bacteria by Means of Sensitive Absorption Spectrophotometrj'^ L. N. M. DuYSENS 164 Oxidation of Cytochromes upon Near-Infrared Irradiation of Chroma- tiutn John M. Olson 174 The Reactions of RhodospiriUimi rubriim Extract with Cytochrome c LuciLE Smith 179 On the Time Sequence of Reactions in the Anaerobic Light Effect in Rhodospirilhmi rubrum Britton Chance 184 The Effect of Hydrosulfite on the Anaerobic Light Effect Britton Chance and Lucile Smith 189 The Fast Light Reaction of Extracts and of Inhibited Cell Suspensions Britton Chance, Margareta Baltscheffsky, and Lucile Smith 192 PART IV: "Dark" Reactions 1 . Fixation of Carbon Dioxide The "Background" CO2 Fixation Occurring in Green Cells and Its Pos- sible Relation to the Mechanism of Photosynthesis Shigetoh Miyachi, Toyoyasu Hirokawa, and Hiroshi Tamiya . 205 Some New Preillumination Experiments with Carbon-14 Hiroshi Tamiya, Shigetoh Miyachi, and Toyoyasu Hirokawa. 213 Photosynthesis by the Etiolated Plant after Exposure to Light N. E. Tolbert 224 Excretion of Glycolic Acid by ChloreUa during Photosynthesis N. 1']. Tolbert and L. P. Zill 228 2. Photoreduction loith Various Reductants Oxygen Evolution and Photoreduction by Adapted Scenedesmus Leonard Horwitz and F. L. Allen 232 CONTENTS XUl Photoreduction in Ochromonas malhamensis Wolf Vishniac and George H. Reazin, Jr 239 Manganese as a Cofactor in Photosynthetic Oxygon Evolution Erich Kessler 218 3. Reduction of Various Oxidants Contributions to the Problem of Photochemical Nitrate Reduction Erich Kessler 250 Certain Effects of Ascorbic Acid on the Reduction of Oxygen in Chloro- plast Preparations Helen M, Habermann and Allan H. Brown 257 4. Reactions in Chloroplasts and Cell Extracts Some Features of the Chloroplast Reaction C. P. Whittingham 263 Natural Inhibitors of the Hill Reaction K. A. Clendenning, T. E. Brown, and E. E. Walldov 274 Light-Dependent Reductions in a Cell-Free System Wolf Vishniac 285 Photosynthetic Carbon Dioxide Fixation by Broken Chloroplasts M. B. Allen, F. R. Whatley, Lawson L. Rosenberg, J. B. Cap- INDALE, and Daniel I. Arnon 288 General Concept of Photosynthesis by Isolated Chloroplasts Daniel I. Arnon, M. B. Allen, and F. R. Whatley 296 5. Phosphate Metabolism Light-Induced Phosphorylation by Cell-Free Preparations of Rhodo- spirillum rubrum Albert W. Frenkel 303 Light-Induced Phosphorylation in Extracts of Purple Sulfur Bacteria Jack W. Newton and Martin D. Kamen 311 A Light-Reversible Carbon Monoxide Inhibition of Isotopic Phosphate Uptake by Photosynthesizing Barley Leaves Albert R. Krall 313 Isolation of a Possible Primary Hydrogen Acceptor from Photosj^n- thesizing Barley Leaves Albert R. Krall 327 Phosphate in the Photosynthetic Cj^cle in Chlorella E. C. Wassink 333 Photosynthetic Phosphorylation by Isolated Spinach Chloroplasts F. R. Whatley, M. B. Allen, and Daniel I. Arnon 340 PART V: Kinetics, Transients, and Induction Phenomena On the Efficiency of Photosynthesis above and below Compensation of Respiration Robert Emerson and Ruth V. Chalmers 349 XIV CONTENTS Report of Some Recent Results at Wageningen I. Transitory Rates B. KoK and C. J. P. Spruit 353 II. Kinetics of I'hotosynthcsis B. KoK and J, A. Businger 35-4 III. Photoinhibition B. KoK and J. A. Businger 357 Mechanism of the Initial Steps in Photosynthesis J. A. Bassham and K. Shibata 366 Chemical-Kinetic Studies of the Hill Reaction RuFUS LuMRY and John D. Spikes 373 Kinetics of the Photosynthetic Incorporation of Radiocarbon S. Aronoff. 392 Transient Phenomena in Leaves as Recorded by a Gas Thermal Con- ductivity Meter Warren L. Butler 399 Transient Changes in Cellular Gas Exchange Robert Emerson and Ruth V. Chalmers 406 Induction Phenomena in Photosynthetic Algae at Low Partial Pres- sures of Oxygen C. P. Whittingham 409 Transients in O2 Evolution bj'^ Chlorella in Light and Darkness I. Phenomena and Methods F. S. Brackett, R. a. Olson, and R. G. Crickard 412 II. Influence of O2 Concentration and Respiration R. A. Olson, F. S. Brackett, and R. G. Crick.\rd 419 Transients in the Carbon Dioxide Gas Exchange of Algae Hans Gaffron 430 Chromatic Transients in Photosynthesis of Red Algae L. R. Blinks 444 Transients in Acid Production by Purple Sulfur Bacteria Daniel D. Hendley 450 PART VI: Formation and Condition of Chlorophyll in the Living Cell Chloroplast Structure and Its Relation to Photosynthesis S. Granick 459 The Natural State of Protochlorophyll James H. C. Smith, Donald W. Kupke, Josef E. Loeffler, Aii- LEN Benitez, Ingrid Ahrne, and Arthur T. Giese 464 The Enzymatic Synthesis of Uroporphyrin Precursors Lawrence Bogorad 475 On Uniformity of Experimental Material Jack Myers 485 Induced Periodicity of Photosynthesis and Respiration in Hydrodicfyon A. PiRSON. . /. 490 Index 501 Part I ABSORPTION, FLUORESCENCE, LUMINESCENCE, AND PHOTOCHEMISTRY OF PIGMENTS IN VITRO The Photochemistry of Chlorophyll in Vitro ROBERT LIMNGSTON, Institute of Technology, University of Min- nesota, Minneapolis, Minnesota It is the primary act, the capture of light energy for chemical pur- poses, which makes photosynthesis the most important and the most fascinating of biological processes. Even if every enzyme, with its attendant substrate, product, and coenzyme, which enters into the biochemical cycles of the living plant were known, we might still be completely ignorant of the primary process which distinguishes photo- synthesis from other biological processes. There is no known straight- forward way by which the nature of the primary act can be deter- mined. In fact, there are very few photochemical reactions of complex molecules in solution whose primary act has been unequivocally de- termined. In view of the difficulty of the problem, it appears worth while to investigate many different photochemical and spectroscopic properties of chlorophyll in solution, in synthetic and natural aggre- gates of chlorophyll molecules, and in the intact cell. The study of dilute solutions of chlorophyll is certainly insufficient to establish the nature of the primary act of photosynthesis. How- ever, knowledge gained from such study should enable the student of photosynthesis to reject many otherwise plausible interpretations of the primary act. This discussion is limited to recent developments of the photochemistry of chlorophyll solutions. Irreversible photo- chemical reactions of chlorophyll and reactions photosensitized by chlorophyll have been excluded. Although the irreversible reactions are of intrinsic interest to organic chemistry, they have no direct re- lation to normal photosynthesis. An adequate discussion of photo- sensitized reactions would require more space than is warranted by their importance. Furthermore, such reactions, occurring in dilute homogeneous solutions, cannot serve as reasonable models of photo- synthesis. 3 4 li. LIVINGSTON ISOMERS, TAUTOMERS, AND ADDITION COMPOUNDS OF CHLOROPHYLL Freed and Saucier (1) have shown that a tautomeric form of chloro- phyll is stable at low temperatures, and they have studied, at inter- mediate temperatures, the tautomeric equilibria. This tautomer- ism is fast even at 150°K. The absorption maxima of the low-tem- perature forms are displaced to the red by about 300 A. The tem- perature at which the low- and high-temperature forms are present in comparable amounts is a sensitive function of the isomeric form of the chlorophyll. For example, the spectra of chlorophylls b and h' are almost identical at room temperature. Reducing the temperature to 75°K. changes both alike, so that the spectra of their low-tem- perature forms are also identical. Plowever, the temperatures at which the tautomers coexist in equal amounts are 180° and 230°K., respec- tively, for chlorophylls h and b'. (More recent work by the same authors (./. Am. Chem.. Soc, 76, 198, 6006 (1954)) indicates that the species that are stable at low temperatures are probably solvates rather than tautomers of the unsolvated molecules.) Chlorophylls a and b form addition compounds with water, alcohols, amines, and other "bases." Similar adducts are formed by bases with metal-complexed porphyrins and chlorins, but not with pheophytins or metal-free porphyrins (2). These facts support Krasnovskii's con- tention (3) that the point of addition is the metal atom of the pigment. Unlike the other pigments, the fluorescent yield of chlorophyll is strongly affected by the addition of the base. Solutions of chloro- phyll in pure dry hydrocarbons, or similar solvents, are practically nonfluorescent. The stability constants of compounds of chloro- phylls, porphyrins, and chlorins with a given base do not differ by a factor of more than seven. As measured by these stability constants the oxygen bases are abnormally strong relative to nitrogen bases (4). At — 80°C. a solution of chlorophyll in isopropylamine is reddish brown, but when the solution is warmed to ordinary temperatures it becomes green and exhibits its normal spectrum (5). Except for a slow side reaction at the higher temperatures, the process is strictly reversible. The absorption spectrum of the low-temperature solution is identical with that of the Molisch phase test intermediate (6). This spectrum is similar to, but not the same as, the spectrum of chloro- phyll in its triplet state, as formed by flash illumination (7). Since there is good reason for the belief that the phase test intermediate PHOTOCHEMISTRY OF CHLOROPHYLL in vUrO 5 is an ion, formed by the loss of a proton from Cio it appears plausible that the stable electronic state of this ion is a triplet. PROPERTIES OF CHLOROPHYLL IN ITS SINGLET EXCITED STATES Chlorophyll, excited by the absorption of blue light, emits only the usual red fluorescence. The molecule goes from its second to its first excited singlet state by a radiationless process of high probability (internal conversion). Since visible fluorescence having a yield as high as 10~' is readily detectable and no blue or green fluorescence has been obser^'ed, the quantum yield, of this fluorescence is less than 10~'. The actual mean life, r = ^r°, where t° is the intrinsic mean life, must be less than 10"' X 10^^ = 10-^^ second. It is, therefore, most improbable that the second excited singlet state pla3^s any direct role in the photochemistry of chlorophyll. In other words, the photochemical action of visible light should be independent of wave- length whenever chlorophyll is the only light-absorbing entity pres- ent. This does not necessarily apply to the absorption of ultraviolet light. Higher excited states may possibly enter directly into photo- chemical reactions, but such processes would not be related to normal photosjTithesis. The intrinsic mean life, t°, of the first excited state of chlorophyll may be calculated (4) from its integrated extinction coefficient for "red" fight. The actual mean life, r, may be obtained (4) from meas- urements of the degree of polarization of fluorescent light in viscous solvents, by the use of Perrin's equation. Since the maximum quan- tum jdelds of fluorescence, (p°, are known (8) for chlorophylls a and h, the intrinsic fife may be obtained from the actual life, as t° = t/(p°. The presently available values are listed in Table I. Weil's value for t° (h) is based upon measurements made with benzyl alcohol as the solvent, assuming that i Fig. 1. Absorption spectra of the singlet and triplet states of chlorophyll in pyridine (1.3 X 10-« m). quenching process in fluid solvents. The study of the reactions of the triplet state has scarcely begun. It is known that molecular oxygen quenches the triplet state efficiently; presumably, at every en- counter. AUylthiourea has little if any effect upon the mean life of the triplet state. Further investigation of these reactions should be of great value in the interpretation of the primary act of photochemical reaction?. 10 R. LIVINGSTON REVERSIBLE PHOTOCHEMICAL REACTIONS OF CHLOROPHYLL Anaerobic solutions of chlorophyll in niothaiiol, (^ther, etc., but not in hydrocarbons show rapid reversible changes in their ab- sorption spectra when they are strongly illuminated. Under experi- mentally realizable conditions, the maximum chan}:;e in the absorp- tion of monochromatic light is less than 1%. Since the percentage change is proportional to the square root of the intensity of the ab- sorbed light, we may conchide that the labile product is a pair of radicals or ions which reform chlorophyll by reeombining. The ab- sorption spectrum of this intermediate has been determined (14) only very crudely, but appears to resemble that of the triplet state (7). Although this reaction is intrinsically interesting and may be related to the primary act of photosensitized reactions, the last publication dealing with the subject appeared in 1950 (16). It was observed by Linschitz (15) that chlorophyll, in a rigid solvent (EPA at liquid nitrogen temperatures) containing a quinone, is photochemically transformed into a product w^hich is stable at low temperatures but reverts immediately to normal chlorophyll when the glass is allowed to w^arm to its softening point. The ab- sorption spectrum of this product is, likewise,, generally similar to that of the triplet state. The earlier finding of Krasnovskii that chlorophyll can be revers- ibly, photochemically reduced by ascorbic acid, when dissolved in pyridine, has been confirmed by A. S. Holt and E. Rabinowitch and by Evstigneev and Gavrilova (17b). The latter authors also observed this reaction in toluene with phenylhydrazine as the reducing agent. They have further shown that two labile intermediates are formed, and have postulated that the maximum at 585 mu is due to a neutral molecule or radical which ionizes in basic solution with the resulting appearance of a maximum at 518 m/z. They measured the fluorescence and low-temperature phosphorescence of these substances. Discussion Rabinowitch: How does the absorption curve of the triplet state compare with that of ionized chlorophj'll? Livingston: The absorption spectra of the ions, of the triplet state, of Kras- novskii's reduced form, and of Linschitz' chlorophyll-quinone intermediate, are similar but not identical. They show distinct differences; but in each case the two singlet peaks disappear, the Soret band is replaced by what looks Uke two PHOTOCHEMISTRY OF CHLOROPHYLL in vUrO 11 separate peaks, and there is apparently a new absorption band in the far red or near infrared. Rabinowitch: There seems to be a definite distinction in that metastable (triplet) chlorophyll a has a large band at 475 m/* and only a slight indication of a hand at 525 m/j (or somewhere in that region), while Krasnovskii's reduced chloro- phyll has a very prominent peak at 515 m/i, but at 475 ni/x it shows only a nega- tive effect (decrease in absorption). The ionized (and also perhaps the oxidized") chlorophyll a, as well as the metastable and the reduced form all have a band at 515-525 m^, but I think there is a definite distinction between the metastable and the reduced form at 475 m^t. This, I think, is important, particularly in con- nection with what we will report later on the changes in the absorption spectrum of illuminated Chlorella cells. Duysens : Mr. Goedheer has made some careful measurements of the polariza- tion spectra of the fluorescence of the chlorophylls, and his interpretation of these measurements is that the Soret band consists of two different bands with transitions perpendicular to each other. There is between the Soret band and the red absorption band of chlorophyll and of bacteriochlorophyll another weaker electronic band, with a transition perpendicular to that of the red absorption band. Rabinowitch: Kuhn has assigned one of the weak chlorophyll bands in the green to a different electronic transition. This could not be confirmed at Utrecht. On the other hand, another band in the yellow region appeared to belong to a separate electronic transition. We, too, could not confirm Kuhn's result, and thought that it may have been due to contamination of his chlorophyll with pheophytin, which has a stronger band in the green than chlorophyll. Linschitz : We have a flash spectrum which shows a peak for the metastable state somewhat further toward the green, at about 530 m/x. Rabinowitch : Do you have a prominent peak at 475? Linschitz: We have a suggestion of a shoulder at 475, but we have a higher peak further out toward the green, where Livingston found a somewhat lower absorption. {Note added in proof: Since this work was prepared for press, Dr. Linschitz com- pleted a series of careful measurements of the absorption spectra of the chloro- phylls a and b in their triplet state. Chloroph.yll a in its transient state has its main peak at 485 m/x and a lesser peak at 385 mp. The absorption extends throughout the visible, being still appreciable at 750 m/z- There does not appear to be a maximimi near 530 mn, nor on the long wavelength side of the normal red peak. — R. L.) References 1. Freed, S., and Sancier, K. M., Science, 114, 275 (1951); ibid., 116, 175 (1952). 2. Livingston, R., and Weil, S., Nature, 170, 750 (1952). 3. Evstigneev, V. B., Gavrilov, W. A., and Krasnovskii, A. A., Doklady Akad. Nauk. S.S.S.R., 70, 261 (1950). 4. Weil, S., Doctoral dissertation, University of Minnesota, Minneapolis, 1952. 5. Freed, S., and Sancier, K. M., Science, 117, G55 (1953). G. Weller, A., ./. Am. Chem. Soc, 76, 5819 (1954). 12 R. LIVINGSTON 7. Livingston, R., /. Am. Chem. Soc, 77, 2179 (1955); Livingston, R., Porter, G., and Windsor, M., Nature, 173, 485 (1954). 8. Forster, L., and Livingston, R., /. Chem. Phys., 20, 1315 (1952). 9. Stupp, R., and Kuhn, H., Helv. Chim. Acta, S6, 2469 (1952). 10. Watson, W. F., and Livingston, R., J. Chem. Phys., 18, 802 (1950). 11. Duysens, L., Nature, 168, 548 (1951). 12. Becker, R., and Kasha, M., /. Am. Chem. Soc, 77, 3669 (1955). 13. Porter, G., Proc. Roy. Soc. (London), A200, 284 (1950); Porter, G., and Windsor, M., /. Chem. Phys., 21, 2088 (1953). 14. Livingston, R., and Ryan, V., J. Am. Chem. Soc, 75, 2176 (1953). 15. Linschitz, H., and Rennert, J., Nature, 169, 193 (1952); also, material pre- sented at the Gatlinburg meeting, 1955. 16. Knight, J., and Livingston, R., /. Phys. & Colloid Chem., 54, 703 (1950). 17. a. Krasnovskii, A., and Voinovskaj^a, K., Doklady Akad. Naijc S.S.S.R., 87, 109 (1952). b. Evstigneev, V., and Gavrilova, V., Doklady Akad. Nauk S.S.S.R., 91, 899 (1953). c. Evstigneev, V., and Gavrilova, V., Doklady Akad. Nauk S.S.S.R., 95, 84, (1954). General References A. Rabinowitch, E., Photosynthesis, Vol. I, Chapter 18. Interscience, New York, 1945. B. Rabinowitch, E., Photosynthesis, Vol. II, Part 1, Chapters 21, 23. Interscience, New York, 1951; Vol. II, Part 2, Chapters 35, 37B, 37C. Interscience, New York, 1956. C. Forster, T., Fluoreszenz Organischer Verhindungen. Vandenhoeck and Rup- recht, Gottingen, 1951. ' L I B R A ^ Y t ^ The Electronic Spectra of Chlorophylls and Related Molecules RALPH S. BECKER, Department of Chemistry, University of Houston, Houston, Texas The present discussion will deal only with recent advances in the knowledge of the absorption and emission spectra of the chlorophylls and their derivatives. The nature of the triplet state, the general significance of quantum yields, the effect of electric and magnetic fields, and the like have been discussed elsewhere. ABSORPTION SPECTRA AND DEDUCTIONS The interpretation of the fluorescence activation of the chlorophylls by Livingston, Watson, and McArdle (3) has been previously dis- cussed ( 1 ) . However, there are several additional factors to be pointed out. Freed and Sancier (5) have shown, from absorption data, the presence of various solvates and temperature-dependent isomers of chlorophylls and related compounds. From these data, it was deter- mined that if solvation occurs at the magnesium atom of chloro- phylls, it also occurs at or near the two corresponding hydrogen atoms of pheophytin. Russian workers (6) emphasized that the differ- ence in activation of fluorescence between pheophytin and chlorophyll was due to the fact that hydration of magnesium was necessary. Thus, the fluorescence of pheophytin, which contains no magnesium, was insensitive to the presence or absence of water vapor. Although pre- viously indicated, with consideration of the work of Freed, it now may be emphasized that the presence of the magnesium may well cause the perturbation of the electronic system of the chlorophyll molecule necessary for the facts concerning fluorescence to become observable. Consequently, the observed results concerning the effects of the presence or absence of water on the fluorescence may be only coincidental facts of interest accompanying the real cause of the 13 14 R, S, BECKER difference in the two molecules, that is, the effect of the presence or absence of magnesium in the molecules. EMISSION SPECTRA AND DEDUCTIONS It has been recently shown that tiie phosphorescence of chlorophyll 6, first reported by Calvin and Dorough (4), is a bona fide emission occurring at 8650 A (2). An analog of chlorophyll, pheophorbide a, has a strong fluorescence and no phosphorescence. However, in the divalent copper complex of pheophorbide a, the fluorescence is com- pletely quenched and only a strong phosphorescence is observed. Moreover, the phosphorescence occurs at 8675 A. The complete quenching of fluorescence in the metallophosphorbide and the like posi- tions of the phosphorescent emissions indicate that the phosphores- cences observed in the pheophorbide complex and in the chlorophyll are the lowest triplet to singlet emissions. Although intrinsic phosphorescence was not unequivocally ob- served from chlorophyll a, there was indication of this type of emission. The difference between chlorophyll a and chlorophyll h is primarily due to the greater sensitivity of chlorophyll a to photodecomposition with accompanying spurious emissions. The difference in photosensi- tivity of chlorophylls a and b was further borne out by Dr. Linschitz in his report at the Conference on Photosynthesis in 1955. Chlorophyll a is being subjected to further investigation. Although the quantum yield of phosphorescence was low (<0.1) for chlorophyll b, the extrapolation to the in vivo system allows im- portant details to alter the value of the quantum yield. Earlier con- siderations by Becker and Kasha (1) indicated the very important possibilities of strong intermolecular spin orbital perturbations which may take place. These factors could substantially increase the quan- tum yield of phosphorescence. The energy available in the lowest excited level corresponding to the wavelength of the phosphorescent emission is about 33 kcal. per mole. The energy available in the lowest singlet state is approxi- mately 43 kcal. per mole. Although the energy available in the singlet state is higher, a consideration of the amount of absorbed energy reemitted from the singlet state deemphasizes this factor. Approximately 90% of the absorbed energy is unaccounted for in terms of reemission; 10% of the absorbed energy reappears as fluores- ELECTRONIC SPECTRA OF CHLOROPHYLLS 15 cence (7). Tlie now confirmed presence of phospliorescence accounts, in part, for the remaining energy and it may, in fact, account for all of it in vivo. Certainly the advantageous properties of the triplet state com- pared to the singlet state such as longer lifetime warrant serious con- sideration and further investigations of the photosynthetic process with the triplet state in mind. Moreover, the energy availability is con- current with the seeming requirements for subsequent chemical re- actions in photosynthesis as offered by recent authors. Practically all work in the field of photosynthesis and photochemistry has been concerned with the minor part of the energy and its possible mode of utilization. It seems more reasonable to be concerned with the fate of 90% rather than with 10% or less of the available energy. Discussion Linschitz : The phosphorescence yield in vitro must be of the order of 70% or so. Could I ask if you are going to account for the chemistry in vitro bj^ the triplet state? Regardless of what happens in vivof In I'ivo do you think that the measure- ment could possibly bring the jdeld up that high? Becker : I won't say yes and I won't say no. Rabinowitch : You mentioned that the photochemical energy yield in photo- synthesis is up to 70%. That does not mean that the j'ield of phosphorescence is 70% in the absence of photos3'nthesis, because the metastable state can be de- stroyed by internal conversion. Linschitz: That is true, but if you put in copper you can bring up the phos- phorescence to ver}- high values. Becker: The chances are the jdeld is not low. I am not saying there is a 100% quantum yield of phosphorescence. All I am saying is that the emission is 100% phosphorescence. In other words, we obtained no fluorescence. Duysens: Is it possible that the quantum yield of phosphorescence is ver}- small — say, 0.1 per cent? Becker: No, it wouldn't be that low. If I had to give an estimate, I would cer- tainly say not less than .50% yield, on the basis of slit widths, exposure times, and plate darkening. Duysens : There is no actual basis except it is a sort of culmination of times of exposure, etc., giving some indication of intensity of emission. Jacobs: I would like to put in a plug for the importance of the h3'dration of metal atoms for reactions in invo. A lot of work has been done on oxygen exchange with water in the in vivo reactions of electron transport systems. It is one of the first changes to disappear when the phosphorylation activity associated with electron transport disappears. I think j'ou underestimate it. This hydration of metal atoms may eventually (;au.se loss of a phenonienon which will turn out to be important in photosynthesis as well as other processes associated with electron transport. This is only a suggestion. 16 K. S. BECKER Weigl : I would like to ii.sk if anyone has tried measuring pliosphorescence in- tensity in vitro during a chlorophyll photosensitized reaction. It seems to me that it is very important to demonstrate that it goes down very appreciably. Becker: I haven't and I don't know of anyone who has. Weigl : It seems to me that ought to be done. Livingston: I would like to make a slightly different interpretation of Dr. Becker's own data. I think these data show that the yield of pliosphorescence, meaning not the transfer to the triplet state but the actual yield of energy as phosphorescence, must be less than 0.001. I am taking this in part from a statement of Kasha's, that the observed intensitj^ of the phosphorescence was less than the infrared contribution to the fluorescence. The infrared contribution of fluorescence is certainly less than a few tenths of a per cent: so if that statement is true, it would show that the phosphorescence has a yield of 0.001 or less. Also you stated that the fluorescence in this infrared region obscured the phospho- rescence; so that the fluorescence intensity must be much greater than the phosphorescence, which again proves that the observable phosphorescence, the omission of phosphorescent radiation, has a quantum yield less than 0.001; nothing like 0.1. References 1. Becker, R. S., and Kasha, M., "Luminescence spectroscopy of molecules and the ph«tosy nthetic sj'stem," pp. 25-45, in The Luminescence of Biological Systems, F. Johnson, ed., American Association for Advancement of Science, Washington, D. C, 1955. 2. Becker, R. S., and Kasha, M., "Luminescence spectroscopy of porphyrin-like molecules including the chlorophylls," /. A7n. Chem. Soc, 77, 3369 (1955). 3. Livingston, R., Watson, W., and McArdle, J., /. Arn. Chem. Soc, 71, 1542 (1949). 4. Calvin, M., and Dorough, G. D., "The possibility of a triplet state intermedi- ate in the photooxidation of a chlorin," J. Am. Chem. Soc, 70, 699 (1948). 5. Freed, S., and Sancier, K., "Solvates of chlorophylls and related substances and their equilibria," J. Am. Chem. Soc, 76, 198 (1954). 6. Evstigneev, V. B., Gavrilova, V., and Krasnovskii, A. A., Doklady Acad. NaukS.S.S.R. 70, 261 (1950). 7. Forster, L., and Livingston, R., "The absolute quantum yields of chlorophyll solutions," J. Chem. Phys., SO, 1315 (1952). Fluorescence Spectra of Protochlorophyll, Chlorophylls c and d, and Their Pheophytins C. S. FRENCH, JAMES H. C. SMITH, and HEMMING I. VIRGIN, Department of Plant Biology, Carnegie Institution of Washington, Stanford, California The fluorescence spectra of pheophytins a and b (1) as well as of pheophytin d in Fig. 1 are very similar to those of the chlorophylls from which they are derived except for a 2 to 13 m^ shift to longer wavelengths. ^ M CHlOROPHYUd In ether// V 650 70 7 50 m|t. WAVELENGTH Fig. 1 . The fluorescence spectra of chlorophjll d and pheophytin. Acid-treated chlorophyll c and protochlorophyll, however, show larger shifts of the main fluorescence band, and a greatly increased height and corresponding shift of the second fluorescence band, as shown in Figs. 2 and 3. The absorption spectra of the same prepara- tions (2) and some other chlorophjdl fluorescence spectra have been published (3) or appeared elsewhere (1). 17 18 C. S. FRENCH, J. H. C. SMITH, H. I. VIRGIN T h A -T-" ■ ■ r- 1 1 CHLOROPHYU-< / \ / ^ y PMtOPHYTIN- \» tthtr / \ ii tibir \ / / \ iM / \ / / \ <>> / v / \ z X \ tM X \ «-< I ^/ \ t/> / 1 \ ^"•^ \ IM / 1 \ \ ac / / \ \ O / / \ \ = / / \ ^"""^v \ Mrf 1 / \ ^v \ «^ ±. I y V L. 1 1 ^^ — 1_:^ 600 650 7 50 mp 700 WAVELENGTH Fig. 2. The fluorescence spectra of chlorophyll c and pheophytin c. 600 650 700 750 mp WAVELENGTH Fig. 3. The fluorescence spectra of protochlorophyll and protopheophytin. References 1. French, C. S., Smith, J. H. C, Virgin, H. I., and Airth, R., "Fluorescence spectrum curves of chlorophylls, pheophytins, phycocyanins, and h3^peri- cin," Plant Physiol, SI, 369-374 (1956). 2. Smith, J. H. C, and Benitez, A., "Chlorophylls: Analysis in plant materials," in Modern Methods of Plant Analysis, K. Paech and M. V. Tracey, eds., Vol. 4, pp. 142-196, Springer, Berlin, 1955. 3. French, C. S., "Fluorescence spectrophotometry of photosynthetic pigments," in Luminescence of Biological Systems, F. H. Johnson, ed., pp. 51-74, Ameri- can Association for the Advancement of Science, Washington, D.C., 1955. General Remarks on Chlorophyll- Sensitized Photochemical Reactions in Vitro JAMES FRAXCK, University of Chicago, Chicago, Illinois, as pre- sented by R. Livingston /rom an abstract The discussion of photochemical reactions sensitized by chlorophyll in vitro shows that the only reactions which proceed with high quantum yields are exothermic or slightly endothermic. More endothermic re- actions have very small yields because these recjuire that the bulk of the excitation energy of the chlorophyll be stored as chemical energy. In the process of sensitization, the excitation energy must be greater than the sum of the heat of activation and the chemically stored energy. Keeping this in mind, one finds it astonishing that the latter type of reaction takes place. However, the difficulty vanishes if one assumes that, for the first photochemical step, the energy of two quanta rather than of one quantum is utilized. The transfer of a hydrogen atom from the chlorophyll to a weak oxidant is an example. We suppose that the hydrogen to be transferred is the one bound to Cio in ring V of chlorophyll. The excitation energy of 41 kcal. (first excited singlet state) permits the transfer only to a strong (not to a weak) oxidant. When chlorophyll is in the lowest triplet state (corre- sponding to approximately 30 kcal.), even the transfer to oxidants as strong as molecular oxygen or quinone in one act becomes energetically impossible. (If one considers an electron transfer instead of an H-atom transfer, the energy situation becomes still more unfavorable.) There is a possibility that one hydrogen atom may be transferred in two chemical steps. However, the assumption which best fits the kinetic data for reactions in vivo is that two quanta are utihzed to excite one chlorophyll molecule into an excited triplet state, whereby between 60 and 70 kcal. become available for one photochemical act — in our case the transfer of one hydrogen atom. That can be achieved by the following steps: (1) Excitation of a chlorophyll molecule by light absorption into the first excited singlet state. (2) Internal transition into the lowest metastable, long-lived triplet state. (3) Excitation of the chlorophyll in its metastable triplet 19 20 J. FKANCK state to the next higher triplet state by sensitized fluorescence. The energy needed for this process comes from a neighboring chlorophyll molecule in the first ex(;ited singlet state. (4) Utilization of the energy difference between the excited triplet and the singlet ground state for photochemistry. I am not competent to discuss the theory of what Franck calls sensitized fluorescence, but let me at least outline the experimental evidence for its existence and the usual criteria for a probable transi- tion of this sort. When chlorophyll absorbs red light it is raised to its first excited (or fluorescent) state, usually with some extra vibrational energy. This excess vibrational energy is quickly lost, bringing the electronically excited molecule into thermal equilibrium with its surroundings. There is a large probability (between V4 and 1 ) that the excited mole- cule will not emit a photon of fluorescence light but will be transferred by a process of internal conversion into its lowest triplet level. In this state, the life of the molecule will be long compared to its life in the fluorescent state. When two similar molecules are close together but not in actual contact, there is a certain probability that, if one is elec- tronically excited, it can lose that energy and the energy of excitation will appear in the second molecule. This can also occur when the molecules are unlike, if certain requirements are satisfied. An empiri- cal requirement is that there should be a strong overlap between the emission band of the first (primarily excited) molecule and the absorp- tion band of the second molecule. The probability of such a transi- tion caused by sensitized fluorescence is great when there is a large overlap between the emission band and the absorption band. For this reason, one might expect a greater probability of such radiationless transfer between unlike molecules than between like molecules. We find that there is a higher probability for energy transfer from chloro- phyll & to o than for a transfer from a to a or h to h because the over- lapping of the fluorescence band of h and the absorption band of a is great. Because the visible absorption spectra of the triplet state and of the Molisch phase test intermediate of chlorophyll are similar, Franck predicted that the triplet state, like the phase-test intermediate, would have an appreciable absorption in the near infrared close to the red maximum of the singlet absorption. This prediction was verified by Linschitz and more recently by Fujimori. There is, there- fore, a strong overlap between the fluorescence of singlet-state chloro- CHLOROPHYLL-SENSITIZED REACTIONS in VltrO 21 phyll and the absorption of its lowest triplet state. This condition strongly favors a radiationless transition of the excitation energy from the first-excited singlet to the lowest triplet level of chlorophyll. If I understand Franck's suggestion, it is that in grana, where the molecules are close together and where they are probably ordered, there is a very high probability of transfer of the energy of excitation from one molecule to another, that eventually this energj^ will end up in a sink and that sink can be a chloroph}^! molecule already in its triplet state. This doubly excited molecule should have a short life, comparable with that of the fluorescent state. It is, therefore, most improbable that the molecule in its excited triplet state could take part in an efficient chemical reaction which is controlled by diffusion. In other words, if this short-li\'ed molecule is to use its energy efficiently in a chemical reaction, it must have come in contact with its reaction partner before it has received its second quantum of energy. Even at the low dye concentrations usually chosen in experiments on photochemical processes in vitro, the probability of exciting the higher triplet state is not too small to allow the low rates of endo- thermal photochemical reactions. If this hypothesis is correct, the fjuantum efficiency of such processes should become much higher whenever chlorophyll is adsorbed on suitable surfaces in such a way that the average distance between the chlorophyll molecules is very small. This should raise the probability of Step 3 without interference from quenching impacts between the chlorophyll molecules. Discussion Weigl : There is a rather stringent restriction on the lifetime of this doubly ex- cited state, unless it is a ver}^ unusual species (for instance, one stabilized by some prior chemical attachment). Internal conversion to the lowest member of any series of states of a given multiplicity is usually very fast — it proceeds in about 10""' to 10"'^ second. Therefore, a doubly excited triplet state would have to engage in chemical reaction within roughly lO'"'^ second, before the molecule drops to its lowest triplet state and the energy of the second quantum can be lost to molecular vibration and heat. Rabinowitch (remarks in proof) : If this were so, no molecules could have a fluorescence yield 0.01%. Brugger (remarks in proof): Prof. Franck believes that the transition tr* —*■ tr, first excited triplet to ground triplet state, resembles s* ^^ s, first excited singlet to ground singlet state. In each case, the transition is to the ground state of a given multiplicity series. One expects the lifetimes of tr* and s*, as well as the fluorescence yields for tr* -^ tr and s* — »- s, to be similar. In addition, tr* can pass non-radiatively to s* just as s* passes to tr. Limiry: In other words: the conditions should be as stringent for the triplet- triplet transition as they are for a singlet-singlet transition. Brugger : Figure 1 may aid in clarifjang some points. 22 J. FRANCK 2) H- tr -OH -^^^ H- R OH -OH -I- H- R OH -OH R OH •OH R OH tr tr ■OH R OH -OH -t-H- R OH •OH R OH 3A) Ox-HH- 3B) Ox-H- tr -OH Ox-'-H- R OH tr ■OH + H- R OH tr -OH R OH -OH — *■ OX---H- R OH tr •OH -I- H- ■OH R OH R OH 4) Ox+H- tr (OxH)- •OH R OH •OH — »► (OxH)- -I- R OH Ox= oxidant -OH R OH Fig. 1 The grouping H — Cio — Cg — OH I I R OH represents a hydrated form of chlorophyll and specifically shows hydration of the carbonyl group in ring V. In reaction sequence ( 1 ) of Fig. 1 , chlorophyll is excited to the first excited singlet state s*. Then, by a nonradiative transition, the singlet state crosses over into the ground triplet state tr. By a process of sensitized fluorescence, as shown in reaction sequence (2), this same molecule in its triplet state tr absorbs a quantum of energy liberated when some nearby chlorophyll molecule in the first excited singlet state s* reverts to its ground singlet state s. This method of raising chlorophyll molecules to the excited triplet state tr* is a unique feature of Dr. Franck's theory. The triplet state has an absorption far enough in the red to make this process feasible. Dr. Franck believes that ad- sorption (reaction (3A)) of some molecule, oxidant and/or enzyme, on the chloro- phyll in the triplet state tr would make it an even better sink for light energy, viz., increase the overlap between singlet fluorescence emission and ground triplet ab- sorption. As a second consequence of such adsorption, one could have the mole- CHLOROPHYLL-SENSITIZED REACTIONS in VltrO 23 rule to be reduced already present. Thus, for photoohemical reactions of chloro- phyll in vitro, Dr. Franck postulates adsorption of the oxidant by chlorophyll in the metastable triplet state, as shown in equation (3A), followed by excitation of the chlorophyll complex to the excited triplet state tr* by sensitized fluorescence and break up of the complex to form a reduced oxidant radical and a chlorophyll radical, as shown in reaction sequence (3B). The chlorophyll radical is presumed to lose an OH and rearrange to form a chlorophyll molecule. In photosynthesis in vivo, Dr. Franck expects that the relatively long-lived triplet state chlorophyll would adsorb (or become associated with) an oxidant, which would eventually be reduced by the H on do, and with an enzyme, possibly a cytochrome, which would carry off one of the OH's on Cg. This process would produce H-Ox- and HO- Enz- radicals and leave the chlorophyll molecule in the ground state. One might expect that a chlorophyll molecule in the excited triplet state tr* would transfer the H-atom directly to an oxidant during a collision, as shown in reaction (4). The state tr* has sufficient energy to reduce quinone, ferricyanide, oxalate, PGA, DPN, etc., by photochemical H-atom transfer. For the usual con- centrations of the molecules, however, the yield would be very low. One has the same problem with the excited triplet state tr* as he has with any higher excited state, namely, if the excited molecule does not collide and react with another molecule within ca. lO^^^ sec, the excitation energy will be dissipated as heat or light and the yield of the photochemical reaction will become small. Dr. Franck does not intend to endow the excited triplet state of chlorophyll tr* with an es- pecially long Ufetime, which is what a good yield by reaction (4) would require. He postulates the reaction sequence (3A) and (3B) in which a reaction complex is formed by the long-lived metastable triplet state tr, thus not only obviating the need for a collision with the short-Uving excited molecule tr* but even increasing the probability of exciting the chlorophyll to the tr* state. Duysens : It seems to me that if you illuminate with weak light, you have a low concentration of triplet; at higher light intensity you would have more triplet. So you would think, if the transfer takes place through induced resonance, that the efficiency of transfer from the excited singlet would be greater at higher light intensity than at a lower light intensity and that would mean the fluo- rescence yield of the chlorophyll would decrease at the high light intensity. Ex- periments indicate, however, that the fluorescence yield is essentially constant. I wonder how this difficulty can be resolved? Brugger : At each light intensity, one should come to a steady state in which the ratio of molecules in the first excited state to those in the ground metastable state is constant. The fluorescence yield is also roughly constant. I expect that the rate of populating the excited metastable state should be proportional to the ir- radiation intensity. It would seem that the percentages of the molecules in the first excited singlet state which (1) fluoresce, (^) populate the ground metastable state, or (3) excite the metastable state by sensitized fluorescence should be roughly the same at all except the lowest light intensities. I do not feel that higher intensities would populate the metastable state disproportionately heavily nor that the efficiency of sensitized fluorescence would be much greater nor that fluo- rescence would be markedly quenched. Rosenberg : May I rephrase Dr. Duysens' question to see whether I understand 24 J- FRANCK what he is asking? I think he is saying that, if you pnpulatf^ (lie ground triplet state to a greater extent, which you presumably do by stronger irradiation, do you increase the drainage from the excited singlet. Duysens: In which case you would lower the fluorescence yield. Rosenberg: But I don't think that that happens to any ajjpreciable extent. The steady-state population of the first fiuores(!eiit state and of the lowest triplet state would both be proportional to the light intensity in close approximation. Duysens : I do not think that Dr. Rosenberg's remark is correct, because the drain is proportional not only to the number of excited singlets but al.so to that of non-excited triplets. Rosenberg : But also, at the same time, you are increasing the total number of quanta absorbed per second so that the fractional yield of the molecules which are fluorescing may remain the same. It is the fractional yield of fluorescence that is constant experimentally. Lumry: If all the processes out of the first excited state are of the first order in the concentration of the molecule in the excited state, then the only way you can change the amount of fluorescence is by changing the crossing point. Unless you are postulating a back reaction from the triplet into the singlet, the fraction of the fluorescence yield isn't going to change, no matter what you do to how many triplet molecules. Duysens : If you have no triplet, there is no loss of fluorescence. Assume the fluorescent yield is, say, 50%. If you have a very efficient transfer, then the fluo- rescent yield is zero. Lumry: This just does not have anything to say about the lifetime of the trip- let or the population. It is a question of whether it can get out of the singlet into the triplet. Duysens : But my contention is that the efficiency of transfer to the triplet state is proportional to its concentration. If it is two million, then you have twice as great a transfer as if you had one million. Then the fluorescence yield of chloro- phyll changes. If the process of transfer is fairly efficient, it may change by, say, 50%. Rosenberg: I think there is a self-balancing mechanism in Franck's picture which answers Duysen's question. In the Franck picture, the metastable state would be formed at a rate proportional to the concentration of the fluorescent state material. It would also di-sappear at a rate proportional to the concentra- tion of the fluorescent state material. Therefore, the steadj'-state concentration should be independent of the light intensity and it should be simply a ratio of rate constants, i.e., the rate constant of formation divided by the rate constant of loss. So we don't have to worry about the variation of jjopulation of the lowest triplet state with the light intensity. Arnold : This is true until the rate is so slow you have to worry about the life- time of this triplet. Rosenberg : Yes, I am forgetting the natural decay of the triplet. Duysens : But this triplet state is supposed to react chemically. Rosenberg : No, only in its excited state. Duysens : So the concentration remains constant? Rosenberg: Yes, as long as you don't worry about the natural decay of the lowest triplet state. CHLOROPHYLL-SENSITIZED REACTIONS m vitrO 25 Rabinowitch {expanded in "proof) : As long as each //• has the chance of gcitting a second quantum by sensitized fluorescence (resonance transfer) during its hfe- time, i.e., as long as every triplet excitation becomes a double excitation, the pro- portion of singlets which will succeed in emitting fluorescence will be independent of light intensit}-. The fluorescence yield will, however, begin to increase when the light intensity becomes so low that, during the lifetime of tr, no quantum will be absorbed by a singlet molecule near enough for the excitation energy to be snatched away by the tr. This intensity must depend on two things — the lifetime of tr, and the range over which ir can grab quanta b\' resonance transfer. If //• lives 10~^ sec. and can grab a second quantum from lU'' singlet neighbors, the critical light in- tensity will be that at which each chlorophyll molecule absorbs a quantum once a second — which corresponds to a quantum flux of the order of 3 X lO^^ photons near the peak of chlorophyll absorption, or to several thousand lux of white light. Experiments indicate no decrease of fluorescence yield with increasing in- tensity in this region. If the phenomenon exists, it must be over-compensated by an increase in yield due to other causes. Weigl : The theor,\' rec(uires that at the very lowest light intensity you would get a ver>' sharp dropping of the quantum yield of whatever reaction evolves. Gaffron: One should perhaps point out that Franck's theory recreates, in a sense, the concept of photosynthetic units. Each time this triplet state appears it becomes momentarily a sink and the nearby excited molecules can deliver their first singlet-state energy into it. Strehler : Does anyone have evidence for the appearance of a band in illumi- nated chlorophyll solutions in the red region of sufficient intensity to absorb fluorescent light from the normal chlorophyll singlet-singlet transition? Rabinowitch : Wasn't it reported todaj^ that the metastable state has an in- creased absorption in the near infrared? Strehler : It was the solvation band that was out further. Linschitz: The absorption spectrum of the metastable state of chlorophyll, as measured in our flash experiments, shows a band in the far red, just beyond the main red band of chlorophyll itself. This new absorption band corresponds to a transition from the ground level of the metastable molecule to what is probably the first excited level, as follows (transition 2). '^ ^ -Ground level of metastable state of chlorophyll -Ground level of chlorophyll The evidence indicating the existence of this band is completely unambiguous. (See Fig. 25 in our paper, for example. ) Weigl : It is in the position required by Franck's theory. Strehler : What is the cross section? Will it be an efficient trap for the energy? It is the total optical cross-sectional overlap, as well as whether it is in the right position, that is important. Can you estimate this? Linschitz : The quantitative measurements in the far red that we have to date 26 J. FRANCK are very crude, but I'd estimate that the total absorption in the new band iH about a fourth of the red chlorophyll band. Strehler : One quarter. Linschitz : Maybe one fourth. You could certainly see it. Bassham : One thing this whole picture is based on is the assumption that one (quantum transfer implies one electron transfer. Many people do not agree with that assumption. I would be interested in some comments in favor of the argument that one cannot transfer one electron by one light quantum. I person- ally believe that you probably do transfer one electron per quantum. Franck's theory would require, among other things, a minimum quantum requirement of 8, if you have to have two light quanta for each electron transfer. Rabinowitch: This obviously is not explained in this abstract because it im- plies, as far as I understand, that you would need to know the energy of the free radical. The transfer of one hydrogen atom creates a free radical. We usually don't know what the energy of this free radical is. Franck says, "The excitation energy of 41 kcal. (first excited singlet state) permits only the transfer to a strong (not a weak) oxidant." That must mean the energy we would assume is needed to trans- fer one hydrogen or one electron to PGA or to TPN. Bassham: For example, the energy of 40 kcal. is equivalent to about 1.7 elec- tron volts per electron. This is sufficient to make hydrogen peroxide from water at a concentration of 10~* molar and at the same time to reduce TPN to reduced TPN. Rabinowitch : You are using the average for the two hydrogens. Bassham : Well, do we know enough about the actual structure of grana to say that it might not have some type of arrangement of semiconductors which makes possible the separation of these charges, providing the electron is of the right potential at the right place? I think this is certainly within the realm of possi- bility. Gaffron: The gist of the entire discussion this morning indicates the singlet state is not directly responsible for the chemical reaction. Thus you cannot work with 41 kcal., you have to lower the available energy to about 30. Bassham : Even with 33 we can just barely make it. Rabinowitch: Well, that ignores the need of going over the radical state. Bassham: Presumably there is some system for getting around the radical state to accomplish this. Reactions that in solution chemistry would require radi- cals are often accomplished in vivo with much lower energy. Brugger : I should mention that Dr. Franck was certain that there was going to be a discussion of electron transfer here. He definitely advocated hydrogen atom transfer. He simply could not see his way clear to agree to electron transfer in this case because, as has been mentioned, you don't have much time. If one transfers electrons and makes hydrogen ions, he has to hydrate the hydrogen ions, or otherwise there isn't enough energy. He felt that one hasn't enough time to hydrate hydrogen ions and not enough light energy to transfer electrons with- out hydrating the hydrogen ion. Bassham: I don't think it takes very long to transfer an electron. Brugger: You have to pay a penalty. The ionization potential for a hydrogen atom is 13 volts. That is, 300 kcal. are required to make a hydrogen ion and an CHLOROPHYLL-SENSITIZED REACTIONS ITl VltrO 27 electron. The hydrogen ion, i.e., a bare proton, can, however, hydrate to forma hydronium ion. This hydration liberates about 270 kcal. Roughly, one can re- move an electron from a hydrogen atom and make a (hydrated) hydrogen ion with 30 to 40 kcal. only if one hydrates the proton. Similar considerations ex- plain why the heat of ionization of water is about 13 kcal.: both the H+ and the 0H~ are hydrated — an exothermic process. Dr. Franck intended to point out that there are no reactions known of the type: (1) dye + h^ > dye* (2) dye* + H^O > dye + H+ + OH- where hv is 30 to 60 kcal. He also wished to mention the reactions: (3) (Fe ++)a<, + h. >(Fe + + +OH-)a<, + H (4) (Fe + + +)aq + h. >(Fe ++)a<, + H+ +0H The second of these does not proceed using visible radiation because there is not time to hydrate the H-*- during the photochemical act and thus to gain the energy of hydration. A quantum of visible light does not have sufficient energy to liber- ate the bare, unhydrated proton. The Franck-Condon principle (conservation of momentum) restricts movements of atomic nuclei during photochemical acts. The energy requirement for the reaction liberating a free proton (H"^) is thus very great. Bassham: The electron is not necessarily taken from a hydrogen atom. It may be taken from water. Brugger : What difference does it make? You still have a bare proton sitting around which you have to hydrate. Bassham : Yes, but we can take several electrons at the same time from water, or we can get away from the radical in between. And, since we have an aggre- gated chlorophyll system where we can have a high population of exited chloro- phylls at the same time, it is conceivable that we don't have to form any radicals in between. Brugger: I don't mean to dodge the issue but I think it would take too long to go into it. Dr. Franck's theory was specifically designed to provide a mechanism whereby hydrogen atoms could be transferred. Rabinowitch: Whenever you transfer electrons, you immediately create an uncomfortable position which requires a subsequent adjustment of the nuclei to fit the new distribution of charges and you lose some of the energy. In other words, electron transfer requires a high activation energy. It is better to trans- fer hydrogen atoms with only the energy needed for the transfer and without extra energy of activation, whther it is one or two atoms at a time. Brugger: I can make one more point w^hile I think of it. Often the electron transfer reactions involve two electrons at a time. The energy is calculated per electron by just dividing the energy by two. One has absolutely no guarantee that the energy required to transfer the first electron is half the sum required to transfer both. So if you can just squeeze by theoretically in transferring one electron at half the energy of the two-electron process, you have no guarantee that this actuallv works. 28 J- FRANCK Bassham: We are not working; with one atom or one molecule at a time. If we accumulate charge at a certain potential, which is necessary, say, for the DPN electrode, that is all we have to ac^complish. There can be a large population of electrons. I might point out that in something like a solar battery, for instance, you can remove an electron from one position. It will by conduction fall into a hole some- where. This creates a charge. The charge can be used for the chemical current or wherever you want to use it. It may be the same in photosynthesis. Strehler : There are some energetic considerations here that Dr. Bassham is not taking into consideration — at least not here. (1) Energy will be dissipated (and a fair amount of it) in the process of oxygen liberation (I assume we're speaking about green plants). (2) In all likelihood that energy has to be available in addition to the energy dissipated to form a triplet state or to store the usable energy in some stabilized pool or intermediate reductant. If this were not the case the re- ductant could react back at high rate. (3) In addition, not all of the energy in the one electron carrier or free radical may be available to do chemistry since radicals generally dismutate to nonradicals with a loss of energy — and DPNH is a radical. I don't believe it is possible at present to formulate binding arguments of a quantitative sort in support of either position. But the arguments raised on purely physical chemical bases, such as given by Franck in his review paper in the Archives of Biochemistry and Biophysics, or as presented in the Phosphorus Symposium, where I developed essentially similar conclusions on entirely inde- pendent basis, cannot simply be dismissed. We must be seriously concerned about the question: Can one quantum of 40 kcal. energy content, considering prob- able losses, transfer an electron from the potential of water to the potential of an acceptor which reacts reversibly with fixed carbon dioxide? Bassham : Take 41 kcal. and throw away 7 or 8 kcal. in the first photochemical step to make a triplet state. Then you still have enough left, as I said before, to create hydrogen peroxide and to reduce TPN+ to TPNH. Then the hydrogen peroxide proceeds spontaneously to water and oxygen. Strehler : How do you make the hydrogen peroxide? By the recombination of OH radicals? Bassham : We don't necessarily go through the OH radical. Perhaps we have a hydrated surface from which we have enough electrons all at one time to allow instantaneous formation of peroxide. I am simply saying that we don't know enough about it to rule out any possibility as long as there is enough energy to carry out the net reaction which is there. Rabinowitch: The free energy of photosynthesis is approximately 120 kcal., or about 30 kcal. per hydrogen atom transferred. If you want to transfer these four hydrogens by four protons from water to something of the same reduction potential as your ultimate acceptor, which is what you want to do in the case of TPN, you have practically nothing to spare. Bassham : I think you misunderstand me. I am not proposing that four quanta will accomplish photosynthesis. I say that we can transfer four electrons. Other electrons have to be used up to make high-energy phosphate to help us out later on. To do what you are suggesting would require a better reducing agent than CHLOROPHYLL-SENSITIZED REACTIONS in vilro 29 TPN. Actually, we don't get by with just transferring four electrons. We have to transfer six or seven electrons and need six or seven quanta. Rabinowitch : For that you have to have a process which would liberate energy after these transfers. Do you want to leave this just to the decomposition of hy- tlrogen peroxide? Bassham : That is s{)ontaneous. Rabinowitch : It evolves energy. Are you going to salvage this energy somehow? Bassham : No, we don't need it. Rabinowitch : For reduced TPX you really need some more energy to shoulder it up. Bassham : We do get that energy when we bvu-n part of the TPNH oxidatively. Rabinowitch : All right, so that would take care of itself, but the fundamental process, the transfer from the level of water to hydrogen, i.e., the level of TPiS'H, requires almost a full quantum itself. Duysens: In one absorbed quantum you have 40 kcal. Of course, a certain amount is lost to form stal)le compounds. Let's say 8 kcal. Then you have 32 kcal. left for reduction of DPX. Transfer of one hydrogen from water to DPN re- quires only 26 kcal. So I think it is possil)le. Rabinowitch : You will have just a very little left. It is just barely possible. Strehler: If you form peroxide as an intermediate you lose another 9 kcal. which are not available for the formation of a potent reductant. Kamen; This is anticipating the argument that I expect tomorrow. The picture that Dr. Bassham has been giving here is very similar to the one that we have been trying to work out in connection with the interaction between chlorophyll and cytochrome. But my feeling is that this discussion is fruitless and we will get nowhere, for the simple reason we are not dealing with pure chlorophyll in photo- synthesis. If you put this chlorophyll molecule on protein, I will venture to say that all these energy requirements will disappear or be changed so that you will have a different picture entirely. Until we find out something about native chloro- phyll and about the biological capacity of it, I think we should postpone all energy considerations entirely. What I am saying is that the cooperative pigment part of the assembly takes up energy which is distributed over a large molecule, the sink of which we don't know. The sink may not be the triplet state at all. It may be a double bond some- where else. It may very well be (I say this with great hesitation) an iron hematin. These are present in great amounts and can do everything that is needed in photo- synthesis, as you will see tomorrow. There are so many processes that are impossible from a strictly physicochemical standpoint but which occur all the time in biological systems, that the argument that something cannot happen because 7 or 8 kcal. are lost is just plain senseless. I think the best thing to do is to table the whole argument and not waste any more time. Brugger: Dr. Franck, I think, certainly never denied any electron transfer processes, but I think he felt if you used photochemistry to move electrons around you had to pay some price. Since you had to do it in a short time you had to use 30 J. FRANCK more energy. It certainly is not the most economiccal way to do the process. He did not deny electron transfer and he certainly believes in semiconductors. If you want to do this efficiently with chlorophyll molecules, as Dr. Ilabino- witch pointed out, you also have to do some moving around and you cannot do this in 10~i2 second. If you have lots of energy, if you use gamma rays, then you can most likely- do it. With 30 electron volts per ion pair, you can do it that fast. If you have 800 or 900 cal., you have plenty to work with. Rabinowitch : This ring V seems to be the key chemically, or maybe it is a sort of general self-hypnosis we are indulging in, feeling that there must be something in this ring which would explain the participation of chlorophyll in photosyn- thesis. Anyhow, it seems also to be significant for the infrared absorption spectrum of chlorophyll. All the changes and variations of chlorophyll which have been re- vealed b}' the visible absorption spectrum also must find their counterpart in the changes in the infrared spectrum. Dr. Holt is going to say something about the infrared spectrum as a means to understanding this structure of chlorophyll and particularly what happens in this one ring. Reversible Spectral Changes in Chlorophyll Solutions, Following Flash Illumination* HENRY LINSCHITZ and EDWIN W. ABRAHAMSON, The Departments of Chemistry, Syracuse University, Syracuse, New York, and Brookhaven National Laboratory, Upton, Long Island, New York Theories of the mechanism of dye-photosensitized reactions, in particular those catalyzed by chlorophyll, have long postulated the intermediacy of metastable excited states of the dye. This assumption was based originally on indirect evidence, such as the maintenance of high photochemical quantum yields even at low substrate concentra- tions, and the lack of fluorescence quenching by photochemically efficient substrates (1). More recently, direct verification of the exist- ence of such long-lived states has been provided by study of reversible spectral changes in chlorophyll solutions, either following intense flash irradiation (2-4) or under steady cross illumination in rigid sol- vents, under conditions precluding bimolecular solute reactions (5). In this report we present observations and remarks on flash-bleach- ing of chlorophyll in fluid solvents. The chief results to date are the verification of the general shape of the transient spectra, observation of a new far red absorption band in the "metastable state," and demonstration of kinetic effects of polar and nonpolar solvents. APPARATUS This is shown schematically in Fig. 1. A beam of light from source "S" was collimated by lens " L," passed through a shutter, the dye solution, thence through a monochromator and onto a photo- multiplier tube. Light transmissions at selected wavelengths were recorded on the oscillograph, immediately preceding and following flash excitation. A major problem in using this technique is to prevent swamping of the measuring light by scattered light from the flash. Sample fluores- cence may also interfere. Toward this end, the monochromator was * Research performed under the auspices of the U. S. Atomic Energy Com- mission. 31 32 H. LINSCHITZ AND E. W. AimAIIAMSON placed in the optical train behind the fla.sh tube-cell assembly, and the distance between the sample cell and entrance slit of the monochroma- tor was made large (120 cm.) to favor the colli mated beam. Scattered light was further decreased by passing the measuring beam through accurately aligned apertures in a series of baffles placed in a light-tight box, completely enclosing the optical path. The whole assembly was mounted on an optical bench. Steady-state bleaching due to the undispersed measuring beam was negligible, compared to that caused by the flash. Provision was made for inserting filters in the beam to SAMPLE BAFFLES FILTER SHUTTER |o-: POWER SUPPLY k FLASH LAMP PULSER SWEEP / TRIGGER *> FLASH TRIGGEF ? X MONOCHROMATOR |i,PHOTOMULTIPLIER Fig. 1. Apparatus for flash-illumination studies. improve the purity of the spectrum provided by the monochromator (a constant-deviation Gaertner instrument). The source ">S" was a 500- watt concentrated filament projection lamp, operated from a stabilized d.-c. supply. The sample cell, 50 mm. long and 13 mm. I.D., carried a side ampoule permitting solutions to be prepared and degassed on the vacuum line, and sealed off. The xenon-filled flash tube was made of 17-mm. Pyrex tubing, wound in a two-turn helix, within which the sample cell was mounted, the whole being surrounded by a cylindrical magnesia-coated reflector. The tube was operated at 4000 volts and 24 mfd., and provided a flash of about 100 microseconds duration. A pulse generator and associated circuitry was used to trigger the sweep of the oscillograph, and after an adjustable delay, to fire the flash tube. PROCEDURE In order to correct for scattered or fluorescent light the following measuring procedure was used. With the shutter open, and sample in SPECTRAL CHANGES IN CHLOROPHYLL SOLUTIONS 83 position, the monochromator slits and amplifier gain were adjusted to give a suitable deflection on the screen at the desired wavelength. The tube was then flashed and a sweep record obtained in which the vertical deflection represents the total of transmitted and scattered light reaching the photocell. The shutter was then closed, and a second flash and sweep recorded, giving just the scattered light. The vertical distance between the two traces then measures the net transmitted light at the selected wavelength and time. A timing sweep, blanked at 10,000 cycles by a calibrated audio oscillator, was also taken for each measurement. Oscillograph deflections were converted to optical densities by calibration records, using either pure solvent or the measured absorption spectra of the test solutions. Linearity of the photometric system was checked occasionally with a calibrated screen. RESULTS Typical results, showing bleaching at the red peak and the appear- ance of new absorption bands in the far red and green, are presented in Fig. 2. It is obvious that flash-illumination leads to drastic changes in absorption spectrum, confirming at least qualitatively the spectro- graphic observations of Livingston d al. Systematic measurements at various monochromator settings enable one to obtain the spectrum of the metastable species. Our data are in general agreement with those of Livingston and co-workers in showing a decreased absorption at the red and blue peaks of chlorophyll and enhanced absorption in the green (around 525 ni/x) and violet. Of particular interest, however, is the new band in the far red (^700 mix), demonstrated in Fig. 2B. It is probable that this is the same band first seen in the low-temperature steady-state bleaching experiments, in rigid solvents (5). However, the lifetime measurement afforded by the flash technique permits us now to assign the far red absorption unequivocally to a metastable species. The e.xistence of a reasonably intense far red band in the metastable state has been predicted by Franck on the basis of a postulated struc- tural analogy between the enolate ion and triplet state of chlorophyll (6,7). In his most recent picture of the mechanism of splitting of water bj^ excited chlorophyll, discussed elsewhere in this volume. Professor Franck has also suggested that the function of this band is to enable chlorophyll, in its metastable form, to become still further 34 H. LINHCHITZ AND K. \\ . ABRAHAMSON excited by resonative transfer of energy from the rest of the chloro- phyll aggregate (()). The flash data show that, in agreement with Franck's postulate, the metastable form of chlorophyll does indeed have an absorption band strategically placed to permit trapping of C Fig. 2. Changes in light transmission through chlorophvll-6 sohitions, following flash illumination. Solvent, deoxygenated 1)5 '/ti methanol; concentration, 1.2 X 10 ~* M; upper sweep, flash plus measuring light; middle sweep, flash alone (base line); lower sweep, 10,000 cycle time marker. A, 655 niM (red peak); B. 700 m/x; C, 525 niM. energy from the first excited singlet levels of adjacent molecules. It is of interest to point out that other " reversibly bleached" chlorophylls, such as the phenylhydrazine reduction product, do not show any appreciable absorption beyond the original red peaks. The curves given here are taken with considerably better signal-to- SPECTRAL CHANGES IN CHLOROPHYLL SOLUTIONS 35 noise ratio and more nearly monochromatic scanning light than was used by Livingston and Ryan in their original flash experiments (2). In addition, the present technique permits determination of the zero- time of the flash and thereby observation of the development as well as decay of the spectral changes. Despite these improvements, defini- tive kinetics are still difficult to obtain, essentially because the flash duration is comparable to the lifetime of the excited species. This limits the experimental accuracy that can be achieved (transmission changes are obtained by diff"erence) and greatly complicates the theoretical treatment. Our apparatus is now being modified to provide much shorter flashes. It has been suggested that a dye-solvent redox reaction might be the mechanism of formation of a metastable species (1,5), and the strong solvation reactions of chlorophyll (8-10) are consistent with this view. To clarify the question, flash experiments were carried out in purified methylcyclohexane, exhaustively dried by distilling over calcium hydride, and then pumping away a large fraction of the sol- vent at low pressure. The extent of dehydration was indicated by the drop in fluorescence of the resulting chlorophyll b solution to 8% of its original A'alue in the undried soh'ent (8) and the appearance in the dry solution of a new band at 665 m^, about eciual in intensity to the usual "wet" band at 655 m^ (9). On flashing, the "dry" band at 665 m^ was found to bleach very markedly. The strong bleaching in com- pletel}' nonpolar solvent, as well as the appearance of a transient spectrum rather similar to that observed in methanol or pyridine, implies that at least one of the metastable states is reached from the excited singlet by an intramolecular process alone. Pheophytin and zinc tetraphenyl porphine also show marked spectral changes after flash illumination. In pure hydrocarbon solvent, the appearance and decay of ab- sorption at 525 mn occurs at the same rate as the bleaching out and recovery of the 665-mM band, the situation apparently being simpler than in polar solvents (4). The reactions in the polar case thus may involve desolvations or keto-enol shifts in the excited molecule, or excited dye-solvent reactions following formation of the first meta- stable product. In this connection, it is of interest that in pyridine, the half-life of the 525 absorption is considerably increased, and the re- versibility of the reaction is much better than in carefully purified 95% methanol. 36 H. LINSCHITZ AND E. W. ABRAHAMSON Further study of the kinetics, using flashes of much shorter dura- tion, is clearly required. References 1. Rabinowitch, E. R., Photosynthesis, Vol. 1, Chap. 18, Interscience, New York, 1945. 2. Livingston, R., and Ryan, V., J. Am. Cheni. Soc, 75, 217G (1953). 3. Livingston, R., Porter, G., and Windsor, M., Nature, 173, 485 (1954). 4. Abrahamson, E. W., and Linschitz, H., ./. Chem. Phys., 23, 2198 (1955). 5. Linschitz, H., and Rennert, J., Nature, 169, 193 (1952). 6. Allen, F., and Franck, J., Arch. Biochem. and Biophys., 58, 124 (1955). 7. Weller, A., /. Am. Chem. Soc, 76, 5819 (1954). 8. Livingston, R., Watson, W., and McArdle, J., J. Am. Chem. Soc, 71, 1542 (1949). 9. Freed, S., and Sancier, K., J. Am.. Chem. Soc, 76, 198 (1954). 10. Livingston, R., and W^eil, S., Nature, 170, 750 (1952). Infrared Spectroscopy of Chlorophyll and Derivatives A. STANLEY HOLT,* Photosynthesis Laboratory, Department oj Botany, University of Illinois, Urbana, Illinois Several reversible chemical and photochemical reactions of chloro- phyll in vitro are known (1-3), but little is known about the groups in the molecule which are involved in these reactions. The study sum- marized below, and presented in detail elsewhere (4), was undertaken in the hope that infrared absorption data may give specific answers. Various derivatives of chlorophyll and ethyl chlorophyllide have been prepared to permit vnieciuivocal assignment of the absorption bands. Of the atomic groups present in the chlorophyll molecule, only the C=0, C=C, — H, N — H, and C — H groups gave infrared absorption bands which could be assigned with reasonable certainty. These bands lie between 3600 and 1600 cm.~'; the spectrum below 1600 cm.~^ proved too complex for definite interpretation. Table I and Fig. 1 summarize the data. Of the three C=0 groups present in pheophytin a and ethyl pheophorbide a, the ester groups at Ct and Cio absorb near 1740 cm.-^, which is the usual frequency found for the C=0 group in an ester (5). The keto group of the cyclo- pentanone ring absorbs at 1697 cm.~^ (in CHCls-solution) or 1706 cm.~i (in CCU-solution). This agrees with the assignment given to this band in the earlier work of Weigl and Livingston (6). The decline of the frequency from 1740 cm.~\ the absorption frequency of the keto group of cyclopentanone itself, to about 1700 cm.~^ must be due to the attachment of the isocyclic ring to the aromatic nucleus. That the 1697 cm.~i band is in fact due to the keto group of the cyclopentanone ring is indicated by its absence from the spectrum of the oxime of ethyl pheophorbide a, and from those of the 2,4-dinitrophenylhydra- zine derivatives of pheophorbide a and pyropheophorbide a. The C=0 in the carboxyl group of the propionic side chain in position 7, present in the two latter compounds, absorbs at 1705 to 1706 cm."~^, and thus * Present address: Division of Applied Biology, National Research Labora- tories, Ottawa, Canada, 37 38 A. S. HOLT c -a «j +j ;-. O O '« c a" _> "-I3 Q C o O 03 -o C pq c o DQ 0) ti c3 o a II o 5-2 rr: C: 00 00 00 sj CO CO CO CO CO CO + CO t^ O "O CO (M "* CO t^ t^ t^ t^ 00 c O 00 00 CO (N CO 1^ I- t^ o CO (M o o o (N iM r-' t^ t^ t^ (N (N iM + 2 O (N O sj 'ti lO iC OJ -f (N CO fe CO CO CO O- o CO CO is CO ^ o a B o O K 0- C5 00 t^ O -H — I CO CO CO CO CO — H >0 O o o o t^ t^ t^ 00 O CO CO -r CO r^ t^ 1^ ^ o o IM CO IM ■3i C5 C-. (N (N C^ o Q, 1 ^ o CO lO 1 Q o CO CO ^H 1^ CO CO Oi CO CO CO IN 00 CO CO CO a ..— « C^ _ 02 o r- c o — " o t3 X K W ,^' ^ U O ^, O o ^ O O Oh -a HO U e c JS -G a a ? ? o 5 o o a -a -^ o — — -^ -c jr — ? P >. _o _3 ti 2 2 — O w -c c c >. P ^ o. o ©■ c" -c ^ J= O ^ :^ INFRARED SPECTROSCOPY OF CHLOROPHYLL 39 X O (N CO O O — C>) (N -H C^ — iC '■£ '-D CO O ^ 00 O 00 — (N — O O Ml I CO CO CO CO (N O CO t^ 23 O 05 o t^ CO ^- ^ CO o o cs 00 00 CR -* O o o -- r^ t^ r^ O t^ o 'I' ro -* i^ t^ t^ O O o O CM (N 1^ t- (N (N O C<3 Oi 1 (N 1 1 i o o 00 CO CO CO CO 1 O O o o .a ^K K K * ^ O O o o o 2; t 1 1 g e i 75. 73. S3 Qj a; *^ '^ oi ^ -O TJ 1 o1 t ^ JZ ^ c >. S S 0^ r- i. — 1 o o s. a:H r^ ■^ -^ -^ O O _2 ^ "^ a a C; Oj t. ^ r o o -C -C o o 0; S ^ a a-c jn -g -C ^ '-^ a D, '^ ("i >. >. o o Poo ^ -C a; a; -S - ^ +j +^ ^ -C *^ >-, >i W W CLi Oi Ch X •C' "^ CM ■< "^ "^ c o CO c^ 00 CO CO CO ^ CO CM H ^ -H CO 03 CM O CO CO -3 .« .i: _!. c a o o o -H 00 •^ CO CO CO _r O U o ffi a o o u CM :2 -o 4= IS J. o -^ § o a o j= _g 3 :g -o ^ 0- W W o 2 -g w o __ r2 --S m 40 A. s. HOLT adds some confusion to the interpretation of the spectra, since it is superimposed on the absorption band of the keto group at C9. We assign the band at 1615 cm. ~^ to the C=C bonds in the aromatic nucleus; this assignment is supported by the strong absorption shown in the same region by the phenylhydrazine derivatives and by the presence of a similar band in the spectrum of bacteriochlorophyll, which contains no vinyl C^C group. The spectra of chlorophyll a and ethjd chlorophyllide a show marked differences, in the region of the C^O bands, from those of the magnesium-free derivatives. These depend on the choice of the solvent. Spectra taken in nonpolar solvents (GCI4 or CS2) or with crystals of chlorophyll a dispersed in Xujol (mineral oil), show an extra band at 1640 to 1650 cm.~i. This band is interpreted by us as indicating a strongly hydrogen-bonded (internally chelated) form of the /3-keto ester in ring V. In polar solvents, e.g., diethyl ether or chloroform, chlorophyll a does not show any such band. Spectra of pheophytin a in pyridine show a band characteristic of a hydrogen-bonded hydroxyl group, between 3100 and 3700 cm.~^ We believe this could be due to enolization of the /3-keto ester, in position 9, and hydrogen bonding between the pyridine nitrogen atom and the enolic O — H group. The aldehyde group of chlorophyll b (or pheophytin h) absorbs at 1663 to 1665 cm.~' ; this band is absent from the spectrum of pheo- phytin a and from that of the 2,4-dinitrophenylhydrazine derivative of ethyl pheophorbide 6. Spectra of chlorophyll h, taken in CS2 or CCI4 solutions and with crystals of chlorophyll h dispersed in Nujol, do not — in contrast to those of chlorophyll a — show a band indicating the occurrence of the chelated enol form of the jS-keto ester in ring V. The absence of this band from the spectra of chlorophyll h must be due to the presence of the oxygen atom in the formyl group in ring II. This electronegative group in position 3 may neutralize the tendency of magnesium to shift negative charge to the oxygen in the keto group in position 9, and thus reduce the tendency of the /3-keto ester for enolization. To summarize: infrared spectra indicate that in the crystalline state and in nonpolar solvents the /3-keto ester group of chlorophyll a is present as a chelated enol. In polar solvents, such as chloroform or diethyl ether, there is no indication of similar enolization . Chlorophyll b, INFRARED SPECTROSCOPY OF CHLOROPHYLL 41 H,C-CH ^ CH, /K /\ /\ H,C-CI \ C 4C-C,H, \ I I / '' C N N C / \ \ HC. Mq CH \ \ / "^l 1,0 d RjOOC-CHg Rg Compound Mg Ri R2 Chlorophyll o + C20H39 — C HOOC— C=0 I I R R II II (h) HO— OC— C=0 + RCHO + HoO > HO— C— C=0* + I I R R H2O + RCOOH (2) Light-Activated Reaction: I 1 (0H-) r n (a) HO— C— C=0 + H,,0 , HOCH COOH R R (b) HOCH- COOH + h^ + O3 > HOCOOH COOH R R (c) HO— C— O— OH COOH + RCHO + H2O > R HOCOOH CO*OH + H2O + RCOOH R It cannot at present be stated whether these chemiluminescences bear a direct relationship to the process in vivo. Such a relationship, if it exists, could be established through the isolation of identical inter- mediates in the two processes. Possibly, these findings may be of in- terest only from the standpoint of chlorophyll chemistry. But they certainly lend support to the hypothesis that chlorophyll can be directly excited by reactions in which it is involved or which take place in its vicinity. Acknowledgments. Thanks are due to Dr. Hans Gaffron and Dr. James Smith for stimulating discussions of some of the results here presented, and to Mr. Charles Soderquist and Mr. John Hanacek for invaluable assistance in the construction of apparatus. This work was supported in part by a grant from the U.S. Atomic Energy Commission. Discussion Liunry : Did j^ou get the emission spectrum? Strehler: The emission spectrum is very close to that of chlorophyll. It may CHEMILUMINESCENCE OF CHLOROPHYLL in vitrO 49 be shifted perhaps 10 to 15 m^ toward the blue, and that would fit with this shift observed here in the absorption spectrum changes, but I think better measure- ments have to be made before one can say anj-thing conclusive about it. Gaffron: The importance of these observations is not only that they permit an evaluation of the reactions of chloroplasts but that they demonstrate how aggregates of chlorophyll seem to be particularly apt to bind substances which, when they react chemically, may cause chemiluminescence. An oxidation appar- ently must occur right at the chlorophyll. Strehler: I should point out that dried chloroplasts treated with ethanol and chloroplasts treated with aldehyde and base do not show this luminescence. You have to remove the chlorophj-ll from its bound form, I believe, in order to obtain this chemiluminescence. Gaffron: That would be understandable. In chloroplasts so many other sub- stances which can't be easily oxidized stick to chlorophyll. Frenkel: What has happened to your efforts to find coenzymes? Is there any chance that you can modify your procedure and use this method for picking up coenzymes? Do you think you can now eliminate that? Strehler: Yes, I think it can be eliminated. It is a relatively sensitive method, but not as good as fireflies for ATP, and we hope to return to these experiments some time, if somebody does not do them in the meantime. References 1. Strehler, B. L., and Arnold, W. A., J. Gen. Physiol., 34, 809 (1951). 2. Arthur, W. E., and Strehler, B. L., Arch. Biochem. and Biophys., 1957, in press. 3. Vishniac, W., and Ochoa, S., /. Biol. Chem., 195, 75 (1952). 4. Strehler, B. L., and Cormier, M. J., Arch. Biochem. and Biophys., ^7, 16 (1953). 5. Strehler, B. L., and Cormier, M. J., /. Biol. Chem., 211, 213 (1954). (■>. Fischer, H., and Stern, A., in Fischer and Orth's Chemie des Pyrrols, Vol. 2, Akadem. Verlagsgesellschaft, Leipzig, 1940. Photoreduction of Synthetic Dyes GERALD OSTER, Polytechnic Institute of Brooklyn, Brooklyn, New Y'ork The work in our laboratory has been concerned mainly with studies of the kinetics of the photoreduction of synthetic dyes in aqueous media using visible light as the source of energy. The ener- gies associated with visible light vary between 2 and 3 electron volts. It would be useful if this energy could be utilized to bring about the reduction of noncolored molecules as well, the dye serving as the re- ceptor of light. In some cases, we have been able to utilize a large fraction of the absorbed energy for this purpose. I wish to summarize for you the results obtained so far in the hope that they may be of help in understanding the photochemical steps in photosynthesis. PHOTOREDUGTIONS It has long been appreciated that some reactions which involve colored substances and which are thermodynamically impossible will proceed, however, in the presence of visible light. For example, Rabino witch (1) has shown that thionine in the presence of acidified ferrous sulfate is photoreduced to give the leuco (colorless) dye, al- though in the absence of light this reaction is thermodynamically im- possible. In this reaction, the ferric ion which is produced oxidizes the leuco dye back to the colored form and a steady state is reached. We have found that a wide variety of other electron donors such as ascorbic acid, ghitathione, cysteine, thiourea, allyl thiourea, and stannous chloride under pH conditions where thionine (or methylene blue) is not reduced in the dark will give a stable leuco dye on irradia- tion with visible light if oxygen is rigorously excluded (2). The leuco dye exhibits a strong yellow fluorescence (phosphorescence in highly viscous media) and has the property of being oxidized to the normal dye even in the absence of oxygen if irradiated with near ultraviolet light. The fading with visible hght, recovery with ultraviolet hght, fading with visible light, and so on can be repeated several times and is reminiscent of certain biological processes (e.g., photoperiodism 50 PHOTOREDUCTTON OF SYNTTHETIC DYES 51 and photoreactivation) where reversals take place on using light of two different wavelengths. A particularly interesting electron donor is the chelating agent, ethylene diamine tetracetic acid (EDTA). Nickerson and Merkel (3) have shown that certain dyes in the presence of this substance are photoreduced by visible light. We have examined the kinetics of reduction of methylene blue in the presence of EDTA and found that the quantum yield varies with pH in the same way as does its degree of ionization and its chelating power (4). The leuco dye, in this case, is not fluorescent and does not revert to the normal form on irradia- tion with ultraviolet Ught. As with all leuco dyes, however, the nor- mal dye is obtained by flushing the solution with oxygen. EDTA is by no means a reducing agent in the ordinary sense. For example, we found that we could effect the photo-reduction of methylene blue in the presence of EDTA even when a ten thousand-fold excess of hy- drogen peroxide is present in the solution. We are now engaged in a more extensive study of the nature of transfer of electrons in photo- chemical reactions invohdng EDTA. The leuco dye is itself a reducing agent and can reduce other sub- stances. For example, leuco methylene blue will reduce tetrazolium salts to their corresponding insoluble formazans (compare refs. 3a and 5). The reduction potential of leuco methylene blue is about zero and it is therefore not a powerful reducing agent. Leuco acriflavine (produced, for example, by photoreduction of acriflavine in the presence of EDTA) is, on the other hand, a very powerful reducing agent (reduction potential about - 1 volt) and, if properly utilized, it should be more effective than acidified zinc as a reducing agent. For example, nitrobenzene is reduced by the leuco dye, the latter then reverting to its normal form. Thus, the dye serves as a sensitizer for the reduction of nitrobenzene and is not consumed in the overall process. If the leuco dye has a suflSciently low reduction potential (about -0.1 volt or less) it will produce free radicals (probably hy- droxyl radicals) on reacting with oxygen. Thus, if some vinyl mono- mer is present the chain polymerization of the monomer will ensue, resulting in a high polymeric product (6). Under conditions where chain termination is suppressed the overall quantum yield for mono- mer conversion may run as high as one billion (7). 52 G. OSTER LONG-LIVED EXCITED STATES We have found that very small amounts of certain substances will retard the photoreduction of dyes. For example 2.5 X lO"* mole per liter of paraphenylenediamine will decrease the rate of photoreduc- tion of eosin (carried out in the presence of allyl thiourea, oxygen being excluded) to one-half its normal value (8). For aqueous solu- tions at room temperature there are 6.6 X 10' encounters per second in a liter of molar solution. Hence, we calculate that the lifetime of the photoreactive dye molecules is at least 10 ~^ second. This is about ten thousand times greater than the lifetime of the first excited sing- let state. Our studies with se^•eral other of the dye-reducing agent combinations show that in all cases the electron donor reacts \v\th dye only when the latter is in the long-lived metastable excited state. Detailed kinetic studies on the photoreduction of fluorescein and its halogenated derivatives (9) show that the steps of the reaction are similar to those proposed for the ethyl chlorophylHde-sensitized oxi- dation of allyl thiourea reaction by Gaffron (10) and the photobleach- ing of chlorophyll in methanol by Livingston (11). Our studies show that not only is the quantum yield of photoreduction decreased by the addition of the dye itself but also a wide variety of other dyes (with or without spectra which overlap the spectra of the original dye) will, even in very small concentration, markedly decrease the quantum yield. The hmiting quantum yield obtained by extrapolating the rates to infinite reducing agent concentration varies markedly even with dyes of the same family. For example, fluorescein has a hmiting quantum yield of 2.0 X 10 "^ whereas for dibromofluorescein the hmiting quantum yield is 12.4 X IQ-^. The limiting quantum yield is determined by the relative number of quanta which are used to obtain the long-lived reactive species to those wasted by fluorescence and internal conversion. Results for eosin, for example, show that on an average about 1 of 11 molecules which are in the first excited singlet state undergoes transition to the metastable state. Of these remaining 10 excited molecules about 1.5 (on an average) revert to the ground state with, the emission of fluorescence, since the quantum yield of eosin is approximately 0.15, and the remainder fall to the ground state by a radiationless transition. PHOTOREDUCTION OF SYNTHETIC DYES 53 DYES IN THE BOUND STATE Most dyes exhibit a spectral shift to long wavelengths when small amounts of a high polymeric material to which the dye binds are added to the solution. In the case of diphenyl and triphenyl methane dyes, the fluorescence of the bound dye is considerably greater than that of the unbound dye (12). Such dyes also exhibit a greater fluo- rescence when in highly viscous media, and it appears that under such conditions the dye molecules in the excited state spend a longer time in the planar configuration (a necessary condition for fluorescence) than they do in low-viscosity media (13). Dyes in the bound state have profoundly different photochemical properties from those of free dyes. For example, dyes of the fluores- cein family are photooxidized irreversibly with a quantum yield in the neighborhood of 10"'* to 10 ""^ On binding, however, the quantum yield is decreased by a factor of about one thousand. Acriflavine under conditions where it is not photoreduced in the unbound state, is readily photoreduced when bound to polymeric acids (14). Tri- phenyl methane dyes when bound to polymeric acids resist reduc- tion in the dark on treatment with strong reducing agents which readily reduce the free dye. With light, however, the situation is reversed. Now the bound dye, in the presence of a mild reducing agent, is photoreduced w^hile the unbound dye remains colored (15). Introduction of certain substances such as nitrobenzene to an aqueous system containing the dye in the bound state and an elec- tron donor inhibits the photoreduction of the dye, the induction period during the irradiation being proportional to the concentra- tion of inhibitor. After the induction period is complete, the rate of the reaction becomes that of the original uninhibited system. It appears that the dye molecules on each polymeric substrate molecule (the local concentration is very high) are acting as a single photo- chemical unit and a few inhibitor molecules can affect several hundred dye molecules. Discussion Rabinowitch: In the case of methylene blue, you also had no quenching of fluorescence? Oster : No, we used too small amounts of allylthiourea. Rabinowitch: What is the quantum yield of fluorescence of methylene blue? Ten per cent? 54 G. OSTER Oster : Much less than 1 %. Wassink : What happens if you leave out the reducing agent? Oster : Nothing. Wassink : Do you observe any long-Hved species of the dye? Oster : My only indications are the fading of the dye or the initiation of poly- merization. Purely chemical criteria. Gaffrou many years ago sensitized the oxi- dation of allylthiourea in methanol by ethyl chlorophyllide; he was following the oxidation of the reducing agent, while I am following the photoreduction of the dye, by the disappearance of its color. Gaffron: At that time we could say only that there must be a long-Hved state; afterwards it was interpreted as the triplet state of the dye. You need only traces of oxygen for sensitized autoxidation to go with a high yield. This can be under- stood only if either the sensitizing dye, or all the substrate undergoing oxidation lives long enough in its excited state. Wassink: In the sensitized reaction, does the sensitizer, chlorophyll itself, change appreciably? Oster: I think, at least for ethyl chlorophyllide (I cannot speak for chloro- phyll), that it is reduced, although I did not see a loss of color. As with the other dyes, the reduced form reduces other substances present. Ethyl chlorophyllide is the only dye which I have studied which does not exhibit a readily discernible loss of color. Rosenberg: How long is the reduced form of your dyes stable in aqueous systems if you keep out oxidizing agents? Oster: Leuco methylene blue in solution where oxygen has been rigorouslj' excluded (and no ultraviolet light is present) seems to remain stable forever. Becker : I would like to point out one important thing. Oxygen in its ground state is paramagnetic, and if oxygen is in intimate contact, so to speak, with the dye molecule, this paramagnetism will allow the triplet state to be occupied to a much higher degree than is allowed ordinarily. A second point is quenching. Collisional quenching you cannot avoid. You cannot effectively remove electronic energy in small portions unless you have a tremendous number of collisions; so, if you want to remove large amounts of electronic energy, you must have a receiving molecule with an electronic fre- quency somewhere near the one that is in the donor. Oster: As a matter of fact, we feel, with no evidence other than what I have indicated here, the reason why certain dyes will inhil)it the reaction and others won't, may be that there is a properly situated long-lived (say triplet) state, which we should be able to pick up spectroscopically. That is something we have to look for. Lumry : Did you try any single-electron reducing agent? Oster : Chromous chloride is in this category. Lumry: The organic substances which you have listed as oxidants all require two electrons for their reduction. Some must pass through rather diffi cult-to-form, one-electron intermediates. Do you believe that two electrons are migrating simul- taneously? Oster : Yes. Limvry • Glutathione in the absence of oxygen or high hydroxyl concentration I'HOTOREDUCTIOX OF SYNTHETIC DYKS 55 has to go through a radical unless the process involves an improbable activated complex containing two SH-containing molecules. Hendricks: This discussion is only on the photochemistry of a particular system, but you might be interested to know that the system is a model for the photoreaction of photoperiodism, which is a photoreduction. Frenkel: What is the maximum energy j-ield in some of these reactions? I mean, actually is there any conversion of light energy into chemical energy? Oster: About 10% in the case of methylene blue. My interest is not so much the energy yield as it is the very strong reduction potential of the leuco dye pro- duced. For example, polarographic studies indicate that leuco acriflavine at high pH values has a reduction potential of —1.33 volt. Weigl : \Miat is the nature of this photoreduced species with its fabulous reduc- ing potential? If it is an excited state of the photoreduced dye, can you get to this state by exciting the photoreduced dye? Oster : The photoreduced species is the normal leuco dye. It is a colorless substance. Some of these dyes you cannot reduce in water by any known agent. Rabinowitch : The difficulty of reducing a substance is not in itself an indica- tion of a lower potential. It may be just an inactive system. It may be a problem of kinetics and not a problem of low potential. Oster : In the case of acriflavine the indications are that it has a very low re- duction potential. Rabinowitch : Your leuco compound does not reduce everything in the world to carbon. Kamen : You emphasize this is only the dye-absorbed polymer, not the free dye. Oster: Not necessarily. For example, acriflavine will not oxidize allylthioiirea unless it is bound to the substrate. Kamen : But you get the same result in the end. Oster : You get a reduced dye, yes. Gaflfron : Apparently it is still more complicated because you cannot expect that the two hydrogens go over at once. So you must have some radical intermediate. Linschitz : Under what conditions do your dyes photosensitize polymerization? Oster : The photoreduced form of the dye in the presence of vinyl monomer is stable as long as oxygen is rigorously excluded. On the introduction of oxygen, however, the monomer is rapidly polymerized. Linschitz : Hence the photoreduced dye is not a free radical. Oster: Yes. Free radicals which initiate polymerization are formed during the reaction between the leuco dye and oxygen. The radicals may be OH radicals or the semiquinone form of the dye. Holt : With ethyl chlorophyllide I tried a system using versene and there is a re- versible color change. Oster: That is very interesting. I want to point out that versene is not what we call a standard reducing agent. As I said before, you can bubble oxygen through it and you don't decompose the material. But it is a reducing agent to light-excited molecules. Lumry: I want to get this point about Krasnovskii's work clarified if I can. 56 O- OSTER In his woik, is the reduction process of chlorophyll a one-electron or a two-electron process? Does anybody know? Rabinowitch {remarks added in -proof): Krasnovskii postulates reduction to a radical (in ionized and neutral form). Linschitz et al, however, found in this sys- tem no paramagnetic resonance effects characteristic of free radicals. Wassink : Have you tried hydrogen gas as a reducing agent for your photoreduc- tions? Oster-.No. Linschitz: Did you say that the chlorophyllide showed no color change but that it was reduced? Oster: I said that I could see no color change, and I would like to ask Dr. GafTron if he saw a color change. Gaffron : No, but this was in the presence of oxygen. Oster : In the absence of oxygen I still did not see a color change. I just looked for it by eye. References 1. Rabinowitch, E., J. Chem. Phys., 8, 551 (1940). 2. Oster, G., and Wotherspoon, N., /. Chem. Phys., 22, 157 (1954). 3. (a) Nickerson, W. J., and Merkel, J. R., Proc. Natl. Acad. Sci. (U.S.), 39, 901 (1953); (b) Merkel, J. R., and Nickerson, W. J., Biochem. et Biophys. Ac/a, /^, 303 (1954). 4. Oster, G., and Wotherspoon, N., to be published. 5. Fujimori, E., J. Chem. Sac. Japan, Pure Chem. Sect., 75, 116 (1954). 6. Oster, G., Nature, 173, 300 (1954). 7. Oster, G. K., Oster, G., and Prati, G., J. Am. Chem. Sac, 79, 595 (1957). 8. Oster, G., and Adelman, A. H., J. Am. Chem. Soc, 78, 913 (1956). 9. Adelman, A. H., and Oster, G., /. Am. Chem. Soc, 78, 3977 (1956). 10. Gaffron, H., Biochem. Z., 264, 251 (1933). See also Rabinowitch, E., in Photo- synthesis, Chapter 18, Interscience, New York, 1945, and Weiss, J., Sym- posium Soc. Dyers and Colourists, p. 135, Manchester, 1949. 11. Knight, J. D., and Livingston, R., /. Phtjs. & Colloid Chem., 64, 703 (1950); Livingston, R., Record Chem. Progr. {Kresge-Hooker Sci. Lib.), 16, 13 (1955). 12. Oster, G., /. Polymer Sci., 16, 235 (1955). 13. Oster, G., and Nishijima, Y., J. Am. Chem. Soc, 78, 1581 (1956). 14. Oster, G., Trans. Faraday Soc, 47, 660 (1951). 15. Oster, G.. and Bellin, J., J. Am. Chem. Soc, 79, 294 (1957). Part II ABSORPTION, SCATTERING, FLUORESCENCE, LUMINESCENCE, AND PRIMARY PHOTO- CHEMICAL PROCESS IN VIVO Methods for Measurement and Analysis of Changes in Light Absorption Occurring upon Illumination of Photosynthesizing Organisms L. N. M. DUYSENS, Biophysical Research Group, Department of Physics, University of Utrecht, Utrecht, Netherlands Upon illumination of photosynthesizing cells a change will occur in the concentration of all intermediates and catalysts participating in photosynthesis. If the changes in concentration result in measur- able variations in absorption, these variations at various wave- lengths provide information about the kind and function of inter- mediates and catalysts in photosynthesis. For colorless cells and their extracts, measurements of spectral changes in absorption of intermediates and catalysts brought about by adding substrates, have proved of utmost importance for the elucidation of biochemical reactions. The magnitude of the change in optical density is proportional to the concentration of cell material. Fairly dense cell suspensions have been used in most experiments. It is not possible, in general, to use concentrations of photosynthe- sizing cells as high as those of colorless cells, because the photosyn- thesizing cells are so strongly colored that they would absorb prac- tically all the measuring Hght. Since the concentration of substances changing their absorption seems to be of the same order of magnitude in photosynthesizing and colorless cells, this may be a reason why, for measurements on photosynthesizing cells, generally an apparatus is required ten to a hundred times more sensitive than the apparatus usable for "colorless" cells. In addition precautions should be taken that the light used for bringing about a change in absorption does not affect the measuring apparatus directly. APPARATUS AND METHODS FOR MEASURING ABSORPTION CHANGES The experimental problem is to measure only those variations in intensity of the monochromatic light beam transmitted by a sus- pension of photosynthesizing cells which are caused by changes in 59 60 L. N. M. DUYSENS optical density ])rought about by illumination. One might try to measure the intensity of the transmitted beam by means of a photo- tube (or multiplier), amplifier, and meter. However, in many experi- ments the changes in intensity must be measured with a precision of the order of 0.01%. In order to do this, a meter of great precision is required; and the intensity of the measuring beam, the sensitivity of the phototube (or multiplier), and the amplifier must be constant within 0.01% — requirements which are difficult to fulfill. Apparatus with compensating beam. The required precision can be achieved by using a compensating light beam from the same light source as the measuring beam. The compensating beam should not change its intensity owing to changes in absorption of the suspen- sions but should deflect the meter or recorder opposite to that of the measuring beam. If the measuring and compensating beam are of about equal intensity, changes in the electronic parts of the apparatus will not cause marked deflections of the meter; appreciable deflections will occur only if the transmission of the suspension changes. The sensitivity of this device increases with the intensity of the measuring (or compensating) beam. In the instruments (1,2) used so far successfully, the measuring and compensating beams were alternated with line-frequency by means of mechanical devices, such as moving mirrors, and both beams impinged upon the same multiplier or phototube. The output was passed through an a-c amplifier and caused a deflection of a phase- and line-frequency-sensitive device. In one type of apparatus (1), the two beams were obtained by splitting the beam leaving the monochromator; the compensating beam bypassed the suspension. A description of such an apparatus will be given below. In the second type of apparatus (2), the compensating beam is ob- tained from a second monochromator, and is also passed through the suspension. The wavelength of the compensating beam is fixed and is selected so that no change in absorption upon illumination occurs. The advantages of the latter arrangement are that it compensates to a certain extent for changes in transmission due to sedimentation of the cells, and that it is relatively easy to catch a larger angle of light scattered by the cells. One disadvantage of the second type of appara- tus is that it is affected by changes in lamp output, when the wave- lengths of measuring and compensating beam are different, since the LIGHT ABSORPTION OF PHOTOSYNTHESIZINO CELLS 61 emission of tungsten is not the same function of temperature at differ- ent wavelengths. It seems also difficult to ascertain whether the ab- sorption of the compensating beam does not change upon illumination. There is in our opinion no reason to assume that an apparatus of the first type is inferior to one of the second type; the former is, however, less complicated and less expensive. Calibration. The calibration of the apparatus required to express deflections as changes in optical density can be carried out in two ways: 1. By measuring the deflection brought about by a known change in the measuring beam. This can be done by inserting a glass plate (1) or wire screen (see below) into the measuring beam. 2. By measuring the photocurrent produced by the measuring beam. This can be done by comparing this current by means of a "Brown converter" with a known current. Various checks on Unearity of amplifier and multipher and on influence of compensating beam are needed to make certain that the results are correct {cf. 3). Apparatus using blank. An apparatus based on the same principle as a conventional spectrophotometer such as the Beckman DU could conceivably be used, provided a "blank" is used of approxi- mately the same absorption as the cells to be measured. However, owing to technical difficulties, the sensitivity generally is not better than about 1 X 10"^ unit of optical density and may in general be insufficient for measuring spectral changes in photosynthetic tissue. Lundegardh (4) used an automatic recording apparatus of this type, but only in a narrow region where the absorption was low so that a dense suspension could be used. Illumination of cells. Two methods of illumination for exciting the changes in absorption can be used. 1. The cells are illuminated in the vessel through which the measur- ing beam passes (1,2). 2. A batch of cells is illuminated outside the absorption vessel and quickly moved into this vessel to replace the nonilluminated cells. The cells are pumped around in a closed circuit, and a transparent part of the circuit at a short distance from the absorption cell is illuminated or darkened (5). In a variation of this method, the flow is stopped after the illuminated cells have passed into the absorp- tion vessel (4). Illumination in the absorption vessel gives the least variation in 62 L. N. M. DUYSENS scattering by the cells and the best time resolution. However, second- order effects on the compensation owing to scattered or fluorescent hght must be eliminated. This may be done (a) by minimizing the scattered and fluorescent Hght by means of filters, (6) by using two sectoriated rotating disks, one in front of the phototube to cut off the hght while the exciting hght is being admitted through a hole in the other disk. MIRRORS HOLES PROJECTION LAMP Q t COMPENSATING m BEAM \. "^^A^ SUSPENSION f MONOCHROMATOR RECORDER AMPLIFIER AND PHASE SENSITIVE RECTIFIER MULTIPLIER Fig. 1. Diagram of apparatus for measuring changes in absorption spectrum of a suspension of photosynthesizing cells. Apparatus for measuring transients. Fast changes may be measured by means of a photomultiplier connected to an oscilloscope. If a measurement is completed in a short time, slow changes in intensity of the hght sources, in multiplier, and in sensitivity of oscillograph amplifier may be neghgible. The contmuous current caused by the measuring beam may be eliminated by means of a condenser coupling or by cancehng it out by a constant countercurrent. Witt (6) succeeded in measuring in this way rapid changes in ab- sorption brought about by hght flashes. (So far, no description of his LIGHT ABSORPTION OF PHOTOSYNTIIESIZING CELLS 63 apparatus has appeared.) Probably lilters were used to prevent ex- citing light from reaching the multiplier. Description of a complete apparatus. A diagram of the apparatus used by us in recent investigations (5,7,8) is shown in Fig. 1. It is a modification of one described before (1). A monochromatic beam is split into two beams by means of a rotating disc consisting of mirrors alternating with holes. The beam passing through the holes in the disc TO B o- TO C O- CAPACITORS IN /J F K = X 1000 M = X 10* Fig. 2. Amplifier and phase- and ffefiuency-sen.sitive rectifier. is the measuring or scanning beam. This beam passes the movable wire screen C and the suspension. The compensating beam bypasses the suspension and causes a deflection of the recorder opposite to that of the scanning beam, keeping the indicator of the recorder on scale. The intensity of the compensating beam can be adjusted by means of a "wedge." In many experiments the intensity of the scanning beam and sensitivity setting of multiplier and amplifier was such that a 64 L. N. M. DUYSENS change in intensity of about 1%, brought about by moving the (cali- brated) wire screen into the beam, caused a deflection of the recorder of 100 to 200 mm. This deflection was compared with the deflection caused by a change in absorption of the suspension upon iUumination with a 500- watt projection lamp, the intensity of which could be varied by varying the lamp voltage. The filters /i and /2 are needed to reduce stray light from the pro- jection lamp. In the absence of these filters, the stray light did not bring about a deflection of the recorder, since it was not interrupted with the frequency 60 c.p.s. ; it was found, however, to cause a de- flection when both the measuring and compensating beams were on. The filters were selected to reduce the stray light to such a low level that the projection lamp caused no deflection when a mastic suspen- sion was used, instead of one of photosynthesizing cells. The electronic part of the apparatus is shown in Fig. 2. The output of the three-stage amplifier is rectified by the double diode 6AL5, filtered by the four-stage filter, which removes frequencies greater than about one, and, via the balanced cathode follower, passed to a Brown recorder. The multiplier voltage and the filament voltage for the first two stages is obtained from batteries, the 6.3-volt filament voltage for the other stages is obtained from a transformer, and the plate voltage is obtained from a stabilized power supply. TYPE OF DATA GIVEN BY MEASUREMENTS OF THE CHANGES IN ABSORPTION Figures 3 and 4 give examples of the type of data obtained. The time course of the change in optical density in Chlorella appears to de- pend upon several factors. Figure 3 shows that the initial increase in optical density at 520 mix upon illumination, which is caused by an unknown pigment (cf. 7) is much greater in an anaerobic medium than in an aerobic medium. However, upon illumination for about 1 minute, the original increase in the anaerobic medium is followed by a decrease. Figure 4 shows the time course at 420 m/j,, where cytochrome / presumably shows a change in absorption (cf. 8) . The changes are much larger, but less rapid, in the presence of carbon dioxide than in its absence, suggesting that carbon dioxide and cytochrome / compete for the same hydrogen donor. A great number of experiments can be made. At each wavelength, LIGHT ABSORPTION OF PHOTOSYNTHESIZING CELLS ()5 Chlorella \520 1 mm Fig. 3. Changes in optical density of Chlorella at 520 m.u upon onset of illumina- tion (upward arrow) and darkening (downward arrow) as a function of time. The relative intensities of the exciting Hght are wiitten near the arrows. Chorella A '420m/j logje .10~^ I air with and without carbon dioxide 1 A /Off Fig. 4. Time courses of changes in optical density of Chlorella at 430 m/* in the presence and absence of carbon dioxide. it is possible to measure the time course upon illumination or darken- ing as a function of the following parameters: 1. The composition and pH of the suspension medium. One may vary, for example, the concentration of carbon dioxide or other nutrients, oxidizing and reducing substances, poisons. 66 L. N. M. ])x:ysens 2. The pretreatment of the culture. Factors such as age, growth medium, and Hght intensity may be varied. 3. The intensity and wavelength of the exciting light and the lengths of light and dark periods. Further data are obtainable by following the time course of changes in optical density caused not by light but by alterations in the me- dium. These measurements are not discussed in this paper. EVALUATION OF THE DATA Among the more useful results which can be derived from measure- ments of the time course at various wavelengths are the difference spectra. These spectra often allow the identification or characteri- zation of the substances changing their absorption. The difference spectrum can be obtained as follows. All external conditions are kept approximately constant, and a periodic sequence of Ught and dark periods is given. It is often found possible to select the external conditions and the patterns of illumination in such a way that at each wavelength the (change in) optical density is a periodic and reproducible function of time. A difference spectrum is then ob- tained by plotting the changes in optical densities which occur be- tween two corresponding times t and t' of the time course graph as a function of the wavelength. This difference spectrum is equal to the difference of the absorption spectra of the cells at the times t and t' . It is the sum of the difference spectra of the intermediates and other substances, which change their absorption spectra. If t is the time at which the illumination period starts and t' is a time between the beginning and end of the illumination period, the difference spectra at various times t' may give a clue to the time sequence in which the intermediates change their absorption spectrum. If, e.g., one inter- mediate changes much more rapidly at the onset of illumination than the others, the difference spectrum obtained for a time t' shortly after the start of illumination is caused only by this intermediate. Another family of difference spectra can be obtained selecting a different value of a parameter, such as the intensity of the exciting light or the composition of the medium. The comparison of difference spectra obtained under various conditions may greatly facilitate the analysis of these spectra in terms of the difference spectra of single intermediates. After such an analysis has been carried out, it is pos- sible to plot the (relative) concentrations of the intermediates (or LIGHT ABSORPTION OF PHOTOSYNTHESIZING CELLS 67 other substances) as a function of a certain parameter, such as the time after onset of illumination or of the intensity of the exciting light. These data may be useful in the elucidation of the mechanism of photosynthesis. Examples of the procedures outlined above can be found in refer- ences 1 and 4 to 9, and in articles in this volume. References 1. Duysens, L. N. M., "Transfer of excitation energy in photosynthesis." Doctoral thesis, Utrecht, 1952. 2. Chance, B., Smith, L., and Castor, L., Biochim. et Biophys. Acta, 12, 289 (1953). 3. Chance, B., and Legallais, V., Rev. Sci. Instr., 22, 634 (1951). 4. Lundegardh, H., Physiol. Plantarum, 7, 375 (1954). 5. Duysens, L. N. M., Carnegie Institution of Washington Yearbook, 62, 154 (1953). 6. Witt, H. T., Z. physik. Chem., 4, 120 (1955). 7. Duysens, L. N. M., Science, 120, 353 (1954). 8. Duysens, L. N. M., Science, 121, 210 (1955). 9. Chance, B., and Smith, L., Nature, 175, 803 (1955). Reversible Bleaching of Chlorophyll in Vivo* J. W. COLEMAN, A. STANLEY HOLT, and EUGENE I. RABINOWITCH, Photosynthesis Research Project, Department of Botany, University of Illinois, Urbana, Illinois It has often been suggested (L3) that in photosynthesis chlorophyll undergoes a reversible change. It could be either: (1) Transformation into a "hiradical," metastable state (an electronic triplet state, with both free valencies on the same atom, or a tautomeric state, with the two valencies at different atoms); (2) reduction, either to a semi- quinone or to a valence-saturated leuco compound; or (3) oxidation, also either to a radical or to a saturated product. Transformation into the metastable state has been suggested as the first step in the internal conversion of excitation energy, which limits the yield of chlorophyll fluorescence to 25% (10) or 33% (7), in vitro, and 2% to 3% in vivo (7). According to Franck (5), photosynthesis probably occurs by reactions of metastable chlorophyll a molecules. According to Livingston and Ryan (12), these molecules are co-re- sponsible for changes in the absorption spectrum of illuminated chloro- phyll solutions in the photostationary state; Livingston and Ryan (12) and Livingston, Porter, and Windsor (11), using condenser flashes with s3Tichronized absorption measurements, found that during an intense flash, up to 90% of chlorophyll (in a IQ-^M solution) can be present in the metastable state, Livingston and Ryan's (12) steady-state experiments indicated bleaching at 403 mn and enhanced absorption at 439.5 to 524.5 mn; whereas their flash results showed bleaching at 468, 470.5, and 477.5 mju, and enhancement of color only at 524.5 m/i. However, according to the newer flash data of Livingston, Porter, and Windsor (11), analyzed by Livingston (9), enhancement extends over the range 450 to 560 m^, with a sharp peak at 475 m/x and a shoulder at 520 m^u. Evstigneev and Gavrilova (4) found that reduced chlorophyll a, obtained by illumination of phenylhydrazine-containing solution in * This work was carried out with the assistance of the Office of Naval Research. 68 REVERSIBLE BLEACHING OP CHLOROPHYLL in VIVO 69 toluene, has absorption bands at 518 m/x and 585 ni/z. Both bands were attributed to a semiquinone: the 518-m/i band to its ion and the 585- m/i band to the nondissociated form.* Krasnovskii (6) has suggested that chlorophyll participates in photosynthesis by reversible reduc- tion to the semiquinone state. Studies of reversible photobleaching of chlorophyll in 02-free methanol (9,10,14) and of its reversible photooxidation by Fe+++ in methanol (15) and by quinone in rigid solvents (8) revealed an en- hanced absorption in the region 450 to 530 m/i, but no sharp bands were detected. The brown intermediate in the phase test (probably, an ionized enol form) of chlorophyll a shows a strong band at 524 m/i, with a shoulder at 486 m/x, and weaker bands at 645 and 683 mn (16). It thus seems that, in vitro, reduced chlorophyll a is characterized by bands at 525 m/x and 585 m/c, metastable chlorophyll a by a band at 475 m/i, and ionized chlorophyll by bands at 486 m/x and 524 m/i. Reversible oxidation increases absorption in the same general region, apparently without producing a sharp new band. The absorption in the red decreases in every case. Duysens (2) (and also Witt (17)) noted that illuminated Chlorella cells show, in addition to spectral changes attributable to the oxida- tion of a cytochrome (and perhaps also to the reduction of a pyridine nucleotide) (1), a sharp rew absorption band at 515 m/x and a some- what smaller "negative" band (i.e., selective decrease of absorption) at 478 m/i. Duysens attributed the two changes to the transformation of an unidentified pigment, whose "dark" form absorbs at 478 m/x and whose "phototropic" form absorbs at 515 m/x. Witt noted the 515-m/t band also in plants exposed to an intense light flash. Duysens observed no change in the absorption in the red region, thus apparently precluding the attribution of the effect at 515 m/i to chlorophyll. Using an apparatus similar in principle to that of Duysens (3) but somewhat different in construction, we have been able to observe a decrease in absorption of illuminated Chlorella cells in the red. In our apparatus, the modulated photomultiplier output was ampli- fied through three sharply tuned and six narrow-band staggered stages; by means of a phase-inverting parallel twin-T tuned network, a considerable portion of the signal was negatively fed back from the fourth stage to the input. The ultrasharp tuning and increased feed- * Linscliitz and co-workers (8A) found no evidence of the presence of free radi- cals in this system by the method of paramagnetic resonance. 70 J. W. COLEMAN, A. S. HOLT, E. I. RAHINOWITCH back were necessitated by difficulty of discriminating between fluc- tuations in the fluorescence excited by the (very intense) actinic light and changes in the (much weaker) measuring light. After the ninth stage of amplification, the signal was rectified, compared, and, by means of a balanced-plate cathode-follower, fed into a Brown recorder (as in Duysens' instrument) . ^f measured =/"(a) Aex=/'(M 650 700 A in rriM 750 Fig. 1. Reversible changes in Chlorella spectrum during illumination. Dashed line: estimated correction for fluorescence. Chlorella cells, grown in our laboratory, were washed, suspended in carbonate, and refrigerated until needed. The cells were used as taken from the refrigerator (optical density of suspensions, 0.45, at 680 mju, corrected for scattering). The actinic light was furnished by a tungsten lamp (1000 watt, GE 1000T20, 120 volts); the entire side of the cuvette was uniformly illuminated. Before using a sample for systematic measurement, a check was made at several selected wave- REVERSIBLE BLEACHING OF CHLOROPHYLL in VIVO 7 1 lengths to see whether the cells showed the normal response to illumi- nation. The apparatus reproduced with excellent agreement the earlier work of Duysens; in addition, it clearly showed absorption changes in the red. A typical differential absorption spectrum is shown by solid line in Fig. 1. At 680 m/i, the optical density of illuminated cells can be 0.25% lower. Exact comparison of this decrease with the increase at 515 mju is not possible because we had to use different exciting Ughts in the two regions. However, since the two effects are of the same order of magnitude, the assumption is permitted that they are both caused by a reversible change in chlorophyll a. Spectroscopically, this change is most similar to that observed by Krasnovskii et al. (and by Evstigneev and Gavrilova) upon reversible reduction of chlorophyll a in vitro. The smaller changes farther in the red (decline of absorption at 710 to 715 niju, and increase at 730 mn), as well as the bleaching at 478 mn (already noted by Duysens), remain to be interpreted. Several reversible changes of chlorophyll may occur at once in the cell, e.g., the formation of metastable triplet molecules may be superimposed on that of the semiquinone. It will be noted, however, that the effect observed in vivo at 475 mju is opposite in sign to that expected from the formation of metastable chlorophyll a. Discussion Duysens: I, too, made some experiments in red while I was working in your laboratory. I used illumination of presumably lower intensity than you did — the same illumination at 515 and 680 ran. It was rather difficult to exclude completely any change at 680 mn, but I think the changes there were less than one-fourth of those at 515 rrifi. So it is possible that the changes which you find occur only at higher light intensity and are not correlated with those at 515 m/x- Rabinowitch: That is a pos.sibility, and I would not say that there is here a di- rect expeiimental disagreement of the kind we are only too familiar with in pho- tosynthesis. However, according to Coleman and Holt, the effects are of the same order of magnitude in the green and in the red; but, since two different kinds of illuminating light were used in these two regions, we as yet cannot compare them exactly. Strehler: I would like to make two points: One, that using Dr. French's ap- paratus we were not able to find any changes on the red side of about 670 mp, while we did get large changes around 648 m^u. Second, apparent changes in trans- mission could be due to fluorescence. How can you rule out the possibility that 72 J. W. COLEMAN, A. S. HOLT, E. I. RABINOWITCH changes in fluorescence yield may occur when you add the cross illumination? Fluorescence caused by the measuring light is modulated, and any fluctuation in it will be picked up by your phase-sensitive detector. If the sense of the change is proper, you will get in this way something that could be interpreted as a decrease in transmission, Rabinowitch: Unless the actinic light caused changes in the spectral composi- tion of fluorescence and not only in its intensity, how could you get in this way both positive and negative effects? However, the commonly made assumption that the spectrum of fluorescence does not change when changes in fluorescence in- tensity are observed in vivo is in need of experimental confirmation. If this happens, the meaning of many data in the literature becomes questionable. Strehler: You are passing through an absorption band here (660 to 720 m/x) going from wavelengths that do excite chlorophyll fluorescence into a spectral region where chlorophyll does not absorb. I am not saying that your effect in the red is necessarily due to this, but this possibility worried us when we measured our 648-m;u band. I believe we can rule it out in our case, because we put a red filter in front of the photomultiplier, which transmitted only the fluorescence, and with this filter in place we didn't observe any significant signals upon croes illuminating. Rabinowitch: One probably has to consider this point more carefully than we did, but it still seems to me that you cannot get in this way both positive and negative effects. Chance: This difficulty would be minimized in the double-beam apparatus we use. If j^ou could change your apparatus corresponding!}^, you may get an ade- quate control. Strehler: Provided the excitation of fluorescence by the two different wave- lengths is the same. Actually, by placing another monochromator after your ab- sorption cell, you can remove most of the fluorescence, since it forms a broad band. Chance: Yes, but you may be unable to get enough light into the analyzing monochromator. Rabinowitch : Of course, when he measured at 515 mju, Holt did use a filter to cut off fluorescence. Strehler: The changes at 515 m^u cannot be due to fluorescence. Rabinowitch: When we measured at 515 m^u, the fluorescence-removing filter did not change the results. This means that at 515 m/i, the effect of modulated fluorescence was insignificant compared to the absorption effect. If the total ob- served effect were much stronger at 515 m^i than at 680 m^, one could suggest that fluorescence was much more significant in the latter region; but, as I said before, Coleman's impression is that the effects in the two regions are about the same; and, if this is so, one can infer that the role of modulated fluorescence is in- significant in the red, too. Linschitz : If we accept your data at their face value, there would still be a prob- lem of showing that the changes at 680 and 515 m/x are correlated. We have found that, in vitro, these two bands may change at different rates. There are probably two different products which cause the change. REVERSIBLE BLEACHING OF (MILOlU)l'IIYLL ill I'ii'O 73 (Remarks added in manuscript) Coleman : The possible influence of changes in modulated fluorescence, pointed out by Dr. Strehler, was taken into account by Dr. Holt and myself, although we did not include the discussion of this point in our report. Since it has been raised by Dr. Strehler, here is a brief summary of the reasons why we considered this in- fluence negligible. In contrast to the constant actinic light, the modulated scanning light does produce a 60-cycle modulated fluorescence to which the photomultiplier can re- spond. In the apparatus as we used it, this modulated fluorescence is compen- sated when the actinic light is off ("darkness"); but, from general knowledge of the fluorescence phenomena in vivo, we must consider it possible — even likel}^ — that switching on the actinic light — changing from "darkness" to "light" — will change the intrinsic capacity of the cells to fluoresce and thus affect (probably, increase) the intensity of the modulated fluorescence. The apparatus will react to such a change in modulated fluorescence as if it were a change (decrease) in absorption in the fluorescence vessel. We thus have : AA (true) = AA (measured) — AA^ where AA^ is the change in the modulated fluorescence reaching the detector, caused by switching on the actinic light. To determine the order of magnitude of AA^, the scanning beam was rerouted, so that it traversed the suspension orthogonally to its usual path. In this wa,y, only a small scattered fraction of the exciting light entered the photomultiplier, instead of the full transmitted beam, as in the usual arrangement. Fluorescence, on the other hand, was collected from the same volume, subtending the same angle at the photomultiplier. First, heat-killed cells were illuminated with the rerouted beam; the scattered scanning light was compensated, and the deflection di (corresponding to 1% change in the beam intensity) was recorded, as usual, with the help of a cali- brated neutral filter. Although di changed slightly when the actinic light was turned on (probably because of scattered actinic light modulated by line ripple voltage), this change was negligible. The suspension of dead cells was then re- placed by one of live cells, having the same optical density, and the procedure was repeated, giving the deflections ^2 in darkness and di when exposed to actinic light. We can expect di to be the same as di, and this was in fact the case within a few per cent. The value of AA^, the increase in fluorescence caused by the actinic light, was calculated from the equation ^, _ K [2ri3 - (dr + d,)] ^^' dTTT, where if is a correction factor (c^l.4) for the loss of intensity of scanning beam caused by additional reflections required for its rerouting. The resulting correction curve, AA^ = /(X), is shown by the dashed line in Fig. 1. At X = 680 mX, AA^ contributes about 7% of A A (measured). Where A A (measured) is smaller, AA^. contributes proportionally more, but in no case is the error large enough to alter the shape of the AA-curve. 74 J. "VV. COLEMAN, A. S. HOLT, E. I. RABINO WITCH References 1. Duj'sens, 1 . N. M., Science, 121, 210 (1955). 2. Duysens, L. N. M., Science, 120, 353 (1954). 3. Duysens, L. N. M., Thesis, Utrecht, 1952. 4. Evstigneev, V. B., and Gavrilova, V. A., Compt. rend. acad. sci. U.R.S.S., 91, 899 (1953). 5. Franck, J., Arch. Biochem. and Biophys., 4^, 190 (1953). 6. Krasnovskii, A. A., Compt. rend. acad. sci. U.R.S.S., 60, 421 (1948). 7. Latimer, P., Thesis, Univ. of Illinois, 1956; Science (in press). 8. Linschitz. H., and Rennert, J., Nature, 169, 193 (1952). 8A. Linschitz, H., et al., Arch. Biochem. and Biophys. (in press). 9. Livingston, R., J. Am. Chem. Soc, 77, 2179 (1955). 10. Livingston, R., /. Phys. Chem., 45, 1312 (1941). 11. Livingston, R., Porter, G., and Windsor, M., Nature, 173, 485 (1954). 12. Livingston, R., and Ryan, V. A., /. Am. Chem. Soc, 75, 2176 (1953). 13. Rabinowitch, E. I., Photosynthesis, Vol. 1, p. 483 ff., Interscience, New York, 1945. 14. Rabinowitch, E. I., and Porret, D., Nature, UO, 321 (1937). 15. Rabinowitch, E. I., and Weiss, J., Proc. Roy. Soc. (London), A162, 251 (1937). 16. Weller, A., /. Am. Chem. Soc, 76, 5819 (1954). 17. Witt, H. T., Naturwiss., S, 72 (1955). Reaction Patterns in the Primary Process of Photosynthesis H. T. WITT, Physikalisch-chemisches Institut der Universitdt Marburtj, Marburg/ Lahn, Germany In the primary process of photosynthesis hght is absorbed in- directly or directly by chlorophyll. During this process the chloro- phyll or chlorophyll complex changes into an unknown excited state. With the aid of the energy of this state water is split in an unknown way into oxygen and hydrogen (1). In a secondary process this hydro- gen reduces carbon dioxide to carbohydrates. You know of the great success in the investigation of this carbon dioxide cycle here in Amer- ica (2). But there is little knowledge about the first step which trans- forms light into free chemical energy. A special state of chlorophyll during photosynthesis has not yet been observed. Furthermore we cannot tell anything about the mechanism of hydrogen production because the usual methods, for instance, the measurement of oxygen production, are not sufficient to detect the fast reaction in the pri- mary process. We have tried to measure reaction patterns in the primary process by looking for fast changes of ahsorption immediately after flashes of light (3). These changes of absorption cannot be detected by the usual type of absorption measurements because the reactions are very fast and the changes very small. There are, furthermore, technical complica- tions due to the scattering of the incident light by the plants. There- fore we have developed a sensitive apparatus by which we could measure absorption changes in times between '-^10 ~^ second and sev- eral seconds.* The experiments were made with various leaves, Chlorella, and chloroplast.«. The photosynthesis was induced by periodic flashes of light. The spectrum of the flash lay between 620 m/x and 700 ni/x. * Changes of the absorption in steadj' light were reported in connection with redox reactions of cytochromes and pyridine nucleotides (4). 75 76 H. T. WITT We were searching for changes of absorption between 400 mn and 580 When photosynthesis is started by flashes of Hght, there is a rapid increase of a new absorption band at ^^515 m/x. At the same time (within the accuracy of measurement) a decrease of absorption takes ,o I- o 01 4j 1 ■ 10 W^ sec time Fig. 1. Change of absorption of C/iZoreZto as function of time. At time < = the Chlorella were lighted by a flash of light, ti = 5 X 10"^ second, td = 0.5 second. Temperature 19 °C. Upper cuive: change of absorption at 515 m/*. The change from to 1 corresponds to an increase of absorption. Lower curve: change of absorption at 475 myu- The change from to 1 corresponds to a decrease of absorp- tion. place with a maximum at about 475 m/x (3) (Fig. 1).* The relative change of absorption is only of the order of a tenth of a per cent. Using flashes as short as 3 X 10~^ second we observe a change of absorption within the duration of the flash. This change must, therefore, take place within the time of 3 X 10 ~^ second or less. In the dark time after the flash, the change of absorption disappears completely within '-^lO"^ second. With an increase of light intensity there is also an increase of the change of absorption (Fig. 2) . At high intensities there is a saturation of the absorption changes. The intensity at the saturation point is of the order of that which suffices to saturate photosynthesis. * In text of all figures t[ means: duration of hght flash, ta: dark time be- tween the flashes. REACTION PATTERNS TN PHOTOSYNTHESIS 77 At teniperatvres near 50°C. the change of absorption decreases with increasing temperature (Fig. 3). At this temperature photo- synthesis also decHnes. With decreasing temperature we can observe by using short saturating flashes that at 515 m/x the fast increase of absorption is independent of temperature (Fig. 4). But in the dark time after the flash the decline of the change of absorption is a function of temperature. Between 30° C. and 5°C. we measured half-times of c .o 5- o -Q C time ' ' 25-/0* sec _j A. — — ^ -• -i — -^ 2 4 6 8 10 lightintensity Fig. 2. Change of absorption of Chlorella at 515 mfi as a function of flash light intensity, ij = 5 X 10~^ second, td = 0.24 second. Temperature 18 °C. The abscissa (broken line) belongs to a curve that connects the maxima of the changes of absorption. 49° 50° 51° 52° 53 °C .o ^ / 9- o -Q CD C -C o time I 1 35 • 10 SeC Fig. 3. Change of absorption of Chlorella at 515 m/j. as a function of time at different temperatures around 50°C. ti = I X lO"* second, ^ — A"^ "E A ^ — ' — in A 1 p ^ ^^^T c o — ^z:^ _ y k- ^ .-*^ c y^"^ my^ a* In C ¥ / a> -S- / 1 t= E / / E - / /i o A 525 Absorption 1 / • A 648 Absorption i o »- / i \ o / / ' * Luminescence | c rJ a> 1 J o 1 7* U) / / c U" E IT 3 \L -J ^ 1 1 100 200 300 Light Intensity (Foot Candles) Fig. 4. Light saturation curves for 525 m^, 648 m/i, and luminescence using a flow system and illuminating the algae for about 0.1 second prior to measurement. 8 11 tn o 2 — / \ ^. a> 1 1 \ ^^^ _— 525 o 1 ^T" 1 i nht \ .- — ^^^ (-\ c 7 L gni 1 ^ — ^p-...^ Dark 1 i 1 "^ 30 90 60 Time (seconds) Fig. 5. Time course of luminescence, 525-mM, and 648-mM absorption in a flow system. are due to the same compound, we have measured the dependence of the two processes on illuminating intensity. The intensity dependence appears to be different for these two wavx'lengths. These results and the time course of the luminescence compared to the absorption spec- trum changes are illustrated in Figs. 4 and 5. The luminescence paral- 94 B. L. STKP^HLER AND V. H. LYNCH lels both the 520- and 648-mM bands in its time course, but appears closer to the G48-iniu band in its i-esponse to Hght intensity variations. It is also clear from these measurements that these two peaks are due to different chemical compounds, since the changes do not parallel each other as a function of the external variables. 8. It was established that the 520-m/i absorption band could not be due to a modulated red fluorescence artifact because there was no HIATTREATED CHlOREllA 250 LIGHT INTENSITY Foot C o n d I e s Fig. (). Light intensity dependence of Chlorella transnaission changes at 525 m/i following partial inactivation by heat. The deflection equals about 3 to 4 O.D.U. X 10 -^ at saturation. difference signal when a red filter was inserted in front of the photo- multiplier and the cells were illuminated in a flow system. The changes could not be due to a modulating green fluorescence because there is little if any green fluorescence in Chlorella. 9. When the cells were subjected to heat treatment and then illuminated an interesting effect was noted. After 4 minutes at 51 °C. only an increase in absorption was noted; and, surprisingly, this increase encompassed the entire spectral region measured. Figure 6 shows the intensity dependence of the process, and Fig. 7 illustrates the wavelength dependence of this phenomenon. Treatment at 60°C. for 4 minutes completely destroyed any reversible changes in absorp- tion. The luminescence was reduced to about 5% of the normal value by the former treatment and completely ol)literated l)y the latter. ABSORPTION SPECTRUM CHANGES 95 10. We also noted a serie« of smaller absorption bands in the region 550 to 660 ni/i (see Fig. 3), perhaps due to cytochromes as earlier sug- gested by Lundegirdh. These results permit the following tentative conclusions: First, there is formed during illumination a compound with an absorption band around 520 mix. In succeeding dark periods, its concentration falls to a level considerably below that of the dark \'alue prior to illumination. This compound can be formed in the dark by thermal chemical reactions as well as by photochemistry, since its 500 550 WAVELENGTH mu 600 TTO Fig. 7. Change in optieal density at different wavelengths during ilhimination following heat treatment of Chlorella for 4 minutes at 51 °C. Incident intensity, 250 foot-candles in a flow system. Temperature, 25 °C. concentration rises again to a higher level in the dark after the initial depression. Second, there is also formed another compound with an absorp- tion maximum aromid 648 m/x different from the one having the 520- m/i band chemically. Both of these substances appear to be related to the luminescence process since their concentration and the lumines- cence intensity exhibit similar kinetic behavior and time courses. Third, it appears unlikely that the 520-m/z band is due to a chloro- phyll a derivative, since there is no comparable change in the chloro- phyll absorption maximum in the red or blue. Rather, it is suggested that this compound may be a flavin-type free radical which is pro- or. B. L. STREHLER AND V. H. LYNCH TABLE I Meas- Worker uring wave- length .Method of ilhiiiiination 'rime of moasurpinent llalf-lifc of eflfect Sien of effect l)u\s('iis 520 Direct (cross illii- niination) During illumination ? + Witt 520 Direct (cross illu- 1 )uriiig C'l. V.ooo + mination) flash illumination sec." Present 520 I )irpct During ilhmiination Several sec." + Present 520 Circulating system — illuminated for 30 sec. After illumination Several sec. Present 520 Circulating system — illuminated for ca. 0.2 sec. After illumination ? + Duysens 480 Direct (cross) illumination During illumination ? Witt 480 Direct (flash) During and after illumi- nation Ca. Vioosec. Present 480 Direct (cross) illumination During ilhmiination Several sec. Present 480 Circulating system — illuminated for 30 sec. After illumination Several sec. + Present 480 Circulating system - — illuminated for 0.2 sec. After illumination 9 Lundeg&rdh 555 Circulating system After illumination ? Present 555 Circulating system — 30 sec. illumi- nation After illumination Several sec. + Present 648 Circulating system — 30 sec. illumi- nation After illumination Several sec. Present 648 Circulating system — 0.2 sec. iUumi- nation After illumination ? Present 660 Circulating sj'stem — 30 sec. illumi- nation After illumination Several sec. " Note: The ca. Vioo sec. half-life is reminiscent of both the Emerson-Arnold time and the short decay luminescence described elsewhere in this volume, whereas the long-lived change in absorption is analogous to the long-lived lumi- nescence. ABSORPTION SPECTRUM CHANGES 97 duced photochemically and is also in chemical equilibrium, in the dark and through enzymatic reactions, with both the photoproduced oxidant and other redox systems in the algae. The 648-m;u band like- wise does not appear to he due to chlorophyll a; it may be a chloro- phyll h derivative but it is also possible that it represents some other type of compoTuid, for example, a cytochrome. Possibly it is the photo- produced oxidant, a precursor of molecular oxygen. Fourth, a partial thermal inactivation produces a change in re- sponse characterized by increases in absorption over the entire wave- length region observed. We cannot account for this phenomenon at present. One cannot say, presently, with any degree of certainty what the compounds involved are. The results reported here and in the earlier literature are tabulated in Table I. To summarize, the simplest interpretation of the present findings consistent with earlier work on luminescence and absorption spec- troscopy is the following : 1 . The absorption bands at 480 and 520 m/z are due, respectively, to the oxidized (X) and reduced (XH) states of the same compound. The assignment of the 520 band to the reduced state follows (a) from Witt's observations on the effect of added hydrogen acceptors which depress the 520 changes and (6) from the fact that the absorption falls below the steady dark level immediately following illumination and then, in the succeeding dark period, rises to the normal dark level. Since it is unlikely that the increase is due to the accumulation in the dark of precursors of molecular oxygen similar to those produced photochemically and since one would a priori expect the primary re- ducing agent to be in djmamic chemical equilibrium with other metabolic pools of hydrogen in the light as well as the dark, it is probable that the 520-m/i band is due to a reduced compound. 2. The 648-m/x band may represent the oxidized form of the pre- cursor of Oo, that is, the photochemically produced oxidant (YOH). S. During illumination there is always an increase in the concentra- tion of XH over its steady dark value, but in the succeeding dark period it may fall below the dark level at steady state, because XH can be consumed by a concurrently produced oxidant. Such a picture is also adequate to explain the qualitative aspects of the induction effects in absorption spectrum changes and in luminescence. Such an inequality between reductant and oxidant could arise be- 98 B. L. STUEIILEI! AND V. IT. LYNCH cause the rediictant i.s presuma])l.v being consumed ])y othei- hydrogen acceptors in the photosynthetic clieniical chain while the oxidant disappears mainly in a reaction producing molecular oxygen. If the rates for disappearance of YOH are such that the steady-state concen- tration of oxidant in prolonged light is higher than that of the re- ductant the changes observed would occur. 4. There should always be an increase in reductant during the initial phase of illumination because the reductant and oxidant are approximately equal in concentration at the onset of photochemistry. These assumptions can explain all the observed results with normal algae. Schematically the relationships postulated are as follows: X(475) + Y(660) + HoO + hv Ch 520 + 648 XH YOH P.S. Luminescence ^ O2 Production ^ Photosynthesis X + Y + H2O + energy 4XH] [YOH] -'[YOH]" -[Z] [XH] [A?] /4O2 Acknowledgment. This work was performed while the senior author was a visiting investigator at the Carnegie Institution of Washington, Department of Plant Biology, Stanford, California, during the summer of 1955 and was supported in part by a Grant from the U.S. Atomic Energy Commission. The authors wish gratefully to acknowledge the helpful assistance of Mr. L. R. Kruger in the building and modification of apparatuses and the per- tinent and useful guidance and comments of Drs. C. S. French and James H. C. Smith. Discussion Brown : Does this hot bath cause any precipitation? Strehler: Yes, I think so because we got occasional noninstrumental noise pulses, probably due to the interruption of the measuring beam by agglomerates of cells after heating. Frenkel : I don't quite understand how Dr. Strehler's theory can account for emission. Of course, you assume that the light that is reemitted comes from a substance closely related to chlorophyll. ABSORPTION SPECTRUM CHANGES 99 Strehler: I am quite certain it is emitted from chlorophyll, but that sa3's noth- ing about whether the energy for exciting it comes from a reaction involving the chlorophyll molecule. Frenkel: Then you could have a recombination of two molecules other than chlorophyll? Strehler: I didn't say that both of them were other than chlorophyll, did I? Duysens: I maj^ perhaps make a suggestion. If reduced riboflavin in the chloroplast tries to emit light, it will be unable to emit it because the chlorophyll traps the excitation energy and you will get chlorophyll a luminescence. So it is quite possible there is luminescence of riboflavin. Strehler: Riboflavin is known to be involved in other luminescences as the emitting molecule. Light emitted by flavin could certainly be trapped by an}^ other absorber present, and what is a more likely absorber in green plants than chlorophyll? I hope j'ou don't get the impression that I have evidence for this hypothesis. I am simply saying that this is another possibility. Rabinowitch : It sounds to me like why make it simple when it can be made complicated. Strehler: In support of the hypothesis it has been shown with firefly extracts that added fluorescent dyes in a relatively low concentration change completely the color of the light emitted by firefly extracts. If one chemiluminescent system can do it, certainly another can. Becker: You can talk about this until you are blue in the face, but I think it is important that you realize you cannot say anything about this energy transfer unless you realize that the molecules that are going to receive it have energy values of the same comparable value. Duysens: They are there. The fluorescence spectrum of riboflavin overlaps the continuous absorption spectrum of chlorophyll. Strehler: The riboflavin emission band is quite broad. It extends from about 485 to 620 m;u, and, although this region does not encompass a major band of chlorophyll, there certainlj^ is enough absorption there, particularly if the mole- cules are in intimate association, to permit energy transfer. References 1. Strehler, B., and Arnold, W., J. Gen. Physiol, 34, 809 (1951). 2. Chance, B., /. Biol. Chem., 202, 397 (1953). 3. Duysens, L., Science, 120, 353 (1954). 4. Lundeg&rdh, H., P/i?/sio/.PtoMtomm, 7, 375 (1954). 5. Witt, H., Nalurwiss., 3, 72 (1955). 6. French, C, Carnegie Institution of Washington Year Book, 53, 182 (1954). 7. Strehler, B. L., and Lynch, V. H., Arch. Biochem. and Biophys., 1957, in press. Selective Scattering of Light by Pigment-Containing Plant Cells* PAUL LATIMER and EUGENE L RABIXOWITCH, Photosynthesis Re- search Project, Department of Botany, University of Illinois, Urbana, Illinois We have observed a strong spectral selectivity in the scattering of light by pigmented algal cells. Sharp maxima in the scattering occur on the long- wavelength side of absorption bands (c/. Figs. 2, 3, 4). The experimental apparatus is shown diagrammatically in Fig. L The cell suspensions were illuminated with light from a Bausch and Lomb grating monochromator (3.3-m/i band half-width). The in- tensity of light scattered at approximately 90° (more precisely, 75° to 105°) to the incident beam, 7^, was measured with a photo- multiplier tube and was compared with that of the incident light, /o, by replacing the cell suspension with a white (MgO) surface. The scattered light collected by the measuring device was of the order of 0.1% of the incident light. In the range of cell concentrations used (corresponding to optical densities of ~0.01 to 0.05 per 1-cm. path in the maxima of the absorption bands), the amount of scattering per cell did not vary with cell concentration. A correction was made for attenuation of the incident and scat- tered beams in the suspension by cells other than the primary scatter- ing cell itself. This was accomplished by attenuating the incident light in the measurement of /o by means of a cell suspension of prop- erly adjusted optical density which was placed immediately before the MgO surface. No correction was made, however, for absorption within the scattering cell itself, since this would require a detailed knowledge of the structure and optical properties of the cell. The asymmetry of the scattering curves in relation to the absorption curves clearly shows, however, that the presence of scattering maxima * This work was carried out under a grant fiom the Office of Naval Research with some assistance from tlie National Science Foundation Grant 1398 (Dr. Emerson). The paper is based on a dissertation submitted by Paul Latimer in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Graduate College of the University of Illinois, 1956. 100 SELECTIVE SCATTERING OF LIGHT 101 is not primarily due to selective absorption of nonselectively scattered light within the scattering cell. cells which scotter observed light Suspension vessel (cleor plexiglos, outside surfaces painted black 25/e" i. Position of filter for fluorescence correction meosurements Photomultiplier tube Fig. 1. Arrangement of apparatus for measuring scattered light as a function of wavelength. I 5 - .m - LlI < o - • — • SCATTERING, LIVE GREEN CELLS J X — K SCATTERING, BLEACHED CELLS I - ■^x c — ABSORPTION, GREEN CELLS \ - — X — - \ A X r ^. / \ - \ '^x / \ r\ ~ -\ \ / \ / \ ~ -\ X [ \ 1 \ ~ :\ ^ ''~'\ V [ \ - - X. . \ \ 1 T - ^x / - Ny "*~"Nc-x. / \ - - >w x*.^ J \ _ ^V^^^^x-x J \^^_ /\ y ^ \ ^-•-»_^''~x-x^x-x^x^ - / L v^ - / - ^ f\ '- y >> o \ /^ \ ^^ a. o c 4> \ / o" \ " ~ o o \> — "oo*^ O \ ~ ~ x: o Ta rtOOOO***'^^ ^ h ~ 1 1 o 1 1 1 1 1 1 o 1 1 1 1 1 1 1 1 iiTii 1 1 1 1 1 1 1 1 1 1 1 1 iV'iwi i -5 .to 3 -, cc t-i- ^\< CD ^ CD < 360 400 450 650 700 - I 750 500 550 600 WAVE LENGTH (m>i) Fig. 2. Scattering and absorption of the green alga Chlorella. The intensity of light scattered at 90° to the incident beam is shown as a function of wavelength. Absorption was determined with an integrating sphere. The absorption maxima of the several pigments are indicated on the wavelength scale. The bleached cells were obtained by extraction with hot methanol. 102 p. LATIMER AND E. I. RABINOWITCH Some pigment fluorescence could be measured together with the scattered hght. In spectral regions away from the fluorescence bands, colored glass filters were used to prevent this contamination. Close to and in the fluorescence bands, it was necessary to measure the sum of the scattered light and fluorescence, and to correct for fluorescence. This correction could be based either on the different spectral com- positions of the two components or on their difi"erent degrees of polari- 400 450 500 550 600 650 WAVE LENGTH m/i. 700 750 Fig. 3. Scattering and absorption of the diatom Navicula minima. zation. The first method failed near the peak of the fluorescence band; the second had no such limitation. To determine the fluorescence correction with colored filters, the transmission factor of a filter for scattered light, Tg, was determined at each wavelength with a white surface in the beam, and that for fluorescence, Tf, by exciting the fluorescence wdth light which ap- propriate filter combinations could prevent from entering the de- tector. From the observed average transmission factor for com- bined scattered light and fluorescence, Ts,f, the fraction of detector response due to scattering, x, was found by means of the eciuation: TsX+ Tr{l~x) = Ts.f This method of determining the fhiorescence correction was used by us with two tj^pes of colored filters — one which preferentially trans- SELECTIVE SCATTERING OF LICJIIT 103 mitt'ed light of longer wavelengths (sharp cut-off red filter), and another which transmitted mainly light of shorter wavelengths (in- frared-absorbing blue filter). Polaroid plates were used in the second method (described b}^ Brice, Nutting, and Hawler (1)). Corrections found in all three ways were in excellent agreement everywhere except at 680 to 700 m/x, where the different methods led to corrected scattering curves which deviated by up to 10% from the average ones shown in the figures. X — • SCATTERING .— ABSORPTION 400 450 500 550 600 650 WAVE LENGTH (m>i) 700 750 Fig. 4. Scattering and absorption of the blue-green alga Synechocysiis. Investigations of crystalHne chlorophyll by Jacobs (2) and of suspensions of colloidal chlorophyll by one of the authors had re- \-ealed that these systems scatter light with a spectral selecti-\-ity similar to that shown in our (;urves for algal, cells. The light scattered by the cells originates from two sources: (/ ) colorless structures, which scatter with a relatively uniform wave- length dependence (not unlike the scattering by bleached cells, shown in Fig. 2), and (2) highly pigmented chloroplasts (or grana), which scatter light with a strong spectral selectivity. The (juantitative in- terpretation of the scattering by pigmented cellular components re- quires a rather complex theory, such as that of Mie (3). Qualitati\-ely, the capacity to scatter light should increase sharply with the index 104 p. LATIMER AND E. I. RABINOWITCH of refraction. According to the classical theory of dispersion, the index of refraction of an absorbing medium should reach a maximum on the low-frequency side of an absorption band. This is in fact where scattering maxima have been observed by us. A similar spectral selectivity has been reported recently by Goed- heer (4) and by Menke and Menke (5) in the double refraction of chloroplasts from Mougeotia. The quantity which was measured i]i those cases is the difference between the indi(;es of refraction of the material for light beams polarized in two different planes; scattering depejids on the average index of refraction. The scattering of protons by atomic nuclei shows the same type of selectivity for proton energies in the neighborhood of the resonance levels (6). Further study of selective scattering may supply information about the packing and arrangement of pigment molecules in biological systems, and may even reveal the presence of pigments which are not easily found by other techniques. As an example, it is seen in Fig. 2 that the selective scattering produced by the carotenoids in Chlorclla is much more conspicuous than their contribution to the absorption spectra. The latter is so effectively obscured by chlorophyll absorption that the very existence of carotenoids as such in green cells has been doubted (7). We wish to thank Professor James Franck and Professor Robert Emerson for helpful suggestions and Ruth V. Chalmers for growing the algal cells. TRANSMISSION CHANGES IN ILLUMINATED CELL SUSPENSIONS We have heard several reports describing small changes in the trans- mission of collimated beams by cell suspensions. It may be worth while to draw attention to the fact that changes in scattering, as well as changes in true absorption, could lead to observations of this type. The figures in the present paper show that, in some regions, light scattered by cells varies shai-ply with wavelength. The physical struc- ture of a cell, in which the selective scatteruig originates, may ])e altered by exposure to light, or by chemical agents. In either case, changes in selective scattering could result and lead to selective changes in transmission. Some investigators have found variations in transmission after changes in temperature (Witt), or after the addi- I SELECTIVE SCATTERINf} OP LHillT lOf) tioti of ITCN (Bishop), which they attribute to changes in scattering. We can also refer to tlie report of Shil)ata (see page 171). Th(^ portion of the nnabsorl)e(I light which does not reach the detector l)ecaiise of scattering, depenrls on the geometry of the apparatus and on (he absorption of light by (he cells. Actually, the contril)ution of scattering to the total attenuation of the; incid(Mit beam often is of (he same order of magnitude as that of absorption. Therefore, some of the reported variations in transmission, which have been interpreted as changes in absorption, could be the result of equally small relative changes in scattering. Experimental evi- dence is needed to eliminate this possibility or to evaluate the con- tributions of absorption and scattering to the observed effect. Discussion Weigl : Similar, although .somewhat clearer-cut, anomalou.s dispersion curves have been ohsorved in specular reftectioti from solid dyes. Rabinowitch : \V hat particularly interested me in this work is that one gets such strong effects from jninor pigments. The example of carotenoids shows that one can detect the presence of some pigm(;nts easier by scattering than by absorp- tion. One sees scattering peaks wheie ones doesn't .see even a shoulder on the ab- sorption curve. Perhaps, when an adecjuate theory is worked out, one will be able to make conclusions as to the state of the different pigments — how densely they are packed, etc. I hope, therefore, that others will try to play with this effect, both theoretically and expeiimentally. Another remark that I want to make does go back to the first Gatlinburg con- ference. Three years ago, when Jacobs presented our study of the crystal spectra. Kasha and Commoner attacked his conclu.sions rather sharply, saying that the band shift we had observed may be due entirely to .selective scattering. We were convinced that this was not .so, but could not jjrove it at that time by direct ex- perimental evidence; so the criticism scared us a little and caused us to look more closely into the cjucstion of scattering, first in crystal suspen.sions and subse- quently in live cells. We found that our crystal spectra were not .significantly affected by .scattering, but that selective scattering did in fact occur^both in crystals and in living organisms. Latimer's study is thus a consequence of the di.scussions at the preceding Gat- linburg conference. Chance: Anomalous dispersion effects, of course, are not unique to Chlorella. An anomalous band at 370 m/z is well known in erythrocytes, and I think it is largely exjilained by a rapid change of the refractive index. With regard to Jiatimer's remarks concerning the possible role of .scattering in measurements of .small absorption changes, we have worried about this when measuring intracellular oi)tical density changes. We made light scattering and transmission comparisons, which showed that, in the cytochrome system^ 106 1'. LATIMER AND E. I. KABIKOWITCH effect of scattering is negligible, since the pigment concentration is not sufficient to give steep gradients of the refractive index. We think therefore that our meas- urements were not influenced by scattering. As a further control, one can add a pigment of known optical properties to the scattering suspension and see if the scattering distorts its spectrum. (We used cytochrome for this purpose.) One can then add a detergent which will clarify the suspension; with whole cells, one may also add lysogen, if they are susceptible to its action. We have done this in every case, and compared the spectra before adding the chelating agent and after, and the spectra before and after adding lysogen. Chemical tests also are of assist- ance: in Rhodospinlluni, where phototaxis occurs, we would be worried, if it were not for the fact that we did get similar effects with oxygen as with light, and that the spectral changes observed corresponded to those in isolated cyto- chromes. I completely agree with Latimer's remarks when it comes to small transmission changes in highly pigmented particles, e.g., Chlorella. Benson: I think that the technique Dr. Shibata has developed, using opal glass, may be useful in testing for this effect. Latimer: Instead of using opal glass, which collects most of the scattered light, one could also use an integrating sphere, which collects all of it. Frenkel: Dr. Shibata, did your results with opal glass and with the sphere agree? Shibata : I believe that you can get better results using the opal glass, although the differences are small. The opal glass method requires no special instruments. James Smith : In very fine suspensions of protochlorophyll in its natural state we got no improvement at all with opal glass. Do you think that this was due to the smallness of the particles? Shibata : I think so. Duysens: I should like to remark that Latimer measured light perpendicular to the incident beam. For most dilute suspensions, 95% of scattering is in the direction of the beam. Latimer: I find the same effect at angles from 45 to 135 degrees. Duysens: Most of the scattering occurs within 35 degrees (or less) to the inci- dent beam. References 1. Brice, B. A., Nutting, G. C, and Hawler, M., J. Am. Chem. Soc, 75, 824 (1953). 2. Jacobs, E. E., and Holt, A. S., /. Chem. Phys., 20, 1326 (1952). 3. Mie, G., Ann. Physik, 25, 'ill (1908). 4. Goedheer, J. C., Biochim. et Biophys. Acta, 16, 471 (1955). 5. Menke, W., and Menke, G., Z. Naturforsch., 10b, 416 (1955). 6. Bender, R. S., Shoemaker, F. C, Kaufman, S. G., and Bouricius, G. M. B., Phys. Rev., 76, 213 (1949). 7. Lubimenko, V. N., Rev. g&n. botan., 39, 547 (1927). The Absolute Quantum Yields of Fluorescence of Photosynthetically Active Pigments* PAUL LATIMER, THOMAS T. BANNISTER, and EUGENE I. RABINOWITCH, Photosynthesis Research Project, Department of Botany, University oj Illinois, Urbana, Illinois Knowledge of the fluorescence yields of photosynthetically active pigments is important for the understanding of their photochemical activity and the probability of resonance energy transfer between these molecules in the hving cell. We have measured the absolute quantum yields of fluorescence of chlorophyll a, ethyl chlorophyllide a, phycocyanin and phycoerythrin in solution, and of chlorophyll a and phycocyanin in the living cell, using the integrating sphere techniciue (1). Prins (2) found the quantum yield of fluorescence, ^, of chlorophyll (a + 6?) in solution to be about 10%, whereas Forster and Livingston (1) found that of chlorophyll a to be about 25%. Wassink and co- workers (3,4) reported, for chlorophyll in various organisms, cp values of 0.1% to 0.3%. Arnold and Oppenheimer (5) estimated, for phy- cocyanin in solution,

. LATIMEK, T. T. BANNISTER, E. I. RABINO WITCH and Norris (7). The phycobilin pigments were kept at about 5°C. during the purification and the fluorescence measurements. Fluores- cein (Eastman, Reagent Grade) was used without further purification. The algal cells, upon remov^al from the culture line, were sus- pended in carbonate buffer #9 and exposed for V2 to 2 hours to visible white light, with an intensity of about 3000 ergs/ (cm. ^ sec.) (^-100 lux). Less than 30 seconds elapsed between the removal of the cells from this light and the measurements — a procedure which eliminated well known transients in the intensity of fluorescence following a dark period. The temperature of the live cells and of pigments other than phycobilins was about 25°C. Prof. Robert Emerson and Mrs. Ruth V. Golvonometer Photomulliplier tube (RCA 6217) Filter opaque to exciting light Baffle Vessel containing Fluorescing material White matt surface-* ^Grating Monochromotor Fluorescence' { Bausch and Lomb) Ulbricht or Integrating Sphere Fig. 1. Arrangement of apparatus for measuring fluorescence. Chalmers* grew the algal cells and determined the light intensities corresponding to compensation and to one-half saturation of photo- synthesis. DilTerent cells were used in the studies of fluorescence and photosynthesis, but they were of the same strain and identically cultured. The light intensities were measured by means of a photomultiplier tube (RCA 6217): The spectral sensitivity curve of the detector sys- tem was determined with a bolometer. By measuring the response of the tube to monochromatic light projected into the sphere, rather than to light falling directly on the tube, we obtained a calibration which accounted not only for the selective sensitivity of the tube but also for changes in the reflectivity of the sphere walls with wave- length. The apparatus is shown in Fig. 1. As an indirect check of this calibration, we measured the fluores- Working on Xiitional Science Foundation Contract G-Ky!)S. ABSOLUTE QUANTIAf YIELDS OF FLUORESCENCE 109 cence yield of chlorophyll a and of phycocyanin in solution as a func- tion of the wavelength of the exciting light. For each of the two pigments, the quantum yields were found to be constant (within ±6%) for excitation between 405 m/x and a wavelength slightly beyond the red absorption maximum. The similarity of the results obtained with the two dissimilar pigments and the agreement of these findings with those obtained by Forster and Livingston in the study of chlorophyll fluorescence may be regarded as indirect evidence that there were no serious errors in the calibration curve. P3,0r o ^ 20 > E c o a O 1.0- 4 Fluorescence observed through Schott RG9 • . ... RG5 o • ■ ■ Corning 2403 I I I I I I 1 4 6 8 10 Absorption (Loglo/I) 12 Fig. 2. Fluorescence of Chlorella cells. Quantum jdeld as function of cell concentration. (Log ly/I is propor- tional to concentration. Path length of exciting beam = 51 mm.) Yields extrapolated linearly to infinite dilu- tion by method of least squares. Average intensity of exciting light = 50 ergs/(cm.^ sec). Xex. = '436 m/i. 350 40- •s 3 e20 lO- H "i I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I ''-lO I 23456 Log Incident Energy (ergs/cm^ sec ) Fig. 3. Fluorescence of Chlorella cells. Fluorescence as function of intensity of exciting beam (averaged over its path in vessel). X„x. = 436 m/*. The average sensitivity of the detector to the fluorescence emitted by the different pigments and cells was determined from the calibra- tion curve of the detector system and the known fluorescence spec- tra. In order to correct for the reabsorption of fluorescence, the yields of fluorescence of several suspensions (or solutions) of different con- centration were measured, and the results extrapolated to infinite dilution. This method permits an accurate correction for the re- absorption of the fluorescence in solutions^ but it does not account for self-absorption within the emitting cells in cell suspensions. The latter 110 p. LATIMER, T. T. BANNISTER, E. I. RABINOWITOH tD 03 0) w (V t) GO ll o 3 PQ H E CO •?-■ 3 >> s. • r3 n CO « T^ o aj ^^ • c 5 c — O OQ CO 03 e ^^^ ,— V '— ~ (S 1-H o -^ ^ ^ c 1-H '■^ c T3 3 ■5}< o ^1 3 m 3 b£ >< s bC rH [/3 C 3 ■IM CO o3 ^ 05 h^l ^ -a T3 ^ — ■o *^ 3 C 00 T tH "2 a 3 Id GQ 03 > =2 tD C? £ 3 oo t-. CO o =2 S -C 3 tc C -3 -1^ s -Hi CO -3 42 ^ o s 1^ 3 o 'ih 3 ^ Oi ^ 3 ^ < fe ^x^o < -< fe p- < -< > << < < oc o iC t-. O CO 00 <— « 1 , >. >, 43 43 w w a; CO a; o 3 3 O e 3 0) O CO --H 3 r^, -3 ^-^ ' — ^ O. 3 e O O o >^ 43 43 UJ r> >> 3 n 73 ^— ' 43 W o CO CO = - ^ ^ = CO CO 1 3 3 42 "tS — -^ d CO 43 (. 'S e . .CO a, Ss ^ 'S Oh *j2 X s. "3 - : g CO §. O 43 3 1 'b" 1^ CD C fe:^ ^ o a ,_^ CO "S o « _> i.s _B si "— - e 43 o o w S> IS ^^ §3 0) 43 3 » -a o3 3 < O O 43 -1-3 3 <3 o QO u a 3 42 03 43 a O 43 JirS; CO O 3. s ^ .2 ?* agog •i 'c -3 :§ • 0) Hi 43 3 a> Q. Q. o 3 03 43 O 3 Ih CO --I CO o in CO CO 55 to O u ^ 00 O i^- O Pk ■^ 43 Oh >^0 ^Q, Ph > 1=' r. ■ I I' 3 03 O ^ ^43 Oh 2 o 43 § 42 -S 03 B CO 3 o (h o a, o3 Ih X ABSOLUTE QUANTUM YIELDS OF FLUORESCENCE 1 1 1 correction was evaluated by using various sharp cut-off red filters to separate the fluorescence from the incident light, thus obtaining v^alues for fluorescence differently influenced by self-absorption. Three filters of various shades of red were used. Total fluorescence yields were computed by dividing the value obtained with each filter, by the fraction of the total fluorescence transmitted by it (de- termined from the fluorescence spectrum and the transmission curve of the filter) . The three values of the total yield at zero concentration obtained in this way can be in turn extrapolated to zero reabsorption (by plotting them as a linear function of the slopes of the straight lines fitting the three sets of experimental points). We arrive in this way at a \'alue of ^ unaff'ected by all forms of self -absorption, "internal" as well as "external." The quantum yields of fluorescence of chlorophyll in Chlorella, excited with an intensity of 50 ergs/(cm.^ sec.) at X = 436 m^u are plotted in Fig. 2 as a function of the concentration of the algal sus- pension. Using the above-described procedure to correct for self- absorption, we obtained for the limiting quantum yield of fluores- cence, at this intensity of excitation, a value of 2.7% — about ten times that reported by Wassink and co-workers — confirming the pre- diction of Duysens. The time available for the migration of excitation energy between chlorophyll molecules is, therefore, ten times as long as it had been assumed to be on the basis of earlier data. The quantum yield of fluorescence of chlorophyll in plant cells has been generall}'' assumed to be independent of the intensity of the ex- citing light in w^eak light, but to increase at high, photosynthesis- saturating intensities (12 j. We have found, however, that the yield is also a function of the intensity of the exciting light at intensities as low as 0.01 of that required for compensation of respiration by photo- synthesis. Figure 3 indicates that fluorescence yield, and thus also the average lifetime of excited chloroph^'ll in cells, is about 50% longer at compensation than it is at very low mtensities. (Similar observa- tions were reported by Brugger at this Symposiinu.) References 1. Forster, L. 8., and Livingston, R., J. Chem. Phys., 20, 1315 (1952). 2. Prins, J. A., Nature, 134, -157 (1934). 3. Vermeulen, D., Wassink, E. C, and Reman, G. H., Enzymohgia, 4, 254 (1937). 4. Wassink, E. C, and Kersten, J. A. H., Enzymologia, 11, 282 (1944). 112 p. LATIMER, T. T. BANNISTER, E. I. RABINOWITCH 6. Arnold, W., and Oppenheimer, J. R., J. Gen. Physiol, 33, 423 (1950). 6. Duysens, L. N. M., Thesis, University of Utrecht, 1952. 7. Haxo, F., O'hEocha, C, and Norris, P., Arch. Biochem. and Biophys., 54, 163 (1955). 8. Vavilov, S. I., Z. Phy.'iik, 22, 266 (1924). 9. Hellstrom, H., Arkiv Kemi Mineral. GeoL, A12, 17 (1937). 10. Uml)erger, J., and LaMer, V. K., /. Am. Chem. Soc, 67, 1099 (1945). 11. Ghosh, I. C, and Sen-Gupta, S. B., Z. physik. Chem., B4I, 117 (1938). 12. Rabinowitch, E. I., Photosynthesis, Vol. 2, Part 1, p. 1047. Interscience, New York, 1951. Fluorescence Yield of Chlorophyll in Chlorella as a Function of Light Intensity* JOHN E. BRUGGER, University of Chicago (Pels Fund), Chicago, Illinois It is known that Franck's explanation of the so-called Kok effect has been based on the assumption that photochemical reduction of intermediates of respiration is involved. The question arises whether there might be a corresponding effect on the fluorescence. In examin- ing the published curves of McAlister, we found that it is un- certain whether the data for fluorescence intensity versus irradia- tion extrapolate linearly to the origin. It was decided to investigate this point further. We have found that the fluorescence yield is in- deed lower in the intensity range below compensation of respiration than in the region between compensation and saturation, but no attempt has been made to find out whether the transition of the yield occurs exactly at compensation. This effect varies in its strength with external conditions, as does the Kok effect. However, unlike the somewhat erratic appearance of the latter, the break of the fluores- cence yield in the neighborhood of compensation was visible in all our curves so long as photosynthetic activity was not suppressed. An incandescent lamp with blue filters was the source of irradia- tion. Light intensity was varied by inserting neutral density filters in the beam. Fluorescence was measured with a 1P22 multiplier photo- tube equipped with an appropriate red filter. The photosignal was amplified and recorded. The light-detecting system responded linearly with signal input. To locate compensation and saturation intensities on the illumination scale, a modified Warburg manometer was used with which chlorophyll fluorescence as well as oxygen evolu- tion could be followed simultaneously. In most measurements, the manometer was replaced by a sample holder having the same geometry but fitted with a sintered glass bottom so that various gas mixtures could be passed through the suspension. One-milliliter samples of -^0.02% Chlorella were used. These had a transmission (1-cm. cuvette) of 55% at 4400 A and 85% at 7800 A. The algae * This work was partially sponsored by the Office of Naval Research. 113 114 J. E. BRUGGER were usually suspended in water, though buffers and other solutions were also used. Water was preferred because it simplified the problem of attaining an equiUbrium with the flushing gas and thus facilitated the changing from one gas mixture to another. COg °2 N2 / Y- - 100 / y lY- - 0.5 - 99.5 Y > / y/^ m- 2 - - 98.0 y^ U- 2 - 0.5 97.5 y^ I - 4 20 - 76.0 • • /- I y^ ^.-^ y^ ^ ^ ^ X" S^^ •f " C^ ""^ IRRADIATION INTENSITY Fig. 1. Fluorescence intensity versus irradiation intensit}- for Chlorella in water swept with various gas mixtures. 4 The anomalies of fluorescence yield during induction periods, pre- viously reported by others, were observed. However, all measure- ments reported here are steady-state values. Generally the fluores- cence was measured by starting at low intensities of irradiation and FLUORESCENCE YIELD OF CHLOROPHYLL 115 o o CO UJ q: o 3 proceeding stepwise to higher ones. When the intensities used were not too great, it was possible to retrace the curve by going from high to low intensities and also to check random points. Figure 1 shows a fluorescence intensity versus irradiation trace tor Chlorella in water swept continuously with 4% carbon dioxide in air. There is a break in the curve at low intensities. One observes that LOW CO2 HIGH COc C= Region of Compensotion S= Onset of Saturation IRRADIATION Fig. 2. Fluorescence intensity versus irradiation intensity for Chlorella in water swept with ordinaty and COa-enriched air. the curve between compensatioii and saturation, if extrapolated, would pass below zero on the ordinate scale. There is some uncer- tainty about the exact shape of the curve at low intensities — most probably there is a gently decreasing slope with diminishing irradia- tion, but a relatively sharp bend in the region of compensation. The slope at approximately half compensation can be seen to be roughly half that of the value between compensation and saturation. Above saturation, the well-known rise of the fluorescence was observed. The slope below compensation was somewhat influenced by the IIG J. E. BRUGGER culturiiig conditions, previous history of the sample, etc. Algae cul- tured in a glucose-rich medium showed a lowered initial section of the curve. Algae grown in Knop's solution with ordinary air, as well as those cultured in an NH4"*"-rich medium, showed steeper slopes than the algae grown with COj-enriched air. No extended study was made of the effect of the medium in which the algae were suspended during the measurement. Work is being continued along these lines, however. It was found most convenient to store the algae in the dark in carbon dioxide-free air — in this way several experiments could be done on the same batch of algae. For control, the first set of measurements was repeated at the end. When one compares the curves obtained for 4% carbon dioxide in air with those in unenriched air, he observes that (Fig. 2) the principal difference is the earlier onset of saturation. When the car- bon dioxide level is maintained at 4% but the oxygen concentration is reduced (Fig. 1), one observes that the slope at high intensities be- comes steeper and the break in the curve occurs progressively earlier. At the same time the initial slope (below compensation) becomes steeper. In the curve (Fig. 1) for 2% carbon dioxide in nitrogen (oxygen less than 10~^ mm. Hg), there is essentially a merging of the two breaks observed in the curve measured in 4% carbon dioxide in air. The shape of these curves depends on the degree of anaerobiosis one maintains. It was necessary to use very high flow rates for the CO2-N2 sweeping gas. When carbon dioxide is entirely absent but the nitrogen contains 0.5% oxygen, the initial slope is only slightly different from the slope at intermediate and higher intensities. The position of the bend varied with the time the algae were exposed to the gas mixture. In nitrogen containing no carbon dioxide or oxygen (less than 10 ~^ mm. Hg), a linear dependence of fluorescence intensity on irradiation is obtained. Low temperature increased the fluorescence and straightened the curves. The results obtained under irradiation with green and yellow light did not appear to differ significantly from those obtained with blue Hght. In addition qualitatively similar results were obtained with Scenedesmus ohliquus D3 and Ankistrodesmus braunii. Studies of the effects of poisons have not been completed and will not be discussed in this paper. Among those tried — phenylurethane, iodoacetate, cyanide, carbon monoxide — the effects produced were understandable within the framework of the interpretation which follows. In general, FLl^OKESCENCE YIELD OF CHLOROPHYLL 117 in the presence of poisons, the fluorescence intensity was increased and the })ends tended to disappear. A more complete presentation of these experiments will be made at a later time. The theoretical conclusions are incorporated in Dr. I'ranck's paper on the "Photochemical Part of Photosynthesis" (seepage 142). Discussion Lumry : I should like to mention some of our preliminary^ results. We have been trying to find out if the fluorescence yield approaches zero as the light intensity ap- proaches zero. Our measurements suggest that the yield changes with light in- tensity at much lower intensities than has been previously thought. Introductory Remarks on the Luminescence of Photosynthetic Organisms* BERNARD L. STREHLER, Biochemistry Departme7it, University of Chicago {Fels Fund), Chicago, Illinois In 19ol it was discovered (1) that green plants emit a low inten- sity chemikmiinescence when they are illuminated. This luminescence persists for some time (up to 200 minutes) after the plants are re- moved from the light (2). Inasmuch as this process undoubtedly represents a minute reversal of early steps in the conversion of light energy to chemical energy and since it is easily measurable, the phenomenon furnishes a useful tool for the study of the photochem- istry of photosynthesis. Moreover, a solution of the intermediary chemistry leading to this bioluminescence in all likelihood will also have a direct bearing on the understanding of its converse process, photosynthesis. The salient features of this phenomenon are as follows. PHYSICAL PARAMETERS OF PROCESS 1 . The emitting molecule is chlorophyll a, as is shown by the color of the emitted light, which is identical within experimental error to the fluorescent light emitted (owing to chlorophyll a) by green plants (3). 2. The efficiency of different wavelengths of exciting light in pro- ducing luminescence parallels their efficiency in promoting photo- synthesis (1). The action spectra of green plants (Chlorella) parallels chlorophyll absorption, whereas in certain blue-green algae the ac- tion spectrum follows that of phycocyanin absorption (4). Tn this case fight absorbed by chlorophyll is inactive (or highly inefficient) in promoting either photosynthesis (5) or chemilumescence, although the emitting molecule is chlorophyll (4). 3. Brown, red, green and blue-green algae and all green plants tested (about 50 different species, including conifers), exhibit the phe- * This work was supported in part by a grant from the United States Atomic Energy Commission. 118 LUMINESCENCE OF PHOTOSYNTHETIC ORGANISMS 119 nomenon (4,6). It has also been shown that photosynthetic bacteria emit a hnninescence of a similar nature at longer wavelengths, al- though this process has not been intensively investigated (4,7). 4. The intensity of the luminescence is quite low — of the order 10~* to 10""'^ of the illuminating intensity below saturation, i.e., ca. 10~^ to 10~^ of the fluorescence intensity. EVIDENCE ON THE ENZYMATIC NATURE OF THE PROCESS 1 . The long-lived luminescence increases with increasing illuminat- ing intensity up to a certain value and thereafter remains essentially constant; i.e., it displays a typical saturation as does an enzymatic reaction (1). 2. At unphysiological temperatures the luminescence is rever- sibly destroyed if the heat treatment is brief. Exposure to 45° to 50° C. for a few minutes irreversibly destroys the luminescence (1). 5. Rate of luminescence as a function of temperature shows a typi- cal enzymatic temperature optimum at ca. 37 °C. (the temperature optimum for photosynthesis) with a heat of activation for the process of CO. 10 to 15 kcal. (1). That this activation energy is characteristic of the limiinescent substrate was established by measuring in a flow system the temperature dependence of the emission process after illumination at different temperatures (1). The intensity of light emission was strongly dependent on the measuring temperature and essentially independent of the temperature at which the plants were illuminated (between 0° and 40°C.). That the formation of luminescent substrate does involve some activation energy was shown by illumination at liquid nitrogen temperature, after which no luminescence is observable upon thawing. 4. Various metabolic inhibitors, including V.\. light, produce strong changes in luminescence (1,8,9). DECAY CHARACTERISTICS 1. The luminescence does not decay according to any simple kinetic formulation (1,2,6). About one-half of the luminescence dis- appears in 0.1 second, after w^hich it decays more slowly. 2. At least two definite components are distinguishable in the decay curves (see Fig. 1) : a. A fast decaying component of ca. 0.01 second duration that 120 li. L. STREHLER does not appear to saturate at high Ught intensity (note resemblance to Emerson-Arnold time). h. A slowly decaying limiinescence that reaches saturation at moderate incident light intensity. 8. The time course of the luminescence induction period is dif- ferent for different portions of the decay curve (6). The long-term 180 Milliseconds Time between Illumination and Meosurement 360 Fig. 1. The decay of Chlorella luminescence measured in a phosphoroscope Integrated intensity constant at ca. 300 foot-candles. Temperature, 25°C. decay possesses a marked induction maximum whereas the fast de- cay component shows a much less marked maximum and actually in- creases with time of illumination. RELATION TO PHOTOSYNTHESIS 1. In intact plants the intensity of luminescence is influenced by the presence or absence of CO2 and by anaerobic conditions (1,9). CO2 depresses luminescence, as might be expected if it removes photochemically produced reductant. Prolonged exposure to anaerobic conditions inhibits the luminescence (9) . In a parallel manner, the luminescence of chloroplasts is inhibited by the addition of hydrogen acceptors such as the Hill oxidants, ferricyanide, or quinone (6,9). 2. The luminescence is also exhibited by chloroplasts and saturates parallel to the Hill reaction rate as a function of incident light inten- LUMINESCENCE OF PHOTOSYNTHETIC ORGANISMS 121 sity. The ability of chloroplasts to luminesce disappears during pro- longed storage of chloroplasts at — 20°C., as does the Hill reaction. However, lyophilized powders retain activity apparently indefinitely. 3. With the exception of a gradual increase in the level of lumi- nescence under certain conditions during the first few minutes of illumination chloroplasts do not exhibit any induction effects (6). By contrast, the luminescence of intact plants evidences striking fluctuations in intensity during the first few minutes of illumination (1,9). These induction effects are very similar in their time course to a number of other transients in photosynthesis, e.g., fluorescence (10,11), ATP concentration (12,13), Oo hberation (14), and CO2 fixation (see Fig. 2). 525^ Density CO2 Fixotion Luminescence (Flow System Ca O.Ssec ofteri [ATP] Florescence Luminescence (.0033 seconds after i lluminotion) 12 3 4 Time (Minutes) Fig. 2. A comparison of induction curves for various processes connected with photosynthesis. Note that the time required for steady state to be reached is between 90 and 120 seconds for all these processes. The ATP curve represents concentration, not rate of turnover. The CO2 curve, on the other hand, is cal- culated from the slope of the total fixed C^Oj as a function of time of illumination (Strehler, unpublished experiments). These facts strongly suggest the following conclusions: 1. The luminescence of green plants is due to an enzymatically catalyzed recombination of early photoproducts in photosynthesis, probably the primary reducing and oxidizing agent or the very next reactants in the sequence of O2 liberation and CO2 reduction. 2. Conditions which would be expected to cause an accumulation of these photoproducts increase the luminescence and, converselj^, 122 B. L. STREHLER conditions which should promote iitiUzation or decrease production of these products decrease the intensity of luminescence. 3. The induction curves are due to changes in concentration of a numlier of intermediates in the chain of reactions leading from the photochemical events to CO2 fixation, including the formation and utilization of ATP. 4. The decay curves suggest that two molecular species (probably on the reductant side) are capable of eliciting luminescence. The short-lived component may be regarded as the primary photoproduct and the longer lived component as a substance derived from the first and representing a later hydrogen or OH carrier in the photosyn- thetic sequence. It is interesting that the short- and long-lived com- ponents behave kinetically similarly to the various constants derived from flashing light experiments and thus mixtures of different life- time intermediates may be responsible for the discrepant results reported by various authors (15-18) under differing conditions of illumination, flash duration, etc. Acknowledgments. I wish to acknowledge with thanks many stimulating discussions with Drs. William Arnold, James Franck, and Hans Gaffron on the subject material here summarized. It should be emphasized that some of the detailed interpretations here set forth are in conflict with parallel inter- pretations of Drs. Franck and Brugger of this laboratory, although we are in agreement about the general nature of the process. Discussion Rabinowitch : Does this mean that one curve is first order and the other curve is zero order? Strehler: The best information (Arnold's work) would indicate that, with a very brief flash, one obtains a second-order decay. But, when one uses a longer flash, he obtains something that is neither first nor second order but in between. Wassink: I would like to ask just one question; namely, how does any con- sideration of the triplet state of the chlorophyll come into this? Strehler : I don't think it is necessary, on the basis of the information we have, to postulate anything about triplet states. I think that should await a better understanding of what the natures of the intermediates are. Wassink : If a triplet state is involved, it must come in between your chemistry and luminescence, just as it must for fluorescence. Strehler: I believe so — yes. "Wassink: You don't see any objection? Strehler: I see no objection, but I probably am not (lualified to make the judg- ment. LUMINESCENCE OF PHOTOSYNTHETIC ORGANISMS 123 Bassham : I would like to ask a couple of questions. First, what was the effect of ox^-gen, if any, on the chemihiminescence? Second, does the heat of activation refer to the long component or to the short component or to both? Third, can the short component possibl}- be, rather than chemiluminescence, a long-lived metastable state of some kind in the pigment? Strehler : I am not sure it is possible to give a definite answer to any one of these (luestions. Arnold and I did not find an appreciable effect of anaerobiosis on luminescence. After incubation of the algae for some time under extreme anaerobic conditions, however, interesting effects have been observed, as shown bj- Brugger and Franck. We measured the heat of activation for the long component. Cer- tainly the short component could be due to a long-lived metastable component. Linschitz: Two comments. First, if this luminescence is actually due to some metastable product, I claim the product can not be a molecule in the triplet state, since the radiative lifetime of the triplet would not be of the proper order of magnitude. Second, there may also be an activation energy for the formation of a metastable product. This activation energy may possiblj^ be associated with the falling off of the quantum yield on the long-wave side of the absorption. Strehler : But, as 3'ou pointed out at one time, the sharpness indicates that there is not too much activation energy involved. Linschitz: That is right. Lumry: I wonder if there is any forward reaction or any other reaction by which you could lose high-energy forms. Strehler: Inasmuch as the yield of fluorescence is of the order of 1% of the light absorbed, excited singlet states derived by any other mechanism would probably have the same probability therefore of emitting light, namely 1 in 100. The rate of the reaction that we are studying is probably of the order of 100 times greater than what we deduce directly from the chemiluminescence intensity. Linschitz: In other words, you are saying that if there are any other reactions they are small compared to the reaction — Strehler : No, I w-ould not say that. I think that all of the reactions involved in photosynthesis which make use of the same intermediates would have an influence on the concentration of luminescent intermediates and therefore on the 3-ield of luminescence. There may also be recombinations of primary oxidant and re- ductant at other sites that do not lead to luminescence. Linschitz: T was trying to find out what the activation energy means here. If you have other non-luminescence-producing processes leading to a destruction of your high-energy substances, then the activation energy becomes related to the luminescence in a comphcated way. Strehler: You have emphasized a serious objection to glib interpretations of activation energy calculations. Rabinowitch : 1 wonder how long the fluorescence due to return from the triplet to the singlet state survives. If the activation energy — that is, the difference in energy between triplet and excited singlet — is about 10 kcal., then one can calcu- late how much emission there should be when molecules in the phosphorescent state return to the fluorescent state. Is this "delayed emission" large enough to actually be a comjjonent of luminescence? Of course, the luminescence can bo reabsorbed by the chlorophyll and thus tend to maintain the concentration of 124 B. L. STREHLER metastable states — a partly self-perpetuating reaction which involves no chem- istry. Strehler: That is certainly a possibility. The only thing about trying to deter- mine the rate of back reaction is that one has to assimie that the fluorescent yield from all chlorophyll molecules is identical. It may be that the very molecules that are involved in the photochemistry have a much different quantum yield for fluorescence than do the average — higher or lower. Rabinowitch: Still, did 3'ou ever look whether red phosphorescence occurs in chlorophyll solutions? Strehler: We have never been able to detect with the quantum counter a luminescence of chlorophyll in anaerobic solutions after illumination. There is a chemiluminescence of chlorophyll in solution which I will mention, verj' briefly, later in this session. This luminescence does have a longer lifetime but is probably only of interest from a chemical standpoint. Weigl: If the reactions of this primary product are enzymatic, as they almost surely must be, it ought to be possible somehow to find a specific inhibitor for one or another of the steps and to increase the luminescence by adding inhibitors. Have you had any luck with that? Strehler: Cyanide, azide, phenylurethane — a variety of substances increase the limiinescence up to almost twofold before they begin to inhibit. We have made the argument earlier that this is consistent with the idea of the pile-up of intermedi- ates. Hydroxylamine has a specificity of action which most of the other inhibitors do not have, in that it seems .selectively to depress the long-lived luminescence but to inhibit the short-lived luminescence to a much lesser e.xtent. Gaffron : I remember that at least in the first experiment with hydro.xylamine N'our concentrations were rather high. Do you have one in the physiological range? Strehler : Yes, hydroxylamine is effective at around 10^^ mole in intact Chlorella for inhibiting the long luminescence. Frank Allen : What is the possibility that this luminescence might be coupled with the decomposition of peroxide? Strehler: I think that the primary oxidant might be something perhaps analo- gous to a peroxide. One just can't answer your question until one has the system in extract. James Smith: If the luminescence spectrum is the same as the fluorescence spectrum of chlorophyll, doesn't this practically eliminate the need to look for a pigment that absorbs at longer wavelengths than chlorophyll in the radiation transfer? Strehler : Except for antistokes sensitization. James Smith : It seems to me that, if the reaction products react and re-activate the chlorophjdl, it must be the end of the line so far as the radiation transfer is concerned. Strehler : I think that is a good argument. What do you think, Dr. Arnold? Arnold : I have thought about it. I am not sure that a triplet state of chlorophyll is not produced by the back reaction, and that the triplet crosses over into an excited singlet which emits the luminescence. James Smith: If a pigment other than chlorophyll were e.xcited in the back reaction, wouldn't you get a luminescence of this pigment? LUMINESCENCE OF PHOTOSYNTHETIC ORGANISMS 125 Arnold : It is difficult to say that 3'ou do not, because there are no photomulti- pliers sensitive for long wavelengths. Suppose it is considerably far into the red. The curve that Davidson and I published, giving the identity of the spectrum from the delayed light and the fluorescent light, was submitted to our statistical analysis panel in Oak Ridge. If you use one test, the two curves are certainly the same. If you use another test, they are systematically different and there is a little radiation in the red. James Smith : This seems to me to be a very important thing to determine. Amon: I was interested in your remarks about the longevity of this effect in various chloroplasts and in your comment that lyophilized chloroplasts retain this activity but the freshly prepared chloroplasts do not. That seems to fit with the observation that, if lyophilized chloroplasts are kept for a long time, only a small activity in the Hill reaction is observed. In your case, if I followed it correctly, the luminescence effect accounts for about 1% or perhaps less of the radiation energy absorbed. Strehler : Much less than that. Amon: Then, would you not require a large level of photochemical activity retention by the chloroplast preparation in order to observe the effect? Strehler: As a rough estimate, I would say that the yield in the lyophilized preparations is not less than V4 of what reasonabl.y fresh chloroplasts emit. Amon: I think the point raised by Dr. Rabinowitch, as to why this effect persists, could be explained by the fact that you need only a portion of the photo- chemical level of activity, and only a small portion at that, in order to observe the effect. Now, turning to questions on the physiological side, have you ever tried to increase this hght emission? As I understand, you have a certain proportion of the Hght which is not being absorbed or utilized, but thrown back. You work in the absence of any normal electron acceptor, that is, without anj' CO2 fixation reaction and in the absence of Hill reagents. Strehler : In the chloroplast studies, yes. Amon: Then, you are re-emitting a very small portion of the absorbed light energy, while the chloroplast has a much higher capacit}-. Rabinowitch : How about oxygen? Isn't oxygen a Hill reagent? Strehler : The preparation may be reducing and thus exchanging oxygen. Amon : Can you increase this percentage of radiant energy that is thrown back by adding electron acceptors? Of course, if you have oxj^gen you are providing a sink for electrons. Can you increase it in some other way? It seems to me it would be a very interesting effect. Strehler: If you have any ideas on how it might be done, we will be glad to try them. Wassink : Just one additional comment on the discussion between Dr. Lumry and Dr. Strehler, which I think is also pertinent to a comment that Dr. Arnon made. Is it true that the properties of this chemiluminescence are very much the same as the fluorescence both in the time course and in their relations to environ- ment? One would expect that any sort of agent that would increase fluorescence might also be expected to increase chemiluminescence. 126 B. L. STREHLER Strehler: There are some agents, for instance hydroxylamine, that increase the fluorescence only slightly but almost completely destroy the long-lived lumines- cence. Luminescence is primarily distinguished from fluorescence kinetically in that it saturates as a function of intensitj'. Rosenberg : With respec^t to the point that Dr. Smith raised, do you think it would be fair to say that the luminescence data are neither more nor less con- vincing than the fluorescence data themselves for the definition of chlorophyll as being a molecule at the end of a chain? We have known from fluorescence data all along that this is presumably the one, and we would be very embarrassed if we found that the luminescence has a spectrum further to the red. I thought possibly if there were another pigment further at the end of the chain beyond chlorophyll it would perhaps be unprofitable to have a fluorescence spectrum so close to that of chlorophyll that one could not distinguish them. Arnold : This test for identity is very good. They are very similar. If there is extra light to the red of the chlorophyll fluorescence band, it is quite dim and quite close to it. Rosenberg : The point I was trying to make was that even if they are identical — and I certainly have no doubt from your data that they are — people who are looking for an extra pigment will say, "Well, we really don't know what chloro- phyll fluorescence should look like in the cell and perhaps it is not really chloro- phyll fluorescence that everybody has been speaking of." Arnold : It is certainly unproved from the experiments that have been done. James Smith : May I make just one comment on this? You could get chloro- phyll fluorescence out of the cell and still not have it at the end of the chain be- cause in the red algae, as Dr. French and his collaborators have shown, you have fluorescence of the phj'cobilins even though you have fluorescence from the chlorophyll. So not all of the energy is transferred. Is this not right, Stacy? French: Yes. James Smith: I should think you might run into the same situation with chlorophyll itself and another pigment that could receive a fair share of the energy. Kamen: Dr. Duysens, I remember that in the original work you had a transfer to another pigment near the long wavelength band of chlorophyll. Is that cor- rect? Duysens: In the fluorescence spectrum of the red alga Porphyra lacineata, there is a verj- strong fluorescence at 730 m/x. If the fluorescing pigment is a con- centration of only 0.1% of that of chloroph3dl, as we thought, then this fluores- cence probably occurs by transfer from chlorophyll a to this pigment. There is some, although less clear-cut, evidence for fluorescence at 780 m^ in other red and blue-green algae, but not for green or brown algae. Kamen : I was wondering whether there is a possibility of a trace amount of another pigment. Linschitz: 8600 A is the spectral region for the triplet. There maj- be activation at that level. Arnold : I have looked very hard for this 8600 A light without finding a trace of it. Strehler : With respect to what Dr. Kamen had to say, suppose there is as a terminal acceptor an iron porphyrin whose absorption is relatively close to that LUMINESCENCE OF PHOTOSYNTHETIC ORGANISMS 127 of the fluorescence emission of chlorophyll. Since iron porphyrins are relatively nonfluorescent you would not expect to pick up the fluorescent light or any singlet excited state from this compound anyway. References 1. Strehler, B., and Arnold, W., J. Gen. Physiol., 34, 809 (1951). 2. Arnold, \^'., unpublishetl data. 3. Arnold, W., and David.son, J., J. Gen. Phtisiol., 37, (577 (1954). 4. Arnold, W., and Thompson, J., /. Gen. Physiol., 39, 311 (1956). 5. Haxo, F., and Bhnks, L., J. Gen. Physiol., 33, 389 (1950). 6. Arthur, W., and Strehler, B., Arch. Biochem. and Biophys., 1957, in press. 7. Arnold, W., and Segal, J., unpublished data. 8. Strehler, B., Arch. Biochem. and Biophys., 34, 239 (1951). 9. Brugger, J., Ph.D. thesis. University of Chicago, Chicago, 1954. 10. Franck, J., French, C, and Fuck, T., J. Phys. Chem., 45, 978 (1941). 11. Wassink, E., Advances in Enzymol., 11, 91 (1951). 12. Strehler, B., Phosphorus Metabolism, 2, 491-502 (1951). 13. Strehler, B., Arch. Biochem. and Biophys., 43, 67 (1953). 14. Brackett, F., Olson, R., and Crickard, R., /. Gen. Physiol, 36, 563 (1953). 15. Warburg, O., Biochem. Z., 166, 386 (1925). 16. Emerson, R., and Arnold, W., /. Gen. Physiol, 16, 191 (1933). 17. Tamiya, H., Studies from Tokugawa Institute, 6, No. 2 (1949) 18. Kok, B., this volume. Decay of the Delayed Light Emission in Chlorella* WILLIAM ARNOLD, Biology Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee The process of photosynthesis seems to he partially reversible; green plants emit light for some time after they have been illumi- nated. Strehler (this volume) has given reasons for believing that this 10^ 10^ io*H lio'-J < z o in 10- 0.01 01 100 1,000 10.000 \ 10 TIME (sec) Fig. 1. Intensity of the delaj^ed light in arbitrary units as a function of the time in the dark in seconds. delayed light production is the reverse of photosynthesis. The present paper is a study of the decay of the delayed light in Chlorella. In * Work done under U.S. Atomic Energy Commission Contract No. W-7405- eng-26. 128 DECAY OF THE DELAYED LIGHT IN Cfllorclla 129 general, no simple relation has been found between the intensity of the delayed light and the time in the dark; for this reason, the results will be presented only in graphical form. All measurements of the intensity of the delayed light were made with an RCA #6217 Photomultiplier, at the temperature of dry ice, >« O — r- 4 6 TlME(min) Fig. 2. The reciprocal of the square root of the intensity of the delayed light as a function of the time in the dark. Numbers on the curve indicate relative in- tegrated light energy of the exciting flash. and with a vibrating reed electrometer. Since the decay covers a wide range of time, three different methods of study were used. For the middle part of the decay curve the flowing method de- scribed by Strehler and Arnold (1) was used, where the time in the dark is the time that the Chlorella spends in flowing between the illuminated vessel and the measuring vessel. Each cell spends most of the time in the light, making periodic excursions through the pumping system. 130 W. ARNOLD For the last part of the decay curve, the Chlorclla suspension was transferred after being illuminated into a large glass vessel directly in front of the photomultiplier, and the signal was recorded as a func- tion of the time. The suspension was kept at room temperature by a water jacket. A hand-operated shutter made it possible to check the dark current of the photomultiplier from time to time. c 3 w O w 2 w O N 1 = 500.0 3 4 TlME(min) — r 6 Fig. 3. The reciprocal of the square root of the intensity of the delayed Hght as a function of the time in the dark. Numbers on the curve give the relative intensity of the continuous exciting light. For dark times less than a few hundredths of a second, the flowing method is not satisfactory. In order to obtain some information about the beginning of the decay curve a small air-driven high-speed cen- trifuge was converted into a Becquerel Phosphoroscope. This in- strument has two shutters on the same shaft. Each shutter is open a fifth of the time. The two shutters are exactly out of phase. A Chlorella suspension placed between the two shutters and illumi- DECAY OF THE DELAYED LIGHT IX ChlorcUa 131 nated through one of them is excited by flashing hght; the number of flashes per second is given directly by an electronic tachometer. Delaj'ed light emitted by the suspension, after passing through the second shutter, falls on the photomultiplier. The signal thus obtained is proportional to the intensity of the delayed light one half-period after the flash. The results are complicated by the fact that the dark time cannot be changed without changing the duration of the flash. 1=100 500 toco 1500 FLASHES/sec 2000 2500 Fig. 4. The intensity of the delayed light as measured with the phosphoroscope as a function of the number of flashes per second. Numbers on the curve are relative exciting light intensities. Decay curves for the delayed light, at room temperature, between 0.05 and 6000 seconds are given in Fig. 1. The results are plotted on logarithmic scales owing to the wide range covered both as to in- tensity and time. There is one exception to the statement that no simple expression has been found for the intensity of the delayed light as a function of 132 \V. ARNOLD the time in the dark. A Chlorella suspension that has been in the dark for 20 to 30 minutes and then exposed to one single flash of light (from a photographic flash bulb) gives over the region of 10 seconds to 10 minutes, a simple second-order decay curve (Fig. 2). However, if the single flash is replaced by continuous light for a few minutes, the results are quite different (Fig. 3). The intensity of the delayed light, as measured with the phos- phoroscope (Fig. 4), increases with increase in the number of flashes per second up to 500 to 1000, and then becomes independent of the frequency of flashing. It may be that this break in the curve can be associated with the new fast reaction in photosynthesis shown by Kok (this volume) in his kinetic studies. Throughout the entire decay curve for the delayed light, the in- tensity of the exciting light needed to give maximum signal is in- creased as the time in the dark is decreased until, as the two curves in Fig. 4 show, at the shortest dark times used (^'gooo second), there is no evidence of saturation. Discussion Weigl: The second-order decay after flash illumination suggests the electron- hole recombination kinetics commonly observed in crystal phosphors and photocon- ductors. Debye and Edwards (J. Chem. Phys., 20, 236 (1952); Science, 116, 143 (1952)) and Yastrebow {Doklady Akad. Nauk. SSSR, 90, 1015 (1953)) have reported similar afterglow in a variety of organic compounds, including proteins. Recombination appears to be limited by the rate at which electrons return to their radicals, and emission takes place by way of the lowest triplet state of the parent molecules. In a chlorophyll-protein complex, recombination at the protein could conceivably sensitize luminescence of the dye — even though the observed emission comes from the lowest singlet level of the chlorophyll. Arnold : And an electron takes a very long time finding the radical. Weigl: Electron diffusion can give rise to lifetimes of many minutes or even hours, but only at about — 196°C. Deep electron traps or inefficient recombination would have to be demonstrated to make this mechanism seem plausible at room temperature. Rosenberg : The use of a single flash seems to have simplified the kinetics of recombination. Arnold : For the longer times. Rosenberg : In chloroplasts the situation may be easier to investigate where we don't have induction phenomenon and we don't have pools of intermediates that have to be brought up. Is the kinetics simpler for the chloroplast situation? Arnold: I checked the question of whether or not the saturation depended on the time afterwards that you measured, and it is much the same for chloroplasts. DECAY OF THE DELAYED LIGHT IX ChlorcUa 133 But I must emphasize what Strehler said about the transients. Whatever these complicated transients are they depend on the whole photosynthetic system much more than the Hill reaction. Rabinowitch: I was rather confused by the different saturation curves. Can one not suppose that one corresponds to the long component and the short com- ponent of different saturation curves and that the others are combinations? Arnold : Well, I am not convinced that it is justified to break my data into two components. Rabinowitch: Now, the second question or suggestion. Perhaps some of this complicated law of decay can be associated with the following: You get two prod- ucts, H and YOH. If both disappear at the same rate, you have a simple second- order reaction. If either one is pulled away faster than the other, you are going to get something much more complicated and you have no reason to assume that the two are pulled out at the same rate, one by the peroxide — Arnold : Particularly after the plant has been doing photosynthesis in a steady way. Rabinowitch : Yes, so that seems to be a bimolecular reaction. References 1. Strehler, B. L., and Arnold, W. A., "Light production by green plants," J Gen. Physiol, 34, 809-820 (1951). Some Observations on the Chemiluminescence of Algae* JOHN E. BRUGGER, University of Chicago (Fels Fund), Chicago, Illinois Our investigations of the chemiluminescence of algae were ini- tiated several years ago. At that time Dr. Rosenberg and I were looking for a phosphorescence of chlorophyll dissolved in rigid glasses or adsorbed on a surface. In the course of this work the chemiluminescence, previously reported by Strehler and Arnold, was observed. It was decided to investigate this phenomenon in a manner somewhat different from that employed by them. Instead of a flow technique, we used a phosphoroscopic method. The results are in- terestingly different in many respects. Algae were irradiated with repeated pulses of blue light. Be- tween these flashes, the chemiluminescence was measured. The time sequence was (in milliseconds): irradiate, 1.25; dark, 0.75; measure, 1.25; dark, 0.75. The intensity of the luminescence was determined with a 1P22 multiplier phototube, fitted with a red filter. The photosignal was amplified and recorded (GR DC amplifier and Esterline Angus recorder). At the present time, a quantum counting technique is being used to meter the photosignal. A 3-ml. sample of algae (density: 2 i^l. packed wet cells/ ml. suspension) was placed in a cell having a sintered glass bottom through which various gases could be passed — another variance from the technique of Strehler and Arnold. Chlorella pyrenoidosa were used in the work reported here. However, experiments with Scene- desmus obliquus D3 have given similar results. Irradiation was also undertaken with green, yellow and red light (with and without ad- mixtures of blue light). Within the uncertainties introduced due to hght leak and other experimental difhcuities, the results are quali- tatively the same as those obtained with blue light. As Arnold and * This work was aided by a contract between the Office of Naval Research, Department of the Navy, and the University of Chicago (Contract ONR 432- (00)). ;34 CHEMILUMINESCENCE OF ALGAE 135 co-workers have shown, the spectral distribution of chemilumines- cence is experimentally indistinguishable from that of the fluores- cence of chlorophyll but is several orders of magnitude less intense. The curves presented are not decay curves in the accustomed sense. They are graphs of the intensity of chemiluminescence followed dur- ing a period of irradiation — with rapidly pulsed light. 18 — UJ 16 (_) — CHLORELLA S 14 o — N2 -WATER UJ ' '- i 10 3 8 /^ UJ 6 X o 4 — \y 2 — 1 1 1 1 1 1 1 1 1 1 1 1 12 3 4 5 IRRADIATION (MINUTES) Fig. 1. Time course of delayed luminescence during irradiation of Chlorella in water. Nitrogen atmosphere. The Chlorella were frequently suspended in water. This reduced the problem of equilibration with the gas used in sweeping and facili- tated rapid changes of sweeping gas composition. Experiments were also conducted using various buffer mixtures to make certain that the aqueous mediinn introduced no artifacts. Figure 1 shows the variation of the intensity of chemiluminescence during the course of the irradiation. Chlorella were suspended in water and swept with oxygen-free (less than 10~^ mm. Hg) nitrogen. The rapid initial rise is followed by a deep minimum after which the luminescence rises to a steady-state intensity, which could be maintained for a con- siderable period of time. The dotted line at about intensity 1.5 indi- cates background and stray signal. Figure 2 shows the chemilumines- cence curve for Chlorella suspended in nutrient and swept with 2% carbon dioxide in air. Irregularities such as those observed at 10 to 20 seconds are customarily found. The chemiluminescence then falls to a very low level. The ratio of the initial spike height to the steady- state level is 10 to 15: 1. 130 J. E. BRUGGER In the case of Chlorella in water swept with pure oxygen, we ob- tained the curve shown in Fig. 3. Note that the steady state is similar in intensity to that observed in pure nitrogen. Addition of small amounts (short bursts) of carbon dioxide to the flowing gas tem- 18 — 16 , — CHLORELLA LlJ ^ 14 — 2% CO2 IN AIR- -NUTRIENT LlJ O 12 — lU [ z lU — ^ \ => 8 — \ _i \ 2 fi \ UJ ^v X , N. o 4 ^\^ 2 ""i 1 1 1 1 1 1 Mil 1 2 3 4 5 6 IRRADIATION (MINUTES) Fig. 2. Time course of delayed luminescence during irradiation of Chlorella in nutrient solution. 2% carbon dioxide in air atmosphere. 16 yl4 5 12 o i 8 6 4 2 5 UJ X o CHLORELLA t \ O2 - WATER H = "BURST" 0.2% CO2 t \ 3 4 5 6 IRRADIATION (MINUTES) Fig. 3. Time course of delayed luminescence dming irradiation of Chlorella in water. Oxygen atmosphere. Effect of transient addition of carbon dioxide. porarily reduced the chemiluminescence intensity, which then quickly returned to its previous value. There was no change in pH of the sohition due to the burst of carbon dioxide. The measurement was made by adapting the sensitive pH meter of Professor Gaffron. Something different was observed, as shown in Fig. 4, when a short admixture of carbon dioxide was made when the streaming gas was pure nitrogen. The luminescence was at first depressed but it later rose CHEMILUMINESCENXE OF ALGAE 137 to a higher level than it had maintained previously. It was not possi- ble to make the luminescence rise indefinitely with successive bursts. In Fig. 4 it will also be observed that the initial portion of the curve is different from that of Fig. 1. The carbon dioxide effect was demon- is 16 iij Sl2 to 1 10 §8 _i 1 6 UJ ^ 4 o 2 CHLORELLA Ng -WATER I I = "BURST" 0.2% CO2 4 5 6 7 IRRADIATION (MINUTES) 10 Fig. 4. Time course of delaj^ed luminescence during irradiation of Chlorella in water. Nitrogen atmosphere. Effect of transient addition of carbon dioxide. 18^- uj 16 — o 2 14 — UJ ^ 12- UJ |lO- 3 8- 6 — UJ I 2 — CHLORELLA No -ACID PHOSPHATE = CHANGE TO Oc L 3 4 5 6 IRRADIATION (MINUTES) 7 8 10 Fig. 5. Time course of delayed luminescence during irradiation of Chlorella in phosphate buffer. Nitrogen atmosphere. Effect of change to oxygen atmosphere. strated on the curve of the type in Fig. 1 also. The height of the initial spike depended on the previous history of the sample. One-half hour of anaerobiosis produced a high spike. These induction effects were much reduced when the chemiluminescence was observed after only a short dark period of anaerobiosis. Treatment with oxygen during anaerobiosis in the dark, followed by careful flushing of the oxygen 138 J. E. imUGGER before irradiation, caused the chemiluminescence to increase. The luminescence measured in the presence of carbon dioxide and nitrogen (no oxygen) was slightly greater than that measured with carbon dioxide and air present. Figure 5 shows the effect of a change to pure oxygen following irradiation under pure nitrogen. Similarly, it was observed that the chemiluminescence was higher in the presence of ordinary tank ni- trogen than in the presence of our specially purified gas. 18 16 <^ 14 UJ o 12 — 10^ 8r I 6^ UJ 2 L SCENEDESMUS N2 -NUTRIENT A = I0"2 M PHENYLURETHANE B=NO PHENYLURETHANE I 2 3 4 5 6 IRRADIATION (MINUTES) Fig. 6. Time course of delaj^ed luminescence during irradiation of Scenedesmua in nutrient solution. Nitrogen atmosphere. Effect of A'^-phenylurethane. The conclusion that the chemiluminescence is least intense when photosynthesis is proceeding with great facility is bolstered by studies made with various poisons, inhibitors, and narcotics. A curve made using Scenedesmus to which A^-phenylurethane had been added is sho\vn in Fig. 6. The intensity of the chemiluminescence in the steady state is roughly linear with irradiation intensity under the conditions of our experiments. (We normally operated in the region of near saturating intensities.) This is to be contrasted with the finding of Arnold that the luminescence saturates at low intensities of irradiation. Whether one observes a saturation or a linear rise with increasing irradiation depends on the time delay between irradiation and measurement of the luminescence. In our experiments, the time elapsed is less than a millisecond, whereas in those of Strehler and Arnold it is of the order of tenths of seconds. I shall not discuss these experimental results and observations in CHEMTLUMTNESCEXCE OF ALGAE 139 detail. It is perhaps sufficient to quote from an abstract composed by Professor Ftanck. He prepared tiiese remarks before his illness made it impossible for him to be here. Professor Franck wrote as follows: "Our theory of the photochemical part of photosynthesis presup- poses a main process of photosynthesis in which the energy of two absorption acts is utilized for the transfer of one hydrogen atom to the photosynthetic oxidant PGA and the simultaneous transfer of a hydroxvl to an enzyme. There is a minor process in which DPGA is reduced to -PGAH and phosphate whereby the energy of only one absorption act is used.* The early products of this process are the -PGAH radical and the chlorophyll radical which has lost one H-atom in ring Y and is connected with two hydroxyls bound to Cg. The latter radical is a potential OH donor as long as the OH-carrying enzyme has not removed one of the hydroxyls. This reaction is slow because a relatively high heat of activation is needed . Thus the potential OH donor has a much longer lifetime than the corresponding products of the main process. Back reactions between the potential OH donor of the second process and a -PGAH radical will therefore occur relatively often even if enough active enzyme for the removal of OH is available. The particular back reaction as- sumed is a removal of the H which the PGA has gained in the for- ward process by a reaction with one of the hydroxyls of the chloro- phyll radical. This back reaction results in the re-formation of PGA and of chlorophyll in the enol state. The energy released by the re- action is quite sufficient to excite the chlorophyll. Other factors very favorable for the occurrence of chemiluminescence are that the back reaction must occur in direct contact with chlorophyll and that most probably the enol chlorophyll has a higher fluorescence yield than the keto chlorophyll. "Certainly back reactions will also occur between the two radicals PGAH- and HO-Enz- of process I, especially, if by addition of in- hibitors, their lifetimes are prolonged. How^ever, these back reactions have little chance to excite the chlorophyll because they will occur everywhere in the solution and not just in contact with chlorophyll. This interpretation can explain the differences of the chemilumines- cence observed directly after irradiation and after a dark pause of two-tenths of a second. It is further in agreement with observations * Compare addendum to "A Theorj- of the Photochemical Part of Photoaj'ii- thesis" on page 142. 140 J. E. BRUGGER on the influence of inhibitors on the intensity of chemiiuminescence and on the quenching influence of carbon dioxide." A presentation of other data obtained in this study, as well as a theoretical consideration of all the phenomena of chemiiuminescence, is in preparation. Discussion Rabinowitch: It should be reemphasized that you used pulsed light to irradiate your algae. Arnold: We measured the chemiiuminescence of Krall's barley and found spikes eight times higher than the steadj' state. This agrees with your observa- tions for Chlorella in air containing 2% carbon dioxide. Witt: Did you observe the high spikes when you resumed pulsed irradiation after having kept the Chlorella in the dark for a minute or so? Brugger: No. I obtained the highest spikes with Chlorella samples which had been kept in the dark for an hour or so. If I stopped the irradiation after the steady state had been reached and allowed the Chlorella to remain in the dark for Fig. 7. Recording monochromatic photometer: (front, left to right) Beckman DU monochromator, cell and filter holder fitted with flow-type "Lucite" cell, housing for photomultiplier tube and battery power supply. Not shown: photo- signal amplifier and recorder, device for irradiating algae, apparatus for circulating algal suspension through cell. CHEMILUMINESCENXE OF ALG.^E 141 only a minute, only a very small spike was observed. The previous steadj'-state level was almost immediately- reached. However, after 5 or 10 minutes in the dark, following an irradiation, the rhemiluminescence spike was considerably higher. Gaffron : It appears that the highest level of chemiluminescence you obtained by adding successive bursts of carbon dioxide to Chlorella in oxygen-free and carbon dioxide-free nitrogen is comparable to that which you found for Chlorella in pure oxygen. The steady-state level of chemiluminescence appeared to be highest in oxygen, lower in pure nitrogen, and lowest in the presence of carbon dioxide, with or without oxygen. Brugger: This is correct. Limiry : You told me that Dr. Frank Allen and you had studied changes in the absorption spectrum of Chlorella suspensions containing phenylurethane. Did you look at any other inhibitors for their eflfeet on the spectrum? We have looked at a number of inhibitors which do affect the light reaction and have detected no changes in spectrum at all at high light intensities. We included agents which quench the fluorescence. I wonder whether there is any indication that these agents have a permanent absorptive effect on the chlorophyll. Brugger: Dr. Allen and I built a recording monochromatic photometer by modii\-ing a Beckman DU spectrophotometer. We constructed a special cell and cell holder and a detecting, ampUf^-ing and recording unit, consisting of a multi- plier phototube, Leeds and Xorthrup DC microammeter, and Speedomax re- corder. Our electronics *ere such that we bucked most of the photosignal and amplified only the residual. We were able to repeat Dr. Duysens' observations on changes in absorption. Chlorella in water containing 3 X 10 "^ M phenylurethane appeared to show a slight shift to the red in the absorption spectrum. We beheve that the complicating feature of the enhanced fluorescence due to the phenylure- thane was eliminated by judicious use of filters. As there is a definitely tentative character to our results, I do not wish to discuss them at length. A Theory of the Photochemical Part of Photosynthesis JAMES FRANCK, * University of Chicago {Fels Fund) , Chicago, Illinois The fluorescence yield of chlorophyll a in green plants is low and, very roughly speaking, constant over the range of illumination in- tensities from that which compensates respiration to the one which produces saturation. When secondary processes involving molecular oxygen are avoided, the fluorescence yield remains unaltered well into the region of saturation. This indicates that the processes principally responsible for the limitation of the fluorescence yield are those associated with the excitation of the chlorophyll molecules rather than with the utilization of the excitation energy for photochemical purposes. The transition of the excited chlorophyll molecule into the metastable triplet state is the principal cause of the low fluorescence yield. Since the energy of this state is not sufficient to transfer one hydrogen atom from water to the usual photosynthetic oxidant at a sufficiently great rate (from a donor capable after dehydrogenation to initiate a series of reactions leading to oxygen evolution), we conclude that the cooperation of two absorp- tion acts is needed. For several reasons, we are convinced that phos- phoglyceric acid itself is the primary H acceptor. The hydroxyl of the water will, according to our picture, be transferred to an enzyme in- volved in photosynthetic oxygen production. Since chlorophyll hy- drates easily, we introduce the assumption that the oxygen on Cg in the keto form of chlorophyll a will be hydrated. Then the steps by which an H atom is transferred to the PGA are : 1. Excitation of the chlorophyll molecule to its first excited singlet state. 2. Transition to the metastable triplet state. 3. Adsorption of PGA and of the enzyme which acts as OH acceptor by chlorophyll in its long-lived metastable state. This process occmrs because the metastable state of chlorophyll has qualities simi- lar to those of a bi-radical. * .\s presented by J. E. Brugj^er. 142 PHOTOCHEMICAL PART OF PHOTOSYNTHESIS 143 4. Transition of the chlorophyll from its metastable triplet state to the first excited triplet state via the process of sensitized fluores- cence. A molecule of chlorophyll in the first excited singlet state is degraded to the ground state, while another molecule in the lowest metastable state is raised to the first excited metastable state. 5. Utilization of the stored excitation energy for the transfer of the H atom bound to Cio to the PGA and of one of the hydroxyls bound to Cg to the enzyme. The formation of a double bond between Cg and Cio is part of this process and leaves the chlorophyll in the ground state in its enol form. A number of enzymatic dark reactions follow the production of the two radicals. We enumerate: {T) The chlorophyll must revert to its hydrated keto form. (^) Several enzymatic reactions are con- nected with the oxygen evolution. (5) The PGAH radicals are, in this theory, supposed to dismute enzymatically into triose and PGA. Considerable energy will be released in this process and can be used for ATP production. The energy stored in ATP is supposed to be utilized in the Calvin cycle, by which the carbon dioxide acceptor, ribulose diphosphate, is made fro trimose. Practically the same process is postulated for the main course of the reduction of Hill reagents. Since they are stronger oxidants, a second process can contribute to their reduction to some degree. The stronger oxidants, like quinone or oxygen, are quenchers of the chlorophyll fluorescence. Since the concentrations of the Hill reagents used in these experiments are low, the quenching impacts are not frequent enough to reduce the fluorescence intensity by more than 20% to 30%. The quenching is due to the utihzation of the excitation energy for an H transfer to the oxidant. The energy available in one chlorophyll molecule in the singlet state is sufficient for this reaction. Thus only two steps occur in this process: excitation of the chloro- phyll to the first excited singlet state and transfer of an H atom to the oxidant, for instance, to the quinone. The result will be the for- mation of a semiquinone and of a chlorophyll radical, which still re- tains both its hydroxyls. The energy needed to remove one of the hydroxyls from this radical is small because this removal permits the double bond between Cg and Cio to close. However, a heat of activation is required for the transfer of one of the OH's to the enzyme. In photo- synthesis, too, the second process may play a minor but significant role. The energy relations will permit a hydrogen transfer in one act 144 J. FRANCK to diphosphoglyceric acid* if the energy stored in one of the phos- phate bonds is utihzed for the reaction. According to this theory, diphosphoglyceric acid is not the ordinary oxidant of photosynthesis, but it will contribute to photosynthesis when it is present in a high enough concentration as a respiratory intermediate. We believe that this is the case at illumination intensities below the compensation point and, as Dr. Brugger has reported, a corresponding quenching of the chlorophyll fluorescence has been found there (see page 113). ADDENDUM The theory outlined above has, in the meantime, been useful for a better understanding of certain other observations with photosyn- thesizing cells. On the other hand, two changes in the discussion of the "minor" process of photosynthesis and reduction of Hill reagents seem to be indicated : 1. This "one quantum process" for the transfer of an H atom from the chlorophyll in its first excited singlet state to the oxidant becomes, in effect, a two-quantum process above a certain low intensity of irradiation, since under those conditions the energy of a second quantum will be utilized for the transfer of one hydroxyl to the enzyme. Only at quite low intensities will a thermal fluc- tuation sufficient for the hydroxyl transfer occur before an excita- tion of the radical takes place (mostly by energy transfer, as in sensitized fluorescence). These deductions are based on observations of the chlorophyll fluorescence and chemiluminescence in living cells. 2. While reduction of PGA coupled with the transfer of one hy- droxyl to the OH accepting enzyme undoubtedly needs the energy of two quanta per hydrogen transferred, indications are that PGA (and not only DPGA) can quench the chlorophyll fluorescence by utilizing the excitation energy for the transfer of the H atom on Cio to the photosynthetic oxidant. In that case, the heat of activation for the H transfer (and the corresponding loss of energy) must be con- siderably smaller than for the OH transfer. Furthermore, the bond between the hydrogen atom and Cio must be weaker than was anticipated by us. This hypothesis about the heat of activatior is quite plausible. That the bond between the H and Cio is weaker than normal has been deduced quite early by Conant from chemical evidence. * A precursor of PGA in the respiratory cycle. PHOTOCHEMICAL PART OF PHOTOSYNTHESIS 145 Discussion Gaflfron : We have expressed regret that Dr. Franck could not be here. This, I feel, applies in particular to our discussions on the chemiluminescence data. What has been missing is a more thorough debate in biochemical terms on what Dr. Franck thinks is the most logical interpretation of Strehler's and Arnold's, as well as Brugger's, observations. Dr. Franck's main contention is that there is no necessity to introduce a hydrogen transferring enzyme for the photochemical reduction of PGA (or whatever carboxylated substance becomes reduced). The task of an enzj^me is to facilitate reactions which otherwise require too much activation energy. In the photochemical reaction we have an excess of energy to take care of activation. Dr. Franck believes the majority of observations available to date indicate that a photochemical hydrogen transfer happens, indeed without a transferring agent, directly between the chlorophyll complex and an attached ultimate ac- ceptor. Such a direct transfer could also be written in the form of an electron movement though, he contends, this entails more conceptual difficulties than is generally recognized. What the biochemist should consider carefully is whether it is wrong, for experimental reasons, to assume that a reducible intermediate, say of the Calvin-Benson cycle, is directly attached to the chlorophyll complex. By means of the one or two quantum process (which one is here irrelevant) the metabolic intermediate receives one hydrogen and becomes a radical. Then it has to wait for the opportunity to get a second hydrogen. During this time it can react back and produce luminescence. Franck belongs among those scientists who be- lieve that a theory becomes respectable only after it has been worked out and re- vised to a point where no experimental contradictions are left that are fairly obvious. I remind you of Dr. Brugger's curve showing the chemiluminescence of Chlorella in nitrogen during steady-state illumination. The luminescence is rela- tively low. Then one puts in a little carbon dioxide (see Fig. 4 on page 137). Since carbon dioxide causes photosjmthesis to start, energy is drained away and one gets a depression in the luminescence immediately followed by an increase. Upon another addition of a little carbon dioxide, the luminescence again goes down and then up. It can be pushed up in steps by adding small amounts of car- bon dioxide to the maximum the luminescence can reach under any steadj'-state conditions. There is a staircase effect due to short-lasting additions of carbon dioxide. This is typical for one of those observations which require an explanation consistent with the biochemistry and the physics of the process before one can say the matter is understood. Higher luminescence means a higher concentration of photochemically produced radicals. Here they seem to increase in proportion to the amounts of C02-dependent intermediates, whenever the latter are de- prived of the opportunity to complete the cycle. Bassham : I think we can explain this easily in terms of intermediate-reduced enzymes; and, secondly, I would point out that as far as photochemical reduction of PGA is concerned, we can reduce PGA in the dark following preillumination. Gaffron: In this case PGA is formed by carboxylation and some of it is cer- tainly reduced in the dark. Does this prove it is exactly the same reaction which goes on in the light? 146 J. FRANCK Benson : But you can do it in the dark just as fast as in the light. Bassham: Why call on a second way when you already have one which works? Gaffron : Because nuiii.y other data do not quite fit this rather plausible assumption. I am a biochemist, too. My way of thinking is like yours, and it is only from contact with Prof. Franck that I am learning to distrust this simple way of explaining by analogy. From the biochemical viewpoint, we have here an obvious explanation of how the light can be used very efficiently, but the data of the photochemists unfortunately do not quite jibe with the idea of a primarj^ intermediate hydrogen acceptor which in turn reduces everything else. Dr. Franck chooses to worry about the existent difficulties. Some of us may believe it is unnecessar}' to worry. But the fact remains that there are some reliable data which are difficult to explain unless one assumes that at least one component of the Benson-Calvin cycle sticks very close to the chlorophyll itself. Part III THE POSSIBLE ROLE OF CYTOCHROMES Hematin Compounds in the Metabolism of Photosynthetic Tissues MARTIN D. KAMEN, Edward Mallinckrodt Institute of Radiology, Washington Medical School, St. Louis, Missouri Two salient facts, established only recently, provide the basis for this discussion. One is that all photosynthetic tissues contain hematin compounds. The second concerns the effect of the light energy utilized in photosynthesis. This energy sets up a steady state in which the major change relative to the dark steady state is a shift in the direc- tion of greater oxidation of one or more hematin compounds. In the case of green plants and algae, it is not difficult to limit the data to be considered to those known to be unique to photosynthetic tissue. In these organisms, the sites of photometabolism are well- defined subcellular particles ("chloroplasts") about which a large body of knowledge is steadily accumulating. Much less well defined are the corresponding entities in the bacteria. This uncertainty is not serious in the case of the obligate photoanaerobes, because in these systems growth is possible only by photosynthesis under strictly anaerobic conditions. In the facultative bacteria, there is evidence that cell particles much like chloroplasts exist in at least two species, Rhodo- spirillum ruhrurn and Rhodopseudomonas spheroides (31,33,37). There is a great deal of significance in the fact that hematin com- pounds are found both in green plant chloroplasts and in large amounts in all varieties of the photosynthetic bacteria. This signifi- cance lies in the fact that, whereas the green plants, algae, and the photosynthetic bacteria cover practically the whole range of living metabolic patterns, they share no property except that of using light energy for growth. Furthermore, until now only two classes of compounds have been found in all photosjmthetic systems. These are the photoactive pigment complexes — the chlorophylls, carotenoids, etc. — and the hematin compounds. The significance of the observation that absorption of light energy results in oxidation of component hematins lies in the apparent magnitude of these changes, in both amount and duration, which 149 150 M. D. KAMEN dwarf changes in other components expected to participate in energy transfer during the early phase of photosynthesis. In elaboration of these remarks this report will be concerned, first, with the data relating to distribution and nature of hematin com- pounds (21) in photosynthetic tissue, and, second, with the observa- tions which have been made on the steady state of oxidation of cell components during photosynthesis. OCCURRENCE AND NATURE OF HEMATIN COMPOUNDS IN GREEN PLANT CHLOROPLASTS Two hematin compounds appear to be unique to the photosyn- thetic apparatus in green plants. One of these, cytochrome /, was the first chloroplast hematin compoimd to be detected (30). Its isolation and properties have been described exhaustively by Hill and his collaborators (8,15,16). Cytochrome / is widelj'' distributed in the higher plants and algae. It appears to be an integral part of the chloroplast structure inas- much as it cannot be separated from the chloroplast unless organic solvents are used to split off lipid. Cytochrome / is estimated to account for at least one-third of the hematin of the chloroplast and is present in the ratio of 1 mole for approximately 400 moles chloro- phyll (8,16,24). It has been possible to obtain a preparation purified some 300-fold over the state in which cytochrome / is initially ex- tracted from the leaf, using ammoniacal ethanol. The yield of puri- fied material is about 13% of the hematin originally present in the crude extract. Cytochrome / is classified as a "modified" cytochrome of the c type because of its spectrochemical characteristics, the spectra of the hemochromogens derived from it by reaction with alkaline cyanide or pyridine, and its relatively high potential. However, it will not function as a substrate for the classical cytochrome c oxidase. In fact, no system spectrochemically analogous to cytochrome a or as appears to be present in the chloroplast, although it exists in the cyto- plasm outside the chloroplast along with normal cytochrome c. Oxidase activity responding to respiratory inhibitors has been re- ported in chloroplasts (29) but such activity is not proof for the presence of cytochrome a. Cytochrome / will couple with cytochrome c nonenzymatically so that it can be oxidized slowly when incubated I HEMATIN COMPOUNDS IX FHOTOMETAHOLTSM 151 in air with cytochrome c oxidase and catalytic amounts of cytochrome c. The other chloroplast hematin compound is a "6" type cytochrome, called 5g by Hill (14). This hematin compound appears to be asso- ciated with the chloroplasts of a variety of plants. Like cytochrome /, it is bound in the chloroplast structure firmly enough to withstand extraction bj'' aqueous solvents. Cytochrome / and cytochrome b& cover a range of redox potential which is more than half of the total required to span the difference between the hydrogen and oxygen electrodes at physiological pH (14). Thus, these two compounds provide the chloroplast with an electron transfer system which, in principle, could carry out oxida- tions in either direction over a large fraction of the physiological range. HEMATIN COMPOUNDS OF FACULTATIVE PHOTO SYNTHETIC BACTERIA The nonsulfur purple bacteria are classified at present into two genera, Rhodospirillum and Rhodopseudomonas (35). Representative species contain large amounts of soluble hematin compounds which can be described as modified "c" cytochromes (19). Like cytochrome /, all these heme proteins show high positive redox potentials. Again like cytochrome /, thej^ are not substrates for a cytochrome a type oxidase, and they can be coupled nonenz\^matically with cyto- chrome c. The heme protein from R. ruhrum has been studied most inten- TABLE I. Some Properties of R. rubrum-Cytochxome c Compared with Mam- malian Cytochrome c (38) R. rubrum-cytocYiTOTne c Mammalian cytochrome c Classification Stability Modified c Resistant to extremes of heat and acidity None None Autoxidation Reaction with CO on reduc- tion Isoelectric point (pH) Electrophoretic mobility pH Cathodic, 3.1 X 10 ~* 7 (cm.Vvolt-sec.) Absorption on IRC-50, pH 7 None Approximately 7 Standard c Resistant to extremes of heat and acidity None None 10 Cathodic, 8.2 X 10"' Stronglj- absorbed 152 M. D. KAMEN sively (10,19,30). Its properties, which are t5^pical of many "c" cytochromes in photosynthetic organisms, are exhibited in Table I. The facultative bacteria yield these proteins readily upon treatment with warm trichloroacetic acid as in the Keilin-Hartree procedure, and can be purified in the same manner as mammalian cytochrome c (19,38). However, they are noticeably more labile than their mam- malian analogs and can be obtained in better yield with less dena- turation by use of less drastic methods (19,20). In R. ruhrum and R. spheroides, most of the cytochrome "c" appears to be loosely bound and extractable. There also is a fraction which is tightly bound to the particles and resists extraction with aqueous systems just as do the chloroplast hematins. There is some evidence that cytochromes of the "h" type are pres- ent in the chromatophores of the facultative bacteria. However, only one such hematin compound has been characterized (38). A remarkable new type of hematin compound, as yet unclassified, has been observed which is distributed generally in the facultative bacteria (19,38). Its properties are hsted in Table II. This compound displays the spectroscopic properties of a compound of the myo- globin type and shows, as well, an ability to form a CO complex in the reduced form. Nonetheless, it yields hemochromogens with spectra like those formed from a typical cytochrome c. The redox potential lies in the region normally attributed to hemoglobins or cytochrome b. It is clear that this hematin compound cannot be classified (at least, HO far) in any of the common categories of the iion porphyrin pro- teins. The enzymatic properties of this compound are also remarkable TABLE II. Some Properties of R. rubruni Yellow Pigment (38) Classification Hybrid Stability Thermostable, slowly denatured at low pH Autoxidation Rapid Reaction with CO in reduced form Forms CO complex Absorption maxima (reduced) 550, 423 m/i (no beta band) Absorption maxima (oxidized) 640, 490-500, 393 mjw £o'(pH7) ~0.1 to -1-0.15 mj* Reduced pyridine hemochromogen Spectroscopically identical with that ob- tained from mammalian cytochrome c Reduced cyanide hemochromogen Spectroscopically identical with that ob- tained from mammalian cj^tochrome c HEMATIN COMPOUNDS IN PHOTOMETABOLISM 153 (38). It can act as a substrate for either the bacterial or mammalian cytochrome c reductase (25). Since it is rapidly autoxidizable and forms a CO compound in the reduced state it could function as a terminal respiratory oxidase. Another interesting property of this compound is its ability to reduce the bacterial cytochrome c non- enzymatically. In this respect it behaves in a manner reminiscent of mammalian cytochrome c with the bacterial cytochrome c or cyto- chrome /. In other words, it provides a sort of oxidase for the cyto- chrome c of the bacteria as well as for the low potential electron do- nors present in the oxidative system. There is no satisfactory name for this compound at present. It has been called variously "pseudohemoglobin," "yellow pigment," and "CO-binding pigment." The difficulty of naming and classifying it can be resolved only by precise determination of its functions in bacterial metabolism. HEMATIN COMPOUNDS OF STRICTLY ANAEROBIC PHOTOSYNTHETIG BACTERIA The green sulfur bacteria and the purple sulfur bacteria make up two divisions of the group of photosynthetic bacteria (34) . They are marked off from the other division, the nonsulfur purple bacteria, be- cause they can grow only as photoanaerobes. In the absence of light, they cannot use energy derived aerobically or anaerobically for growth. Thus, they provide a unique opportunity for the study of the function of hematin compounds in systems uncomplicated by path- ways for utilization of energy other than that available in photo- sjmthesis. In the green bacteria, only one genus — Chlorohium — is recognized (22). The few species known are differentiated from each other by characteristic substrate requirements for growth. Three hematin compounds are known. The first to be isolated in cell-free extracts is an iron porphyrin protein from C. limicola (17). Two others have been detected in C. thiosulfaticum (13). The properties of C. cytochrome- 554 (1) and the C. limicola pigment are summarized in Table III. It is seen that these pigments while resembling a cytochrome of the "c" type show redox potentials more characteristic of cytochromes of the "6" type. One hematin compound, isolated and purified from the purple sul- fur bacterium, Chromatium, strain D, has been studied (26). Purifi- 154 M. I). KAMEN TABLE III. Cytochrome Components of Green Sulfur Bacteria (13,17) C. limicola, cytochrome-553 : see (17) Absorption maxima (reduced) 553, 520, 415 mju Reaction with CO in reduced form None Autoxidation Slow C. ihiosulfaiicum, cytochrome-554 (1); see (13) Absorption maxima (reduced) 554, 523, 417 m;u Absorption maxima (NO-compound-oxidized) 573, 537 m/j. Reaction with CO in reduced form None Autoxidation Slow Fe content 0.37% Hematin content 3.1% £'o'(pH7) ~ +0.160 volt Concentration in cells '^0.1% dry weight cation beyond that achieved by exhaustive electrophoresis of crude ammonium sulfate fractions has not been obtained. The properties of the purest preparation obtained are shown in Table IV. At least two protein components are present. One of these binds a heme group similar to that found in mammalian cytochrome c. There is nonheme iron apparently present in some as yet unspeci- fied form. While the compound is slowly autoxidizable and exhibits spectroscopic behavior reminiscent of cytochrome c, it has a low redox potential. Thus, it, too, falls into no recognized single class of TABLE IV. Properties of Chromatium Cytochrome (26) Absorption maxima (mp) Ferrocytochrome 552, 525, 418 (shoulder at 423) Ferricytochrome — , — , 406 Reduced cyanide hemochi'omogen 553, 527, 419 Reduced pyridine hemochromogen 551, 519, 412 Porphyrin in ether 632, 575, 532, 503, 375 Ratios of ferrocytochrome maxima 270/552 8.57 418/552 9.8 552/525 116 Anodic mobility, pH 6.0 5.9 X 10-« cm.Vvolt-sec. " " , pH 7.8 6.26 X 10-6 cm.Vvolt-sec. 8.43 X 10-5 cm.Vvolt-sec. Fe content, % protein 0.12 Ratio ; Fe/Heme 2 . to 2 . 7 ^'o (pH 7) -0.04 volt HEMATIN COMPOUNDS IN PHOTOMETABOLISM 155 iron porphyrin protein but is a hybrid compound hke the cyto- chrome of the green bacteria and the yellow pigment of the faculta- tive bacteria. No enzymic function has been found for this compound, just as in the case of cytochrome /. Cytochrome c reductase fails to activate the Chroniatium pigment, although extracts of Chromatium contain re- ductase acti\'ity when cytochrome c is present as a substrate (27). It is apparent that the photochemical apparatus in photosjaithetic systems, regardless of overall metabolic pattern, contains hematin compounds. Not only do these hematin compounds appear to be unique but they also do not occur outside the photochemical ap- paratus under conditions where there is no close coupling between normal respiration and photosynthesis. THE FUNCTION OF HEMATIN COMPOUNDS IN PHOTOSYNTHESIS In green plant photosynthesis, the net result of the photochemical act is the transport of four electrons against a potential drop of 1.2 volts, which is the difference betAveen the potentials of the hydrogen and oxygen electrodes at physiological pH. If the unitary theory of photosynthesis is adhered to (34), a similar process may be supposed to occur in bacterial photosynthesis, although no oxygen evolution concomitant with reduction at the level of the hydrogen electrode can be demonstrated. In 1939, Hill proposed (see 15) that the initial phase of the photo- chemical process did not push electrons all the way from the oxygen potential to the hydrogen potential, but only part way. A back oxi- dation was invoked to provide the additional energy storage. He suggested that the substrate for the back oxidation was a part of the reduced material formed in the initial photochemical act, and the H acceptor was a part of the oxygen liberated by the partial movement of electrons away from the potential of oxygen. He also raised the possibility that the transport system for this movement of electrons consisted of respiratory pigments similar to, but not identi- cal with, those normal to respiratory systems. It was inferred that this photorespiratory system was separated in the cell, spatially or otherwise, so that it could function in close coupling with the for- ward reaction of light absorption and act independently of normal respiration. At the time of these proposals, no basis existed for the notion of chloroplast respiratory pigments. As we have seen, however, 156 M. D. KAMEN there is now considerable experimental backing for the existence of such entities. The notion of a back oxidation has intrigued many writers and has appeared in one form or another too often to recount here. (See, for example, 2,3,23,39.) Davenport and Hill (8) have elaborated the concept of a back oxidation coupled to the photochemical process in terms of the hematin compound, cytochrome /, and some other oxidizing agent which is reduced in the photochemical act. They point out that the potential for cytochrome / lies at a point more negative than that of the oxygen electrode by an amount which, for the movement of four electrons, is precisely equivalent to the energy of one quantum in the characteristic red emission spectrum of chlorophyll. The next drop in potential brings one to the region at the limit of the Hill chloroplast reaction and to the potential of cyto- chrome be- Here, again, the potential difference results in an energy change for four electrons corresponding to the energy in one quantum of red light. Gaffron and Rosenberg (12) have summarized the experimental findings which indicate that no back oxidation occurs involving molecular oxygen as such. However, preoccupation with oxygen as the H acceptor in such a process is not requisite to the Hill postulate. The oxidizing system generated in the light can be a substance other than oxygen, such as a complex organic free radical which produces oxygen irreversibly and partly back reacts in the manner suggested by Hill. It is fortunate that techniques now exist for making a start in de- termining the state of cell constituents in intact cells during metabo- hsm. These are dynamic spectrophotometric methods, which have been employed by Duysens (9), Lundegdrdh (24), and others, and have reached a notable degree of development in the hands of Britton Chance and his co-workers (4-7) . Duysens first observed that, in intact cell suspensions of R. ruhrum., there were changes in optical density brought about by a transition from a steady state in the dark to a steady state in the light. These changes coincided with the difference spectrum between oxidized and reduced forms of a cytochrome type compound or compounds. Vernon and Kamen had showTi about the same time (37) that the R. ruhrum cytochrome c could be photochemically oxidized by air in an en- zymic reaction, and that compounds at potentials more negative than HEMATIN COMPOUNDS IN PHOTOMETABOLISM 157 that for the cytochrome component were not reactive (37). Shortly, thereafter, Chance and Smith (6,7) conducted a more exhaustive spectrophotometric analysis of the action spectra exhibited by R. ruhrum cell suspensions under a variety of experimental conditions. In agreement with the other workers, they found that the net effect of illumination was an overall oxidation of one or more hematin compounds. In this facultative bacterium, which displays competi- tion between dark aerobic oxygen uptake and light anaerobic metabo- hsm, the effect of Ught under anaerobic conditions was qualitatively the same as that in the dark when oxygen was admitted. Chance and Smith (6) have proposed a scheme in which the hema- tin compounds {R. ruhrum cytochrome c and yellow pigment) act as a bridge for electron transfer between the products of photolysis and the systems concerned with reduction of CO2 and other substrates. Their basic assumption is that the hematin chain which reacts with oxygen can also react with the products of photolysis. They place the yellow pigment at the end of the respiratory chain. The redox po- tential in the isolated form appears to be negative by more than 100 mv. compared to the cytochrome c component (38). This fact would seem to disagree with the notion of the yellow pigment as a terminal oxidase. However, it has not been established that the yellow pigment has not undergone some modification during the procedures em- ployed for extraction. Kamen and Vernon (18) have shown that, in R. ruhrum, the reductase activity far exceeds the oxidase activity, using bacterial cytochrome c or mammalian cytochrome c as substrate. Hence the cytochrome c should be largely in the reduced state in the dark under aerobic conditions, as well as anaerobic conditions. Chance and Smith (6) find that addition of phenyl mercuriacetate causes a rapid oxidation of the cytochrome c. This is in agreement with the known inhibition of reductase activity by this agent. In the absence of this inhibitor, they suppose that the photooxidant can react with the yellow pigment and thus pull electrons away from the cytochrome in a manner analogous to that observed by Vernon and Kamen for the coupled oxidation of cytochrome c by air in the presence of yellow pigment (38). These latter workers have also shown that the photoenzymic oxi- dation of cytochrome c is not inhibited by cyanide as is the oxidase of the aerobic system in the dark (37). Similarly, Chance and Smith 158 M. O. KAMEN (6) note that the hght-induced oxidation is not affected by cyanid(% and in addition is insensitive to CO. The isolated pigment has been described previously as forming a CO-complex, but apparently this compound may not be effective in preventing light oxidation because of the well-known dissociation of such heme-CO compounds by light. Both groups of workers postulate a competition between the photo- lytic oxidant and oxygen, with the site of the competition in the hematin chain. Chance and Smith (6) propose that the shift to the oxidized state is the result of the preferential attack of the oxidase by the photooxidant. Kamen and Vernon (18) have observed that the photooxidation in air catatyzed by their photooxidase proceeds at a rate at light satura- tion (which is attained at rather low light intensity) some five to ten times faster than the rate of the dark oxidation under the same con- ditions. They argue by analogy from this observation that competi- tion occurs at the cytochrome c le\^el. In this scheme, the bacterial cytochrome would act as the substrate for the two oxidative systems and the light oxidase would compete successfully for the cytochrome c during illumination. The shift from reduced to oxidized cytochrome, on their viewpoint, results from the fact that while the dark oxidase cannot keep up with the reductase in the absence of light, the in- creased rate of oxidation possible with the light oxidase favors a greater degree of oxidation in the light. The complexities which arise in the interpretation of shifts in oxi- dation states in faculative tissues should be obviated by similar ex- periments in tissues of strictly photoanaerobic character. A start has been made by Olsen and Chance with Chromatium (28). Their pre- liminary findings again indicate that practically the entire action spec- trum found can be accounted for in terms of photooxidation of the hematin system as isolated and characterized by Newton and Kamen, described above (26). This finding, even though preliminary, provides strong evidence that the hematin system of photosynthetic tissue is coupled to the photochemical act, because no other function for the hematin system is available in such a tissue. The bacterium can- not use a chemosynthetic oxidation apart from photometabolism to any useful purpose, and in fact, is inactivated and finally rendered nonviable by air. A number of observations on chloroplasts and intact green plant tissues have been reported which parallel those recorded for the bac- HEMATIX COMPOUNDS IX PHOTOMETABOLISM 159 teria in addition to the observations reported previously in this s3^mposiiim. Lundegardh has claimed that, in Chlorella, illumination causes an oxidation of cytochrome / and a barely detectable but significant reduction of some h cytochrome (24). He has suggested a scheme similar in many respects to that provided by Hill, in which electron flow down the respiratory chain in the dark is reversed and run backward through a "special cytochrome system in which cyto- chrome / is one of the Hnks" (24). Recently, Hill has remarked (14) that in leaves of certain "golden" varieties of plants a sharp band corresponding to reduced cytochrome be, appears on illumination in vivo, which can only be obtained from the chloroplasts isolated from the same som-ces by treatment with hydrosulfite. There is also an indication that, at the same time the h component is reduced, the / component is oxidized. Here, as in the bacteria, no components other than the hematin compounds in photosynthetic tissues show changes in oxidation state of this magnitude upon illumination. It is evident that this phase of research is in its infancy but Hke most infants it can be regarded with optimism for the future. The acceptance of the notion that a back oxidation involving a chain of hematin compounds occurs closely coupled with the photo- chemical act would normally invite speculation about a subsequent or concomitant phosphorylation. Such a phosphorylation would be analogous to the coupling of phosphorylation and oxidation in nor- mal respiration. However, no speculation is required, as experimental data are at hand which appear to indicate that the photochemical act can be coupled to phosphorylation under conditions where a normal respiratory oxidation cannot occur (i.e., under strictly anaerobic conditions). Frenkel (11), working with chromatophores from R. ruhriim, has demonstrated that light under strictly anaerobic condi- tions induces a disappearance of inorganic phosphate in the presence of ADP which can be accounted for as newly formed ATP. In well-washed chromatophores this light-dependent phosphoryla- tion is not accompanied by appreciable dark phosphorylation. It is insensitive to the usual respiratory inhibitors and does not require oxygen. In this respect, the light phosphorylation parallels the photo- oxidation of cytochrome c by the same system (37). It is of interest that 2,6-dichlorophenol-indophenol which can act as a substrate in the photooxidation in air (37) suppresses the light anaerobic phos- phorylation (11). Similar phenomena are noted in chloroplast sus- 160 M. D. IL^MEN pensions (1) and also in the chromatophores derived from the strict anaerobe, Chromatium (27). Thus, it appears that in both the plant and bacterial systems, sufficient separation of the photolytic products can be achieved in cell-free extracts to obtain a back oxidation in which useful biochemical energy can be stored. In this connection, it is of interest to note that Wessels (40) has suggested a scheme for participation of hematin compounds in which phosphorylation is the end result of the light process. He suggests that vitamin K or a compound analogous to it is the natural Hill reagent in the chloroplast. According to his scheme, partial reoxidation of photochemically reduced vitamin K by cytochrome c (or cytochrome /) generates a high-energy phosphate bond(s) which may cooperate in the reduction of DPN (possibly proceeding via diaphorase). The oxidative phosphorylation of the reduced vitamin K by the cyto- chrome produces a tautomeric form of vitamin K in the phosphoryl- ated state. The splitting of this compound could result in the formation of the more stable para-quinone structure of the vitamin, thus making the phosphate bond in this tautomeric structure energy- rich and available for conversion of ADP to ATP. Newton and Kamen have attempted to test this hypothesis of the involvement of vitamin K in the anaerobic system from Chromatium. The results were negative (27). Menadione, an analog of vitamin K, appears to be involved in the anaerobic phosphorylation by chloroplasts, however (1). It should be noted that although chloroplasts contain large amounts of vitamin K, there appears to be no appreciable amount of this vitamin in the anaerobic photosynthetic bacterium, Chromatium (27). In all of this discussion there is a hint that the "splittmg of water" as such may not be involved in the initial light reaction, but can occur instead as a result of the transfer and storage of energy during dark back oxidations of the type which have been postulated. In other words, single quantum events are assumed to mediate the production of ATP and other high-energy compounds. The evolution of oxygen is imagined to occur as a result of the gradual accumulation of a store of oxidized material. In the case of the bacteria, the oxi- dants are removed by reaction with exogenous H donors. It is necessary in this connection to bring forward one more bit of speculation. The hematin compounds resemble chlorophyll and the other magnesium porphyrins closely enough so that direct excitation HEMATIN COMPOUNDS IN PHOTOMETABOLISM 161 of the hematin compound by inductive resonance can occur. The physical disposition of the cytochrome and other hematin com- ponents of the chloroplast is not known, but from the studies of vari- ous investigators (see, for instance, ref. 41) it appears that the chloro- plasts consist of laminae in which fatty layers alternate with the aqueous phase. In the interface, the chlorophyll molecules lie close- packed on end, oriented so that the phytol chain dips into the lipid phase and the porphyrin-polar end binds to a protein in the aqueous phase. The hematin compounds would be expected to bind to protein in the aqueous phase, perhaps to the same protein as that which holds the chlorophyll. Thus there could be a bridge of hematin compounds connecting the chlorophyll-lipoprotein with the systems directly contiguous to the chloroplast w^hich are involved in the secondary processes of CO2 reduction, etc. Light energy absorbed by a given chlorophyll molecule would be likely to migrate through the interface until it found a cluster of hematin compounds, whereupon the whole energy of the quantum would become available to the hematin compounds. It is not difficult to imagine that absorption of the relatively large energy equivalent to one quantum would excite the heme protein to a state in which a dismutation reaction resulted. The result would be the production of partially dissociated ferroheme with one or more free Fe valences which could reduce H acceptors such as DPN at or close to the potential of the hydrogen electrode. After reduction, the ferriheme could "relax" into its original bound state with an "electron deficient region" left in the hydration en- velope of the hematin compound. The continuation of this process would accumulate electron-deficient compounds in the phase sur- rounding the protein which could act as oxidants for back reactions and also as precursors of free oxygen. A fragment of such a mechanism is part of a recent suggestion by Warburg invoking a "photo-dissocia- tion" of a heavy-metal binding compound coupled with a back oxida- tion of the dissociation products (39). Acknowledgment. In conclusion, it is a pleasure to acknowledge the continued support of the C. F. Kettering Foundation, which in large part has made possible many of the studies which provide the frame- work for this report. 1C2 M. D. KAMEN Discussion Amon : Did you get (cytochrome c from the whole cells or trom particles isoluteti from the cell? Kamen : Fii-st, from the whole cell by sonic treatment. In this treatment, the supernatant becomes pink and you get most of your cytochrome that way. There is, however, always a certain amount of cytochrome left in the particles, which requires further treatment to get it out, and there is always some that does not come out at all. Amon: My question implied: what is the evidence for the statement that, in these bacteria, most of the cytochrome is in the particles? Kamen : I did not say that most of it is in the particles. I said that in this bac- terium most of it is easily extractable. Amon : Is that true of all photosynthetic bacteria or just of certain kinds? Kamen : Only for these two we know about. Lucile Smith: If you prepare the particles carefully, you can find all the cytochromes in them. It is only when you prepare them with the sonic vibrator that you get the C3^tochromes in solution. Kamen : There is an enormous amount of cytochromes in these organisms by comparison with normal aerobic tissue. The amount is considerable even by com- parison with live mitochondria. [See also Discussion following paper by Albert R. Krall, pp. 320-325.] References i. Anion, ]:>. I., Whatley, F. R., and Allen, M. B., ./. Am. Chem. Soc, 76, 6324 (1V».54). 2 Bassham, J. A., Shibata, K., and Calvin, M., Biochim. et Biophys. Acta, 17, 382 (1955). 3. Burk, D., and Warburg, O., Natiirwiss., 24, 1 (1950). 4. Chance, B., J. Biol. Chem., 202, 397 (1953). 5. Chance, B., Science, 120, 7G7 (1954). 6. Chance, B., and Smith, I., Nature, 175, 803 (1955). 7. Chance, B., Smith, L., and Castor, L., Biochim. el Biophys. Acta, 12, 289 (1953). 8. Davenport, H. E., and Hill, R., Proc. Roy. Soc. (London), B139, 327 (1952). 9. Duysens, L. M. N., Nature, 173, 692 (1954). 10. Elsden, S. R., Kamen, M. D., and Vernon, L. P., J. Am. Chem. Soc, 75, 6347 (1953). 11. Frenkel, A., J. Am.. Chem. Soc, 76, 5568 (1954). 12. Gaffron, H., and Rosenberg, J., Natnrwiss., 42, 354 (1955). 13. Gibson, J., and Larsen, H., Biochem. J. (London), 60, xxvii (1955) 14. Hill, R., Nature, 174, 501 (1954). 15. Hill, R., Symposia Soc. Exptl. Biol, 5, 222 (1951). 16. Hill, R., and Scarisbrick, R., New Phytoloqist, 50, 98 (1951). 17. Kamen, M. D., and Vernon, L. P., J. BacterioL, 67, 617 (1954). 18. Kamen, M. D., and Vernon, L. P., /. Biol. Chem., 211, 663 (1954). HEMATIN COMPOUNDS IX PHOTOMETABOLISM 163 19. Kamen, M. D., and Vernon, L. P., Biochim. et Biophys. Ada, 17, 10 (1955). 20. Kamen, M. D., and Takeda, Y., Biochim. et Biophys. Acta, 21, 518 (1956). 21. Keilin, D., Proc. Roy. Soc. (London), B98, 312 (1925). 22. Larsen, H., /. Bacteriol, 64, 187 (1952). 23. Lipmann, F., and Tuttle, L. C, /. Biol. Chem., 158, 505 (1945). 24. Lundeg&rdh, H., Physiol. Plantarum, 7, 375 (1954). 25. Mahler, H. R., Sarkar, N. R., Vernon, L. P., and Alberty, R. A., J. Biol. Chem., 199, 585 (1952). 26. Newton, J. W., and Kamen, M. D., Arch. Biochem. and Biophys., 58, 246 (1955); also Biochim. et Biophys. Acta, 21, 71 (1956). 27. Newton, J. W., and Kamen, M. D., unpublished observations. 28. Olsen, J., and Chance, B., private communication. 29. Rosenberg, A. J., and Doucet, G., Compt. rend., 229, 391 (1949). 30. Scarisbrick, R., and Hill, R., Biochem. Soc. Proc, 37, xxii (1943). 31. Schachman, H. K., Pardee, A. B., and Stanier, R. Y., Arch. Biochem. and Biophys., 38, 245 (1952). 32. Smith, I., Bacteriol. Revs., 18, 106 (1954). 33. Thomas, J. B., Koninkl. Ned. Akad. & Wetenschap. Proc, Ser. C, 55, 207 (1952). 34. Van Niel, C. B., Advances in EnzymoL, 1, 263 (1941). 35. Van Niel, C. B., Bacteriol. Revs., 8, 1 (1944). 36. Vernon, L. P., Arch. Biochem. and Biophys., 43, 492 (1953). 37. Vernon, L. P., and Kamen, M. D., Arch. Biochem. and Biophys., 41, 122 (1954). 38. Vernon, L. P., and Kamen, M. D., /. Biol. Chem., 211, 643 (1954). 39. Warburg, O., Naturwiss., 42, 449 (1955). 40. Wessels, J. S. C, Rec trav. chim., 73, 529 (1954). 41. Wolken, J. J., and Schwertz, F. A., /. Gen. Physiol., 37, 111 (1953). Investigations in the Photosynthetic Mechanism of Purple Bacteria by Means of Sensitive Absorption Spectrophotometry L. N. M. DUYSENS,* Biophysical Research Group, Department of Physics, University of Utrecht, Utrecht, Netherlands METHODS AND MATERIALS Apparatus. The apparatus used is described in another paper in this volume. Bacieria-Rhodo spirillum ruhrum strain 1 was obtained through the courtesy of Dr. C. B. van Niel, strain 4 from the Biophysical Re- search Group, Utrecht. Both strains were grown for 1 or 2 days in incandescent light, in stoppered test tubes completely filled with 1% Difco bactopeptone, strain 4 with the addition of V2% sodium chloride. All experiments were done at room temperature: 18 to 24 °C. An aqueous extract of the bacteria was prepared in the homogenizer according to French (c/. Milner et at. (1) ). The changes in absorption were measured in a 1-cm. Beckman cell; the optical densities of the bacterial suspensions at 880 m^u minus the optical densities at 960 m;u (to correct for scattering) were about 1.0 as measured with a Beckman DU spectrophotometer. These wavelengths were selected because the infrared maximum of bacteriochlorophyll in intact cells is located at 880 m/^, while the intrinsic absorption of the extracts is negligible at 960 m/x. CHANGE IN ABSORPTION AT ONE WAVELENGTH The time course of the changes in absorption appeared to depend upon the suspension medium and the intensity of the exciting light. Figure 1 shows the changes in optical density of a suspension of Rhodo spirillum ruhrum strain 4 at 430 m/i in anaerobic peptone. For intensities of the order of magnitude at which saturation of photo- synthesis just occurs, the absorption decreases upon illumination * Future address: Biophysical Laboratory, Nieuvvsteeg, State University, Leiden, Netherlands. 164 PHOTOSYNTHETIC MECHANISM OF PURPLE BACTERIA ] 65 (graphs I a, II a) until a steady-state value is reached within a min- ute; upon darkening the change is reversed (I e and II e). At very high intensities more complicated curves were observed (graphs III and IV). A decrease in absorption a was followed by an increase h, which in its turn was followed by a decrease c. Upon darkening a de- crease d occurred followed by an increase e. If the light was left on, the decrease c went on for minutes. In the dark this decrease was slowly reversed. The much faster changes a, b, d and the fast part of e took place also after long illumination had produced an appreciable, not yet reversed, "bleaching" c. It seems that the changes a and the fast part of e of graphs III and IV are caused by the same pigment as the changes a and e in graphs I and II; h and d are presumably caused by a different pig- ment. Since the change c is slow, it is probably not caused by a photosynthetic catalyst, and, since the other changes seem to take place independently of c, it may be left out of consideration in the dis- cussion of the changes a, b, d, and the fast part of e. The bottom part of Fig. 1 shows that at 530 m/z changes corre- sponding to a and e are small (graph V) and that the changes b and d are pronounced (graph VI). It should not be concluded, however, that at 530 niyu the changes b and d are greater than at 430 m/x, since at 430 m/x these changes are counteracted by a and e. DIFFERENCE SPECTRA Under most experimental conditions the time course of the change in absorption is a monotonously increasing or decreasing function of the time, which approaches a steady-state value. Examples of such changes are graphs I and II of Fig. 1. If this steady-state value (at a certain constant exciting intensity) is plotted as a function of the wavelength of the measuring light, a "difference" spectrum is ob- tained. The shapes of the difference spectra of Rhodospirillum appeared to depend on the suspension medium. Three different spectra are represented: 1. The difference spectrum in anaerobic peptone (top curve of Fig. 2). 2. That in aerobic distilled water (bottom curve of Fig. 2). 3. That in anaerobic distilled water (Fig. 3). 166 L. N. M. DUYSENS 10 37 160 350 quanta _i I 1 2 3 it minutes Fig. 1. Changes in optical density upon illumination of a RhodospiriUum sus- pension in anaerobic peptone at 430 (top figure) and 530 mju (bottom figure). Onset of illumination is indicated by an upward pointing arrow and darkening by a downward pointing arrow. The numbers written at the bottom of the figure are the approximate nimaber of quanta absorbed per minute per bacteriochlorophyll molecule. Fig. 2. Difference s])ectra of RhodospiriUum. strain 1 in anaerobic peptone (top figure) and in aerobic distilled water (bottom figure). The left part in aerobic water is for strain 4, the right part for strain 1 , and the middle part for an aqueous extract of strain 1 ; two parts were multiplied by factors to get an approximately continuous connection between the curves. PHOTOSYNTHETIC MECHANISM OF PURPLE BACTERIA 167 The difference spectrum in anaerobic peptone is similar to (albeit not identical with) the difference of the absorption spectra of oxi- dized and reduced cytochrome c. This indicates that upon illumina- tion a cytochrome becomes more oxidized. Until its identity has been + 20 + 15 + 10 + 5 Fig. 3. Difference spectrum of Rhodospir ilium ruhrum, left 1 day in water after flushing with hydrogen ("anaerobic water"). Exciting light was a band at 540 van with intensity of L2 X 10^ erga/(cm.^ sec). established, we shall call this cytochrome "cytochrome 428" after the wavelength of the maximum in its difference spectrum. In aerobic distilled water there is a positive maximum (increase in absorption) at 432 m^u. This is not caused by reduction of cyto- chrome 428 or of Rhodospirillum cytochrome c (3), since the band is too broad and occurs at another wavelength. In the infrared, there is a 168 L. N. M. DUYSENS decrease at 880 myu and an increase at 790 m^u. The infrared part of the spectrum is probably caused by a change in bacteriochlorophyll (2). Since this type of difference spectrum occurs in the presence of oxygen or oxidizing conditions and disappears under reducing conditions (c/. 4) , the change may be caused by an oxidation of bacterio chloro- phyll. If a small part of the bacteriochlorophyll takes part in this re- action, then the difference spectrum indicates that the 880-mAi Fig. 4. Changes in optical density of Rhodospirillum rubrum, at 420 and 430 myu as function of the intensity of the exciting Hght in water flushed with hydrogen. peak is shifted to 790 mfj, and the near ultraviolet band to 430 m/i. These shifts may be interpreted as indicating the dehydrogenation of one of the two reduced pyrrole nuclei in bacteriochlorophyll (cf. 5). In anaerobic distilled water the negative maximum occurs at 420 mfj. (Fig. 3), suggesting the oxidation of Rhodospirillum rubrum cytochrome c, a pigment isolated by Vernon and Kamen (3). The hump at about 430 m/j, suggests the oxidation of cytochrome 428. Figure 4 shows that the change at 430 is saturated at a much lower PHOTOSYNTHETIC ,MK( IIANISM OF PURPLE BACTERIA 100 intensity than the change at 420 m/i. Thus the peak at 420 and the hump at 430 m/z must be caused by different pigments — a conchi- sion consistent with the suggestion that the changes at 430 and 420 are mainly caused by cytochromes 428 and c, respectively. Since thus cytochrome 428 seems to attain saturation at a lower intensity than cytochrome c, it may further be concluded that in this experi- ment oxidation of cytochrome c is presumably not caused by oxi- dized cytochrome 428. CONCLUSIONS The experiments reported indicate that cytochrome 428 partici- pates in photosynthesis, since its oxidation-reduction state is shifted considerably towards the oxidized side upon onset of photosynthesis. Also, cytochrome c can be oxidized by light, but this oxidation was observed by us only under slightly unphysiological conditions. It is possible, as Chance and Smith (6) concluded, that it occurs to a small extent also in photosynthesizing cells. This conclusion was based on the assumption that the minor band at 550 m^u in the difference spectrum was caused by cytochrome c. It may, however, have been caused by a band of cytochrome 428. A curve similar to that of Fig. 4 but measured at 550 mju would probably estabUsh this point. Vernon and Kamen (3) extracted three cytochromes, one of which was c, from Rhodospirilhim strain 1 and determined the absorption spectra in the oxidized and reduced state. The difference spectrum (oxidized minus reduced) of one of these cytochromes had a maximum at 428 m/jL. However, the half-width of the band was 27 m/j., which is different from the 12-mfj. half- width of cytochrome 428. The dif- ference spectrum of the third cytochrome was also different from that of cytochrome 428. The suggestion (4b) that cytochrome 428 was oxidized not only by Ught, but also by bubbling air through an anaerobic suspension, was confirmed by Chance and Smith (6). The changes, interpreted above as caused by oxidation of bac- teriochlorophyll, were observed only under conditions in which cy- tochrome 428 was in the oxidized state. It is possible, although not yet proved, that the changes b and d occurring at high exciting in- tensities in anaerobic peptone (Fig. 1, graphs III and IV) were also caused by oxidation of bacteriochlorophyll. 170 L. N. M. DUYSENS The scheme of Fig. 5 is an attempt to explain various observations. A small part of the bacteriochlorophyll is oxidized by light and simul- taneously a rechieed com])ound H, wliich may 1)C a reduced pyridine nucleotide, is formed. Oxidized bacteriochlorophyll oxidizes one or more reduced cytochromes. The oxidized cytochromes cause the oxi- dation of the substrate and of a small part of H, the last reaction leading to the formation of adenosine triphosphate (ATP), which assists in the reduction of COo. The reactions leading to the reduction ^ lacterio- hv^ J ^ . oxidized chlorophyll > ^ bacterio-^ j^ ' chlorophyll [HJ- ■'- ADP+P oxidized 'cytochrome substrate y^ -—[CH20I -^ATP reduced -cytochrome - ^>-~., oxidized substrate Fig. 5. H3'pothetical scheme indicating role of "active" bacteriochlorophyll and of cytochromes in photosynthesis of Rhodospirillum rubrum. of COo may be analogous to those in algae. The postulated formation of ATP is in accordance with Frenkel's (7) finding that illuminated extracts of strain 1 produce ATP from ADP and inorganic phos- phate. It should be stressed that the scheme is provisional and in- complete. Acknowledgment. Most experiments reported in this paper were carried out in the Department of Plant Biology, Carnegie Institution of Washington, Stanford, California. I wish to thank Dr. C. S. French for continued interest, advice and valuable assistance, Dr. J. H. C. Smith and other members of the Carnegie Institution for friendly advice, and Mr. J. J. Stekert for technical assistance. Discussion Whittingham : Hill has suggested there is a second cytochrome in Chlorella and Porphyridium which, in the reduced form, has the sharp absorption at 563. You would assume that, if this were reduced in algae and cj^tochrome were oxidized, you would see some change at 563. PHOTOSYNTHETIC MECHANISM OF PURPLE BACTERIA I 7 1 Duysens : The difference spectra of Forp/ii/ridiutn measured so far do not show a change in that region. Neither do the difference spectra of Chlorella. These experi- ments do not exclude that cytochrome b is participating in photosynthesis, since the changes in the difference si)ectrum may have been too small to be observable under the present experimental conditions. Becker: WTiat is the present precision in measuring these differences? Duysens: The highest precision obtained is 2 X 10~^ unit in optical density, when everything is all right. With the verj^ dense suspensions the precision is less because, at lower intensi- ties, the signal-to-noise ratio is lower. I think that, with the present apparatus, no greater precision than that mentioned can be obtained. It is not only the noise of the primary phototube current but also other changes which cause variations of the indicating apparatus. Shibata: We have done experiments just on temperature-produced changes in absorption. With Chlorella, we observed some changes like those caused by illumination, for instance, in the 525 m^u band. W^e have also recorded temperature difference spectra for Rhodospirillum. With this bacterium, there appeared to be a slight change between 820 and 880 m/x, which in some ways is comparable to that observed in light-produced difference spectra. I do not wish to say that the differ- ence spectrum produced by illumination is actually a temperature effect, but I cannot exclude that a part may actually be caused by temperature. In our experi- ments, we maintained one sample at 10°C. and the other at 40°C. For the infra- red measurements, we used a sulfide detector. For the work in the visible, we modified a model 11 Gary spectrophotometer. I would like to ask Dr. Duysens whether he noticed anj' temperature effects similar to ours. Duysens: In our experiments, there was no general increase in temperature exceeding 5° or 6°C. even after hours of periodic illumination. Our illumination- produced difference spectra changes were quite rapid. Thej' were also reversible within a few seconds. Shibata : I do not implj^ that your entire effects are due to temperature changes, but I do not see that you have eliminated temperature effects. Bear in mind that the temperature within the grana during illumination may be considerabty above that of the surrounding medium. Duysens : I can only give indirect evidence that the temperature increase during my experiments had little influence, if any. If there was a pronounced effect, one should certainly have seen an increasing change during illumination, because the temperature would be changing during that time. Benson: The grana would warm up immediatel3^ They would then transfer the heat to the solution. I do not see that the temperature of the grana should rise continually. Bassham : One might reach a steady state of temperature where there is a warm area around the grana and the heat is conducted away at a steady rate. It is not inconceivable to me that there would be an initial warming up followed by the attainment of a lower steady-state temperature. Much light energy is being converted into heat within the small volume of the grana. Duysens : This distribution of the heat will take place, I believe, in 0.01 second or so. If this is true, then j'ou would expect changes in the spectrum to talsc i)lace 172 L. N. M. DUYSENS within a time of that order. In general the changes take place in a time of one or several seconds. [Except for Witt's effects. — Ed.] Frank Allen : I would not expect the temperature effect to be so selective over the si)ectrum. Rabinowitch : Dr. Shibata, I wonder whether your effects are not the result of scattering. Your changes might result from so-called anomalous dispersion, which is associated with changes m refractive index in the vicinity of an absorption peak. When you change the temperature, even slightly, you do change the index of refraction of the colored particles, in this case, of the grana. Frenkel : I feel that the 40° C. temperature maintained by Dr. Shibata in his work is rather unphysiological. I would be quite cautious in interpreting the effects. Though some cultures have been acclimatized to 40° C, normal cultures can only stand about 30° C. Shibata : I believe the Rhodospirillum were unharmed by the temperature. Duysens : I think we have evidence which shows that the decrease in absorption at 880 m/x in Rhodospirillum is not a temperature effect under the conditions of the experiment. If the change in transmission caused by illumination is measured first with the bacteria in distilled water and then in a culture medium, we get strikingly different results. In the first case there is a very great change m transmission; in the second a slight one. If the change were due to a temperature effect, you would not expect this difference because the temperature effect should be about the same for the two cultures. Myers: Is that a reversible effect? Duysens : Completely reversible. Chance : Did it occur to you that the metabolism would increase at the higher temperature and might therefore give rise to the absorption band that j^ou observe between 400 and 450? Also, some physical properties of the bacteria might change with the rate of metabolism. Do you think that j'ou are recording the intensification of an absorption band already pre.sent or the shift of a band position with temperature? Wassink : Is there any harm in assuming that, since these obviously are conse- quences of the related metabolism, they may be brought about as well by light as by temperature? I mean by a different pathway probably, but just by attacking the same compounds, let's say cytochrome or whatever they are, that are involved in these oxidation-reduction changes. French : I think everybody would agree that that happens. Chance : I would like to concur in a suggestion made by Dr. Duysens. We rou- tinely use a gray filter of a density of 4 or 5 between a half-voltage tungsten lamp and cell in order to cause a measurable effect. This light intensity might be of the same order of magnitude as the light which you get from the Gary spectropho- tometer, and perhaps sufficient to cause a light effect. If the hotter cells have a sensitivity to light different from that of the colder cells, you might well get just the effects you record: an effect of temperature upon the photochemical reaction and not upon the absorption spectrum per se. Rabinowitch: Let's say, in this light, the clilorophyll molecule absorbs a PHOTOSYNTHETIC MECHANISM OF PURPLE BACTERIA 173 quantum of about 2 volts once in 100 seconds, about 0.02 of an electron volt per second. Distributing this energy over 100 molecules gives you a temperature rise of only about a fraction of a degree per second. In a few seconds, you can not raise the temperature of a granum more than a degree, if that. Strehler: In our circulating system, there was no evidence of a temperature change greater than a few tenths of a degree — for the whole suspension, of course. References 1. Milner, H. W., Lawrence, N. S., and French, C. S., Science, 111, 633 (1955). 2. Duysens, L. N. M., "Transfer of excitation energy in photosj^nthesis," Doc- toral thesis, Utrecht, 1952. 3. Vernon, L. P., and Kamen, M. D., /. Biol. Chew., 211, 643 (1954). 4. (a) Duysens, L. N. M., Carnegie Institution of Washington Yearbook, 52, 157 (1953); 53, (b) 166 (1954). 5. Duysens, L. N. M., Huiskamp, W. J., Vos, J. J., and van der Hart, J. M., Biochim. et Biophys. Acta, 19, 188 (1956). 6. Chance, B., and Smith, L., Nature, 175, 803 (1955). 7. Frenkel, A. W., J. Am. Chem. Soc, 76, 5568 (1954). Oxidation of Cytochromes upon Near- Infrared Irradiation of Chromatium JOHN M. OLSON,* Johnson Research Foundation, University of Pennsyl- vania, Philadelphia, Pennsylvania The first observation of the oxidation of cytochromes in photo- synthetic bacteria upon illumination was made by Duysens (1) in his studies of absorption spectrum changes in Rhodo spirillum ruhrum. Subsequently, Chance and Smith (2) confirmed and rein- interpreted Duysens' observation with a more detailed study of the cytochromes involved. My observations were made on the photo- synthetic sulfur bacterium, Chromatium D, furnished by Dr. M. D. Kamen. The bacteria were cultured at 29 °C. in an inorganic medium con- taining sodium sulfide, sodium thiosulfate, and sodium bicarbonate as substrates. Glass-stoppered culture bottles were illuminated by tungsten lamps. In each experiment on the effects of illumination a concentrated suspension of bacteria in growth medium was observed by means of a double-beam spectrophotometer (3). The sample could be irradiated by a near-infrared beam (X > 700 m^u) perpendic- ular to the measuring beams. In other experiments two samples in the dark were compared in a recording spectrophotometer (4). All experiments were performed at room temperature (23° to 27 °C.). Since the suspensions were aerobic after the centrifugation and resuspension necessary for concentrating the bacteria, some sus- pensions were covered with paraffin oil to obtain anaerobic samples. After 15 to 60 minutes under oil in the dark the absorption of a sam- ple at 422 m/z slowly increased to a new steady-state value. The sam- ple was known to be anaerobic when irradiation caused a large diphasic decrease in absorption (Fig. 1) in contrast to the small monophasic decrease in absorption characteristic of aerobic samples. The change in absorption spectrum of anaerobic suspensions upon irradiation for about 3 minutes is given by curve A in Fig. 2. This difference spectrum has an alpha peak, a beta peak, and a gamma * National Science Foundation Predoctoral Fellow. 174 OXIDATION OF CYTOCHROMES 175 peak characteristic of cytochrome difference spectra, reduced minus oxidized* The direction of the change shows that the oxidized form of some cytochrome or combination of cytochromes has a higher con- centration when the bacteria are irradiated than when the bacteria are in the dark. In this respect Chromatium is similar to Rhodospiril- lum rubrum (1,2). However, in the case of Chromatium this dif- ference spectrum does not indicate whether more than one cyto- chrome is involved. Dark Steody State Level Light on 428-440 mi Increasing Optical Density Fig. 1. Typical anaerobic light effect observed at wavelengths between 420 and 430 niyu. In the above record the change in optical density at 428 m^i minus the change at 440 m^ is shown. The kinetics of this anaerobic light effect answer in part the ciues- tion of the number of cytochromes involved. The first and second phases of the anaerobic light effect shown in Fig. 1 at 428 m/x were also recorded at several other wa^'elengths. The spectrum of the first phase is given by curve B in Fig. 2 ; the spectrum of the second phase is given by curve C. The positions of the peaks are fisted in Table I. The diphasic character of the anaerobic light effect suggests that a chain of intermediates is involved in the photosj^nthetic trans- fer of oxidizing equivalents to substrates. Furthermore, the dis- tinct differences between the spectra of the first and second phases indicate that more than one cytochrome is involved. For example, the presence of the .560-m/i shoulder in the spectrum of the second phase demonstrates that a 6-type cytochrome is involved in the second phase in addition to a c-type cytochrome (553-m/i peak) in- volved in both phases. Experiments performed with a recording spectrophotometer re- vealed additional reactions of the Chromatium cytochromes in the 176 J. M. OLSON -oe O20i -04- • +£I4 400 420 4 40 Wavelength imy) 460 520 540 560 Wavelength (my) 580 Fig. 2. Changes in optical density of anaerobic Chromatium suspensions upon irradiation with near infrared. Curve A is the difference between the final irradiated stead,v state and the dark steady state. Curve B is the difference be- tween the intermediate irradiated steady state and the dark steady state. Curve C is the difference between the final steady state and the intermediate steady state during irradiation. Curve C lies above curve B at 553 mju, but lies below curve B at 423 m/x, because a different sample was used for the wavelength interval, 510 to 580 m/i, than was used for the interval, 390 to 460 m/i. dark. The existence of a CO-binding pigment similar to that found in Rhodo spirillum rubrum (2) was shown by the addition of carbon monoxide to a suspension in which the cytochromes had been re- duced by sodium dithionite. The spectrum of the difference between TABLE I. Peaks and Troughs of Difference Spectra. Positions of Shoulders Are Shown in Parentheses Anaerobic light effect Total change First phase Second phase Anaerobic minus aerobic 423 mM (420)425 422(426) 424 524 m^ ? ? 524 553 mfi 553 553(560) 552 Aerobic light effect CO minus reduced 422 ? ? Peak 416 Troughs 432 551-552 OXIDATION OP CYTOCHROMES 177 a reduced sample with CO and a reduced sample without CO has a sharp peak and a shallow trough in the Soret region. (See Table I.) This CO-binding pigment may be identical to Vernon and Kamen's (5) Chromatium pseudohemoglobin. Although Chromatium is an obligate anaerobe, at least one of its cytochromes becomes oxidized in the presence of oxygen. The spec- trum of the difference between an anaerobic sample and an aerobic sample is roughly similar to curve A in Fig. 2, but the respective peaks have slightly different maxima as shown in Table I. Also the change in optical density at 422 m^ upon aeration is only about 90% of the total anaerobic Ught effect. However, irradiation of an aerobic suspension gives an additional decrease in optical density. This aerobic Ught effect is about 30% of the total anaerobic light effect. These observations indicate that the cytochromes of Chromatium are oxidized during irradiation whether the suspension is aerobic or anaerobic. The fact that oxygen completely inhibits the growth of this bacterium cannot be explained simply in terms of its reactions with the cytochromes. The inhibition of growth may be caused by oxygen "poisoning" or bj^ the reaction of the substrates with oxygen outside the bacteria. The conclusions may be summarized briefly. Chromatium contains at least three hemoproteins : a CO-binding pigment, a c-type cyto- chrome, and a 6-type cytochrome. The c-type cytochrome has been partially purified by Vernon and Kamen (5). The anaerobic light effect clearly involves the c-type cytochrome and the 6-type cyto- chrome. The CO-binding pigment is probably involved also, but its appearance in the anaerobic light effect is masked. In the presence of oxygen the c-type cytochrome is oxidized. This oxidation probably requires the mediation of the CO-binding pigment, but the anaerobic minus aerobic difference spectrum does not specifically indicate changes in the CO-binding pigment or in the 6-type cytochrome upon aeration. Van Niel's (6) mechanism for photosynthesis in purple bacteria is given by the following equations : 4(H20 + h.; > H + OH) (1) 4H + CO2 > (CH2O) + H2O (2) 2(2 OH + H,A > 2H2O + A) (3) 178 J. M. OLSON The spectrophotometric studies of Rhodo spirillum ruhrum and Chromatium support the concept that a cytochrome system mediates the overall reaction between "OH" and H2A shown in equation 3. Discussion Frenkel: I noticed that the reduction in the dark was relatively slow after the light was turned off. I was wondering if the rate was appreciably different from that in the Rhodospirillum which Dr. Chance talked about. Olson : Yes, it is appreciably different. In my illumination experiments, I often had to wait as long as 10 minutes in order to be sure that the trace had returned to the base line because of that very slow reduction. Frenkel : How long did Dr. Chance's experiment take? Chance : A matter of 30 seconds was adequate. Frenkel : Have you any explanation for that? Olson: I think that the cytochrome reductase system is perhaps weaker in Chromatium than in Rhodospirillum. Linschitz : Was there any fast and slow phase in CO? Olson: I do not know, because I took only the dark steady-state difference spectrum between reduced samples with and without CO. I observed no kinetics. Chance : It looks as if, when you turn the light on the CO-binding pigment, this 430-band is o.xidized and, in addition, some cytochrome c, indicated by the sharp 552-mM band. The bands that come in slowly are probably due to the remainder of the cytochrome c and to cytochrome b. References 1. Duysens, L. N. M., Nature, 173, 692 (1954). 2. Chance, B., and Smith, L., Nature, 175, 803 fl955). 3. Chance, B., Rev. Sci. Instr., 22, 634 (1951). 4. Yang, C. C, and Legallais, V., Rev. Sci. Instr., 25, 801 (1954). 5. Vernon, L. P., and Kamen, M. D., J. Biol. Chem., 211, 643 (1954). 6. Van Niel, C. B., Advances in EnzymoL, 1, 324 (1941). i The Reactions of Rhodospirillum rubrum Extract with Cytochrome c LUCILE SMITH,* Molteno Institute, University of Cambridge, Cambridge, England The experiments reported investigated (a) the possible role in the respiratory chain of the soluble cytochrome co which has been isolated from the photosynthetic bacterium Rhodospirillum rubrum (1) and (6) the reactions of extracts of the bacteria on illumination. These experiments represent a continuation of the work of Duysens (2), Vernon and Kamen (3-5), and Chance and Smith (6), which led to the suggestion that one or more cytochromes may be oxidized or reduced in a photochemical reaction. It was hoped that the experi- ments might shed some hght on the oxidation-reduction reactions of the cytochromes, as well as on the possible involvement of these reactions in photosynthesis. DARK REACTIONS OF EXTRACTS OF RHODOSPIRILLUM RUBRUM An extract of Rhodospirillum rubrum was prepared by grinding washed cells with powdered glass. This preparation showed oxygen uptake on addition of a number of substrates. From this extract a particulate fraction was isolated by high-speed centrifugation which took up oxygen only on addition of succinate; the O2 of the particulate suspension oxidizing succinate approached that of the whole cells. Both the extract and the particles contained the chlorophyll and carotenoids of the bacteria. When the extract (or particulate suspension) was added to reduced mammalian cytochrome c or the bacterial cytochrome co in the dark, a slow oxidation of the cytochrome was observed, as found by Vernon and Kamen (3). The rate of oxidation of mammalian cytochrome c was more rapid than that of the bacterial cytochrome C2. With the * Exchange fellow of the British and American Cancer Societies. Present address: Johnson Foundation for Medical Physics, University of Pennsylvania, Philadelphia, Pa. 179 180 L. SMITH washed particle suspension, there was no reduction of cytochrome c in the dark in the presence of cyanide, which inhibits the dark oxidase, but the cytochrome was rapidly reduced on addition of succinate. Thus the soluble cytochrome seemed to be available for reaction with the enzymes on the particles. Quantitative measurements were made to determine whether the rate of oxidation of the cytochrome c could account for the observed rate of oxygen uptake of the same suspension while oxidizing succinate in the dark, as is observed to be the case with the mammalian respira- tory sj^stem. Table I shows a comparison of the observed rate of oxygen uptake of the particles oxidizing succinate and the rate of oxygen uptake calculated from the rate of oxidation of cytochrome c, assuming that one molecule of O2 is equivalent to four molecules of cytochrome c. The data show that the rate of oxidation of 15 nM cytochrome c could account for less than 10% of the respiration; and the rate of oxidation of bacterial cytochrome C2 is even slower. The data argue against the presence in the bacteria of a cytochrome d dark oxidase system similar to that of mammalian tissues. TABLE I. Comparison of Observed Rate of Oxygen Uptake of Rhodospirillum rubrum Extract with Oxj^gen Uptake Calculated from the Rate of Oxidation of Reduced Mammalian Cytochrome c Oxygen uptake calcul xidation of cytochrome jlated from the rate of 3me c, assuming 1 O2 dehydrogenases [See Discussion following paper by Britton Chance, Margareta Baltscheffsky, and Lucile Smith, pp. 197-202.] References 1. Duysens, L. M. N., Nature, 173, 692 (1954). 2. Chance, B., and Smith, L., Nature, 175, 803 (1955). 3. Castor, L. N., and Chance, B., J. Biol. Chem., 217, 453 (1955). 4. Olson, J., this volume. The Effect of Hydrosulfite on the Anaerobic Light Effect BRITTON CHANCE and LUCILE SMITH, Johnson Foundation for Medical Physics, University of Pennsylvania, Philadelphia, Pennsylvania In the course of our studies of the effect of infrared illumination upon the steady-state oxidation-reduction levels of the cytochromes of R. rubrum, we have attempted to determine whether the oxida- tion of cytochromes is prevented by the well-known reducing agent, sodium hydrosulfite (1). A typical experiment is sho^ai in Fig. 1. The tracings were made in the double-beam spectrophotometer by recording the differences of optical density at 430 as compared with 440 ni/z. In the left-hand trace, the aerobic cells exhaust the oxygen dissolved in the solution and become anaerobic. This causes reduction of the hemoprotein ab- sorbing at 430 m/i, as is indicated by the downward sweep of the trace. Illumination with infrared light causes partial oxidation of this pigment and darkness leads to a restoration of the reduced state. This suspension is then shaken vigorously in order to aerate it and is then replaced in the spectrophotometer. Proof of adequate aeration is afforded by the identical levels of optical density in the initial por- tions of the two records. Addition of sohd sodium hydrosulfite to give a final concentration of roughly 1 mM now causes a rapid reduction that proceeds considerably farther than in the first recording. When the trace has reached an end point, the light is turned on, and again an oxidation of the hemoprotein is obtained. Quantitatively, the optical density change on reduction is 18% greater in the presence of hydrosulfite, while the optical density change on infrared illumina- tion increases 5%. In order to show that an excess of hydrosulfite is present, we made a second addition of hydrosulfite to such a solution and still obtained the oxidation reaction upon illumination. At higher hydrosulfite levels the light effect is diminished. Other evidence in favor of an excess of hydrosulfite is based on the fact that the addition of a low concentration of ox3^gen l)y stirring the solution caused no transient oxidation of the cytochromes. 189 190 B. CHANCE AND L. SMITH In order to guard against the possibility that the optical density changes caused by illumination of the solution in Fig. 1 (right) involve different pigments from those of Fig. 1 (left), we have repeated the experiment at various wavelengths and have thereby obtained the difference spectrum plotted in Fig. 2. In this spectrum, the optical density changes recorded on turning the light off are plotted. They Reduction due to respirotion Reduction due to hydrosulfite Fig. 1. A comparison of the anaerobic light effect in a suspension of 22. rubrum in which anaerobiosis has been caused by respiration (left) or by hydrosulfite addition (right) . Optical density increases at 430 m/i corresponding to a down- ward deflection of the trace. correspond to increases of optical density. This spectrum is essentially the same as that obtained in the absence of hydrosulfite; the same pigments are involved in both cases. +0.004 (in excess SO" ) 380 420 500 460 Fig. 2. The difference spectrum caused by infrared illumination of li rubrum suspensions under the conditions of Fig. 1, right. EFFECT OF HYDROSULFITE 191 DISCUSSION Hydrosiilfite readily reduces the ferric forms of hemoglobin, myoglobin, cytochromes, and peroxidases. The only well-documented exception is intact catalase (2). Pigments isolated from R. rubrum are reduced (3). Thus we must seek an explanation for the apparent insensitivity of the light effect in R. rubrum. There are significant differences in the kinetics of reduction of the hemoprotein in the two records of Fig. 1. On the left, the reduction starts slowly and accelerates to a rapid rate. On the right, the reaction starts at the maximum rate, which is more rapid than in the figure on the left. Thereafter a fairly slow and prolonged reaction ensues which pro- ceeds 18% farther than in the left figure. We attribute the first rapid phase to a combination of a direct reaction of hemoprotein on the cell surface with the reducing agent and to the spontaneous reduction caused by exhaustion of oxygen. The slower phase of the reduction reaction is attributed to the slower penetration of hy- drosulfite to more remote parts of the bacterial cell. It is significant that appreciably more hemoprotein is reduced on hydrosulfite addition and that slightly more is available for oxidation by light than in the absence of hydrosulfite. The result that the light-mediated oxidation of hemoprotein proceeds even in the presence of hydrosulfite gives strong support to the idea that oxygen is not evolved in this photosynthetic process (4,5) and further indicates that transfer of oxygen in a bound form from chlorophyll to hemoprotein is unhkely. A hypothesis in accord with these data is that the oxidant is produced within a structure of chlorophyll and hemoprotein that is inaccessible to the concentra- tion of reductant used here. [See Discussion following paper by Britton Chance, Margareta Baltscheffsky, and Lucile Smith, pp. 197-202.] References 1. Chance, B., and Smith, L., Nature, 175, 803 (1955). 2. KeUin, D., and Hartree, E. F., Proc. Roy. Soc. (London), B, 119, 141 (1936). 3. Vernon, L. P., and Kamen, M., /. Biol. Chem., 211, 643 (1954). 4. Van Niel, C. B., Advances in EnzymoL, 1, 236 (1941). 5. Johnston, J. A., and Brown, A. H., Plant Physiol, 29, 177 (1954). The Fast Light Reaction of Extracts and of Inhibited Cell Suspensions BRITTON CHANCE, MARGARETA BAT.TSCHEFFSKY, and LUCILE SMITH, Johnson Fotwdationfor Medical Physics, University of Pennsylvania, Philadelphia , Pennsylvan ia As indicated in the preceding paper, the hemoprotein absorption bands, that disappear upon infrared illumination of anaerobic whole- cell suspensions of R. ruhrum, do so in a two-step reaction that in- volves different members in the sequence of electron transfer reac- tions. It was also found that an effect of the opposite sign is recorded when the cells are treated with phenyl mercuric acetate. In this case a strong absorption band in the region of 430 m/x appears when the aerobic suspension is illuminated and this was then interpreted as a photoreduction of hemoprotein (1). On the other hand, Duysens, who has found that aerobic cells treated for some time with distilled water give spectroscopic responses to infrared illumination similar to those caused by phenyl mercuric acetate, proposes that the absorp- tion band that appears upon infrared illumination is caused by an oxidation product of bacteriochlorophyll instead of by a reduced hemoprotein (2). Extracts of R. rubrum prepared according to Frenkel (3) also show spectroscopic effects upon illumination, which resemble those observed in whole cells treated with phenyl mercuric acetate. This paper compares the various conditions that lead to the appearance of an absorption band in the region of 430 m/x upon infra- red illumination and includes some discussion on the nature of the compound observed. METHOD The double-beam spectrophotometer used in the studies of the whole-cell suspensions was employed (1). The cross-illumination sys- tem used two Wratten 88A filters and a gray filter of density 1 to 5 depending upon the infrared illumination needed to give maximal spectroscopic effect. The monochromatic light intensity was main- tained as low as practicable. 192 FAST LIGHT REACTION 193 Preparation of the materials. The cells were grown as described previously (1). The preparation of the extracts follows as closely as possible the procedures outlined by Frenkel (3). The phosphoryla- tive activity usually corresponded to 22 nM Pyhour for a preparation that gave an optical density reading of 1.0 at 800 m/i. The reactions were carried out in a 0.2 M glycyl-glycine medium of pH 7.4. RESULTS Phenyl mercuric acetate treatment of whole-cell suspensions. The first indication of the fast light reaction was obtained by a treat- ment of the whole cell suspension of R. ruhrum with 0.1 mM phenyl mercuric acetate (1). The rather complex kinetics of this reaction are Fig. 1. The time course of the development of phenyl mercuric acetate inhibi- tor in a suspension of R. ruhrum. The traces also show the effects of infrared illu- mination upon the inhibited cells. indicated by the tracing of Fig. 1. The inhibitor is added to the anaerobic cell suspension and, after a 2-minute lag, the reaction that involves a large decrease of optical density at 430 m/x is com- plete. This decrease is caused by an oxidation of cytochromes as the dehydrogenase activity is inhibited; and, as previous studies showed, considerably more cytochrome Co is oxidized under these conditions (1). Infrared illumination causes an abrupt increase of absorption followed by a much slower reaction and the same sequence of reac- tions is recorded when the light is turned off. The spectrum corre- sponding to the fast phase of the light reaction is plotted in Fig. 2 and consists of a sharp peak at 430 ni/z* and a trough at 385 m/z. * One of us (L. Smith) has restudied this effect and obtains a much broader peak at 432 to 436 m/n. 194 B. CHANCE, M. BALTSCHEFFSKY, L. SMITH Duysens reports that distilled water treatment of the whole cells gives a similar light effect (2) . After incubation with the inhibitor, the kinetics of the fast light reaction is recorded without interference from the slow reaction as e Z +0.005 e Q. O -0.005- Opticol density at 430 m^ increases as light IS turned on 430 m>j 370 400 430 460 Fig. 2. The difference spectra corresponding to the rapid phase of the reaction caused by infrared illumination in Fig. 1. shown in Fig. 3. The records show clearly (1) that no "instan- taneous" processes are involved; (2) that the kinetics of the light reaction differ from those of the dark reaction, the light reaction being ^^/A^^-s-^-^*^ 430-470m;j log Io/I=0.005- Increase of O.D. at 430m>ii 10 sec *i Temp = 7''C -"'^sc^ ^ Tennp. = ~22°C Fig. 3. The kinetics of the effect of infrared illumination upon R. rubrum cells inhibited by 0.1 mM phenyl mercuric acetate. considerably more rapid than the dark reaction, as would be ex- pected if the effect of light is to displace a chemical equilibrium; (5) that temperature affects both the light and dark reactions, al- FAST LIGHT REACTION 195 though the effect upon the latter is considerably greater. Thus the 430 m;u peak is not due to a direct excitation of chlorophyll. Studies of extracts of R. ruhrum. Light effects observed in extracts of R. ruhrum prepared as described here are shown in Fig. 4. Without any additions to the extract, the light effect is similar to that recorded in Fig. 3 for the whole cells treated with phenyl mer- curic acetate. Starting at the left-hand side of the record, a rapid in- crease in absorption at 430 myu occurs when the suspension is illumi- I50;jM on DPNH 340-374m>j 430- 470m;j log lo/I ^0.005 TT Increased O.D. at 430nT>jjM DPNH Fig. 5. The effect of infrared illumi- nation upon the kinetics of DPNH utilization by an extract of R. rubrum. An increase of optical density at 340 m/i corresponds to an upward deflection. nated, and a rapid decrease to the original base Hne is observed in other experiments when the light is turned off. The peak of the difference spectrum for this effect lies at a slightly longer wavelength than that recorded in Fig. 2, but the general nature of the effect is very similar. If now a reducing agent such as DPNH is added, a large increase of optical density is recorded. The peak of the difference spectrum for the changes caused by DPNH addition is about 428 m^u. This is the expected value corresponding to the reduction by DPNH of the pigment that Kamen has isolated from R. rubrum and shown to be reducible by DPNH (4). This pigment is probably identical with the terminal oxidase that we term the "CO-binding pigment" (5). Infra- red illumination now causes a further rapid increase of absorption, 196 B. CHANCE, M. BALTSCHEFFSKY, L. SMITH but in this case there then follows a slow decrease of optical density which is attributed to oxidation of those pigments reduced by DPNH. On cessation of illumination, the rapid effect is again completed be- fore the slow one gets under way. The oxidation-reduction level then returns very nearly to its initial value. Effect of light upon DPNH disappearance. In order to determine whether the rate of DPNH disappearance is affected by infrared il- lumination, we have repeated a portion of the experiment of Fig. 4 in Fig. 5 and have measured light absorption changes at 340 m^u. Addition of DPNH causes an increase of optical density at 340 m/x, recorded as an upward deflection. The utilization of DPNH produces a downward deflection. Illumination causes such a very slight slacken- ing of the rate of DPNH utilization that one may raise the question as to whether DPNH participates in the light-induced reaction. In addition, these effects on DPNH kinetics are irregular; in some cases, an acceleration of DPNH disappearance was noted. INTERPRETATION Our current interpretation of the rapid increases of light absorp- tion at 430 mfj, following infrared illumination of the cell extracts or of the whole cells treated with phenyl mercuric acetate is as fol- lows: the increased absorption corresponds (1) to the reduction of a hemoprotein (£) or to the formation of an intermediate compound of chlorophjdl, possibly an oxidation product as suggested by Duysens. On the one hand, the peak in the region 430 to 434 m/x suggests hemo- protein reduction (except for the broadness of the band) but, on the other hand, the trough at 385 m/x suggests a disappearance of chloro- phyll. There are a number of arguments against the simple explanation that the effects at 430 to 434 m^ are caused by conversion of all the chlorophyll into a new compound. The magnitude of the spectroscopic change (about 0.025) corresponds to a very small fraction of the total chlorophyll content (<1%, see Discussion), and is not increased with stronger infrared illumination. Such a small extent of reaction is inconsistent with the kinetics of the light reaction which is 4 to 9 times the speed of the dark reaction (see Fig. 3) and should correspond to a large degree of completion of the reaction, i.e., conversion of all the dark form of the 430 m^ pigment. Thus, if chlorophyll is this dark form, only about 1% of the total cell chlorophyll can participate in FAST LIGHT REACTION 197 this reaction. Interestingly enough, this amount is of the same order of magnitude as the cytochrome content. Some further data obtained, under the conditions of Fig. 3, lead to the possibility that the effect at 430 m^ has somewhat different kinetics from that at 385 m^u ; the back reaction at 7° has roughly twice the initial rate of that at 430 m/i. Thus two or more components may be involved, and various hypoth- eses for this interesting effect are still under active consideration. Discussion Granick : In Fig. 3 (p. 187) you had a 520-Tnfj. band that was going up and down. What was that? Chance : I cannot say it is all due to cytochrome, although a part of it may be attributed to the beta bands of cytochromes b and c; there are cytochromes which do have relatively large beta bands, for example, cytochrome 6<. Some might be due to the 520-mM band seen in Chlorella. But, of course, the 520-mM band of Chlorella appears upon illumination, whereas that in R. rubrum disappears and has no related effects at 475 mju. It is unUkely that the effect is due to heat for there is the same amount of heat aerobically and anaerobically, whereas there is no spectroscopic effect aerobically and there is a large spectroscopic effect anaerobi- cally. If heat were the cause of these effects, you would expect the same spectro- scopic effect aerobically as anaerobically. Granick: When you add hydrosulfite you reduce the cytochrome at the same time that you remove any oxygen that might be there. Was there something else that showed up by the addition of hydrosulfite? Chance : Yes, there was a further reduction of cytochrome at 430 m^. Also this experiment suggests that oxygen is not the oxidant for the light effect. Frenkel : How long does this hydrosulfite remain in solution as such? Does it decompose very rapidly? Chance : It is fairly stable for the three-minute interval. Benson : Does it get into your cells? Chance : Fig. 1 (p. 190) indicates that it does. It certainly can cause a larger and more rapid reduction of the cytochromes than the endogenous process. Strehler : Did you say the O2 is consumed faster than it can be by the endoge- nous process alone? Chance : I compared the slopes of the two traces of Fig. 1 (p. 190) and the reduction is more rapid. Strehler : But there is extracellular oxygen they have to consume. Gaffron : If you bubble nitrogen through, why do you have to add the hydro- sulfite? Chance : We add hydrosulfite to see if light can cause oxidation of cytochromes in the presence of this reducing agent. The hydrosulfite affects pigments that are presumably inside the cell, and therefore should also be able to react with any oxygen produced in the cell. Gaffron: What would a little oxygen produced by illumination do? Do you think this would be taken care of completely? 198 B. CHANCE, M. BALTSCHEFFSKY, L. SMITH Chance : Yes, by hydrosulfite. Gafifron: Nobody would really ever expect to get oxygen in Rhodospirillum rubrum. Brown : On the right-hand side of Fig. 1 (p. 190) you are removing oxygen from the environment around the cell. Right? Chance : Yes. Brown : Therefore, the cells can effect the internal change much more rapidly than if they have to use up all the oxygen in the environment as they do on the left? Chance : Somewhat more rapidly. Brown : You have no evidence, as I see it, that the material penetrates the cell. Chance: Our evidence is the fact that the spectroscopic shift is greater with hydrosulfite than with anaerobiosis. Brown : That is significant then? Chance : The effect is 18% greater with hydrosulfite. With a number of bac- teria hydrosulfite addition reveals pigments that are otherwise not reducible under anaerobiosis. Amon : To pursue this point, this is a quantitative change on the right, isn't it? Chance : Yes, this is a quantitative change. Amon : Could that mean that you have removed the oxygen more completely here in the presence of hydrosulfite? Chance : I think we have reduced pigments not reducible under anaerobiosis alone. Amon : Will you state the reasons why you don't think the difference could result from the more efficient removal of oxygen on the right and less efficient on the left? Chance: The respiratory chain is one of the most efficient oxygen removers known; the reaction has not been shown to be reversible. Gaffron : How far apart are the two records in time? I mean, your base line is not dropping. Chance : I think this was a couple of minutes. In other words, I shook the sus- pension to oxygenate it and re-establish the oxidized level, which checks very closely. The base line was not dropping. Allen: Is there any possibility that the infrared-induced transformation of some of these pigments might be pH-sensitive? Chance : We have some data from yeast cells obtained by changing the internal pH. If these cytochromes are like the ones in yeast, a spectral shift is unlikely. Allen : Dr. Hendley has found rather large "gushes" of acid on illuminating some strains of purple bacteria. These very much resemble the changes in transmission that you get initially on turning the light on. I wonder whether there is a connec- tion between the gushes and your absorption changes. Chance: We would expect a change of pH to affect the pyridine nucleotide because it is in equilibrium with hydrogen ions. In yeast cells we find that the P5Tidine nucleotides are oxidized when acid penetrates the cells, but the cyto- chromes are not oxidized or reduced and thus would not respond spectroscopically to the acid gushes. Unfortunately, we don't get any response at 340 vein with the FAST LIGHT REACTIOxX 199 Rhodosjyirillum rubrum, so could not use DPNH as an indicator of intracellular pH changes. Strehler: Do you get any evidence that a substance other than that from p- phenyl mercuric acetate treated cells is being reduced in the light? Chance : No, not to any appreciable extent in R. rubrum. The spectroscopic effects of Fig. 1 (p. 193) observed in the aerobic cells don't appear until an ap- preciable portion of the cytochrome has been reduced under anaerobiosis. Frenkel: Have you observed any spectral changes in cell-free preparations? Chance: Yes, but they are very different from those observed in whole cells. A photooxidation reaction is observed when you reduce the cell-free preparation with DPNH. Also the cell-free preparations show an aerobic light effect similar to that observed in phenyl mercuric acetate treated whole cells. Frenkel : I just ran an experiment which somewhat reminds me of your reduc- tion in the dark. I refer to the reduction of cytochrome added to a cell-free prepara- tion after the hght period. The reaction appears to be almost completely anaerobic, but there must be a small amount of oxygen present because on illumination the cytochrome is oxidized. I don't know whether it goes as you, or as Vernon and Kamen, postulate. The interesting thing is that, after the light is turned off, the cytochrome is reduced at about twice the initial rate. So, this effect resembles the one you obtained. Chance: Yes, our records show a rapid reduction of the intracellular cjrto- chromes when the light is turned off, and our results with the DPNH-reduced extract are in agreement, too. Frenkel: I just wondered whether this was evidence for the formation of a reducing power in the light. Chance : Reducing power is formed, but our data do not say where. Remember that the dark oxidation of DPNH is not slowed by illumination (Fig. 5, p. 195). Amon : Did I understand you to say that, in cell-free preparations, you observed reduction rather than oxidation? Is that correct? Chance : A change at 430 mp in cell-free preparations is also observed in whole bacteria inhibited by phenyl mercuric acetate. This effect may be due either to a reduction of cytochrome or to an oxidation of bacterial chlorophyll. On the other hand, add a reductant like DPNH to the extract and you will still observe this 430 m/z effect. In this case it wiA be followed by a slow oxidation reaction. Thus one may question whether a reduction occurs. Amon: Will you just define what you mean by 430 effect? Chance: There are two hypotheses: (1) an oxidation of chlorophjdl or (2) a reduction of hemoprotein. Amon : Is the reducing effect the last one? Chance : Yes, although this is a less plausible hypothesis. Frank Allen: Is it possible to calculate in any way the extent of the change that one would expect in chlorophyll under normal conditions and to compare this with the changes that are observed whose nature is undetermined to see if we are at least in the right ball park? Rabinowitch: What you mean by "under normal conditions"? Frank Allen : Any conditions you can do it under. Given an illuminated plant, can we not assume that the changes we observe are related to the number of 200 B. CHANCE, M. BALTSCHEFFSKY, L. SMITH molecules that become transformed? Are these changes of sufficient magnitude? I don't know. Rabinowitch : You want to see the changes in the pigments responsible for the changes of absorption; but if you don't know the absorption coefficient of your pigment you just cannot separate the product into two factors. Therefore, you don't know if you have a very small change of the whole material or if a very small part of the material is transformed into something which has a quite different absorption coefficient. Lucile Smith : Let us say, if the changes that Dr. Duysens gets and possibly the change observed in the extract of Rhodospirillum are changes in chlorophyll, and if you assume that the Soret peak of chlorophyll is similar to that of cyto- chromes in that region, there are the changes of the same order of magnitude as the changes in cytochromes. Duysens: The changes are only 1% of the absorption of bacteriochlorophyll. I think what Dr. Allen wants to ask is whether there is a change of all bacterio- chlorophyll molecules or a change in only a fraction of the bacteriochlorophyll molecules which is very large. At present this question cannot be answered with certainty. Brown : There is a related question. That is, if you attempt to compute potential changes from these spectral differences and you calculate for the total quantity of material present, the change might be very small. If you calculate on the basis of assumed compartmentalization the change can be enormous. I don't think there is any easy way out of that dilemma. Wassink : Just a comment on this. Under normal light intensities only a small fraction of the chlorophyll will be excited at a certain moment, and you cannot assume that in the presence of an acceptor it stays in the excited stage. Rabinowitch: Not in the excited, but in the chemically changed state, yes. Wassink : Can you? Rabinowitch : Yes, why not? It depends on the lifetime. Kamen : Dr. Vernon and I isolated the pigment which is presumably identified with the Rhodospirillum rubrum. It turned out to have a broad band. The region at 550 mju, rather than that at 570 mju, was associated with the 428 m/u peak. It seems to me, therefore, that there should probably not be an anomaly here, at least so far as this is concerned. But it is a broad band and not sharp. It could be that the fast component that you observed, or that the portion of the band at 550 m/i which was fast, could have come from this broad band. You would not see it so easily with the resolution of the apparatus now, but I don't think there is any real anomaly. The hemoproteins in these bacteria are not identical with the classical hemoproteins. Any arguments based on what you would expect from the classical type of oxidase need not be of concern to us now. Chance: On the other hand, surely there would be a sharp 550 band due to cytochrome c. Kamen : And there is. Chance : You also state that the 428 might have a broad alpha band. Kamen : It does have a broad alpha band in the isolated pigment. I don't deny that in the course of the isolation of this pigment we may have changed it con- siderably because we used rather drastic methods. This can also explain the FAST LIGHT REACTION 201 potential of about 200 mv., which may not be correct. However, there is a broad band at 550; and, if we had been more careful in isolating the pigment, we could have sharpened the band. James Smith : The question has been asked about the positions of the bacterio- chlorophyll and bacteriopheoph_ytin absorption bands in the middle of their spectra. In ether the positions of the bands, in m^u, and their specific absorption coefficients (given in parentheses), are as follows: bacteriochlorophyll 577(22.9), 530(3.0), 591.5(52.8); bacteriopheophytin 525(31.9), 495(6.5), 384.5(70.6). As we know, the absorption bands maj'^ lie at other positions in living organisms. + AD 0.140 r ot 550 m;u 0.120 - 0100 ■ 0.080 - 060 0.040 0.020 5min. light 20 40 120 140 160 180 60 80 100 TIME ^MINUTES) Fig. 6. Dark reduction and light oxidation of mammalian cytochrome c by cell-free preparations of Rhodospirillum rubnim showing the phenomenon of accelerated dark reduction after a period of illumination. Frenkel: (comments added in proof): I would like to add some observations on the ability of cell-free preparations of Rhodospirillum ruhrum to reduce added cytochrome c in the dark and on the phenomenon of accelerated dark reduction of cytochrome c after illumination of this system. Cell-free preparations in 0.2 M glj'cylglycine buffer (pH 7.5) with an optical density of 0.3 at 800 m/i are placed in Thunberg tubcis with or without mammalian cytochrome c (0.5 n^l in a final volume of 3.0 ml,) which are then evacuated. In Fig. 6 above, slope A is the initial rate of dark reduction of cj^tochrome (1.4 X 10~' optical density unit per minute at 550 m^u). After 80 minutes in the dark, the tubes are illuminated with in- candescent light for 5 minutes, which brings about oxidation of the added cyto- 202 B. CHANCPl, M. BALTSCHEFFSKY, L. SMITH chrome, presumably by a residual trace of molecular oxygen in a manner described by Vernon and Kamen in 1953. Slope B is the initial rate of subsequent dark reduc- tion of the cytochrome after illumination (2.75 X 10~* optical density unit per minute). Partially purified preparations of the photochemically active particles which are virtually free of cytochrome oxidase activity still show this phenome- non of accelerated dark reduction of cytochrome c after illumination. References 1. Chance, B., and Smith, L., Nature, 175, 803 (1955). 2. Duysens, L. M. N., this volume. 3. Frenkel, J., /. Arn. Cheni. Soc, 76, 5568 (1954). 4. Kamen, M. D., this volume. 5. Castor, L. N., and Chance, B., /. Biol Chem., 217, 453 (1955). Part IV "DARK'' REACTIONS 1, Fixation of Carbon Dioxide 2, Photoreduction with Various Reductants 3, Reduction of Various Oxidants 4, Reactions in Chloroplasts and Cell Extracts 5, Phosphate Metabolism 1. Fixation of Carbon Dioxide The ** Background" CO2 Fixation Occurring in Green Cells and Its Possible Relation to the Mechanism of Photosynthesis SHIGETOH MIYACHI, TOYOYASU HIROIvAWA, and HIROSHI TAMIYA, The Tokugawa Institute for Biological Research, and Department of Botany, Faculty of Science, University of Tokyo When preilluminated algal cells are brought into contact with Ci^02 in the dark, C^^ is fixed in two different ways (1,2). One is a fixation caused by some photogenetic agent (s), and the other is a background fixation which occurs also in nonpreilluminated cells. It may be reasonable to assume that different agents (or systems of agents) are acting in these two kinds of CO2 fixation. Let us denote the photogenetic agent (s) by R and the agent (s) acting in the back- ground fixation by D. It has been shown (3,4) that, whereas the re- action between R and C^^02 proceeds very fast and is completed within about 30 seconds at 25 °C., the reaction between D and €^"02 is considerably slower. If the radioactivity fixed by D is denoted by r, it is almost a linear function of the time of contact of cells with C^*02 in the dark; namely, dr/dt = k[D], or r = kt[D]. The relative concentration of D in nonpreilluminated cells can be gauged by measuring the radioactivity (r) fixed within a definite time of con- tact of cells with 0^*02 in the dark. The level of R in preilluminated cells can be determined by measuring the C^* fixation in 30 seconds since, within such a short period, the C^* fixed by the background capacity is negligible. If the exposure to C'^02 is longer, say 5 minutes, background fixation becomes significant and the C^* fixed will be the sum of a rapid light-induced action of R and a slower almost linear background fixation attributable to D. By subtracting the radio- activity fixed in 30 seconds from that fixed in 5 minutes, we can de- termine the background C^^Oa-fixing capacity (on 5 minutes basis) existing in preilluminated cells. Using these methods, we investi- 205 206 S. MIYACHI, T. HIUOKAW A, H. TAMIYA gated the process of background CO2 fixation as it is influenced by- oxygen and light. The experimental organism was Chlorella ellipsoidea and the ap- paratus used was virtually the same as that employed by Benson et at. (2, see also 4). Algal cells were suspended in 0.02 M phosphate buffer of pH 7, and the final concentration of labeled NaHCOs given to the algae was 6.9 to 9.6 X 10-W. The experimental temperature was 25°C. EFFECT OF OXYGEN It was found, in conformity with the earlier findings of Brown, Fager, and Gaffron (5,6), that the level of D in the dark was mark- cpm 20 40 TIME IN MINUTES Fig. 1. Dark Ci^02-fixation by algal cells that have been kept dark (without pro- vision of CO2), first in N2 and later in O2 atmosphere. Ordinate: cpm of C'^ fixed in 20 minutes' contact with Ci^02 in the dark; abscissa: time of preincubation in the dark with flushing of N2 and O2 as indicated. edly increased when the cells were exposed to oxygen. In the ex- periment shown by Fig. 1, the algal suspension was first aerated with nitrogen, and at the expiration of 20 minutes the nitrogen was promptly replaced with oxygen, all procedures being carried out in the dark without provision of CO2. At various intervals during this process, aliquots of algal suspension were brought into contact with C"02 in the dark, and the radioactivity fixed in 20 minutes was measured. It was found that at the transition from N? to O2 atmos- phere, there occurred an abrupt increase in the relative D level, which, however, gradually tapered off and eventually attained a THE BACKGROUND CO2 FIXATION 207 certain steady level. The final r-valiie established in O2 atmosphere was always higher than the steady r-level obtained in the N2 at- mosphere. In the particular experiment illustrated by Fig. 1 the level obtained in Oj was exceptionally low since the value was generally from 2 to 10 times higher than the r level in N2 (c/. Fig. 4). This factor depended upon the prehistory of the algae. EFFECT OF LIGHT The effect of light upon the D level was investigated with the results shown in Fig. 2. The cells were first kept in the dark (flushed 20 40 SO TIME IN MINUTES Fig. 2. Effects of transition from dark to light and from light to dark upon the background C"02-fixing capacity. Measurement of background Ci*02-fixing capacity {n) was made by subtracting R (i.e., the radioactivity fixed in 30-second contact with Ci^02 in the dark) from R + r;, (i.e., the radioactivity fixed in 5 min- utes under the same condition). The insets show, on a magnified scale, the events occurring at earlier stages of illumination. with N2) for 50 minutes to make them adapt to the darkness and to anaerobic conditions. Immediately after the Hght was turned on, the 208 S. MIYACIII. T. HIROKAWA, H. TAMIYA development of C**-fixing capacity was measured, by transferring at intervals aliquots of cell suspension into C^^O-z solution kept in the dark, and by measuring the C^^ fixed within 30 seconds (R), on the one hand, and that fixed within 5 minutes (R+n), on the other. After 60 minutes of illumination, when the 7^ level was at the stationary state, the light was turned off, and by the same method as above, the course of decay of C^^-fixing power was traced. As reported elsewhere (3), the R level showed an induction period with two depressions at the beginning of illumination. When the light was turned off at the sta- tionary state in the light, the decay of R occurred immediately, as it has already been observed by Calvin and Benson (1). A quite dif- ferent phenomenon was observed in the decay-curve which was measured by 5-minute C^^02 fixation; immediately after cessation of illumination, there occurred an abrupt temporary increase of the C'*-fixing capacity, after which the capacity decayed steeply to attain eventually a steady level. On the upper part of the figure, the difference between the (R+r^) curve and the R curve is plotted. This difference curve, representing the time course of rg, shows us: (1) that on illuminating the dark- adapted algae, the background fixing capacity decreased temporarily, then increased, and eventually attained a steady level, which was considerably higher than the steady level observed in the darkness; and (2) that when the light was turned off, the background ca- pacity increased abruptly, and after attaining a certain maximum value, it gradually decreased to a final level which was lower than the steady level observed during the illumination. PHOTOCHEMICAL CO2 OUTBURST FROM THE PRODUCT OF BACKGROUND CO2 FIXATION An interesting phenomenon was observed when, during the course of dark C^^02 fixation, the cells were subjected to brief illumination. The experimental results are presented in Fig. 3. To the algal sus- pension which has previously been kept dark for 25 minutes in N2 atmosphere, C'^02 was added and the course of dark C'^ fixation was followed. As shown by the curve on the lowest part of the figure, the C* fixation proceeded almost linearly wdth time when the dark- ness was continued uninterrupted. When, however, during the course of this process, the algae were exposed to a 1 -second light THE BACKGROUND CO2 FIXATION 209 flash, there occurred a distinct break in the course of C^^-fixation, and the subseciuent ascension of the curve started from a point which was lower than that observed immediately before the flashing of light. If the duration of flashing light was prolonged to 3 seconds, the break of the curve became less abrupt, but the lowering of the subsequent cpiii 120 80 40 OJ X o 10 SLC. I LASH II, TIME (IN MINUTES) OF CONTACT OF CELLS WITH C'*02 IN THE DARK ( IN N2 ATMOSPHERE ) Fig. 3. C"02 outburst from the product of background 0^*0^ fixation occurring in N2 atmosphere. C^^Oo was supplied in the dark after the cells were dark- adapted for 25 minutes in N2 atmosphere. Light flashes of indicated duration were applied at the points shown by arrows. curve was more pronounced than was the case with the 1-second flash. When the flash duration was 10 seconds, the break of the curve practically disappeared, and the subsequent part of the curve showed a peculiar course, ascending with increasing steepness with the lapse of time. The break in C^'' fixation time course caused by 1- or 3- second flashes must be taken as evidence that a part of C^* fixed 210 S. MIYACHI, T. HIROKAWA, H. TAMIYA during the preceding period was removed by the effect of brief il- lumination. Conceivably, there occurred a decarboxylation of some carboxylated compound formed by the preceding background C^^02 fixation. The fact that the breaking effect became more and more 5 10 15 20 25 30 TIME ( IN MINUTES) OF CONTACT OF CELLS WITH C'*02 IN THE DARK 'IN 02 ATMOSPHERE) Fig. 4. Effect of light flash upon the background of C'^Oz fixation occurring in O2 atmosphere. C'''0- was supplied in the dark after the cells were dark-adapted for 25 minutes in O2 atmosphere. Flashing light of indicated din-ation was applied at the point shown by the arrow. obscure with the increase of flash duration may be due to the can- celing of the effect by the formation of light-induced C'*02-fixing power, which became stronger at longer flash duration. In the experiment described above, C'^02 was supplied to the algal suspension in the absence of oxygen. When similar experiments were performed in O2 atmosphere, the results obtained were entirely dif- ferent. As may be seen from the results shown in Fig. 4, there was no THE BACKGROUND CO2 FIXATION 21 1 indication at all for the occurrence of photochemical decarboxylation when algae were given C'*02 in O2 atmosphere. DISCUSSION AND SUMMARY We have seen that the C^^02 fixation by green cells in the dark occurs much faster in O2 atmosphere than in N2 atmosphere. By applying flashing light lasting 1 to 3 seconds it w^as demonstrated that a part of C* fixed in N2 was photochemically decarboxylated, whereas no such hght effect was detectable when C^'*02 fixation occurred in O2 atmosphere. These facts indicate that the processes of dark CO2 fixation are different under aerobic and anaerobic condi- tions. Indeed, it was observed by Brown, Fager, and Gaffron (6) that the products of background C^*02 fixation were different accord- ing to the presence or absence of oxygen; namely, whereas the C" fixed under anaerobic condition was all found in the water-soluble fraction, those fixed under aerobic condition were both in water- soluble and water-insoluble fractions, being incorporated even m fats, starch, and proteins. In view of these observations our experi- mental results may be explamed as follows. Under anaerobic condi- tions the products of background CO2 fixation may remain mostly in the state of rather simple carboxylated compounds, at least part of which are sensitive to photodecarboxylation. In the presence of oxygen, these compounds will be subject to further transformations, the primary step of which may be a reduction or exchange of car- boxyl groups, so that the decarboxylating effect of light, if any, will become undetectable by the tracer technique we applied. By the "subtracting method" we used, it was shown (1) that when algal cells w^ere illuminated after prolonged darkness, the background C02-fixing capacity decreased temporarily, and then increased gradually to attain a steady level which was considerably higher than the steady level observed in the darkness, and (2) that when the light was turned off after continued illumination, the background C02-fixing capacity increased temporarily, and, after attaining a certain maximum value, it gradually decreased to a final level which was lower than the steady level observed during the illumination. The temporary decrease of C02-fixing capacity occurring immedi- ately after transition from dark to light, as well as the photochemical decarboxylation of some product of background CO2 fixation, may 212 S. MIYACHI, T. HIROKAWA, H. TAMIYA probably be related to the CO2 outburst observed by Emerson and Lewis (7) and to the lag of CO2 uptake occurring at the beginning of illumination, which was observed by McAlister and Myers (8-10), Aufdemgarten (11, 12), and van der Veen (13, 14). These authors have also observed that when light was turned off after prolonged illumination there occurred a "gulp" of CO2 or a lag of CO2 produc- tion, which may be accounted for by the temporary increase of back- ground C02-fixing capacity occurring immediately after the transi- tion from light to dark. The fact that the background C02-fixing capacity is largely influenced by light may be taken as an evidence for its being related directly or indirectly to the mechanism of normal photosynthesis. Acknowledgment. The experiments reported here were assisted by grants from the Ministry of Education and from the Rockefeller Foundation. The authors are grateful for the advice of Dr. A. H. Brown in the preparation of this paper. References 1. Calvin, M., and Benson, A. A., Science, 107, 476 (1948). 2. Benson, A. A., Calvin, M., Haas, V. A., Aronoff, S., Hall, A. G., Bassham, J. A., and Weigl, J. W., "C" in photosjmthesis," in Photosynthesis in Plants, J. Franck and W. E. Loomis, eds., p. 381. Iowa State College Press, Ames, 1949. :i Tamiya, H., Miyachi, S., and Hirokawa, T., "Some new observations in preillumination experiments using Carbon 14," in this volume, p. 213. 4. Miyachi, S., Izawa, S., and Tamiya, H., J. Biochem. (Japan), 4^, 221 (1955). 5. Brown, A. H., Fager, E. W., and Gaffron, H., "Kinetics of a photochemical intermediate in photosynthesis," in Photosynthesis in Plants, J. Franck and W. E. Loomis, eds., p. 403. Iowa State College Press, Ames, 1949. 6. Brown, A. H., Fager, E. W., and Gaffron, H., Arch. Biochem., 19, 407 (1948). 7. Emerson, R., and Lewis, C. M., A7n. J. Botany, 28, 789 (1941). 8. McAlister, E. D., J. Gen. Physiol., 22, 613 (1939), 9. McAUster, E. D., and Myers, J., Smithsonian Misc. Collections, 99, No. 6, 1 (1940). 10. McAUster, E. D., and Myers, J., Science, 92, 241 (1940). 11. Aufdemgarten, H., Planta, 29, 643 (1939). 12. Aufdemgarten, H., Planta, SO, 343 (1939). 13. van der Veen, R., Physiol. Plantarum, 2, 217 (1949). 14. van der Veen, R., Physiol. Plantarum, 2, 287 (1949). Some New Preillumination Experiments with Carbon- 14 HIROSHI TAMIYA, SHIGETOH MIYACHI, and TOYOYASU HIRO- IvAWA, The Tokugawa Institute for Biological Research and Department oj Botany, Faculty of Science, University of Tokyo Using C^^ as a tracer Calvin and Benson (1,2) found that the CO2- fixing capacity of green cells increased markedly when the cells were illuminated immediately prior to provision of CO2 in the dark ("pre- illumination"). This C02-fixing capacity gradually decreased in i_JO_„S..J 10 20 30 40 TIME (IN MINUTES) OF PRE -ILLUMINATION Fig. 1. Induction phenomenon in the process of formation of R in the light. Ordinate: cpm of C^* fixed in 30 seconds in the dark; abscissa: the time which the cells spent in the light and dark, as indicated, prior to the provision of C'^02. Experiments illustrated in the upper and lower parts of the figure were performed with different algal samples. darkness, and it was inferred that during illumination the concen- tration of the agent in cells responsible for the CO2 fixation is de- termined by relative rates of photochemical formation and non- 213 214 H. TAMIYA, K. MIYACHI, T. HIROKAWA photochemical decay. We have made studies on the processes of for- mation and decay of this agent which, for convenience, we shall de- note as R. In this paper we report the results of experiments which lead to the conclusion that R is formed photochemically concomitant with oxygen evolution, reacts with oxygen and quinone, functions as a reducing agent in CO2 fixation, and is involved in the cyanide inhibi- tion of photosynthesis. 10 20 30 " 50 TIME IN MINUTES 60 70 Fig. 2 Effect of air on the course of formation of R in the Ught. Air was intro duced 45 minutes before the beginning of illumination. The experimental organism was Chlorella ellipsoidea, w^hich was grown as previously described. In most cases the experimental tem- perature was 25 °C. Preillumination was accomplished while the sus- pension was flushed with N2. The C^*02 was provided in the dark in an atmosphere of N2. We were able to confirm that nearly all the effect of preillumina- tion is achieved within 30 seconds of light and that background fix- ation by nonpreilluminated cells is negligible in short time intervals. Consequently the C^'' fixed in 30 seconds of light is taken as a meas- ure of R. Figure 1 shows that, after prolonged darkness, R was formed with a characteristic induction period. Two transient depressions in the time course of R increase may be noted — the first occurring at about the 10th to the 15th second, the other between the 2nd and 5th minute. The steady-state R level obtained in the light was about 5 to 7 X 10~^ molar C02-equivalent per gram dry weight of algae. (;arbon-14 preillitmination experimknts 2 1 f) 20 40 60 80 100 TIME IN MINUTES Fig. 3. Effect of air on the course of formation of R in the light. Temperature: 20 °C. Curve A: algal suspension was flushed with N2 throughout the experi- ment; Curve B: algal suspension was flushed first with N2 and later (at the time indicated by the arrow) with air; Curve C: algal suspension was flushed with air from the beginning of illumination. 20 40 TIME IN MINUTES Fig. 4. Effect (in the light) of varying concentrations of O2 on R pre-formed in N2 atmosphere. Decay of R in the dark normally was complete in some 15 minutes. In the presence of oxygen R was considerably depressed. Figure 2 shows that the initial course of R formation (including the char- 216 H. TAMIYA, S. iMIYACIIl, T. IIIIU)KA\\ A TIME IN MINUTES Fig. 5. Effect of O, on the decay of R in the dark. Temperature: 20 °C. Abscissa: the time which the cells spent in the light and dark, and in N2 and O2, as indicated, prior to the provision of C^''02. - LIGHT QUINONE ADDED V NONE 40 GO 60 TIME IN MINUTES ion Fig. 6. Effect (in the light) of varjnng concentrations of quinone on the i?-level which had previously attained the stationary value in the absence of quinone. acteristic second transient) was not affected by oxygen but the final R level was less in air than in N2. If R was formed anaerobically and the suspension later flushed with air the same final steady-state value of R was obtained as shown by Fig. 3. The effect of oxygen in depress- CARBON-14 PREILLUMINATION EXPERIMENTS 217 ing the R level increased with O2 partial pressure (Fig. 4) and was reversible. Figure 5 shows that the decay of R was markedly accelerated by the presence of O2. Quinone reacts with R in much the same manner as does O2 except that, in high quinone concentration, inhibition is irreversible (Fig. 6). The C02-fixing power of cells is not restored even by repeated wash- ings wuth phosphate buffer after light exposure in the presence of 20 40 50 GO TIME IN MINUTES Fig. 7. Effect of quinone (lO"''-' M) on the decay of R in the dark. The algae were first illuminated for 50 minutes in the absence of quinone, and simultane- ously with turning off the light, quinone was added and the subsequent fate of R was followed by measuring 30-second Ci*02-fixation in the dark. 2 X 10-* M quinone. With 5 X 10"* M quinone (Fig. 7) the decay of the R level in the dark is greatly accelerated, as is the case for the analogous experiment using O2 instead of quinone. When we investigated the effect of cyanide on the processes of for- mation and decay of R, it appeared that this poison did not affect formation of R, except in so far as the second induction was eliminated. This latter observation corresponds to the observa- tions of McAlister and Myers (3) and of Aufdemgarten (4) on the influence of cyanide on the induction phenomenon occurring in the photosynthetic CO2 uptake. However, when R was formed by pre- illumination for 50 minutes in the absence of cyanide and subse- quently measured by exposure of the cells to C'^O? in the presence of varying amounts of cyanide, it was observed, in confirmation of 218 II. TAMIYA, S. MIYACHl, T. HIROKAWA Gaffron et at. (5), that the initial rate of dark CO2 incorporation was cyanide-sensitive. Upon reducing the C"02 fixation time in the dark to 10 seconds (instead of 30 seconds) the initial rate of CO2 fixation was determined in the presence of varying amounts of cyanide and at various times after the end of illumination. During the course of 1 I 2 3 4- s TIME IN MINUTES IN THE DARK Fig. 8. Effect of cyanide on the decay of R in the dark. Algae were first illumi- nated for 50 minutes in the absence of cyanide; and simultaneously with turning off the light, cyanide was added and the subsequent fate of R was followed by measuring C* fixed in 10 seconds in the dark. this 10-second exposure to C^^02 it was determined by other experi- ments that fixation was linear with time. Figure 8 shows that the decay of COj-fixing capacity was an increasing function of cyanide concentration. The upper part of Fig. 8 is a semilogarithmic plot of these data indicating approximately first-order decay. It can be shown from kinetic considerations that, irrespective of the difference in the rate of reaction between R and C^'^Oo in the CARBON-14 PREILLUMINATION EXPERIMENTS 210 dark, the tangents of the curves in the above-mentioned semilogarith- mic plot represent rates of decay of R in the darkness. The effect of cj^anide on the decay of R can be determined by comparing the tangents of these curves. Once the magnitude of the effect of cyanide upon the decay of R is kno\Mi, the suppressing effect of cyanide on the steady-state R level can be estimated, since this level is de- termined by the relative rates of formation and decay. The results of this computation are sho^^^l in Fig. 9. For the sake of comparison, corresponding data are included showing the effect of cyanide on the 10"* KT* 10"' I0-' CONCENTRATION CM) OF CYANIDE Fig. 9. Concentration-inhibition curves of cyanide acting upon (i) the R-CO2- reaction, (ii) stationary i2-level in the Hght, (iii) normal photosynthesis, measured manometrically under the condition of hght- and C02-saturation, (iv) catalytic H2O2 decomposition by intact Chlorella cells, measured titrimetrically with KMn04 after bringing the cells into contact with 0.015 M H2O2 for 30 seconds. All meas- urements were made at 25°C. and pH 7.0. The degree of inhibition represents the value 1 — vg/v, where Vg and v are the reaction rates or the i2-levels in the presence and absence, respectively, of cyanide. reaction of R with C^^02, on the catalytic decomposition of H2O2 by intact Chlorella cells, and on normal photosynthesis under conditions of light and CO2 saturation. From these data w-e conclude that: (a) cyanide has a dual effect upon R; first an acceleration of its decay tending to lower the steady-state level of R in the hght, and second, a hindrance of the reaction between R and C^''02, and (h) the sensi- tivity of the R level toward cyanide is greater than that of the R-CO2 reaction, and indeed it is approximately the same as those of photo- synthesis and of the catalytic decomposition of H2O2. We have seen that there are three kinds of reagents whose presence entails more or less marked decrease of the concentration of R in the 220 H. TAMIYA, S. MIYACHI, T. HIROKAWA algae, viz., (1) CO2, (^) oxidants such as quinone or oxygen, and (5) cyanide. It is unlikely that these agents of entirely different nature would act upon R by the same chemical mechanism. The paradoxical facts can, however, receive a coherent explanation if we make the following assumptions: a. R is a reducing agent which is formed photochemically from its precursor P (an oxidized form of R) by a reaction accompanied by a liberation of oxygen, such as: p _f. H2O _J!f!^ R (or PH2) + V2O2 (I) h. R reacts with oxygen in the manner : /e + O2 -^ p + H2O2 (II) followed by the catalytic decomposition of H2O2: catalase „ , „ /txx\ H2O2 > H2O + V2O2 (HI) and, in so far as catalase functions normally, the latter process oc- curs much faster than the former, so that the overall reaction pro- ceeds according to: R + V2O2 -> P + H2O (IV) c. R reacts also with other oxidizing agents such as quinone (Q) and H2O2 in the manner: R+Q-*P + QB.t (V) R + H2O2 -*P + 2H2O (VI) d. In the presence of CO2, R is involved in the following sequence of reactions which were postulated by Bassham, Benson et at. (2) :* CO2 + RDP -* 2PGA (VII) 2PGA + 2/2 ^ 2TP + 2P (VIII) 2TP > V5RDP (IX) V5RDP + V6CO2 -* VsPGA (VII) and among these reactions (VIII) is the rate-determining step,t so that the overall reaction proceeds according to: R + VsCO. -^ P + VaPGA (X) * Abbreviations used: RDP for ribulose-diphosphate, PGA for phospho- glyceric acid, and TP for triose-phosphate. t Relative rates of these reactions are assumed to be in the order: (VII) > (IX) > (VIII). CARBON-14 PREILLUMINATION EXPERIMENTS 221 e. Under ordinary experimental conditions (i.e., in the concentra- tion range occurring in ordinary experiments), the reactivity of re- agents acting upon R increases progressively in the order: 0224-hour lag in photo- 220 -N. K. TOIilSKFtT synthesis during the greening of an etiolated plant and the stepwise acquisition of the ability to synthesize the photosynthetic organic products. The induction period in photosynthesis by a fully grown plant may be a measure of the rapidity with which a plant can build up the pool size of carbon compounds provided all the enzymes are present. The slow stepwise build-up of photosynthesis during green- ing of an etiolated plant would appear to be limited by other proc- esses such as synthesis of adequate amounts of enzymes to catalyze some of these reactions. Discussion Rosenberg : Is there any evidence from your work, or from previous work in the literature, that in the early stages of development of the photosynthesis ap- paratus the ratio of CO2 to oxygen is higher than the steady-state value, as might be the case if reduction to triose did not occur? Tolbert: No. Rather a literature comparison indicates that oxygen evolution begins before CO2 fixation, which would make this ratio smaller than the steady- state value. Gaffron : Did you say that oxygen development starts earlier than CO2 fixation? Tolbert: Unfortunately, all the investigators concerned have used different plants. Dr. Thomas, in Utrecht, has measured oxygen evolution during the greening process, and both he and Dr. Smith get oxygen evolution fairly soon after chlorophyll formation begins. James Smith: It took pretty nearlj'^ an hour to get very much though. Tolbert : This would still be sooner than CO2 fixation, which did not begin for 4 hours. Gaffron : It is very interesting that the plant learns to do full photosynthesis by steps. Fuller : I think that is a very important point. There are other ways that you can control enzyme formation in photosynthesis. For instance, j^ou can divorce enzyme formation from the greening process of the plant. We have grown Chlorella variegata in the light on organic substrate, and it does not produce any active carboxylation enzyme. If the endogenous organic substrate is removed then the enzyme is rapidly formed. We have also found that by growing Euglena in the presence of acetate in the light, where no chlorophyll is lost, carboxylation activity is strongly suppressed. There are other ways of inactivating photosynthetic en- zymes, so it is not the greening process itself but it is the pathways of metabolism, as Dr. Tolbert points out, that control enz3ane formation. James Smith : I might say that steps in the oxygen evolution can be built up in the dark. If you illuminate for 5 minutes and transform the protochlorophjdl to chlorophyll, then put the leaf back in the dark for 2 hours, you get only a very slight amount of oxygen evolution. But if you illuminate the leaf again in the air and then measure the oxygen it just streams off. So there is a second photochemical activation in this. In continuous light, this goes on all the time. But there is some- PHOTOSYNTHESIS BY THE ETIOLATED PLANT 227 thing that goes on in the dark that is not photochemical which then can be trans- formed by the photochemical trigger. Tolbert: An additional point of interest is that the etiolated plant and the greening process are not sensitive to ionizing radiation. You cannot prevent greening and normal development of the photosynthesis apparatus by massive dosages of ganmia radiation of the order of magnitude of 100,000 to 300,000 r. Limiry : Is there any evidence of spectral chlorophyll changes? As you convert protochlorophyll to chlorophyll can you increase the density of the chlorophyll? James Smith : Well, you do get this change that I showed here, and whether it is a change in the chlorophyll or whether it is a change in the environment that causes the change in the spectrum, I don't know. Lximry : Is it ruled out that the spectrum shifts that we have observed in chloro- phyll in vivo are due to organization of chlorophyll molecules? Must this now be assumed to be a function of the absorptive act for each chlorophyll molecule or can it be due to the nearby neighboring chlorophyll molecules? I was very much worried by Rabinowitch's and Jacob's conclusion (which I think was their conclusion anyway) in the Journal of Chemical Physics that this is not the kind of change— that we would observe too big a change in the spec- trum, too much of a shift toward the red, if this was due to dipole-dipole inter- action between neighboring chlorophylls and the like. References 1. Tolbert, N. E., and Gailey, F. B., "Carbon dioxide fixation by etiolated plants after exposure to white light," Plant Physiol, SO, 491-499 (1955). 2. Smith, J. H. C, "The development of chlorophyll and oxygen-evolving power in etiolated barley leaves when illuminated," Plant Physiol, 29, 143-148 (1954). 3. Blaauw-Jansen, G., Komen, J. C, and Thomas, J. B., "On the relation be- tween the formation of assimilatory pigments and the rate of photosynthesis in etiolated oat seedlings," Biochim. et Biophys. Acta, 6, 179-185 (1950). 4. Irving, A. A., "The beginning of photosynthesis and the development of chloro- phyll," Ann. Botany {London), 24, 805-819 (1910). 5. Tolbert, N. E., and Cohan, M. S., "Activation of glycolic acid oxidase in plants," J. Biol Chem., 204, 639-648 (1953). 6. Schou, L., Benson, A. A., Bassham, J. A., and Calvin, M., "The path of carbon in photosynthesis. XI. The role of glycoUc acid," Physiol. Plantarurn, 3, 487- 495 (1950). 7. Tolbert, N. E., and Cohan, M. S., "Products formed from glycoUc acid in plants," J. Biol. Chem., 204, 649-654 (1953). I Excretion of Glycolic Acid by Chlorella during Photosynthesis* N. E. TOLBERT and L. P. ZILL, Biology Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee Chlorella cells that had fixed NaHC^^Os for a few minutes were found to excrete 3% to 10% of the fixed C^* as glycolate-C^* into the aqueous nutrient medium. Glycolate was the only organic anion ex- creted which could be detected and analyzed by paper and column chromatography. Analysis of the soluble cellular constituents by paper chromatography revealed the expected photosynthetic prod- ucts inside the cell. During steady-state photosynthesis, 10 to 100 times as much glycolate was in the supernatant as inside the cells. Normally, 3 to 8 mg. of glycolate per liter of cultures was obtained from actively growing algae. Lewinf has recently found about twice as much glycolate in the media from cultures of Chlamydomonas. The rapid excretion of glycolate by Chlorella is dependent on all the factors influencing active bicarbonate fixation: (a) the presence of bicarbonate, (b) aerobic conditions, and (c) light for active photo- synthesis, since there must be a net bicarbonate uptake. In addition, the younger cultures in the more active phases of growth and photo- synthesis excrete a proportionately larger amount of glycolate. In media above pH 5.5, the excretion is at a constant percentage rate of the total fixed, whereas in more acid cultures the excretion of glyco- late decreases until at pH 2.5 it becomes zero. Glycolate excretion and absorption may represent a glycolate- bicarbonate anionic exchange across the Chlorella cell wall without the necessity of a similar cationic shift. Since bicarbonate uptake during photosynthesis and excretion during respiration represent a major ionic movement, a balance of it with some other diffusible anion would lessen an undesirable loss or absorption of cations. If a Donnan * Work performed under U.S. Atomic Energy Commission contract No. W- 7405-eng-26. A detailed report on the identification of glycolic acid and environ- mental conditions affecting this excretion is being published elsewhere. t LeA\iu, R. A., personal communication. 228 EXCRETION OF GLYCOLIC ACID 229 equilibrium exists across the cell membrane, a change in bicarbonate ion concentration should result in an opposite movement of glyco- late ion to help restore the following equality: (glycolate-)ceii (HC03~)ceii (glycolate ) medium (HCO3 ) medium The enormous rate of glycolate-C^^ excretion during active bicar- bonate uptake of photosynthesis indicates that this organic anion is uniquely able to respond quickly to the upset of equihbrium when bi- carbonate ion in the cell suddenly diminishes in concentration and more of this anion moves in from the medium. That glycolate should be the anion functioning in this manner may be the result of several factors. It is a small organic acid, and yet it is one of the strongest acids associated with the cell. It is somewhat unique in being readily available from the photosynthetic carbon cycle. In this respect, it is believed to be formed from a side oxidation pathway from this cycle, starting from a C2 complex at the oxidation level of glycolaldehyde. The excreted glycolate does not accumulate in the Chlorella culture medium in large amounts, but rather it is rapidly reabsorbed by the algae when one or more of the conditions for its excretion are not met. This accumulation inside the cell, in relatively large amounts, is about equal to the total normally found in the medium, but it should not persist for a long time, since glycolate is metabolized to glycine and serine (1). This factor could contribute to induction effects in photo- synthesis after periods of darkness of about 1 hour or longer. Little glycolate would then be left in the cell to exchange rapidly for bi- carbonate ion. When the algae are placed in the hght, a photosyn- thetic induction period would ensue while the glycolate pool was re- plenished by a slow rate of photosynthesis or from reduction of gly- oxylate arising from glycine (2) or from isocitrate (3,4). Other consequences of glycolate excretion during photosynthesis, and its reabsorption and accumulation inside Chlorella during dark respiration, have not yet been evaluated. This bicarbonate-glycolate shift may be involved in acid bursts, pH shifts, CO2 bursts, and quantum efficiency as calculated from short-time or flashing-light experiments. 230 N. K. TOLBERT AND L. P. ZILL Discussion Gibbs : I want to coranient on the glycolic acid excretion, .since I remember that at the botanical meetings last summer Dr. Ranson had some very interesting data on this point. He showed that with normal CO2 concentrations and Bnjophyllum one obtains a typical radiochromatogram as shown by many workers. However, as one increases COs concentration up to 30% and 40% the compounds which one normally finds labeled, that is, the free sugars, sugar phosphates, etc., disappear and the only compound which contains tracer is malic acid. As a consequence one might say the carboxylating enzyme is apparently more sensitive to CO2 concentration than the malic enzyme. One wonders if the ribulose diphosphate that accumulates is partially converted to a diose which cannot carboxylate since the carboxylation enzyme is inhibited so that it spills out of the cell as gly- colic acid. Tolbert: Ranson's work was at 30% to 40% CO2 partial pressures and he did not report glycolic formation under those conditions. The bicarbonate-glycolate shift occurs at low CO2 partial pressure, in the range of 1% CO: to 0.3% or lower. Even at 2, 3, or 4% CO2 in air, variations in the path of the carbon have not been reported for Chlorella. Thus glycolic formation and excretion would not appear to arise from abnormal breakdown of ribulose diphosphate, but rather this is a nor- mal pathway of metabolism from the pentose phosphates. Gibbs: If one incubates pentose phosphate with (I hate to mention the word) cell-free preparations under anaerobic conditions, the pentose phosphate is con- verted to hexose phosphate. However, we found this summer if one does this under aerobic conditions the diose can be split off and converted to glycolic acid. We are attempting to purify this enzyme which takes a C2 piece away from pentose phosphate and causes oxidation to form glycolic acid. Thus there is good evi- dence that glycolic acid can be formed directly from the so-called "active glycol- aldehyde" piece of pentose phosphate. Aronoff: It is interesting to note the excretion of glycolic acid during photo- synthesis. However measurements on CO2 exchange by roots, which presumably would be in a high atmosphere of CO2, showed the presence of considerable gly- colic acid in the root. Tolbert : With roots the net exchange is that of bicarbonate or CO2 arising from respiration and passing out of the root. According to our hypothesis of bicarbon- ate-glycolate shift, there should thus result an accumulation of glycolate inside the root cells. M. B. Allen: Excretion of organic acids by Chlamydomonas might be of interest here, although this is not a transient phenomenon. In contrast to Chlorella, which excretes glycolic acid and then reabsorbs it, growing cultures of six species of Chlamydomonas have been found to accumulate soluble organic material in the culture medium (Allen, M. B., Arch. Microbiol., in press). Formation of soluble products paralleled the growth of the alga and was favored by high light intensity and nitrate as nitrate source. Ten to forty per cent of the organic ma- terial synthesized by the algae appeared in the medium in soluble form. Examination of the culture filtrates showed that the soluble organic material consisted partly of polysaccharide and partly of organic acids. The acids included glycoHc, oxalic, and a keto acid which may be pyruvic. » EXCRETION OF GLYCOLIC ACID 231 Tolbert: For Chlorella the C'^ labeled in the glycolate of the medium ap- proaches a constant value after about 10 minutes. We are dealing with a dynamic exchange of bicarbonate and glycolate across the cell wall and not an excretion and continued accumulation in the medium of large amounts of glycolate. Amon : I don't know of any evidence for bicarbonate absorption by Chlorella. But work has been done with roots, and roots absorb very little bicarbonate. Is there any evidence that there is any bicarbonate absorption of the sort in algae? Tolbert : Not exactly. The evidence just presented was that glycolate excretion occurred in medium at pH 5.5 or above and only in the presence of bicarbonate ion. References 1. Tolbert, N. E., and Cohan, M. S., "Products formed from glycoUc acid in plants," /. Biol. Chem., W4, 649-654 (1953). 2. Zelitch, I., "The isolation and action of crystalline glyoxylic acid reductase from tobacco leaves," /. Biol. Chem., 216, 553-575 (1955). 3. Smith, R. A., and Gunsalus, I. C, "Isocitritase: a new tricarboxylic acid cleav- age system," /. Am. Chem. Soc, 76, 5002-5003 (1954). 4. Olson, J. A., "The d-isocitric lyase system: the formation of glyoxylic and succinic acids from d-isocitric acid," Nature, 174, 695-696 (1954). 2. Phot or eduction with Various Reductants Oxygen Evolution and Photoreduction by Adapted Scenedesmus* LEONARD HORWITZ and F. L. ALLEN, f Research Institutes, University of Chicago (Pels Fund), Chicago, Illinois The reaction steps on the oxygen-hberating side of the photosyn- t.lietic machinery are totally unknown. It was therefore of consider- able interest when Gaffron reported that Scenedesmus ohliquus, when suitably subjected to anaerobic incubation, became so adapted that now irradiation with weak light in an atmosphere of hydrogen and carbon dioxide resulted, not in the evolution of oxygen, but in the absorption of two hydrogen molecules for each carbon dioxide mole- cule consumed. In addition, he found that this adaptation allowed the algae to perform the oxyhydrogen reaction, in which, at suffi- ciently low oxygen tensions, the three gases' oxygen, hydrogen, and carbon dioxide, are normally consumed in the dark in the ratio 2 :6: L He considered the possibility that the photochemical uptake of hydrogen and carbon dioxide by adapted Scenedesmus consisted of no more than photosynthesis and the oxyhydrogen reaction running concurrently (this suggestion was also made later by Laisen et al. (1)) according to the following equations: (A) Photosynthesis CO2 + H2O + light -> (CH2O) + O2 (B) Oxyhydrogen reaction 62 + 2H2 -* 2H2O (C) Dark CO2 reduction qCOi + 2gH2 — g(CH20) + gHjO (D) (1 + ?)C02 + 2(1 + 9)H2 + light -* (1 + gXCHjO) + (1 + g)H20 * Work supported by the Atomic Energy Commission, contract nimiber AT(ll-)-239, the Office of Naval Research, contract numbers NOnr-432(00)nr 119-272 and nr 160-030, and the Fels Fund. The isotopically enriched oxygen was prepared by Dr. A. O. Nier under a grant from the American Cancer Society through the Committee of Growth of the National Research Council. t Present address: Arthur D. Little, Inc., Cambridge, Massachusetts. 232 OXYGEN EVOLUTION ANU PHOTOKEDUCTION 233 Reaction (A) is simply ordinary pliotosyntliesis as it is usually written. Reaction (B) is the oxyhydrogen reaction which supplies energy to accomplish the coupled carbon dioxide reduction indi- cated by reaction (C). The efficiency of the coupling between (B) and (C) is measured by q, which is normally V'2. The overall result of (A) + (B) + (C) is reaction (D) which differs from (E) below only by the additional gas consumption {q) due to the coupling of (B) to (C). However, on the basis of the evidence available to him, Gaffron concluded that the light process he observed in adapted Scenedesrmis was a true photoreduction, essentially like the process occurring in various photosynthetic bacteria in which molecular oxygen ap- parently does not play any role at all. (E) CO2 + 2H2 + light — (CH2O) + H2O There were indications, however, in the work of Franck, Prings- heim, and Lad (2) that the situation was more complicated. At the suggestion of Gaffron, we have repeated and extended their ex- periments. The results indicate that although adapted algae can perform a true photoreduction they do not necessarily lose the ca- pacity to evolve oxygen. True photoreduction with hydrogen and photosynthesis can exist side by side in adapted algae, and either may predominate depending upon conditions. EXPERIMENTS WITH THE FRANCK-PRINGSHEIM OXYGEN APPARATUS A careful comparison was made, using Scenedesnms ohliquus, strain D3, between the curve of oxygen evolution versus light intensity ob- tained in an atmosphere of hydrogen and that obtained in an at- mosphere of nitrogen, both with 2% carbon dioxide. The rate of oxygen evolution was measured, as in the work of Franck et al. (2), by the method of phosphorescence quenching. Oxygen-free carrier gas passing over or through a cell suspension flushes away any oxygen that it may produce. After passing through a liquid nitrogen trap to remove moisture, it flows over a dye adsorbed on silica gel, whose phosphorescence is a measure of the amount of oxygen the carrier gas contains. The algae were used at concentrations ranging from 0.0002% to 0.02%, and were flushed with oxygen-free carrier gas for a minimum 234 L. HOKWITZ AND F. I.. ALLKN of 3 to 4 hours, often much longer, before beginning measurements. The partial pressure of oxygen in the carrier gas before entering the vessel was shown to be below 10-« mm. Hg. There is no doubt that HYDROGEN + 2% CO2 30 40 SO LIGHT INTENSITY 60 70 Fig. 1. Data obtained with the Franck-Pringsheim oxygen apparatus, compar- ing oxygen evohition from Scenedesmus D3 incubated under hydrogen plus 2% carbon dioxide, and under nitrogen plus 2% carbon dioxide. Relative light intensity is plotted on the abscissa and oxygen pressure X 10* millimeters of mercury, in the gas stream emerging from the reaction vessel, is plotted on the ordinate. this partial pressure of oxygen is sufficiently low to permit adapta- tion. Uniform illumination of all the cells was obtained by using green OXYGEN EVOLUTION AND PHOTOREDUCTION 235 light and only a relatively thin depth of very dilute suspension. The intensities were varied by interposing neutral screens in the beam, and were monitored with a Weston photocell and suitable galvanom- eter. As is shown in Fig. 1, Scenedesmus evolves oxygen upon weak illumination, even after several hours of dark incubation in an atmos- phere of pure hjdrogen plus 2% carbon dioxide. However, at equal Ught intensities, the rate is always greater if the carrier gas is nitrogen rather than hydrogen. In obtaining these curves, the light intensity was increased stepwise from zero to the highest value shown on the graph. At this intensity the rate of oxygen evolution (in hydrogen plus carbon dioxide) corresponds approximately to one volume of oxygen per volume of algae per hour. With such low levels of oxygen production, the possibility of deadaptation is certainly re- mote. The data of Fig. 1 clearly indicate that an active hydrogenase does not exclude the photosynthetic evolution of oxygen. However, be- cause of the limitations in the apparatus used, it was not possible to estimate the extent of true photoreduction. It was, therefore, nec- essary to supplement these experiments with mass spectrometer and manometric ones. THE MASS SPECTROMETER EXPERIMENTS The instrument used in these experiments has been described by Brown et al. (3). Conventional Warburg vessels were attached to a male joint containing a very fine leak through which gas from the at- mosphere above the algal suspension could enter the spectrometer analyzer tube at a very slow rate. In this way, there was continuous sampling of the gas phase in the Warburg vessel. At intervals of 1.8 minutes analyses were obtained for the four gases: oxygen 32, oxygen 34, carbon dioxide 44, and carbon dioxide 45. Whereas in the Franck-Pringsheim apparatus only net oxygen pro- duction could be followed, in the mass spectrometer it was possible to measure oxygen and carbon dioxide consumption and production separately. Another desirable feature of the mass spectrometer was that it was possible to demonstrate adaptation during the course of the experiments by showing that carbon dioxide production is small (see column 2 of Table I) compared to the respiratory- rate (0.6 to 230 L. HORWITZ AND F. L. ALLEN 0.8 cell volume per hour), although large amounts of oxygen are being consumed (in the oxyhydrogen reaction). TABLE I. Mass Spectrometer Data on the Gas Exchange of Adapted Scenedes- mus (Temperature is 20.6°C.) (1) (2) (3) (4) (5) % of ultraviolet Respiration, O2 production, inhibition of Experiment cell vols./hr. cell vols./hr. CO2/O2 photosynthesis 080255-(2) 0.16 0.73 1.2 280454-(5) 0.14 1.8 1.0 30 280454-(7) 0.05 0.62 0.86 70 300454-(3) 0.29 3.0 0.78 .300454-(5) 0.11 1.8 0.90 39 270155-(3) 0.13 0.84 1.2 63 A ratio indicative of the relative amounts of true photoreduction and photosynthesis which the algae are carrying on is that of carbon dioxide consumed photochemically to oxygen produced in the light. If there is only photosynthesis, this figure is close to 1 and if there is only true photoreduction, it is infinity. Combinations of the two processes will give intermediate values. The ratio has been calculated by assuming that the carbon dioxide fixation concomitant with the oxyhydrogen reaction occurs in the light as well as in the dark. The results are tabulated in column 4 of Table I. In some of the experiments an ultraviolet treatment had just pre- \-iously been administered to the algae. However, there is no reason to believe that this should invalidate the results. The data on carbon dioxide production indicate the cells were adapted, and previous work (4) has indicated that ultraviolet treatment stabilizes the adapted state. The data obtained with the mass spectrometer and summarized in Table I give no reliable evidence for the existence of a true photo- reduction in normal, adapted algae. They indicate that the photo- chemical activity of these algae is mainly, if not completely, photo- synthetic. Nonetheless, since the measurements with the mass spectrometer were, of necessity, made in the presence of appreciable amounts of oxygen (0.04% to 0.24%), and oxygen tension could easily be an important variable in determining the extent of photore- duction, it was desirable to check for the presence of a true photo- OXYGEN EVOLUTION AND PHOTOREDITCTION 237 reduction under more completely anaerobic conditions. This was done with Warburg manometers, as is explained in the next section. MANOMETRIG OBSERVATIONS Larsen, Yocum, and van Niel (1) have pointed out that, if the photochemical consumption of hydrogen and carbon dioxide by normal adapted Scenedesmus consists of no more than photosynthesis and the oxyhydrogen reaction running concurrently, then this com- plex process should have a quantum yield for carbon dioxide uptake that is 50% higher than the quantum yield of photosynthesis alone. Another requirement of this kind of metabolism, however, is that the rate of carbon dioxide uptake not exceed the maximum rate of oxygen uptake in the oxyhydrogen reaction by more than 50%. Manometric evidence presented here indicates that this requirement is violated and that, therefore, a true photoreduction must exist in normal adapted cells. Under conditions where oxygen tension is not limiting, the rate of oxygen uptake in the oxyhydrogen reaction at 28.3°C. by Scenedes- mus Ds is 3.6 cell volumes per hour. The maximum rate of carbon dioxide uptake in a process consisting of photosynthesis and oxy- hydrogen reaction running concurrently is, therefore, 5.4 cell volumes per hour. However, it is possible manometrically to achieve rates of carbon dioxide uptake in the light that reach and even exceed 8 cell volumes per hour before deadaptation sets in. True photoreduc- tion must, therefore, accoimt for at least 30% of the photochemical carbon dioxide uptake under these conditions. CONCLUSIONS Our data require an interpretation more complex than is provided by heretofore prevailing views on the nature of the photochemical ac- tivity of adapted Scenedesmus under hydrogen and carbon dioxide. The present experiments make it clear that adapted Scenedesmus in the presence of hydrogen and carbon dioxide can, under certain conditions, perform photosynthesis at a rate which may even ap- proach or reach that necessary to account for all its photochemical activity. Under these circumstances a correspondingly large part of the photochemical uptake of hydrogen and carbon dioxide by adapted Scenedesm,us consists of the combination of photosynthesis and the 238 L. HORWITZ AND F. L. ALLEN oxyhydrogen reaction running concurrently. On the other hand, under two of the hmited number of circumstances explored in this and preceding papers, it is possible to demonstrate a true photoreduc- tion. Gaffron had previously shown that in the presence of certain poisons, a true photoreduction occurs with oxygen production com- pletely excluded. We were able to demonstrate a true photoreduction under the conditions obtaining in Warburg manometers, as discussed in the preceding section. These facts, of necessity, lead to the conclusion that the nature of the photochemical uptake of hydrogen and carbon dioxide by adapted Scenedesmus is variable, and depends on several factors which apparently control a sensitive balance between photo- reduction and photosynthesis. Depending on the circumstances, there can be almost any proportion of photosynthesis and photoreduction in adapted algae, ranging from exclusive photoreduction to combina- tions of photoreduction, photosynthesis, and the oxyhydrogen reac- tion, to photosynthesis and the oxyhydrogen reaction running con- currently with photoreduction practically excluded. References 1. Larsen, H., Yocum, C. S., and van Niel, C. B., J. Gen. Physiol, 36, 161 (1953). 2. Franck, J., Pringsheim, P., and Lad, D. T., Arch. Biochem., 7, 103 (1945). 3. Brown, A. H., Nier, A. O. C, and van Norman, R. W., Plant Physiol., 27, 320 (1952). 4. Holt, A. S., Brooks, I. A., and Arnold, W. A., J. Gen. Physiol, 34, 627 (1951). Photoreduction in Ochromonas malhamensis* WOLF VISHNIAC and GEORGE H. REAZIN, JR., Department of Microbi- ology, Yale University, New Haven, Connecticut, and Research Department, Joseph E Seagram & Sons, Inc., Louisville, Kentucky The chrysophyte flagellate Ochromonas malhamensis has been isolated and grown in pure culture by Pringsheim (1), and a synthetic medium for its cultivation has been devised by Hutner (2). Having thus become available as a laboratory subject, the physiology of Ochromonas has been studied by Meyers (3), Reazin (4), and Weis (5). It has been found that the organism fails to grow in the light unless a suitable organic substrate, such as glucose or acetate, is present. This observation suggested to Meyers (3) that Ochiomonas contained insufficient chlorophyll to carry out photosynthesis at a rate required for growth and therefore multiplies by oxidative assimila- tion of an organic substrate. The ability of Ochromonas to grow under initially anaerobic conditions in the light renders this interpretation less plausible. An alternative interpretation is that the light-depend- ent metabolism of this alga may be in part a photoreduction, that is, a bacterial type of photosynthesis. This type of photosynthesis is characterized by the oxidation of an external hydrogen donor in place of oxygen evolution. Photoreduction in algae was first observed by Gaffron (6), who found that hydrogen-adapted algae in dim light fixed CO2 with the simultaneous uptake of hydrogen. Frenkel (7) has extended this ob- servation to a wide variety of algae. However, in the instances studied photoreduction in algae was found only in resting thalli or cell suspensions; growth of algae by photoreduction was not ob- served. To investigate the possibility that Ochromonas might grow by photoreduction, the alga was grown in the continuous culture apparatus described by Benson e,t at. (8) on a synthetic medium. Two types of experiments were performed: manometric determina- * E.xperiments were performed at the Brookhaven National Laboratory. Thanks are due to Dr. M. Gibbs, in whose laboratory the work was carried out, nnd to Dr. TJ. C. Fuller, whose culture apparatus \va.s use 22.0 mg. C of Ochromonas formed 28 . 9 28 . 5 Ochro7nonas was grown in closed flasks for 5 days at 30°C. Atmosphere: Initially 95% N2 + 5% CO2, no net formation of O2 occurred. Illumination: ca. 1000 lux, fluorescent light. Glucose was determined by anthrone reagent, glycerol by periodate titration, and total carbon as BaCOs after AgNOa + K2SO5 com- bustion. No significant changes occurred in the dark. C formed) /(mg. C used) = 0.98. A similar culture grown on glycerol consumed 22.0 mg. of carbon and formed cell material to the extent of 28.5 mg. of carbon, which gives a ratio (mg. C formed)/(mg. C used) = 1.29. In other words, when grown on glycerol, Ochromonas fixed more GO2 than could be provided by the oxidation of glycerol. The ratio 1.29/0.98 = 1.31 corresponds approximately to the ratio between the number of hydrogens per carbon in glycerol to the num- ber of hydrogens per carbon in glucose (2.66/2 = 1.33). Inasmuch as the amount of organic carbon formed during growth appears to be a function of the reduction level of the substrate, this experiment sug- gests that Ochromonas can grow by photoreduction. It is possible that this alga is limited in its ability to evolve oxygen. 242 W. VISHNIAC AND G. H. REAZIN, JR. Discussion Frenkel : Is the assumption of the ratio of isopropanol used to CO^ fixed de- pendent upon the product that is formed? Vishniac : If you assume that carbohydrate was formed, then two moles of iso- propyl alcohol would have to be oxidized to acetone to fix CO2. If the end prod- ucts are different then the ratio of isopropyl alcohol used to CO2 fixed will change, but it would not change the general interpretation of this experiment. Aronoff : Is it not possible that, even with glucose, you had CO2 reduction? Vishniac: Yes. We can assume, for instance, that all the organic carbon is formed by CO2 fixation, but this goes along with oxidation of the glucose. Aronoff: And having been at Brookhaven, were you not testing this with isotopes? Vishniac : No, because I think that as far as these data are concerned we are dealing with net results and are not at a point where isotopes would help in the interpretation. Frenkel : I was wondering if Dr. Gaffron wants to make some remarks about his experiments on the effect of glucose on photoreduction. Gaffron: Glucose or glucose derivatives are capable of replacing molecular hjrdrogen as hydrogen donor during photoreduction in Scenedesmus. I think we have experiments which show that Chlorella might be trained to do this trick too, though for lack of a hydrogenase it cannot utilize free hydrogen. However, as 3^et I do not have a mass spectrograph. So we expect from Dr. Brown the next data on this. References 1. Priugsheim, E. G., Quart. J. Microscop. Soc, 93, 71 (1952). 2. Hutner, S. H., Provasoli, L., and FiKus, J., Ann. N.Y. Acad. Set., 66, 852 (1953). 3. Meyers, J., personal communication. 4. Reazin, G. H., Jr., Am. J. Botany, 4I, 771 (1954). 5. Weis, D., Thesis, Yale University, New Haven, 1955. 6. Gaffron, H., Atn. J. Botany, 27, 273 (1940). 7. Frenkel, A. W., and Rieger, C., Nature, 167, 1030 (1951). 8. Benson, A. A., et al., in Photosynthesis in Plants, J. Franck and W. E. Loomis, eds., p. 381. Iowa State College Press, Ames, 1949. 9. Foster, J., /. BacterioL, 47, 355 (1944). 10. Schatz, A., /. Gen. MicroUoL, 6, 329 (1952). Manganese as a Cofactor in Photosynthetic Oxygen Evolution ERICH KESSLER,* Research Institutes {Pels Fund), University of Chicago, Chicago, Illinois Manganese deficiency has been known for a long time to be a powerful inhibitor of photosynthesis in algae (1). Its effect on metab- olism is unique in so far as only photosynthesis is affected and not respiration or chlorophyll formation (2,3). The inhibition of photo- synthesis can be relieved within a short time by an addition of man- ganese (1-3); the degree of inhibition is independent of light in- tensity (2,3). The latter observation suggests a similarity between manganese deficiency and poisoning with hydroxylamine, a substance known to be a specific inhibitor of the oxygen-evolving process of photosynthesis (4,5). Therefore, Pirson suggested that manganese might perhaps be involved in the oxygen-liberating system of photo- synthesis (2). The process of photoreduction found in certain algae adapted to hydrogen (6) affords a possibility to test this hypothesis. In this re- action, the photochemical process and the reduction of CO2 are pre- sumably the same as in ordinary photosynthesis; only the evolution of oxygen is replaced by a reaction between its precursors and molec- ular hydrogen activated by hydrogenase (6). Therefore, an in- hibitor specific for oxygen evolution in photosynthesis should not inhibit photoreduction and should, in addition, prevent the reversion of photoreduction to photosynthesis with evolution of oxygen which normally occurs at higher light intensities. This was shown to be the case in algae poisoned with hydroxylamine, o-phenanthroline, and phthiocol (4,7). The green alga, Ankistrodesmus hraunii, can be adapted to hy- drogen within a few hours in an atmosphere of hydrogen -|- 4% * This work was performed during the tenure of a fellowship awarded by the National Academy of Sciences and International Cooperation Administration. Permanent address: Botanisches Institut, Universitat Marburg, Marburg/Lahn, Germany. 243 244 E. KESSLER CO2. The light metabolism of this alga, grown with and without man- ganese, is shown in Fig. 1. Photosynthesis is inhibited to about one- fourth of the normal rate in manganese-deficient cells, the degree of inhibition being independent of light intensity. By contrast, photo- reduction is not at all inhibited by manganese deficiency; its rate in UJ o z < X o UI cc LJ CC a. in 1 1 1 NORMAL 1 ^ +40 ^ — ^ — I ^ t- ^^ ^ z j,^ ^ y^ > j/''^ J* I/) ^^^ yr ^^ y^ +20 - ^y^ y^ — X ^^ y^ -Mn Q. _/" / • c i Vi^ —m — — •"/' • ^Tj m in ^^ / \ Y^ / E V\ y NORMAL E \ ifc,^ ^^^^ -20 - \ z \ \ 1- \ ^ -40 — \ y-Mn — Q ^^ Ixl ^V^ 01 ^V ^ X^ v' (— ^^■V. j"^ g-60 - ••^''^ - a 1 1 1 1 1000 2000 3000 LIGHT INTENSITY (LUX) 4000 Fig. 1. Rates of photosynthesis and photoreduction at various Hght intensities in normal and in manganese-deficient Anki^trodesmus braunii. Manometric readings correspond to evolution of oxygen (corrected for respiration) in photo- synthesis and to absorption of hydrogen and caibon dioxide in photoreduction. Experimental conditions: Warburg No. 9 buffer, pH 9, gas phase air, for photo- synthesis; 0.02 M phosphate buffer, pH 6.5, gas phase 96% H2 + 4% CO2, puri- fied by means of a "Deoxo" cartridge, for photoreduction; temperature 25°C., 10 mm.' cells (3 mg. dry weight) per vessel. Preceding dark period: 4 hours; periods of 15 minutes for each light intensity. the light-limited range is even somewhat higher in the deficient cells {cf. 8). At higher light intensities extremely high rates of photo- reduction are observed with the deficient algae, whereas in normal organisms deadaptation and evolution of oxygen occur. Thus the action of manganese deficiency strongly resembles the effect of hy- droxylamine, o-phenanthroline, and phthiocol (4,7). At high con- centrations these poisons, however, are more efficient in stabilizing MANGANESE IN OXYGEN EVOLUTION 245 photoreduction at very high light intensities than is manganese de- ficiency. This is quite understandable since the concentration of poison required inhibits photosynthesis completely, whereas man- ganese deficiency does not. Other differences between poisoned and deficient algae are the following: (7) The above-mentioned poisons, if added prior to adaptation, will inhibit the process of adaptation to hydrogen (4,7). With manganese-deficient algae, no such inhibition of adaptation is observed. (^) High concentrations of poisons reduce the quantum yield of photoreduction (i.e., the rate at low light in- tensities) up to 50% (4,7). With increasing strength of manganese deficiency, however, up to an age of five weeks for the deficient cul- tures, the rate of photoreduction in the light-limited range remains absolutely constant. At the same time, the protection of photoreduc- tion against deadaptation at higher light intensities increases steadily. (By contrast, the yield of photosynthesis drops to about one-fourth during the first 7 to 10 days of deficiency.) Later the manganese- deficient cells become chlorotic, the yield of photoreduction de- creases, and the algae deteriorate rapidly. An addition of manganese to deficient cells results under aerobic conditions in the well-known increase in photosynthesis (1-3); with adapted algae, however, a decrease of photoreduction occurs be- cause deadaptation is accelerated and starts at lower light intensi- ties. In the light-limited range, the amounts of CO2 reduced per unit of time are approximately the same for normal algae in photosynthesis and photoreduction and for manganese-deficient algae in photoreduc- tion. These deficient algae, however, can reduce only one-fourth as much CO2 in aerobic photosynthesis. Studies performed with algae deficient in phosphate and in iron (c/. 2) show that the effect of manganese is quite specific for this ion. Phosphate deficiency inhibits photosynthesis and photoreduc- tion to about the same degree; the percentage inhibition of photo- synthesis increases with increasing light intensity. The effect of phosphorus deficiency resembles the action of dinitrophenol on photoreduction as observed by Gaffron with Scenedesmus D3 (4). Iron deficiency inhibits photoreduction much more strongly than photosynthesis. In this connection it should be mentioned that photoreduction has been found to l^e very sensitive toward cyanide and carbon monoxide (4). 246 E. KESSLER Previously it was suggested (3,9) that manganese might be pri- marily involved in the basic photochemical reaction of photosyn- thesis. However, from our results it follows that its main role in photosynthesis must be concerned with oxygen evolution, the only process of photosynthesis which is not needed in photoreduction. Also the results of Gerretsen (10) with chloroplasts from higher plants suggested the importance of manganese for the formation of peroxides. Recently Kenten and Mann (11) have demonstrated a light-dependent oxidation of manganese in chloroplast preparations and have discussed its possible role in photosynthesis. It might well be that there is an additional and perhaps smaller requirement for manganese in CO2 reduction, as suggested by the work of Arnon et al. (12) with chloroplasts. This, however, could be detected only at saturating light intensities which cannot be reached in photoreduc- tion because of deadaptation. The hypothetical scheme I might serve to explain the observation that the inhibition of photosynthesis by manganese deficiency is "^ »H,0 HYDROGENASE ^ Kv + X+Y+KO XH—»CCO^- REDUCTION) Scheme I. independent of light intensity, although an enzymatic reaction, namely, oxygen evolution, is affected primarily. We assume that manganese is specifically involved in the transformation of the first oxidized product of photosynthesis, YOH, to a peroxide, Z(0H)2, which eventually gives off oxygen {cf. 6). Normally, in the presence of manganese in algae not adapted to hydrogen, all the YOH formed will be disposed of to evolve oxygen. In adapted algae, however, most of the YOH reacts with hydrogen activated by hydrogenase to MANGANESE IN OXYGEN EVOLUTION 247 form water; under these conditions only a comparatively small frac- tion of the YOH goes via Z(0H)2 to O2 (13). With increasing light intensity, more Z(0H)2 is formed, which eventually oxidizes the hy- drogenase and thereby causes deadaptation to normal photosynthesis with evolution of oxygen. In the case of manganese deficiency, the formation of Z(0H)2 is inhibited; thus deadaptation of photoreduc- tion is prevented or at least considerably delayed, and all the YOH formed will be disposed of by the hydrogenase reaction, thereby avoiding the step requirijig manganese in the evolution of oxygen. In unadapted algae, how^ever, this possibility does not exist. Therefore, the YOH may react back causing chemiluminescence (14) or oxidize cellular substances. This latter reaction may lead to the formation of a "narcotic" (15), which, in turn, will inhibit the photochemical reac- tion, thus reducing the rate of photosynthesis at all light intensities. Discussion Aronoff: How many generations occur after starting a manganese-deficient culture? Kessler : Growth is not completely inhibited by manganese deficiency. It is only retarded, but I don't know how many generations there are. Jacobs : Can the inhibition of photosynthesis be relieved by the addition of manganese? Kessler : Yes. If one adds manganese to deficient cells, the rate of photosyn- thesis will go up to the normal level. The speed of recovery depends upon the age of the deficient culture. In the beginning it goes very fast. With strong deficiency, however, the rate of photosynthesis goes up to the normal level within about 10 hours or so. But if one adds manganese in photoreduction, then the addition of manganese will inhibit photoreduction and will accelerate deadaptation. Amon: On this point of time that Dr. Kessler has answered, I would like to say that we are able to confirm that under certain conditions the minus manganese ceils can restore their full photosynthetic rate equal to that of the plus-manganese cells within 20 minutes after the addition of manganese. This is far less time than required for any new cell division or even formation of a protein. Rosenberg : I wonder if you have any data on the concentration of manganese in chloroplasts in normal cells and in deficient cells. Kessler : No, I don't have any analytical data on the manganese content of the cells. I also haven't done any experiments on the Hill reaction, but I know from unpubhshed work of Clendenning that the Hill reaction is also inhibited by lack of manganese, and that its rate can be increased by the addition of manganese. Also Gerretsen has shown that manganese is somehow involved in peroxide formation in isolated chloroplasts. However, it is not quite clear from his data what his results have to do with normal photosj'nthesis. He worked with isolated chloroplasts in the absence of Hill reagents, measured the oxidation-reduction 248 E. KESSLEK potential, and found that it changed in a very marked way upon the addition of manganese, suggesting that peroxide is formed in the presence of manganese but not in its absence. Spikes: In isolated, well-washed chloroplasts we can find practically no manga- nese. Kessler : Probably the chloroplasts of nondeficient plants contain an amount of manganese sufficient for the maximum rate of the Hill reaction. Spikes : If it is there, it must be in a very tightly bound form, because 20 or 30 washings of isolated chloroplasts do not result in any marked decrease in the rate of the Hill reaction. Amon : I think that the evidence in support of the role of manganese in oxygen evolution is quite substantial. However, this would not necessarily exclude the participation of manganese in CO2 fixation. Aren't we assuming that the COj metabolism in photoreduction is similar to that in photosynthesis. I take it that there are no chemical data to tell us that they are in fact similar. James Smith : Long ago, when we treated sunflower leaves with carbon dioxide we could demonstrate that manganese was one of the carbon dioxide acceptors. When we treated leaves with carbon dioxide in order to see what compounds were soluble we could extract manganese, magnesium, and calcium, and we could demonstrate b}^ radiocarbon experiments that the manganese certainly entered into this and was part of the reservoir for the carbon dioxide uptake. I don't know how this fits into your picture, but it may have something to do with it. Amon : The conclusion I was leading up to was this and this again is an assump- tion: If the CO2 fixation pattern during photoreduction were altered bj^ manga- nese deficiency in such a way that light would not be able to accomplish deadapta- tion, this would still be compatible with the results that you presented. Now let me add, as I stated earlier, that we ourselves have also found that oxy- gen evolution is influenced by manganese deficiency and that the effect is reversi- ble, but at the moment the other possibility should be kept open at least as a sub- sidiary point. Jacobs: Can the requirement for manganese be replaced by higher levels of magnesium, cobalt, or nickel? Kessler: Cobalt and nickel are not present in our culture medium except for those impurities which are always introduced with iron and zinc, and the other nutrient components, but quite a high concentration of magnesium is present in both the normal and the deficient medium. References 1. Pirson, A., Z. Botan., SI, 193 (1937). 2. Pirson, A., Tichy, C, and Wilhelmi, G., Planta, Jfi, 199 (1952). 3. Arnon, D. I., Vlll' Congr. intern. Botan., Paris, Sect. 1 1, 73 (1954). 4. Gaffron, H., J. Gen. Physiol., 26, 195 (1942). 5. Weller, S., and Franck, J., J. Phys. Chem., 45, 1359 (1941). 6. Gaffron, H., Biol. Revs. Cambridge Phil. Soc, 19, 1 (1944). 7. Gaffron, H., J. Gen. Physiol, 28, 269 (1945). 8. Kessler, E., Arch. Biochem. and Biophys., 59, 527 (1955). MANGANESE IN OXYGEN EVOLUTION 249 9. Bergmann, L., Flora {Jena), U2, 493 (1955). 10. Gerretsen, F. C, Playit and Soil, 2, 159 (1950). 11. Kenten, R. H., and Mann, P. J. G., Biochem. J. (London), 61, 279 (1955). 12. Allen, M. B., Arnon, D. I., Capindale, J. B., Whatley, F. R., and Durham L. J., J. Am. Chem. Soc, 77, 4149 (1955). 13. Horwitz, L., and Allen, F. L., Arch. Biochem. and Biophys., 66, 45 (1957). 14. Strehler, B. L., and Arnold, W., J. Gen. Physiol., 34, 809 (1951). 15. Shiau, Y. G., and Franck, J., Arch. Biochem., 14, 253 (1947). 3. Reduction of Various Oxidants Contributions to the Problem of Photochemical Nitrate Reduction ERICH KESSLER,* Research Institutes (Pels Fund), University of Chicago, Chicago, Illinois It has been known for a long time that light has a strong accelerat- ing influence upon nitrate reduction in green plants (1-3). In spite of