7 BACTERIAL PHOTOSYNTHESIS Contribution No. 1 12 of The Charles F. Kettering Research Laboratory ^' Q "rr BACTERIAL PHOTOSYNTHESIS Edited by HOWARD GEST lVashini>ton University St. Louis, Missouri ANTHONY SAN PIETRO Charles F. Kettering Research Laboratory Yellow Springs, Ohio LEO P. VERNON Charles F. Kettering Research Laboratory Yellow Springs, Ohio A SYMPOSIUM SPONSORED BY The Charles F. Kettering Research Laboratory THE ANTIOCH PRESS J^ YELLOW SPRINGS. OHIO • 1963 Copyright 1963 by The Charles F. Kettering Research Laboratory Library of Congress Catalog Card Number 63-19302 Lithographed in U.S.A. by ED^VARDS BROTHERS. INC Ann Arbor, Michigan INTRODUCTION This book contains the papers presented at a small, invitational Symposium on Bacterial Photosynthesis held in Yellow Springs, Ohio on March 18-20, 1963. The Charles F. Kettering Research Laboratory was very pleased to serve as host for this conference. Research progress on bacterial systems is moving rapidly and a review of the present state of knowledge seemed appropriate. The organizing com- mittee therefore invited some fifty-five overseas and American in- vestigators to meet and exchange information at a small, informal meeting held under the Foundation's auspices. The staff of the Charles F. Kettering Research Laooratory was stimulated immensely by the conference; we hope these papers will serve as a point of departure for additional photosynthetic investigations, E. W. Kettering Left to right: H. Gest, H. Gaffron, C. B. van Niel, R. Hill, L. P. Vernon, A. San Pietro. To KEES In recognition of his pioneering research on bacterial photosynthesis, this volume is dedicated to Dr. C. B. van Niel, Herzstein Professor of Biology, Stanford University. CONTENTS Introduction v E. W. Kettering Preface xiii List of Participants xv List of Abbreviations xvi OPENING ADDRESS Van Niel's Theory: Thirty Years After 3 Hans Gaffrou I. COMPONENTS, STRUCTURE, AND FUNCTION OF THE BACTERIAL PHOTOCHEMICAL APPARATUS Carotenoids of Photosynthetic Bacteria— Distribution, Structure, and Biosynthesis 19 Synn0ve Liaaen Jensen Tetrapyrroles in Photosynthetic Bacteria 35 June Lascelles A Note on the Effect of Inhibitors of Electron Transport and Phosphorylation on Photopigment Synthesis in RJiodopseudonwnas spheroides 53 W. R. Sistrom The Heme Proteins of Photosynthetic Bacteria 61 Martin D. Kamen The Structure of the Photosynthetic Apparatus in the Green and Purple Sulfur Bacteria 71 R. C. Fuller, S. F. Conti, and D. B. Mellin Some Observations on the Organization of the Photosynthetic Apparatus in Purple and Green Bacteria 89 Germaine Cohen -Bazire Isolation of Photochemically Active Chromatophores from Rhodospirillum Molischianum HI Donald D. Hickman, Albert W. Frenkel, and Konsta)itine Cost Isolation of Bacteriopheophytin-Containing Particles from Rlwdospirillum rubrum 115 Torn Kihara and Albert W. Frenkel Structure and Function in Bacterial Photosynthesis 121 Hoirard Gest and Subir K. Bose II. METABOLISM AND PHYSIOLOGY Metabolic Aspects of Bacterial Photosynthesis 129 Howard Gest Biochemical Basis for the Obligate Photoautotrophy of Green Bacteria of the Genus CMorobium 151 Robert M. Smillie and W. R. Evans Some Observations Concerning the Purification and Properties of the Aerobic Phosphorylation System of Rhodospirillum nibrnni Extracts 161 David M. Geller Metabolism of Photosynthetic Bacteria. II. Certain Aspects of Cyclic and Noncyclic Photophosphorylation in Rhodospirillum rubrum 17 5 M. Nozaki, K. Tagaiva, and Daniel I. Anion Photophosphorylation in Rhodospirillum rubrum. About the Electron Transport Chain and the Phosphorylation Reactions o . . . . 195 Margareta Baltscheffsky and Hcrrick Baltscheffsky Light-Induced and Dark Steps of Bacterial Photophos- phorylation 201 Mitsuo Nishi)nura The Effect of Ubiquinone2 on Photophosphorylation in Particles Obtained from Rhodospirillum rubrum Grown in Media Containing Diphenylamine 217 Harry Rudney Photosynthetic Phosphorylation with Bacterial Chromatophores: Catalysis by a Naturally Occurring Factor (Phosphodoxin) 223 C. C. Black and A. San Pictro III. ELECTRON TRANSPORT Photooxidation and Photoreduction Reactions Catalyzed by Chromatophores of Purple Photosynthetic Bacteria .... 235 Leo P. Vernon Effect of Reduced 2,6-Dichlorophenolindophenol upon the Light-Induced Absorbancy Changes in Rhodospirillum rubrum Chromatophores: A Coupled Reduction of Ubiquinone 269 Howard Bales and Leo P. Vernon Electron Transport System in Facultative Photoheterotroph: Rhodospirillum rubrum . . 275 T. Horio and J. Yamashita Physiology of Bacterial Chromatophores 307 J. W. Newton Nonheme Iron Proteins and Chromatium Iron Protein 315 Robert G. Bartsch The Respiratory System of Rhodomicrobium vaimielii 327 S. Morila and S. F. Conti IV. PHOTOCHEMICAL CONSIDERATIONS Primary Quantum Conversion: Electron Spin Resonance Evidence 335 R. H. Ruby and M. Calvin Effect of Reduced 2,6-Dichlorophenolindophenol and N,N,N',N'-Tetramethyl-/)-phenylenediamine on the Light-Induced Electron Spin Resonance Signal Observed with Rhodospirillum nibnon 343 John J. Heise and Leo P. Vernon The Light-Induced Electron Spin Resonance Signals Observed in the Green Bacterium Chloropscudomonas ethylicum 351 Christiaan Sybesma and John J. Heise Mechanisms of Light -Activated Electron Transport in Bacteria: The Effect of Viscosity on Reaction Rates 357 Britton Chance, Mitsuo Nishinmra, S. B. Roy and Heinz Schley er A Kinetic Analysis of the Light Responses of Photosynthetic Bacteria and Plants 369 Britton Chance Photochemical Reaction Centers in Photosynthetic Tissues . . . 377 Roderick K. Clayton Photochemistry of Bacteriochlorophyll 397 J. C. Goedheer Energy Transfer and Cytochrome Oxidation in Green Bacteria. 413 John M. Olson wid Christiaan Sybesma The Protein-Chlorophyll-770 Complex from Green Bacteria . . 423 John M. Olson, David Filmer, Roger Radloff, Carol A. Roinano, and Christiaan Sybesma Light-Induced Absorbancy Changes in Rhodomicrobium vannielii 433 John M. Olso)i and Sigehiro Morita SUMMARY Extemporary Remarks by Way of Summary 445 Martin D. Kamen APPENDIX A Brief Survey of the Photosynthetic Bacteria 4 59 C. B. van Niel Composition of Bacterial Chromatophores 469 J. W. Neivton Spectroscopic Properties of Purified Cytochromes of Photosynthetic Bacteria 475 Robert G. Bartsch Absorption Spectra of Photosynthetic Bacteria and their Chlorophylls 495 R. K. Clayton Media for Anaerobic Growth of Photosynthetic Bacteria 501 Snbir K. Bose Bibliography on Metabolism of Photosynthetic Bacteria 511 INDEX 521 PREFACE The past ten years have witnessed a rapidly increasing tempo of research on bacterial photosynthesis. This is perhaps attributable, in part, to the fact that modern developments in microbiology and bio- chemistry have demonstrated the potential advantages of using bacteria as the source of experimental systems for investigation of numerous basic biological phenomena. The relatively large, and possibly unique, range of metabolic capacities shown by the photosynthetic bacteria has added to their appeal for such studies. There is little doubt, how- ever, that the main stimulus for closer scrutiny of these organisms stems from the desire to understand the "comparative biochemistry" of photosynthesis in greater depth. Although considerable evidence has accumulated showing close similarities between green plant and bac- terial photosynthesis, investigators have long been intrigued with the reasons for, and possible implications of, the differences observed between the two processes. Indeed, when a fundamental research ad- vance is made with either type of photosynthetic system, pertinent reexamination of the other soon follows. This pattern of cross- checking, which has become more prominent in recent years, has unquestionably facilitated progress in elucidation of the mysteries of photosynthesis. In the past, however, symposia on this important topic have been concerned primarily with green plant systems and only secondarily with photosynthetic bacteria. This realization and the conviction that an up-to-date review of the problem would stimulate further progress led to organization of the present symposium, devoted exclusively to the bacterial process. Inevitably, we were faced with the usual dilem- mas posed by the attempt to arrange a meeting at which all investiga- tors actively working in the field would be present and able to ex- change ideas and viewpoints freely under informal circumstances. It is our hope that any shortcomings in this respect will be ameliorated by our efforts to make the proceedings of the symposium available to the scientific community at the earliest possible time. A number of animated controversies developed during the course of the meeting and this we interpret as one of the signs of its success. Groups of participants directly interested in the debated questions met informally, as time permitted, with the aim of resolving basic issues. Our deepest gratitude goes to Dr. Martin Kamen who undertook the formidable task of presenting their conclusions to the symposium audience during his summarizing remarks. xiv PREFACE It is a pleasure to acknowledge the valuable services of the follow- ing conveners: Dr. W. Arnold, Dr. C, S. French, Dr. R. Y, Stanier, and Dr. C. B. van Niel. Special notes of gratitude are due to Dr. R. K. Clayton and Miss Jane Finney for their editorial assistance and to Mr. Justin C. Crawford for his capable efforts in making the necessary arrangements. The editors are also indebted to those who contributed to the appendices, which contain frequently required experimental data and bibliographies of areas which could not be adequately covered due to lack of time. We are particularly grateful to Antioch College for providing an auditorium and other attractive facilities for the symposium, and to the Charles F. Kettering Foundation for financially supporting the symposium. The Editors SYMPOSIUM PARTICIPANTS William Arnold Daniel I. Arnon Howard Bales Herrick Baltscheffsky Robert Bartsch Helmut Beinert C. C. Black Lawrence Bogorad Subir Bose Thomas Brown Bruce Burnham Warren Butler J. E. Carnahan Britton Chance R. K, Clayton Germaine Cohen-Bazire S. F. Conti J. C. Crawford C. Stacy French Albert W. Frenkel Keelin T. Fry R. C. Fuller Hans Gaffron David Geller Howard Gest Martin Gibbs J. C. Goedheer Norman Good Govindjee John J. Heise Robin Hill G. Hind S. Hood Takekazu Horio S. Ichimura Andre T. Jagendorf Synnp've Liaaen Jensen Martin Kamen Bacon Ke D. L. Keister Bessel Kok H. V. Knorr June Lascelles R. A. Lazzarini Gene Lindstrom Berger Mayne Bernard S. Meyer R. M. Miller S. Morita Jack W. Newton Mitsuo Nishimura John Olson John J. Ormerod Norbert Pfennig Jose Ramirez Ronald Ruby Harry Rudney A. San Pietro Kenneth Sauer G. R. Seely W. R. Sistrom Robert Smillie Lucille Smith Roger Y. Stanier Carroll A. Swanson Christiaan Sybesma Kunio Tagawa R. W. Treharne C. B. van Niel E. H. Vause L. P. Vernon Wolf Vishniac W. Zaugg LIST OF ABBREVIATIONS A(OD) absorbance (optical density) ADP adenosine diphosphate ATP adenosine triphosphate BChl bacteriochlorophyll Chi chlorophyll pCMB p-chloromercuribenzoate Cyt cytochrome DCMU 3 - (3 -4 -dichlorophenyl) -1,1- dimethylurea DPIP, DPIPH2 2,6-dichlorophenolindophenol and its reduced form EDTA ethylenediaminetetraacetic acid ev electron-volt ESR (EPR) electron spin resonance FAD, ] FADH2 flavin adenine dinucleotide and its reduced form FMN, FMNH2 flavin mononucleotide and its reduced form IDP inosine diphosphate ITP inosine triphosphate mv millivolt NAD, 1 SIADH (DPN, DPNH) nicotinamide adenine dinucleotide and its reduced form NADP, , NADPH (TPN, TPNH) nicotinamide adenine dinucleotide phosphate and its reduced form PCMB /)-chloromercuribenzoate Pi orthophosphate PMS (MPM) phenazine methosulfate (methyl phenazonium methosulfate) PPNR photosynthetic pyridine nucleotide reductase PQ plastoquinone RHP Rhodospirillum riibruin heme protein (cytochromoid C) Tris tris(hydroxymethyl)aminomethane UQ (C( :)Q) ubiquinone H. GafTron addressing the opening session of the symposium. -J^ *^^ OPENING ADDRESS Opening Address VAN NIEL'S THEORY: THIRTY YEARS AFTER HANS GAFFRON The Florida State University, Institute of Molecular Biophysics (Fels Fund) Tallahassee, Florida Ladies and Gentlemen: Considering that we are, as I thought we would be, a gathering of experts, there is really not much excuse for an opening address. The purpose of such an address is probably to remind the participants of the few major problems which the meeting is about, in order that we do not lose sight of them when we begin to discuss the countless de- tails and ramifications into which atopic like Bacterial Photosynthesis must of necessity be subdivided. On the other hand we have had lately symposia and meetings contributing to the problem of photosynthesis at the rate of one or two per year. It is unlikely, therefore, that any one of us could have lost sight of the major problem. Even with modern teamwork, progress in terms of essential new discoveries is not so fast that a proposed meeting here and there could not be skipped. But the fact that we are all here shows that the idea of the organizing committee to hold this particular symposium must nevertheless have struck today's guests as an attractive proposition. Two reasons can be pointed out immediately. Though we have been acquainted with the Kettering Foundation as a place of research in our field since the days of Inman, Albers and Knorr, Rothemund and later of Clendenning and Eyster, the Laboratory has lately undergone a re- building and an expansion which has moved it into the front line of modern research on the photochemistry in living cells. One attraction must have been the desire to visit the Kettering Laboratory, and the second was the idea to single out the phototrophic bacteria for special consideration. This plan has automatically brought together not only the young keen minds for whom history begins after 1945, but also those of us who, in a much more leisurely way than is the fashion today, began once upon a time to investigate those reactions which still provide so much material for lively discussions, A look at the elegantly done Symposium program has sharpened our anticipation of the coming intellectual pleasures, Mr, Kettering and Dr, Vernon deserve thanks indeed for having called us together. And then there was the prospect that we might have among us our good colleague, the eminent and wise scholar Cornelis Benardus van Niel of Pacific Grove, whom his friends and pupils call Kees, Actually 4 OPENING ADDRESS I did not believe he would show up— too many of our meetings during past years had to be held without him. But to my surprise, and to everybody's pleasure, he arrived yesterday evening. Our program promises us the description of quite a number of new observations, experimental techniques and contributions for or against certain hypotheses aimed at explaining the particular kind of metabol- ism that sets the purple bacteria apart from the green plant. Hardly any one of us who were around twenty years ago would have believed that van Niel's idea of a photolysis of water as the core of the photosynthesis problem could still elicit a vivid discussion today. For the green plants it had been proven as correct by Hill's reaction. And as a reasonable interpretation also for the anaerobic photo- metabolism of purple bacteria there was the indirect evidence of the adaptable hydrogenase- containing algae. Purple bacteria furnished van Niel the key to the first generally convincing picture of the photosynthetic process in terms of modern metabolic ideas. And purple bacteria are now believed to provide clear evidence that a photolysis of water— water as an intermediate hydrogen donor— should not be accepted as part of the hypothetical picture for bacterial photosynthesis. That is, van Niel's generalization of 1935 is disallowed. It is about this question mainly that I would like to speak to you. Usually after thirty years a theory ought to have been transformed into fact or replaced by a better one. With van Niel's theory it so happened that after ten years there were no doubts left that the oxygen of photo- synthesis originates from water. This we have accepted as fact. I propose toshowthat, like any good scientific theory which managed to live in these hectic times for thirty years, van Niel's extended version is still useful. A truly good theory never dies— it only becomes more refined. This may makeit more difficult to explain and to teach- but it does not render the simpler version wrong. It is often repeated that one new fact which does not fit destroys a hypothesis. This is not true. As long as this new observation does not give birth to a better theory— and better is by definition the more encompassing view— it should be noted but treated as if with a little more thought and patience it may soon find its place within the existing order. We have accepted the proposition that light will split, oxidize, de- hydrogenate, or photolyze water in green plants, because on the face of so much evidence we cannot explain from where else the oxygen could originate. On the other hand, purple bacteria do not evolve oxy- gen. Why should we assume that water is involved in the photochemical process, even as an intermediate and incomplete process, when there is as yet no incontrovertible evidence that the assumption is warranted? How sound a viewpoint— and what a dull one. As I pointed out recently somewhere else, the mechanism to release oxygen from water with VAN KIEL'S THEORY: THIRTY YEARS AFTER 5 eight quanta is too remarkable and complex a mechanism not to have a long evolutionary history. And there are too many parallels in the be- havior of photosynthetic bacteria and plants not to be intrigued by what I am willing to call the more interesting and therefore more rewarding hypothetical proposition. And perhaps I am biased because it took me once so long to recognize its elegance. Certainly thirty years ago I simply could not see why I should accept van Niel's proposition that organic substances serve purple bacteria exclusively as hydrogen donors (just like H2S, sulfur or hydrogen) for the reduction of carbon dioxide to carbohydrate, and thence to bacterial substance. My OAvn results with purple bacteria did not show this at all. Quite independently (never having heard of van Niel)Ihad started about 1929 on investigations on purple bacteria after Warburg had mentioned to me that Stalfeld had told himof these strange organisms. As a chemist I had never looked at a microbe before and knew only Warburg's great discovery— the alga ChloreUa. Soon I discovered that the reddish microbes behaved quite differently from green plants. They refused to do photosynthesis but evidently ate organic acids in the light without further ado, either with a stoichiometrically determined amount of carbon dioxide, or, if available, also with hydrogen. The product of the photometabolism was partly a substance (C4H6O2) (which Hans Fisher later depolymerized into crotonic acid) and for the greater part just more bacteria. Later, when working with Chromatium, the purple sul- fur bacteria, I found that they produced lots of H2S in the dark, particu- larly when previously illuminated in presence of butyrate. So I con- cluded that the light reactions withsulfur were reversible and that this was the mechanism by which they were able to attack organic sub- stances. Many of you will remember that van Niel challenged this vigor- ously. Years later Henley in my laboratory confirmed the fermentative sulfide formation from internally stored sulfur but not from sulfate. My observation of a particular accelerating effect of added sulfate was indeed, as van Niel had shown, a nonspecific salt effect. In 1935 van Niel extended his special theory so that it included also the metabolism of the heterotrophic purple bacteria. In this paper he quotes Gaffron's statement that photosynthesis of the purple bacteria involves the cooperation of a larger number of molecules and that several intermediate reactions occur before the first stable reaction products appear. Van Niel then wrote, "This statement seems to con- tain an argument against a unified concept of photosynthesis in green plants and purple bacteria." Because I could not see eye to eye with a Dr. Roelofsen, working in Kluyver's laboratory, van Niel had spent a year in Holland devising a good number of experiments to prove con- vincingly that sulfur bacteria can use organic substances directly as hydrogen donors, just like the Athiorhodaceae, He then came to Berlin to see me. we set up one or two experiments, they were absolutely 6 OPENING ADDRESS convincing, and I conceded defeat quickly so that we could go sight- seeing. Soon a long publication appeared in which van Niel explained every one of my experiments according to his views. My experiments were perfectly reproducible. But except for some evidence that the photochemistry with aliphatic acids was much more complex than the simple overall equation of green plant photosynthesis allowed for, I had no rational hypothesis at all. It was only a few years later, after Hill's chloroplast reactions and mainly on account of my own photo reduction experiments, that I understood fully the power of van Niel's concept. Because I firmly believe that the mere description of new phenomena is the lesser half of any scientific task and that facts, unless they can be used to supporter to revise current theoret- ical opinion, remain just memorable curiosities until someone pro- vides the theoretical connection with existing knowledge, I would like to point out the following. Considering that van Niel's was the first comprehensive and fruitful theory of photosynthesis which had been proposed until that time, he had to make the attempt to keep the funda- mental principle intact. It has been very agreeable, satisfying, and flattering, of course, that recently Stanier and Doudoroff did prove that the reactions in purple bacteria I had written about really exist. And I can only recom- mend warmly the technique of staying alive long enough to see such vindications happen. But does this invalidate van Niel's general con- cept of photosynthesis as it applies to the metabolism of purple bac- teria? I do not think so for a moment. Here I would like to digress with a remark on the importance of schools and the influence of masters. When I first met van Niel in 1935 I was absolutely under the spell of Warburg. He was twenty years older, in the prime of his productivity with a dozen fundamental dis- coveries already to his credit and many more ahead of him. He was then an implacable foe of Wieland's dehydrogenation theory. He be- lieved (as most of you are well aware he still does) in the direct photochemical decomposition of carbon dioxide, and this was sufficient to put a block into my brain. The Wieland-Kluyver-van Niel way of looking at the same factual material I rejected for reasons only a psychologist may be able to explain to us in the future. I tell this story because the same strange power is, as you know, still at work today. And it means that conscientious teachers should, after having convinced their pupils of their own viewpoint, challenge them to find faults with the ruling theory of the laboratory and help them even in this exercise. Another digression is on the lucky choice of material to work with. If Willstatter and Stoll had studied minced spinach extracts instead of minced sunflower leaves they might have found what Hill found twenty years later. The fabulous success of Warburg's choice of Chlorella we need only to mention in passing. If I had started with Chromatium in- stead of with Athiorhodaceae the confusing conflict with van Niel might VAN NIEL'S THEORY: TfflRTY YEARS AFTER 7 not have arisen. On the other hand, as I pointed out, the wrong theory exerts perhaps an even stronger inhibitory influence. And then there is costly apparatus— it too can hinder progress when you believe that because it was acquired on a special government grant it has to be used for its money's worth. It made a great impression on me when a new-fangled very expensive light source (I believe specially built by Siemens) had arrived at Warburg's laboratory. The great man and his assistants spent an afternoon trying to make it work. When it turned out that the thing was a disappointment, Warburg coolly said, "Negelein— store it in the attic," and never looked at it again. To return to thephotolysisof water. What! have been able to under- stand about our experiments on adapted algae during the past twenty years I managed on the basis of the idea of internal back reaction involving water. That is, internal oxido- reductions involving hydrogen donors have to be assumed. As you know, during recentyears the question concerning the nature of the primary process has been progressively restricted to a smaller and smaller field of enquiry (Compare Fig. 1). The entire carbon assimilation mechanism has been cleared away as a typical example of orthodox biochemistry upon which plants and photosynthetic bacteria have no exclusive property right. And a look at our program will tell you that the discussion has narrowed to the question of how much phosphorylation, TPN reduction, and oxygen evolution have to do with the light energy conversion process. Thus the difference between bac- teria and plants brings us to the question: What happens in the pigment complex of bacteriochlorophyll which distinguishes the end result from that we find in the green plants? Let us enumerate what purple bacteria and plants have in common. 1) A chlorophyll a type pigment. The bacterial form contains two more hydrogens and handles light quanta at a discount of ten kilocalories per mole quanta as compared with the green chlorophyll, because the singlet absorption band lies around X. 890 mfi. 2) Different percentages of one chlorophyll a are distributed among several binding sites, as attested to by the various humps in the main absorption bands seen in living cells. 3) Plants and bacteria contain carotenes and quinones, not quite identical chemically but very likely serving the same func- tions. 4) Not only the aerobic plants as shown by Hill, but also the obligate anaerobic purple bacteria have, as Kamenandhis co-workers discovered, cytochromes— not one, but at least two, and with oxidation- reduction potentials that are 0.2 volts apart. This is true for plants as well as bacteria, and we ought to remember this when we come to dis- cuss the role of cytochromes, 5) Photophosphorylation, The observa- tion of a light-induced phosphate turnover in intact purple and green cells preceded by several years the demonstration that a respiratory type of phosphorylation occurs in cell-free preparations. Light and water replace, as van Niel would say, the role which a hydrogen donor OPENING ADDRESS I. Carboxylotion CO2 + C ,^C5C00h( \ -r II Reduction Hexoje EI Back-Reoction / Photo-Phosphorylofion ITPN Fact, i H,0 3ZI Dehydrogenotion Photoxydotion H,A H--,H-S;S;Asc.A.;RH, Hq Substitute Reductions COp + RH— RCOOH— RCO Dye Dye H2 Quinone — *• Hydroqumone Flavin *■ Flavin H2 Cyt.c^* — ► Cyt.c^* 02 ► H2O2 TPN * TPN(HJ 1 r"^ nz Chlorophyll complex ■ XHg *-XH-^Chl. al X > 685 m/-( roH ^\ ~ P \\ . Accessory Pigments — -;(0H)-— YOH —Chi. a.E X < 685 m/u 3ZI Evolution of Oxygen ^Mn"^ -02 Cytochrome? Plcstoquinone? Fig. 1. Sum of the recognized partial enzymatic systems which together con- stitute the mechanics of photosynthesis. All except the chlorophyll complex are able to function without light. and oxygen play in the corresponding ordinary dark metabolism. Thus a cell-free system can be set up which converts light energy into typical phosphate energy: ATP, The experiment works with green or purple chromatophores— there is no difference, 6) Identical or very similar enzymes serve as reducing agents for the carbon dioxide re- duction system and for the initial carboxylation, 7) Photosynthetic units exist in both types of organisms, according to Arnold's experi- ments. This implies at least two differently bound molecules of the same pigment class in bacteria just as well as inplants: the photon- collecting and the photon- converting molecules of which the latter, according to Franck, must have its main absorption at the long wave- length end of the complex spectrum shown by living chlorophyll, 8) Transfer of energy from light-absorbing accessory pigments happens in both classes of organisms. To these eight points others equally important may soon be added. In short, there is so much in common that the problem of utilization of water as hydrogen donor might better VAN NIEL'S THEORY: THIRTY YEARS AFTER 9 be approached by caHing attention to the differences. Which, and how important, are they? Is the color difference important? Probably not, because the green bacteria use quanta of very nearly the same energy level as the green plants, yet do not evolve oxygen. Are the enzymes important that permit the purple and green bac- teria to use inorganic or organic hydrogen donors for their photo- synthesis? Again I believe the answer is no, because we have now a dozen or more typical oxygen-releasing algae which can skip the oxygen part and do a bacterial photo reduction with hydrogen just like obligate anaerobes. Furthermore, Pringsheim andWiessner discovered a green flagellate that, at best, shows only a marginal capacity for normal photosynthesis. Like a purple bacterium, it cannot grow at all in the light when presented only with carbon dioxide and water. It cannot even grow like Rho do spirillum with acetate and oxygen in the dark. It grows exclusively with acetate in the light. In short, the more we look around the more we find gradual transi- tions between the photochemical capacities of a typical purple bac- terium and that of the most specialized oxygen- evolving green plants. The only clear-cut difference I know, in respect to the ability to release oxygen, is the need for manganese first shown by Kessler in our laboratory, and confirmed for the Hill reaction by Eyster in this laboratory, Wiessner later found that purple bacteria thrive on a low manganese diet which forces the algae to turn heterotroph for dear life. Then there is a difference in the quinones which might be im- portant. Finally the absence of an Emerson effect in green algae after adaptation to hydrogen, a condition that parallels again, as I have just learned, the behavior of purple bacteria. In these van Niel and Blinks could find no Emerson effect. The spectral investigations of Duysens, Kok, Rabinowitch; the gas exchange measurements by French, Myers, Govindjee; and whoever may by now have acquired the proper mo nochromators have established that in photosynthesis two or more pigments which are activated by distinctly different wavelengths have to cooperate. The products of the two separate photochemical processes are chemical substances, not physical states, because, according to Myers and French, they live for seconds. And they both must be present in order that oxygen can appear, and without oxygen evolution there is no reduction under anaerobic conditions. The two-pigment problem has now merged with the two-quanta- per-hydrogen problem, first clearly enunciated by Franck and Herzf eld many years ago. Hill, Kok, Duysens and Witt, all independently, gave reasons why one part of the tandem pigment should be assigned to the reduction, the other to the oxidation side. A year ago Dr. Franck paid us a visit. Dr, Clayton happened to be there too. Naturally we discussed the Emerson effect, about which Franck had published some very in- 10 OPENING ADDRESS teresting theoretical propositions. Franck suggested that the long wavelength pigment was likely to be oxidized more easily than the rest of the chlorophyll. If this were truly so, it would follow that our algae might be able to produce oxygen in the dark red under anaerobic conditions. Bishop and I made the experiment and it turned out that an adaptation to photo reduction was necessary. Anaerobic treatment was not enough to make the dark red radiation efficient. I reported this at the Paris meeting last summer. Of course this confirms the assumptions of Kok, et al. If it is possible to set the entire carbon dioxide reduction machinery in motion, as we have done with that part of the twin-pigment system which for itself alone cannot produce oxygen, then one is tempted to relegate the photolysis of water entirely to the other twin. Instead of making things easy, the now popular schemes of Kok, Duysens, etc., introduce, however, two problems at once— or, rather, make them especially conspicuous. According to well-known measurements of quantum yields, photo- reduction on the one hand turns out to be singularly inefficient and uses at least twice or three times as many quanta as it ought to, while the evolution of oxygen per se might become, in these schemes, a four-quanta process. Instead of continuing to speculate how near or far from the double primary process van Niel's photolysis may be found, we may look at some of the available experiments. Arnon, in the Proceedings of the National Academy of Sciences, has in a charming way enumerated the stages of the strip tease which have followed each other for a century to get at the true first stable chem- ical products of the primary process. The accent is on stable, as I would like to emphasize. First there was starch. Then numerous in- termediates known from the respiratory sugar metabolism. Then specifically PGA, then finally ATP, And here the strip comes dan- gerously close to the end, because the photochemically reduced enzyme PPNR of San Pietro, now rebaptized ferredoxin, is stable only in the absence of oxygen. Thus the first stable primary product appears to be reduced TPN. This is also the beginning of the back flow of electrons (hydrogen) which leads to phosphorylation, the only chemical energy source which, according to a new generalized concept, is necessary under anaerobic conditions to give us the metabolism of purple bac- teria or adapted algae. Twenty years ago Gaffron and Rubin observed a photochemical evolution of hydrogen from adapted algae when they were put in nitro- gen or helium, without carbon dioxide and preferably in the presence of dinitrophenol. Some time later, Kamen, Breghoff andGest discovered a photohydrogen evolution which they considered as being much more significant than the reaction in green algae because it was so much VAN NIEL'S THEORY: THIRTY YEARS AFTER 11 bigger. A while ago Bishop in our laboratory noted that a mutant of Scenedesmus , which cannot evolve oxygen but handles photo reduction very well, never produced hydrogen under conditions when we would expect it to do so. In normal Scenedesmus evolution of hydrogen was found to be sensitive to typical oxygen evolution inhibition. Putting the old and the new experiments together, it seemed clear why light- induced evolution of hydrogen is poor in algae and good in bacteria. The purple bacteria are set to use organic material, and if there is a separation of |H| and [OH I the latter can easily be used to oxidize the organic substrate. This is supported by Gest's recent studies where hydrogen and carbon dioxide appear simultaneously and in proper proportion. The green plants, on the other hand, have either put up a permeability barrier against, or lost the enzymes for, the same organic hydrogen donors. This apparently has been one condition for the efficient evolution of free oxygen. What happens in the green plants is that there will be back reactions whenever the reducing in- termediates of the light reaction, such as reduced ferredoxin, are not utilized. Some of these back reactions may be the ones that produce cyclic phosphorylation. Assume now that we activate the hydrogenase in a normal green plant, replace the hydrogen by nitrogen, remove the carbon dioxide, poison the phosphorylating back reactions, and illuminate. What do we get? A complete and direct photolysis of water by light. Hydrogen and oxygen are evolved in about the right proportions and in impres- sive amounts. If the oxygen evolution is stopped specifically by any of the three possible methods— poison, manganese deficiency, or mutation— there is no hydrogen evolution. No hydrogen without oxygen. It follows in our opinion that this is the reaction nearest to the photochemical process which gives you stable, usable products (Fig. 2). Thus we have reached the end of the line. This is what light is able to produce. Once there is nascent hydrogen and nascent oxygen, everything that follows falls under the category of enzymatic dark reactions. Dr. Bishop will present these results at the Atlantic City Federation meeting. Let us come back to the two-pigment problem and van Niel's theory. The basic two-pigment skeleton of a general photo synthetic system thus looks like the scheme of Fig, 3, It is of utmost importance, of course, to fill in the arrows correctly with the proper enzymes, metals, proton and electron transfer agents. Only then shall we know for sure how the photolysis of water proceeds with two quanta. The water splitting is the result of the whole sequence and there may be no one particular place to point to where we can say: here photo- lysis happens. There are too many spots (at any one-electron transfer metal catalysis, for instance) where we have to formulate the reaction with the aid of water to balance the charges. 12 OPENING ADDRESS 80 70 60 Q \ii o 50 o (C '^ 40 CO I 30 20 10 I 750 LUX CONTROL ©. .ALKALINE- \^i/ PYR06ALL0L PALLADIUM •- 40 80 120 160 TIME (MINUTES) 200 240 Fig. 2. Photolysis of water in Scenedesmus as a simple manometric experiment. Hydrogen and oxygen are produced in about the ratio 2:1 if photosynthesis, photoreduction and phosphorylation are suppressed. Adapted algae are illumi- nated in absence of CO2, under nitrogen and in presence of an inhibitor of phosphorylation. The mixed character of the gas evolved becomes apparent by the way it reacts after the light has been turned off. Unless a water molecule is shown to be firmly bound as part of a molecule in such a way that it is not exchangeable, metabolic reactions in aqueous media will not distinguish one water molecule from another. When we speak of the photolysis of water, we do not mean that the emerging hydrogen and oxygen molecules stem from one and the same molecule as earlier hypotheses assumed. Nobody ever suggested this for electrolysis, for instance. Movements of electrons and holes, or charge transfer, are part of the mechanisms of reversible oxido re- ductions which in turn are the wheels of the apparatus which brings VAN KIEL'S THEORY: THIRTY YEARS AFTER 13 hi/| hi/2 X>685m/x X<685m/i hydrogen I / oxygen or [H] — Chi I -/*-"— Chi n— [oh] or / \ ^ reduction \ oxidation [OH]or (+)-. [h] or (- Fig. 3. The basic mechanism of photosynthesis which in a two pigment system produces overall what amounts to a photolysis of water. about the final result, the photolysis of water. This final result is not the appearance of positive holes on the one side, or a corresponding stream of electrons ejected on the other, but the appearance of the elements hydrogen and oxygen, or in their place such permanent chemical changes as may be brought about metabolically by either hydrogen or oxygen. To achieve present day photosynthesis, it was necessary in the course of natural evolution to prevent as far as possible all types of intermediary re-oxidations. Free oxygen had to be eliminated as a waste product. And also the release of free hydrogen had to be pre- vented in order that it could be used instead in intermediate forms for synthetic reactions. Shall we believe now that the purple bacteria just have only one half of the system and that nowhere water enters into their photo- chemical mechanism? This would mean, if we think of evolution, that the plants arose by doubling the arrangement of the purple bacteria. If so, this would leave us with the dilemma of the eight quanta in photo reduction mentioned above, Duysens, in desperation, believes the quantum number to be an accident. Let me point out that the obligate anaerobic phototrophic bacteria, as well 3.8 Chlamydobotrys oiVvings- heim and Wiessner, have to grow while the light is shining. No fer- mentation supports growth in the dark. In the green plants we keep respiration and growth on a separate energy balance sheet, while with the bacteria our energy measurements include everything. This may equalize the energy requirements of plants and bacteria in an accidental way. But a look at the regularity of Larsen's results in his beautiful measurements of the quantum requirements in green and red bacteria makes the chance hypothesis appear rather weak. Regardless of the substrate, H2, H2S or thiosulfate, and the correspondingly different 14 OPENING ADDRESS efficiency in energy utilization, the quantum number remained about nine, and constant. If purple bacteria and adapted algae use only one half of the mechanism shown in Fig. 3 which is required for full photosynthesis— i.e., for the complete photolysis of water— then there is no obvious spot for the noncyclic phosphorylation which everybody likes now to place at the junction of the two-pigment systems. The bacteria are forced to put extra energy into cyclic phosphorylation. This too might explain an equal quantum number per carbon dioxide reduced. If we accept at all the idea that green plant photosynthesis evolved from the simpler system still to be found in purple bacteria, it is not so unreasonable to believe that the twin-pigment arrangement is al- ready present in the latter. Only the differences in potential between the left and right halves of the tandem arrangement did not yet shift far enough apart to allow for a spontaneous dismutation of [OH] into free oxygen with the aid of newly added enzymes such as the manganese- containing one I mentioned earlier. Complicated as this all sounds, I need not remind you that in reality it will eventually turn out to be more intricate still. There are the various sets of accessory pigments, for instance, which seem to be attached either to the bright red or to the dark red absorbing chlorophyll. Why under these circum- stances van Niel's simple concept of an intermediary photolysis of water should have aroused so much opposition, I cannot see. If a theory explains a series of rather diverse observations in a consistent manner, and makes sense from the point of view of increased metabolic potentialities, I prefer it to a disjointed set of explanations of which each one does not reach further than a narrowly circumscribed set of experimental conditions. I am quite confident that despite the great number of new phenomena we are going to discuss during the coming session, there will not be one observation which clearly demands that we abandon van Niel's hypothesis of 1935. In other words, at the end of this week it will be just as much alive as during the years gone by. Top: C. S. French; right: C. Black, R. A. Lazzarini, R. Bartsch; left: N. Good, G. Hind. I COMPONENTS, STRUCTURE, and FUNCTION of the BACTERIAL PHOTOCHEMICAL APPARATUS CAROTENOIDS OF PHOTOSYNTHETIC BACTERIA DISTRIBUTION. STRUCTURE AND BIOSYNTHESIS SYNN0VE LIAAEN JENSEN Institute of Organic Chemistry The Technical University of Norivay, Trondhcim A characteristic feature of all photosynthetic bacteria is their con- tent of yellow to violet carotenoid pigments, which contribute to the spectacular colours occasionally exhibited by the bacterial cell. Elegant studies of such bacteria, performed by R, Y, Stanier, L, N. M, Duysens and others, have demonstrated the function of the carote- noids as protectors against photo- oxidation and as auxiliary absorbers of radiant energy for photosynthesis and phototaxis. Other functions have been claimed, but not rigidly proven. Pertinent reviews in this field are available (see, for example, 1,2,3), In most of the photosynthetic bacteria the biosynthesis of carotenoids proceeds in such a manner that considerable amounts of intermediates can be isolated in addition to end products. The carotenoids of these organisms therefore represent a unique array of compounds which are biochemically and structurally very closely related. In addition, con- ditions can be created under which otherwise inaccessible intermediates accumulate, and their interconversions can be studied. Mainly for the above reasons investigations on photosynthetic bacteria have con- tributed much to the solution of problems connected with the biosyn- thesis of this class of natural products. The number of known carotenoid pigments has increased significantly as a result of study of the pigment complex of such bacteria. Despite the fact that the carotenoids of photosynthetic bacteria are chemically closely related, sufficient variation does occur to make them useful for the characterization of these bacteria. The present paper will be limited to recent progress in our know- ledge of the distribution, chemical structure, and biosynthesis of coloured carotenoids of photosynthetic bacteria— a topic which reflects the special interest of the author. I hope the limitation is justified by the rapid and perhaps unexpected expansion of this field during the last few years. Much of the work to be discussed has not yet been pub- lished, although it is in press. 19 20 THE BACTERIAL PHOTOCHEMICAL APPARATUS The carotenoids of photosynthetic bacteria are, with a few excep- tions, aliphatic. They often carry tertiary hydroxy! and methoxyl groups located in the 1,1' -positions, and sometimes contain conjugated keto-groups. The chromophore, which consists of a variable number of conjugated double bonds, causing the yellow to pink-blue colour characteristic of these pigments, is often located rather unsymmetri- cally in the molecule. In addition, isolated double bonds may be present in agreement with the formal composition of the carotenoids by com- bination of eight isoprene units followed by dehydrogenations. As an example is shown the assumed structure (4) of OH-spheroidenone (Formula I), (I) Table 1 gives a summary of the characteristic structural features of carotenoids in the photosynthetic units of various groups of organ- isms. The carotenoid pigments of the photosynthetic bacteria are distinguished from those of algae and higher plants by their aliphatic and frequently unsymmetric nature and the presence in the molecule of tertiary hydroxyl or methoxyl groups. The photosynthetic tissue of the algae and higher plants generally contain bicyclic carotenoids of the a- and /3- carotene type, often substituted in the 3-positions with sec- ondary hydroxyl groups; epoxy- carotenoids are here quite abundant. From photosynthetic bacteria there have been isolated to date 32 different carotenoids (including colourless forms). To the majority of these pigments fairly reliable chemical structures have been ascribed. In Table 2 is presented the distribution pattern of coloured carotenoids in 16 speciesof photosynthetic bacteria belonging to six different genera and four families. Aliphatic, hydroxylated and methoxylated carotenoids of what we shall call the normal spirilloxanthin series (involving the seven carotenoids participating in the transformation of lycopene to spirilloxanthin in R. nibrum (6,7), are abundant in the genus Rhodo- spirillum. This series occurs olso in some species oiRhodop seudo- monas, Chromatium, Thio spirillum, 2ccid in Rhodomicrobiuni vanuielii. Keto-carotenoids of the spheroidenone type are restricted to three species of Rhodopseudomonas , whereas another type of keto-carot- enoids is present in some Thiorhodaceae spp. Cyclic carotenoids have not so far been found in pure cultures of any of the Athiorhodaceae or Thiorhodaceae. Derivatives of the monocyclic y- carotene are, so far as is known, restricted to the green bacteria, and y5-carotene has been found only in Rhodomicrolmim vannielii. CARTENOIDS OF PHOTOS YNTHE TIC BACTERIA 21 I o X a b% o w >2 f-l '— ' s > CQ -S ^ a c p 3 o ^ O o "^ ;-, O >! bJO OT^ I- §1 .2 q B- ■S -^ Tert meth gro ^ o "^ N. <3 >> Is, Si t) s^ o ^ s rt W^ ^ "S Uo s> !-, ^ X p 1 o S ^ o 0) Ic §5 I CO xs -c ^ If B CD * -x M CTJ ^J3 * Ph < 22 THE BACTERIAL PHOTOCHEMICAL APPARATUS spiouaiojED pajnoiOD JO jaqiunu jEiox oo in ro CM C^ CO m ID „ „„ auajojEO- (1 „ aua^OJED- t-HO CM 3U9 JO JEO - t CM 2 § 1 2 1 < 1 1 1 3UOU35IO + ^ i . --lTuau73!d' ^ f 1 (2 luauiSid) auoSuiiujEM + ; 1 £ uimuEXOiUJids-oiaM-Z •S s uijqiAOpoqj-oaM-s - 1 "S dnojS suouspiojaqds 3uouapioJ8nds-HO^ CO a = auouapiojaqds r + + + CO = ? dnojS auapiojaqds (A-HO) auapiOJaqds-Ho"' s •■ + + ^ ti => (A) auapiojaiids 5;+ + + ^ - t,- auapiojaqds pajETAiflaiuaa 1 » dnoaS uimuExonuids umjuBXOTTiJids H 1 " 1 uimuExoniJids -mauiapouoiM + CO dnOJS isi'd (181'd-HO) utjqiAOpoqa CO = ^ (iSl'd) ui.iq!Aopoqj-ojpAnuv + + + + + + r- d s U!dopoqj-ojpXqaa-{.'£ + CM 1 -J dnojg auadooAT £ luauiStd ^ 1 -s i E auadooXq-HO-TP £ + CO (auadODAT-HO) uidopoqH J + + + + + + + <:. auadoD/Sq t^ 1 , ^. ^ ^ + +$ + + CO dnOJS auajodsojnaN uiqiUEXojomo Big + E S CO auajodsojnaN s- .1 _ ^ S o" c- c- 2"' 1 il III til a: e; e; 5) f It II ='! 3 1 1 1 1 5 1 1 S 1 ■-5. i li ■3 f 1 1 < 1 1 i i i 3 1 g .5 .£ CARTENOIDS OF PHOTOSYNTHETIC BACTERIA 23 Lycopene is the carotenoid most widely distributed— it is found in 13 of the 16 species. The composition of the carotenoid mixture from Rhodopseiidomonas gelatinosa is the most complex, consisting of 12 coloured components in cells grown under semi-aerobic conditions. In turning to a discussion of the bios5aithesis of these pigments I would like to refer to recent reviews (1,20,21), and to take as the starting point the structural scheme in Fig, 1, suggested for carotenoid biosynthesis in purple bacteria two years ago by Stanier's group (22, 23,7). This pathway offered a simple and logical picture of the carot- enoid biosynthesis in non-sulphur purple bacteria. The scheme was based on a number of studies, viz., the carotenoid composition of mu- tants of RJiodopseiidomonas spheroides with deranged carotenoid syn- thesis (24), the diphenylamine- effect and carotenoid transformations 11 •C, - cnrotcne (IV) 11 ■ ' '.' 3. :hl,.roxnnthin ("> Ip' L^a.^cn 5.!;° OH-V (Mil) 31 A^h^J^orhoJ..^.^r,n(VIm 31 21 Rhixluvibnn (1\) Spir.lloxamhin (XI) Fig. 1. Structural scheme for the pathway of carotenoW biosynthesis in purple bacteria (23,7). 24 THE BACTERIAL PHOTOCHEMICAL APPARATUS TABLE 3. The categories of step -reactions operative in the biosynthesis of coloured carotenoids in photosynthetic bacteria Type no. Reaction type Structural features 1 Dehydrogenation 2 Dehydrogenation 3 Hydration 4 Methylation 5 Oxidation (aerobic) 6 Hydroxylation 7 Oxidation (anaerobic) 8 Cyclization in washed cell suspensions of R. nihrum (6), the effect of oxygen on anaerobic cultures of Rps.spheroides (25,26), and finally the structural determinations of the carotenoids involved (27,7). The biochemical reactions participating in these transformations were grouped into five different categories (23,7), namely, two types of dehydrogenations, a hydration reaction consisting of addition of water to an isopropylidene double bond, methylation of a tertiary hydroxyl group, and finally a type of oxidation involving introduction of a conjugated keto- group. These reaction types are represented by numbers 1-5, depicted in Table 3. Tentative chemical structures, (XIII), (XV) and (XVII a or b), in agreement with the limited data available for these trace compounds, were assigned to the three carotenoids designated by Goodwin as hydroxy- Y, hydroxy-R and P512 (5). The carotenoidP512, now referred to as P518 (11,4), exhibits absorption maxima at longer wavelengths than any other known carotenoid. This carotenoid has now been isolated from Rps. spheroides and Rps. gelatinosa, and its structure appears very likely to be that of a 2-keto-spirilloxanthin (XVIII) (11), CARTENOIDS OF PHOTOSYNTHETIC BACTERIA 25 (XYD) a)R=H R*= b)R=CH3 R'=/^^^,Aw^s CARTENOIDS OF PHOTOSYNTHETIC BACTERIA 33 the unsolved problems related to the distribution, chemical structure, and biosynthesis of carotenoids in photosynthetic bacteria. REFERENCES 1. Stanier, R. Y., Carotenoid pigments: problems of synthesis and function. The Harvey Lectures, 1958-1959, 219 (1960). 2. Goodwin, T. W., The biosynthesis and function of the carotenoid pigments. Advan. EnzynioL, 21, 295 (1959). 3. Clayton, R. K., Phototaxis of purple bacteria. Handbuch der Pflanzenphysi- ologie, 17, Part 1, 371 (1959). " 4. Jensen, S. L., Bacterial carotenoids XL On the constitution of the minor ca.rotenoids oi Rliodopscndo))ioiias. 2. OH-R. Acta Chcin. Scand., 17, 489 (1963). 5. Goodwin, T. W., The carotenoids of photosynthetic bacteria. II. The carot- enoids of a number of non-sulphur purple photosynthetic bacteria (Athior- hodaceae). Arch. Microbiol., 24, 313 (1956). 6. Jensen, S. L., Cohen-Bazire, G., Nakayama, T. O. M., and Stanier, R. Y., The path of carotenoid synthesis in a photosynthetic bacterium. 5ioc/i//». Biophys. Acta, 29, 477 (1958). 7. Jensen, S. L., The constitution of some bacterial carotenoids and their bearing on biosynthetic problems. Kgl . Norske Vid. Selsk. Skr. 1962,^0. 8. 8. Jensen, S. L., Bacterial Carotenoids XII. On the constitution of the minor carotenoids of ft/?or/o/)se??r/o;;/o;/<7s. 3. OH-Y. Acta Chem. Scand., 17, 500 (1963). 9. Goodwin, T. W., Land, D. G., and Sissins, M. E., Studies on carotenogene- sis. 23. The nature of the carotenoids in the photosynthetic bacterium Rhodopsendomonas spheroides (Athiorhodaceae). Biochem. J., 64, 486 (1956). 10. Nakayama, T. O. M., The carotenoids of Rhodopsendomonas . 11. A compara- tive study of mutants and the wild type. Arch. Biochem. Biophys., 75, 356 (1958). 11. Jensen, S. L., Bacterial carotenoids X. On the constitution of the minor carotenoids oi Rhodopsendomonas. 1. P518. Acta Clicni. Scand., 17, 303 (1963). 12. Goodwin, T. W., and Land, D. G., The carotenoids of photosynthetic bac- teria. I. The nature of the carotenoid pigments in a halophilic photosynthetic sulphur bacterium {Chromatinm spp.). Arch. Mikrobiol., 24, 356(1956). 13. Benedict, C. R., Fuller, R. C, and Bergeron, J. A., [I4c]- Acetate incor- poration into the carotenoids of Chromatiuni. Biochim. Biophys. Acta, 54, 525 (1961). 14. Schmidt, K., Die Carotinoide der Thiorhodaceae. 2. Carotinoidzusammen- setzung von Thiospirillum jenensennd Chromatinm vinosnin. Arch. Mikro- biol., 46, 127 (1963). 15. Schmidt, K., Jensen, S. L., und Schlegel, H. G., Die Carotinoide der Thior- hodaceae. 1. Okenon als Hauptcarotinoid von Chromatinm okenii. Arch. Mikrobiol., 46, 117 (1963). 16. Jensen, S. L., und Schmidt, K., Die Carotinoide der Thiorhodaceae. 3. Die Carotinoide von Clironiatinm war mi n ffii. Arcli. Mikrobiol., 46, 138 (1963). 17. Goodwin, T. W., and Land, D. G., Studies in carotenogenesis. 20. Carot- enoids of some species of Chlorobium. Biochem. J., 62, 553 (1956). 18. Volk, W. A., and Pennington, D., The pigments of the photosynthetic bac- terium Rhodomicrobinm uannielii. J. Bacterial., 59, 169 (1950). 34 THE BACTERIAL PHOTOCHEMICAL APPARATUS 19. Conti, S. F., and Benedict, C. R., Carotenoids of Rhodomicrobyiuni vannieUi. J. Bacteriol., 83, 929 (1962). 20. Goodwin, T. W., Biosynthesis and function of carotenoids. Aim. Rev. Plant Physiol., 12, 219 (1961). 21. Goodwin, T. W., Chemistry, biogenesis and physiology of the carotenoids (carotenoids associated with chlorophyll). Haiidhuch der Pflanzenphysi- ologie, 5, 394 (1960). 22. Jensen, S. L., Bacterial carotenoids VIII. A structural scheme for the biosynthesis of coloured carotenoids in the photosynthetic bacterium Rliodo- spirillum vubrum. Acta Clioii. Scand., 15, 1182 (1961). 23. Jensen, S. L., Cohen-Bazire, G., and Stanier, R. Y., Biosynthesis of carot- enoids in purple bacteria: A re-evaluation based on considerations of chem- ical structure. Nature, 192, 1168 (1961). 24. Griffiths, M., and Stanier, R. Y., Some mutational changes in the photo- synthetic pigment system of Rhodopsendonionas spheroides. J. Gen. Micro- biol., 14, 698 (1956). 25. van Niel, C. B., Studies on the pigments of the purple bacteria. III. The yellow and red pigments of Rhodopseudomoiias spheroides. Antonic van Leeuweuhoek, J. Microbiol. Serol., 12, 156 (1947). 26. Cohen-Bazire, G., Sistrom, W. R., and Stanier, R. Y., Kinetic studies of pigment synthesis by non-sulphur purple bacteria. J. Cell. Coiiip. PhvsioL, 49, 25 (1957). 27. Davis, J. B., Jackman, L. M., Siddons, P. T., and Weedon, B. C. L., The structures of phytoene, phytofluene, (^-carotene and neurosporene. Proc. Chem. Soc, 1961, 261. 28. Shneour, E. A., Carotenoid pigment conversion in Rhodopseudomoiias spheroides. Biochim. Biophys . Acta, 62, 534 (1962). 29. Eimhjellen, K. E., and Jensen, S. L., Biosynthesis of carotenoids in Rhodo- pseudomoiias gelatiiiosa. Biochim. Biophys. Acta. In press. 30. Jensen, S. L., Bacterial carotenoids xni. On the constitution of the minor carotenoids of Rhodopseudomoiias . 4. 2-Keto-rhodovibrin. Acta Chem. Scaiid., 17, 555 (1963). 31. Shneour, E. A., The source of oxygen in Rhodopseudomoiias splieroidcs carotenoid pigment conversion. Biochim. Biophys. Acta, 65, 510 (1962). 32. Schlegel, H. G., und Pfennig, N., Die Anreicherungskultur einiger Schwefel- purpurbakterien. Arch. Mikrobiol., 38, 1 (1961). 33. Jackman, L. M., and Jensen, S. L., Bacterial carotenoids IX. The consti- tution of the third member of the P 481-group(3,4-dehydro-rhodopin). Acta Chem. Scand., 15, 2058 (1961). TETRAPYRROLES IN PHOTOSYNTHETIC BACTERIA JUNE LASCELLES Microbiology Unit, Department of Biochemistry, University of Oxford Tetrapyrroles in the form of chlorophylls and cytochromes have a vital function in the transformation of radiant energy into a form available for the metabolism of photosynthetic forms of life. It is therefore important to know how photosynthetic organisms synthesize and control the formation of these substances. The significance of this information is not confined solely to photosynthetic organisms. The unique capacity of photosynthetic bacteria and algae to make tetra- pyrroles has led to their profitable exploitation by biochemists inter- ested in wider aspects of tetrapyrrole metabolism and now it is not unusual to find cultures of photosynthetic bacteria in laboratories pre- viously devoted to ducks and rabbits. In this paper the distribution, biosynthesis, and control of formation of tetrapyrroles is discussed with particular emphasis on those areas where information is still sadly lacking. TYPES OF TETRAPYRROLES FORMED BY PHOTOSYNTHETIC BACTERIA Chlorophylls . These comprise by far the major proportion of tetrapyrrole deriva- tives in photosynthetic bacteria. It seems that bacteriochlorophyll (Fig. 1) is the only form of chlorophyll in the Thiorhodaceae and Athiorhodaceae (see "Note added in proof," p. 52). The structure proposed by Fischer (1) has recently been confirmed (2), In the green sulfur bacteria two different types of chlorophyll have been identified, designated Chlorobium chlorophyll 650 and 660 according to the red absorption maxima of the extracted pigments in ether (3), The structure of these is being currently investigated; the 650 pigment is a derivative of 2-desvinyl-2-Q'-hydroxyethylpyropheo- phorbide a and the 660 compound is a derivative of 5-methyl-2- desvinyl-2-Q'-hydroxyethylpyropheophorbide a. Both pigments lack the type of cyclopentanone ring typical of other chlorophylls and have no carbomethoxyl groups (4,5), Also, the alcohol side chain differs from phytol (C20H39OH) and appears tobetrans-trans-farnesol (C15H25OH) (6). 35 36 THE BACTERIAL PHOTOCHEMICAL APPARATUS ^yV-n< PhytyJ Fig. 1. Bacteriochlorophyll (top) and Chlorophyll a (bottom). TETRAPYRROLES 37 The amount of bacteriochlorophyll in the Athio- and Thiorhodaceae is of a similar order to that found in algae (Table 1), In the Athiorho- daceae, the conditions under which the organisms are grown can cause quite marked changes in the concentration of bacteriochlorophyll. Apart from factors such as light intensity and oxygen tension (to be discussed later) the carbon source also influences bacteriochlorophyll synthesis (7,8), Iron deficiency causes a marked decrease in bacterio- chlorophyll content and this metal seems to be involved in the bio- synthesis of all forms of chlorophyll. Formation of Chlorobmm chloro- phyll is diminished by lack of iron (9) and there are numerous ex- amples of iron-deficiency chlorosis in higher plants (10). TABLE 1 Bacteriochlorophyll content of some photos ynthetic bacteria Organism Carbon source Bacteriochlorophyll (m/imoles/mg dry wt) Athiorhodaceae R. rubnim Malate 13 NCIB no. 8255 Rps. spheroides Malate 12 NCIB no. 8253 Rps . palustris Malate 11 2.1.7 Rps. capsidata Succinate 22 2.3.11 Thiorhodaceae Thiopedia sp. CO2 16 Chromatium D CO2 24 Succinate 33 Data from Kornberg & Lascelles (8) and from unpublished personal observa- tions. All cultures were grown anaerobically under a light intensity of about 250 ft-c and estimations were made when the culture density had attained a density of 0.6-1.0 mg dry wt/ml. Cytochromes and catalase. The contribution to the total tetrapyrroles made by the prosthetic groups of cytochromes and catalase is slight (about 1%) compared with that made by the chlorophylls (Table 2), Even the high catalase mutant of Rps. spheroides (11), of which catalase comprises 5-25% of the dry weight, contains only about 0,5-2,4 m^mole/mg dry weight of tetrapyrrole (calculated as catalase heme), Athiorhodaceae grown in 38 THE BACTERIAL PHOTOCHEMICAL APPARATUS TABLE 2 Iron porphyrins in photosynthetic bacteria Organism Iron porphyrin contributed by Cytochrome b Cytochrome c RHP Catalase R. rubrum Light 0.069 0.11 - 0.18 0.04 - 0.21 <0.001 Dark 0.053 0.04 - 0.12 0.003 - 0.14 <0.001 Rps. spheroid es Light _ 0.11 0.41 <0.001 Dark - 0.09 0.06 0.025 - 0.042 ChroDiatiun? D Light - 0.03 0.06 - Values are calculated from the data of Clayton (12), Geller (13) and Bartsch & Kamen (14), and are expressed as m/imoles of iron porphyrin/mg dry wt of cells. the dark instead of photosynthetically synthesize little or no bacterio- chlorophyll and under these circumstances the hemoproteins become the major tetrapyrroles of the cells. Though there is a notable rise in the catalase, particularly in Rps. spheroides , the total tetrapyrrole does not approach the level found in light-grown cells (Table 2), It is clear, therefore, that the main business of the tetrapyrrole-forming machinery in photosynthetic organisms is directed towards making chlorophylls. Porphyrins and derivatives. In his classical monograph on the Athiorhodaceae, van Niel (15) observed that some species of Rhodopseudonionas excreted apink pig- ment with absorption bands at 610, 565, and 535 m^. This was later identified as a mixture of free porphyrins, coproporphyrin IE (Fig. 2) being the major component, and is commonly found in the medium of stationary phase cultures (16). The amount of porphyrin varies with the species and strain and is most marked under conditions of iron deficiency. There is an inverse relation between the amount of por- phyrin accumulated and the concentration of bacteriochlorophyll in the cells, the latter being favored by addition of iron (16). When oxygen is rigidly excluded from cultures of Athiorhodaceae, colorless porphyrinogens (Fig, 2) accumulate which are converted to porphyrins by autooxidation in the presence of air (17), TETRAPYRROLES 39 COOH I CH, H3C ^ COOH CHj CH2 CH, ' C^ CH,— CH,- COOH HjC-^/^^ "^ ^^.^^CH, — CH,— COOH H.C^1. / C \ CH, " CH2 I CH, COOH CH, COOH Coproporphyrin HI h'^n H^- H2C. CH, CH, I CH, CH, I COOH COOH Coproporphyrinogen HI CH, II CH CH HX^ /^'"-^^ i^^V^CH = CH2 / '"^^ N «N- ^' II ■ C H CH, CH, :OOH COOH Protoporphyrin IX CH, CH, cc Fig. 2. Coproporphyrin III (top), Coproporphyrinogen III (middle), and Protoporphyrin IX (bottom). 40 THE BACTERIAL PHOTOCHEMICAL APPARATUS Smaller amounts of magnesium protoporphyrin monomethyl ester have been observed in Rps. capsulata and Rps. spheroides (18,17, and Lascelles, unpublished observations). Excretion of chlorophyll deriva- tives by mutant strains of Athiorhodaceae is discussed later. The accumulation of porphyrins and derivatives by the Athiorho- daceae is connected with the formation of bacteriochlorophyll. It has not been observed in cultures growing aerobically in the dark, when little or no photosynthetic pigments are formed. Also the inverse re- lation between porphyrin and bacteriochlorophyll, influenced by the iron concentration of the medium, points to an association in the syn- thesis of these pigments. BIOSYNTHESIS OF PORPHYRINS AND CHLOROPHYLLS Pathway to protoporphyrin. The discovery by Granick (19) that mutants of Chlorella blocked in chlorophyll synthesis accumulated free porphyrins gave the first indication that the formation of chlorophylls and hemes proceeded via a common pathway up to the stage of protoporphyrin. The basic outline of the pathway to protoporphyrin is now known (Scheme 1), This has been achieved initially by studies with preparations from avian ery- throcytes (reviewed by Granick & Mauzerall (20)) and it is only com- paratively recently that bacteria, in particular Rps. spheroides , have been exploited (reviewed by Lascelles (16,21)), All the evidence points to a common pathway in animals and photosynthetic bacteria and the exceptional ability of Rps. spheroides to form tetrapyrroles has made it a fruitful source for enzymic studies. Glycine pyridoxal PO^ -^-aminolevulinic -8CO2 Succinyl-CoA -2CO2, -4H coproporphyrin in uroporphyrin HI uroporphyrinogen I -6H uroporphyrin I Hemes, chlorophylls Scheme 1. Path of tetrapyrrole synthesis. TETRAPYRROLES 41 There are stillmajor gaps in the pathway to be filled in, particularly in the area between porphobilinogen and protoporphyrin. Uroporphyrin- ogen and coproporphyrinogen seem firmly established as intermediates and the accumulation of the corresponding oxidized porphyrins in cultures is probably due to spontaneous oxidation of the porphyrinogens. Conversion of porphobilinogen to the uroporphyrinogen in occurs in at least two steps catalysed by different enzyme fractions, one being a porphobilinogen deaminase and the other having an isomerase function (22). An enzyme fraction has been prepared from Rps. spheroides which decarboxylates uroporphyrinogen to coproporphyrinogen (23). This conversion is likely to proceed in several stages and the detection by chromatography of traces of porphyrins with five to seven carboxyl groups in enzymic reaction mixtures, as well as in the porphyrin mix- tures excreted by whole cells, support this. The conversion of copro- porphyrinogen to protoporphyrin has been studied with an enzyme fraction purified twentyfold from beef liver mitochondria (24). The reaction requires oxygen but is not inhibited by cyanide. The mechanism is largely unknown but a tricarboxylic porphyrinogen and protoporphy- rinogen are probable intermediates. There is also evidence that an intermediate is, or can become, covalently bound to protein (25). Under conditions of iron deficiency coproporphyrin HI is always the major porphyrin accumulated by photosynthetic bacteria; protopor- phyrin and chlorophyll derivatives are not found. With adequate iron, the porphyrin output is considerably less (1 to 10% of that with low iron) but under these conditions protoporphyrin (Fig. 2), magnesium protoporphyrin monomethyl ester, and chlorophyll derivatives pre- dominate. This suggests that iron participates at a stage in the con- version of coproporphyrinogen to protoporphyrin. Such a function for iron is supported by whole cell experiments with Rps. spheroides (26,27). Conversion of 5-aminolaevulate (ALA) to coproporphyrin III occurs when iron-deficient cells are incubated anaerobically in the light in the presence of phosphate and Mg2+only; no protoporphyrin is formed. Addition of iron to such suspensions promotes synthesis of protoporphyrin and free heme. Additional evidence for the involvement of iron in protoporphyrin synthesis is provided by the inhibition by o-phenanthroline of the mitochondrial enzyme system that converts coproporphyrinogen to protoporphyrin (24). To establish the function of iron in the conversion of coproporphyrinogen to protoporphyrin, further enzymic studies are clearly needed, but so far there has not been much success in preparing extracts of Rps. spheroides active in this respect. Another interesting aspect of the conversion of coproporphyrinogen to protoporphyrin in the photosynthetic bacteria concerns the nature of the oxidant needed for the oxidative decarboxylation. The enzyme systems from animal tissues show an obligatory requirement for oxygen and alternative electron acceptors have not been demonstrated. 42 THE BACTERIAL PHOTOCHEMICAL APPARATUS Clearly oxygen cannot participate in the reaction in photosynthetic bacteria growing anaerobically and presumably the acceptor is gener- ated by the photosynthetic apparatus. Since i?/)s. s/?/^ero^■(ies is so active in forming protoporphyrin under appropriate conditions it is a promis- ing candidate for enzyme studies; analysisof this system might clarify the general mechanism by which these bacteria perform other reactions which are obligatorily linked to oxygen in aerobic organisms. Formation of iron and magnesium protoporphyrins . Soluble enzyme systems have been purified from animal tissues which catalyse heme synthesis from protoporphyrin and ferrous ions (28,29), and it is probable that similar enzymes are present in bac- teria. Burnham (30) has obtained heme synthesis from protoporphyrin and iron citrate with crude extracts of Rps. spheroides incubated anaerobically in the light with succinate. Ferrichrome or related iron- binding factors may participate coenzymically in this reaction. In the Rps. spheroides systems, ferrichrome replaces iron citrate and ex- periments with Fe^^-labelled ferrichrome have shown that the iron is transferred to protoporphyrin to form labelled heme. Purification of the enzyme is required to establish whether ferrichrome is an ob- ligatory cofactor. The participation of magnesium protoporphyrin as an intermediate in chlorophyll synthesis was indicated many years ago by the isolation of chlorophyll- less mutants of Chlorella which accumulated this metal complex (31). There is no information about the enzymic mechanism of magnesium protoporphyrin synthesis. Synthesis of chlorophylls. The pathway from protoporphyrin to chlorophyll and bacterio- chlorophyll has received little analysis at an enzymic level and present knowledge is derived mainly from the compoiinds accumulating in cultures of mutant strains of Chlorella and Rps. spheroides (Table 3; Scheme 2). An early step is the formation of magnesium proto- porphyrin monomethyl ester. The specific incorporation of C^"^- formate by Chlorella into the methyl ester group of chlorophyll sug- gested that this group is derived from a one-carbon unit (32). In Rps. spheroides l-C^^-methionine labels the methyl group of bacterio- chlorophyll specifically; inhibition of methyl transfer from methionine by ethionine could account for the inhibition of bacteriochlorophyll synthesis by this analogue (33). Chromatophore preparations from this organism form magnesium protoporphyrin monomethyl ester when in- cubated with S-adenosylmethionine and magnesium protoporphyrin (39). Neither protoporphyrin nor the corresponding porphyrinogen is methylated in this system, showing that the biosynthetic sequence is: Protoporphyrin ^ magnesium protoporphyrin *- magnesium protoporphyrin monomethyl ester TETRAPYRROLES 43 TABLE 3 Chlorophyll derivatives accumulated by photosynthetic microorganisms Compound Accumulated by Reference Mg-protoporphyrin Mg-protoporphyrin monomethyl ester Mg-vinylpheoporphyrin a 5 (protochlorophyll- ide a) Protochlorophyll-type pigments Pheophorbide a Chlorella mutant Chlorella mutant Rps. capsulata Rps. spheroides Chlorella mutant Rps. spheroides chlorophyll- less mutants Rps. spheroides carotenoid- less mutants* (34) (34) (17) (18) (35) (36) (37) (38) * This mutant excretes a variety of chlorophyll derivatives of which pheo- phorbide a and a closely related compound are the major components; a minor component has the tetrahydropyrrole ring system of bacteriochlorophyll and is probably bacteriopheophorbide. Mg Me Protoporphyrin »- Mg-protoporphyrin ^Mg-protoporphyrin monomethyl ester donor 4-5 steps holochrome complex Light Chlorophyll a t Chlorophyllide o Mg-vinvlpheoporphyrin a 5 I I t D..-^ r - bacteriochlorophyll pheophorbide a Scheme 2. Possible pathways from protoporphyrin to chlorophylls. 44 THE BACTERIAL PHOTOCHEMICAL APPARATUS Four to five intermediate stages between magnesium protoporphyrin monomethyl ester and magnesium vinyl pheoporphyrin a.^ (MgVP; protochlorophyllide a; see Fig, 3) have been postulated by Granick but these are completely unknown. MgVP is accumulated by a Chlorella mutant, which, unlike the wild type, does not form chlorophyll vinless illuminated. It also accumulates in etiolated leaves treated with ALA in the dark (34), This pigment can exist in three forms (40), Type 1 has an absorption maximum at 631 m/i and is bleached by light; this is the free pigment. Type 2 has an absorption maximum at 650 m/i and is converted in the light to chlorophyllide a; this is attached to the holo- chrome complex studied by Smith (41), Type 3 also has an absorption maximum at 650 m/j. but is not transformed by light; this may be attached to a holochrome which lacks a reducing component. Free MgVP when accumulated by etiolated leaves in the dark from ALA is not utilised for chlorophyll synthesis upon subsequent illumina- tion. For such a transformation to occur the pigment must presumably be attached to the holochrome complex, at least in the higher plants. It appears that free MgVP formed from added ALA cannot enter this complex; it is possible that attachment of the tetrapyrrole component to the holochrome may occur at a stage before MgVP, The protochlorophyll-like pigment isolated from a strain of i?/)s . spheroides , unable to form bacteriochlorophyll, resembles type 1 MgVP; it has an absorption maximum at 623 m/i and it probably lacks the phytol group (36), A similar, possibly identical pigment has been identified spectroscopically in mutants of the same organism (37), Perhaps the greatest mystery in chlorophyll synthesis is the nature of the light reaction which results in reduction of ring D of MgVP, The fully functional holochrome complex presumably contains a light- activated reducing system. Most of the simple algae form chlorophyll in the dark and presumably contain an additional enzyme system which catalyses the reduction without the intervention of light. Bacteria of the Athiorhodaceae family must have a similar type of system since they can form bacteriochlorophyll in the dark provided the oxygen pressure is low (42,43). Protein-bound intermediates may participate in the final stages of bacteriochlorophyll synthesis. The obligatory association of pigment formation and protein synthesis or turnover supports this (43,44,45), It seems likely that the enzymes for these later stages are in the chromatophores; this is suggested by the obser- vation of Tait & Gibson (39) that the magnesium protoporphyrin methylating system is confined to the chromatophore fraction. The accumulation of pheophorbide a (Fig, 3) by the blue-green carotenoidless mutant of Rps. spheroides provides additional evidence that bacteriochlorophyll and chlorophyll synthesis proceeds by a com- mon path, but the significance of the appearance of this magnesium- free pigment is not clear. A compound similar if not identical with it TETRAPYRROLES 45 H,C CH2 — CH3 CH2 COOCH3 COOH Mg vinylpheoporphyrin a5 HX H,C CH, — CH, CH, Pheophorbide a Fig. 3. Magnesium vinylpheoporphyrin 05 (top) and Pheophorbide a (bottom). 46 THE BACTERIAL PHOTOCHEMICAL APPARATUS is accumulated by suspensions of Rps. spheroides incubated with 8- azaguanine in the dark under low aeration (unpublished observations). Griffiths (37) has recently isolated a series of mutants of i?/)S . spheroides which do not form bacteriochlorophyll but which accumulate a magnesium-free pigment with a spectrum similar though not identi- cal with pheophorbide a. Its absorption maxima correspond with those of Chlorohium 650 pheophytin. The slight differences in spectral characteristics from pheophorbide a could denote differences in the side chains of the dihydrotetrapyrrole nucleus though the data do not show whether the compound is phytolated, Esterification withphytol. This may be the final stage in the synthesis of chlorophylls. In higher plants short exposure to light results in formation of the un- esterified chlorophyllide a, which is converted to the esterified chlorophyll a by subsequent incubation in the dark (46). Accumulation of the unphytolated-protochlorophyll pigment and of pheophorbide a by mutants of Rps. spheroides suggests that esterification also occurs at a late stage in bacteriochlorophyll synthesis. It seems likely that the esterification with phytol occurs within the lipoprotein complex of the chromatophore because of the hydro- phobic nature of the phytol residue. CONTROL OF TETRAPYRROLE SYNTHESIS It is obviously to the advantage of organisms which form chloro- phylls to have mechanisms for regulating synthesis of the pigments to fit environmental demands. Not only does the formation of chlorophylls make a considerable drain on glycine and succinyl CoA but organisms making these pigments must also elaborate a formidable array of enzymes for the steps in the biosynthetic pathway. That control mechanisms do exist is shown most clearly in the Athiorhodaceae, which grow either aerobically in the dark or anaerobically in the light and which form bacteriochlorophyll and carotenoids only under photo- synthetic conditions. There is now evidence, though far from complete, that synthesis of the pigments is subject to control by negative feed- back mechanisms and by enzyme repression. Iron and hemin. The excretion of free porphyrins by photosynthetic bacteria under conditions of iron deficiency shows a close resemblance to other bio- synthetic systems where breakdown of a negative feedback control occurs due to inability to form a metabolite which inhibits an early enzymic reaction in the biosynthetic pathway (47). Thus, the accumu- TETRAPYRROLES 47 lation occurs only in the later stages of growth and the quantities of porphyrin formed far exceed the amount of tetrapyrrole, as hemes and bacteriochlorophyll, which are formed with adequate iron (26), Iron acts catalytically in preventing porphyrin formation and this could suggest that it is needed to form a compound which controls an early step in the synthesis of porphyrins by negative feedback inhibition. Work with whole cells and partly purified preparations of ALA synthe- tase of Rps. spheroides suggest that heme may exert such a controlling function (48,49), The enzyme is inhibited by low concentrations of hemin (down to 10"^ M); porphyrin accumulation by intact cells is in- hibited by hemin when glycine and succinate are the substrates but conversion of ALA to porphyrin is unaffected. Under normal conditions hemes within the cell are mostly if not entirely present as hemopro- teins, Hemoproteins (hemoglobin and myoglobin) also inhibit ALA synthetase and it is in this form that hemes may function in the control mechanism. Besides inhibiting the action of ALA synthetase hemin also represses synthesis of this enzyme by growing cultures of Rps. spheroides (50), The conclusion from these various observations is that the intracell- ular level of hemes, probably as hemoproteins, participate in the con- trol of tetrapyrrole synthesis by influencing both the synthesis and the action of the synthetase. The effect of hemin on the synthetase does not, however, satisfactorily account for all the effects of iron observed with Rps. spheroides , in particular the action of iron in promoting bacteriochlorophyll synthesis, Bacteriochlorophyll might also act as a controlling factor. Free bacteriochlorophyll does not inhibit ALA synthetase but this may not be significant since in the cell it exists in combination with the chroma- tophore complex. Gibson et al. (33) have suggested that an intermediate between magnesium protoporphyrin methyl ester and bacteriochloro- phyll may act as a feedback inhibitor of ALA synthetase and their ob- servations on the effect of ethionine on Rps. spheroides support this. This analogue inhibits bacteriochlorophyll synthesis, probably by in- terfering with the methylation step, but stimulates the accumulation of coproporphyrin. Since the path of heme and chlorophyll synthesis is common up to the protoporphyrin stage and since photosynthetic organisms must form both types of tetrapyrrole for photosynthetic development, control mechanisms might be expected to operate at the branch joint leading to iron and magnesium protoporphyrins. Information about this must await knowledge of the enzyme systems catalysing the insertion of the metals into the tetrapyrrole structure. Light intensity and oxygen. On teleological grounds it is an advantage for an organism, which relies onlightfor energy, to be able to increase or decrease its chloro- 48 THE BACTERIAL PHOTOCHEMICAL APPARATUS phyll content in response respectively to decreased or increased light intensity. This is analogous to the response of animals which form more hemoglobin under diminished oxygen tensions. Such adjustments in chlorophyll content occur in the plant kingdom as shown by the higher levels in shade leaves compared with sun leaves, while unicellular algae groAvn in dim light are richer in chloro- phyll than those grown in bright light (10). The elegant experiments of Cohen- Bazire et al. (42) have shown that synthesis of photosynthetic pigments by cultures of Athiorhodaceae is regulated by light intensity. In Rps . spheroides the rate of synthesis of bacteriochlorophyll is inversely proportional to the light intensity and the pigment content of cells grown in dim light (50 ft-c) is about eight times higher than in those grown in bright light (5000 ft-c). On transfer from dim to bright light or vice versa cultures rapidly adjust their pigment level by preferential synthesis or by transient repres- sion of pigment formation. In the Athiorhodaceae oxygen exerts a spectacular control over pigment synthesis as shown by the almost complete absence of bac- teriochlorophyll and carotenoids in organisms grown aerobically in the dark. Introduction of oxygen into cultures growing in the light results in an immediate arrest of pigment synthesis and this is reversed by restoration of anaerobic conditions (42). These experiments suggest that absence of pigment in dark- aerobic cultures might be due to re- pression of their synthesis by oxygen rather than to an obligatory re- quirement for light. This was confirmed by showing that Athiorhodaceae can indeed form bacteriochlorophyll and carotenoids in the dark pro- vided that the oxygen tension is reduced (43). With suspensions of Rps. spheroides forming bacteriochlorophyll in the dark the oxygen tension which permits synthesis is critical and must presumably be sufficient for general metabolism (e.g. to supply ATP by oxidative phosphoryla- tion) yet insufficient to cause repression of pigment formation. In an attempt to understand the mechanism by which oxygen re- presses bacteriochlorophyll synthesis, attention has been given to the key intermediate, succinyl CoA. In organisms such 3.S Rps. spheroides with a tricarboxylic acid cycle, ALA synthetase has to compete for succinyl CoA with enzymes which would pull it through the cycle; utili- sation via the cycle might be favored by high oxygen tensions since there is evidence that oxidation of succinate becomes rate- limiting under anaerobic conditions (51). Increasing the level of the synthetase could favor diversion of the succinyl CoA towards tetrapyrrole syn- thesis. There is in fact a strong correlation between level of the sjm- thetase and ability to form bacteriochlorophyll; the enzyme is five to ten times higher in Rps. spheroides grown anaerobically in the light than when grown aerobically (43). In addition, synthesis of the enzyme is repressed by high oxygen tensions, though, like bac- teriochlorophyll, it is formed at a maximum rate under low oxygen TETRAPYRROLES 49 tension in the dark (43,50), ALA dehydratase is also repressed by oxygen, suggesting that co-ordinate repression by oxygen occurs in the tetrapyrrole pathway. Regulation of bacteriochlorophyll synthesis by oxygen and by light intensity may operate by a similar mechanism (42), In support of this the levels of ALA synthetase and dehydratase in Rps , spheroides are affected by the light intensity just as they are influenced by the oxygen pressure; their rates of formation in growing cultures are inversely proportional to the light intensity (50), These observations suggest that one of the ways in which oxygen and light may influence the formation of bacteriochlorophyll is by repressing synthesis of enzymes concerned in early stages of the biosynthetic pathway. They tell us nothing of the mechanism by which the repressing effect is exerted. Nor do they fully account for all the observed effects of oxygen. Control by enzyme repression only would result, upon the introduction of oxygen, in a gradual fall in the dif- ferential rate of bacteriochlorophyll synthesis by cultures growing in the light; this would occur as the enzymes already present became diluted out. In fact, oxygen produces an immediate and complete stop- page of pigment synthesis (42,50), This suggests that oxygen is inhibit- ing the action of one or more enzymes on the biosynthetic path. Since porphyrins do not accumulate in oxygen- repressed cultures it seems that an early stage is either directly or indirectly inhibited by oxygen. In addition to the effects of oxygen and light on tetrapyrrole forma- tion, consideration must be given to their action on the carotenoids. These respond to the environment in the same way as bacteriochlor- ophyll, yet the biosynthetic pathways have nothing in common, except perhaps for the phytol residual of the bacteriochlorophyll. A full understanding of the response of the pigment system to the environment can only come when we know more about the steps in the formation of the chlorophylls and carotenoids and about the stage where their synthesisbecomesinterwoven with the lipoproteins of the chroma- tophore and chloroplast structures. REFERENCES Fischer, H., and Stern, A., Die Chemie des Pyrrols, Vol. II, part 2. Akad- emische Verlagsgesellschaft M. B. H., Leipzig, 1940. Golden, J. H., Linstead, R. R.,and Whitham, G. M., Chlorophyll and related compounds. VII. The structure of bacteriochlorophyll. J. Chem. Soc, 195S, 1725. Stanier, R., and Smith, J. H. C, The chlorophylls of green bacteria, Bio- chiw. Biophys. Acta, 41, 478 (1960). Holt, A, S,, and Hughes, D. W., Studies of Chlorobium chlorophylls. III. Chlorohium chlorophyll (650). J.Am. Chem. Soc, 83, 499 (1961). 50 THE BACTERIAL PHOTOCHEMICAL APPARATUS 5. Holt, A. S., Hughes, D. W., Kende, H. J., and Purdie, J. W., Studies on Chlorobium chlorophylls. V. Chlorohium chlorophylls (660). J. Am. Chem. Soc, 84, 2835 (1962). 6. Rapoport, H.,and Hamlow, H. P., Chlorobium chlorophyll-660. The esteri- fying alcohol. Biochem. Biophys. Res. Commiin., 6, 134 (1961). 7. Stanier, R. Y., Formation and function of the photosynthetic pigment system in purple bacteria. Brookhaven Symp. Biol., 11, 43 (1959). 8. Kornberg, H. L., and Lascelles, J., The formation of isocitratase by the Athiorhodaceae. J. Gen. Microbiol., 23,511 (1960). 9. Larsen, H., On the microbiology and biochemistry of the photosynthetic green sulfur bacteria. Kgl. Norske Videnskab. Selskabs. Skrifter, 26, 1 (1963). 10. Rabinowitch, E. I., Photosynthesis and related processes (3 vols.). Lnter- science Publishers Inc., New York, 1945, 1951, 1956. 11. Clayton, R. K.,and Smith, C, RJiodopseudomonas spheroides: high catalase and blue-green double rcMtants. Bio chem. Biophys. Res. Commun., 3, 143 (1960). 12. Clayton, R. K. , Aerobiosis and the hemoprotein content of photosynthetic bacteria. Archiv. Mikrobiol., 33, 260 (1959). 13. Geller, D. M., Oxidative phosphorylation in extracts of Rhodospirillum rubrum. J. Biol. Chem., 237, 2947 (1962). 14. Bartsch, R. G., and Kamen, M. D., Isolation and properties of two soluble heme proteins in extracts of the photoanaerobe, Chromatium. J. Biol. Chem., 235, 825 (1960). 15. van Niel, C. B., The culture, general physiology, morphology and classifi- cation of the non-sulfur purple and brown bacteria. Bacteriol. Revs., 8, 1 (1944). 16. Lascelles, J., Tetrapyrrole synthesis in microorganisms, p. 335 in The Bacteria (I. C. Gunsalus and R. Y. Stanier, ed.). Vol. 3. Academic Press Inc., New York, 1962. 17. Cooper, R., The biosynthesis of magnesium protoporphyrin monomethyl ester and coproporphyrinogen by Rhodopsendonwnas capsulata. Biochem. J. 89, 100 (1963). 18. Jones, O. T. G., The production of magnesium protoporphyrin monomethyl ester by Rhodopseiidomonas spheroides . Biocliem. J., S6, 429 (1963). 19. Granick, S., The structural and functional relationships between heme and chlorophyll. Harvey Lectures, 44, 220 (1950). 20. Granick, S., and Mauzerall, D., The metabolism of heme and chlorophyll, p. 525 in Metabolic Pathways (D. M. Greenberg, ed.). Vol. 2. Academic Press Inc., New York, 1961). 21. Lascelles, J., Synthesis of tetrapyrroles by microorganisms. Physiol. Revs., 41, 417 (1961). 22. Bogorad, L., The enzymatic synthesis of porphyrins from porphobilinogen, I, II, HI. J. Biol. Chem., 233,501, 510, 516 (1958). 23. Hoare,D. S.,and Heath, H., The biosynthesis of porphyrins from porphobi- linogen by RJiodopseudomonas spheroides. 2. The partial purification and some properties of porphobilinogen deaminase and uroporphyrinogen de- carboxylase. Biochem. J., 75,679 (1959). 24. Sano, S., and Granick, S., Mitochondrial coproporphyrinogen oxidase and protoporphyrin formation. J. Biol. Chem., 236, 1173 (1961). 25. Porra, R. J., and Falk, J. E., Protein-bound porphyrins associated with protoporphyrin biosynthesis. Biochem. Biophys. Res. Commun., 5, 179 (1961). 26. Lascelles, J., The synthesis of porphyrins and bacteriochlorophyll by cell suspensions ol Rhodopseudomonas spheroides. Biochem. J., 62, 78 (1956). TETRAPYRROLES 51 27. Lascelles, J., An assay of iron protoporphyrin based on the reduction of nitrate by a variant strain of Staphylococcus aureus: synthesis of iron protoporphyrin by suspensions of Rhodopsetidomonas spheroides. J. Gen. Microbiol., 15, 404 (1956). 28. Labbe.R. F.,and Hubbard, N., Metal specificity of the iron protoporphyrin chelating enzyme from rat liver. Biochim. Biophys. Acta, 52, 130 (1961). 29. Yoneyama, Y., Ohyama, H., Sugita, Y., and Yoshikawa, H., Iron-chelating enzyme from duck erythrocytes. Biochim. Biophys. Acta, 62, 261 (1962). 30. Burnham, B. F., Investigations on the action of the iron containing growth factors, sideramines; and iron containing antibiotics, sideromyc ins. J. Gen. Microbiol., 32, 117 (1963). 31. Granick, S., Magnesium protoporphyrin as a precursor of chlorophyll in Chlorella. J. Biol. Chem., 175, 333 (1948). 32. Green, M., Altman, K. 1., Richmond, J. E., and Salomon, K., Role of formate in the biosynthesis of chlorophyll a. Nature, 179, 375 (1957). 33. Gibson, K. D., Neuberger, A., and Tait, G. H., Studies on the synthesis of porphyrins and bacteriochlorophyll by Rhodopseudomonas spheroides. 2. The effects of ethionine and threonine. Biochem. J., 83, 550 (1962). 34. Granick, S., Magnesium protoporphyrin monoester and protoporphyrin monomethyl ester in chlorophyll biosynthesis. J. Biol. Chem., 236, 1168 (1961). 35. Granick, S., Magnesium vinyl pheoporphyrin f/5, another intermediate in the biological synthesis of chlorophyll. J. Biol. Chem., 183, 713 (1950). 36. Stanier, R. Y., and Smith, J. H. C, Protochlorophyll from a purple bac- terium. Carnegie Institute of Washington Year Book, 58, 336 (1959). 37. Griffiths, M., Further mutational changes in the photo synthetic pigment system of Rhodopseudomonas spheroides. J. Gen. Microbiol., 27, 427 (1962). 38. Sistrom, W. R., Griffiths, M., and Stanier, R. Y., A note on the porphyrins excreted by the blue-green mutant of Rhodopseudomonas spheroides. J. Cell. Comp. Physiol., 48, 459 (1956). 39. Tait,G. H., and Gibson, K. D., The enzymic formation of magnesium proto- porphyrin monomethyl ester. Biochim. Biophys. Acta, 52, 614 (1961). 40. Granick, S., The pigments of the biosynthetic chain of chlorophyll and their interactions with light. In: Symposium No. 6, Vth International Congress of Biochemistry, Moscow (1961). Pergamon Press Ltd., Oxford, 1963. p. 176. 41. Smith, J. H. C., Chlorophyll formation and photosynthesis. In: Symposium No. 6, Vth International Congress of Biochemistry , Moscow, 1961. Per- gamon Press Ltd., Oxford, 1963. p. 151. 42. Cohen-Bazire, G., Sistrom, W. R., and Stanier, R. Y., Kinetic studies of pigment synthesis by non-sulfur purple bacteria. J. Cell. Comp. Physiol., 49, 25 (1957). 43. Lascelles, J., The synthesis of enzymes concerned in bacteriochlorophyll formation in growing cultures of Rhodopseudomonas spheroides . J. Gen. Microbiol., 23, 487 (1960). 44. Sistrom, W. R., Observations on the relationship between formation of photopigments and the synthesis of protein in Rhodopseudomonas spheroides. J. Gen. Microbiol., 25, 599 (1962). 45. Bull.M. J., and Lascelles, J., The association of protein synthesis with the formation of pigments in some photo synthetic bacteria. Biochem. J., 87, 15 (1963). 46. Wolff, J. B., and Price, L., Terminal steps of chlorophyll a biosynthesis in higher plants. Arc/i. Biochem. Biophys., 72, 293 (1957). 52 THE BACTERIAL PHOTOCHEMICAL APPARATUS 47. Umbarger, H. E., Feedback control by endproduct inhibition. Cold Spr. Harb. Synip. Quant. Biol., 26, 301 (1961). 48. Burnham, B. F., Evidence for a negative feedback system in the control of porphyrin biosynthesis. Biochem. Biophys . Res. Commun., 7,351 (1962), 49. Burnham, B. F.,and Lascelles, J., Control of porphyrin synthesis through a negative-feedback mechanismrstudies with preparations of 5-aminolaevulate synthetase and dehydratase irom Rhodopseudomonas spheroides. Biochem. J., 87,462 (1963). 50. Lascelles, J., The synthesis of enzymes concerned in bacteriochlorophyll formation in growing cultures of Rhodopseudomonas spheroides . J. Gen. Microbiol., 23, 487 (1960). 51. Wiame.J.M., Le role biosynthetique du cycle des acides tricarboxyliques. Advan. Enzymol., 18, 241 (1957). Note added in proof ipa.ge 35). It has been learned that another pigment designated as bacteriochlorophyll b, has been identified spectroscopically in a Rhodopseudomonas sp. Its structure is unknown (Eimhjellen, Aasmundrud, and Jensen, Biochem, Biophys. Res. Commun., 10,232, 1963). A NOTE ON THE EFFECT OF INHIBITORS OF ELECTRON TRANSPORT AND PHOSPHORY- LATION ON PHOTOPIGMENT SYNTHESIS m RHODOPSEUDOMONAS SPHEROIDES W. R. SISTROM The Biological Laboratories , Harvard University, and The Department of Biology, University of Oregon INTRODUCTION The general features of the control of photopigment synthesis in the photosynthetic bacteria are well known (1,2). The hypothesis orig- inally proposed (1) to account for the changes in the bacteriochlorophyll and carotenoid pigment contents of cells grown at different light in- tensities or oxygen tensions is adequate to a first approximation at least. It does not, however, account satisfactorily for all the kinetic observations (2), The hypothesis assumed that pigment synthesis is controlled by the ratio of the oxidized to the reduced form of an electron carrier. When the carrier becomes oxidized, for example by a sudden increase in light intensity, pigment synthesis is inhibited; conversely, when the carrier is reduced, pigment synthesis is accelerated. More recent experiments have shown that at constant light intensity the bacteriochlorophyll content of a culture depends directly upon its growth rate. For example, when Rhodopseudomonas spheroides is maintained in a chemostat the pigment content of the cells depends upon the dilution rate. If it is assumed that the rate of reduction of the electron carrier which controls pigment synthesis is proportional to the growth rate, then our original hypothesis can account for these re- sults also. It seemed possible that the site of control could be determined by studying the effect of inhibitors of electron transport and phosphoryla- tion on bacteriochlorophyll synthesis. This paper describes in a pre- liminary way some experiments along these lines. The results indicate that this will be a fruitful approach and suggest that pigment synthesis is controlled by the ratio of reduced to oxidized DPN, 53 54 THE BACTERIAL PHOTOCHEMICAL APPARATUS RESULTS AND CONCLUSIONS All experiments were conducted with Rps. spheroides, strain Ga. Medium B of Sistrom (3) was used. The cultures were aerated with 5% CO2 in N2. A lightintensity of 600fcs was used throughout. Two layers of red cellophane were used as a filter to avoid photoreactions with pigmented inhibitors. Bacteriochlorophyll and neurosporene were 100 .8 .3 1 1 1 1 1 1 ^^^ ^A—""'' ^^■^/y"' ^^ y^ " - ^^ ^^ ^ — m^ ^^^'""''^ -£X-^ — — y^^ ^^1:::^— ::r5^'''^^" ~ ^^^ ^jo--""^ Jt^^^O-^" -^^f'j^r^^ - 1 1 1 1 1 1 10 8 ^ E 6 o> 4i OL O ^ o _l X o lU \- o < CD 60 120 180 MINUTES 240 300 360 Fig. 1. Effect of dicoumarol on growth and bacteriochlorophyll synthesis. A culture was grown overnight at a light intensity of 800 fcs; one hour be- fore the start of the experiment the culture was diluted and 90 ml portions placed in rectangular lucite growth vessels and illuminated with 600 fcs. At zero time dicoumarol (6 x 10-5 m) was added to one vessel (open points); the other served as a control (filled points). Bacteriochlorophyll (^ug/ml), DandB; cell mass (OD at 1000 m/i), Oand#; protein (|/g/ml),A and A. INHIBITORS AND PHOTO PIGMENT SYNTHESIS 55 estimated as described previously (1); protein was determined by the Folin-Lowry method. Except in experiments with dinitrophenol, in- crease in protein paralled increase in turbidity at 1000 m/i, and in some experiments protein was not determined. Experiment 1. The effect of dicoumarol on growth and bacterio- chlorophyll synthesis is shown in Fig, 1. The growth rate is reduced to about 50% of the control. Bacteriochlorophyll synthesis is immedi- ately inhibited but eventually recommences. 180 MINUTES Fig. 2. Effect of dinitrophenol (DNP) on growth and bacteriochlorophyll synthesis. The experiment was conducted as described in the legend of Fig. 1, except that at zero time one culture received DNP (4 x 10-5 M). Symbols as in Fig. 1. 56 THE BACTERIAL PHOTOCHEMICAL APPARATUS Experiment 2. The effect of dinitrophenol (DNP) (6 x 10-5 M) is similar to that of dicoumarol (Fig. 2). When lower concentrations of either dicoumarol or DNP were used bacteriochlorophyll synthesis was inhibited but to a lesser degree (Table 1), Experiment 3. Amytal at a concentration of 1,6 x 10-3 M inhibits bacteriochlorophyll synthesis to a lesser extent than growth (Fig. 3); consequently, the specific bacteriochlorophyll content increases. This is in marked contrast to the results with DNP and dicoumarol, TABLE 1 Comparison of the effects of certain inhibitors on growth and bacteriochlorophyll synthesis. T , ., .. ,, , ., Relative Relative bacterio- Inhibitor Molaritv ^, ^ , , , ,, ^ ^ ^ growth rate chlorophyll content none — 100 100 DNP 3.0 X 10-5 62 79 ti 4.0 X 10-5 56 71 II 6.0 X 10-5 60 60 dicoumarol 2.5 X 10-5 72 80 " 5.0 X 10-5 64 71 amytal 1.0 X 10-3 71 111 II 1.6 X 10-3 65 126 atebrin 4.0 X 10-5 55 84 DL-5-methyl- tryptophan 2.5 X 10-4 64 64 In order to avoid the effects of self-shading, the bacteriochlorophyll content was determined in each case when the bacteriochlorophyll concentration was Afig/ml. Experiment 4. Cohen- Bazire and Kunisawa (4) have reported that the specific bacteriochlorophyll content is reduced in cultures grown at a low temperature. Fig. 4a shows the results of an experiment in which a culture was subjected to a sudden decrease in temperature. Bacteriochlorophyll formation is inhibited, but not completely as it is after addition of DNP or dicoumarol. Essentially the same result is obtained when growth is inhibited by 5-methyltryptophan (Fig. 4b). Fig. 5 shows the differential rate of bacteriochlorophyll synthesis in the presence of amytal and of 5-methyl-tryptophan. In Table 1 are shown the relative growth rates and bacteriochlorophyll contents ob- tained with inhibitors. INHIBITORS AND PHOTO PIGMENT SYNTHESIS 57 120 180 240 MINUTES 300 360 Fig. 3. Effect of amj^al on growth and bacteriochlorophyll synthesis. The ex- periment was conducted as described in the legend of Fig. 1, except that at zero time one culture (open points) received amytal (1.6 x 10-3 jv/). Symbols as for Fig. 1. These results can be summarized by saying that amytal mimics the effect of a sudden ^gcrease in light intensity, while DNP and dicoumarol mimic the effect of a sudden increase in light intensity. Bacterio- chlorophyll synthesis is completely, although temporarily, inhibited by the latter compounds. When the growth rate is reduced by lowering the temperature or by addition of 5-methyl-tryptophan there is only a partial inhibition of bacteriochlorophyll formation, Dicoumarol and DNP presumably uncouple phosphorylation and electron transport; consequently, the electron transport chain becomes more oxidized and bacteriochlorophyll synthesis is stopped. However, the specific pigment content will decrease because of continued growth 58 THE BACTERIAL PHOTOCHEMICAL APPARATUS e o o o .6h .4 .3 - ■■ I 1 1 ^ \ 1 - A-- ^- ^^' ^-=^-3-- D — ^ ^ ^t ^z- ^^ --^-'' n*s."^''^ '.g-' ^^'^ r , , , 1 1 1 1 \ 60 120 180 240 MINUTES 300 360 >- X Q_ O a: o _i X o o q: ^ I- U < -2 Fig. 4. Effect of temperature shift and of 5-methyl-tryptophan on growth and bacteriochlorophyll synthesis, (a) Two cultures were grown at 34°, at the time indicated by the arrow one culture (open points) was transferred to 23°, the second culture (filled points) remained at 34°. (b) This experiment was con- ducted as described in the legend for Fig. 1; at time zero DL-5-methyl- tryptophan (50 //g/ml) was added to one culture (open points). Bacteriochloro- phyll (/ig/ml), DandB; cell mass (OD at 1000 m^), Oand#. INHIBITORS AND PHOTO PIGMENT SYNTHESIS 59 5-MT 0.2 AO.D. AT 1000 mfi 0.3 0.4 Fig. 5. Differential rate of bacteriochlorophyll synthesis in the pres- ence of amytal and of 5-methyl-tryptophan {5-MT). The data of Figs. 3 and 4b have been replotted to show the increase in bacteriochloro- phyll relative to increase in cell mass (OD at 1000 m^). 60 THE BACTERIAL PHOTOCHEMICAL APPARATUS of the cells, the supply of photooxidant will accordingly fall and the electron carriers will return to their initial level of reduction and bacteriochlorophyll synthesis will recommence, Amytal is reported to be a poor inhibitor of cyclic photophosphory- lation in Rko do spirillum rubrum chromatophores (5); however, amytal is known to block the oxidation of DPNH by flavoprotein in mitochondria (6), It therefore seems reasonable to ascribe both the inhibition of growth and the stimulation of the differential rate of bacteriochlorophyll synthesis by amytal to a decrease in the rate of DPNH oxidation. In other words, the electron carrier which controls pigment synthesis is DPN. Preliminary results indicate that 2- w-nonyl- hydroxy quinoline-N- oxide and SN5949 have the same effect on bacteriochlorophyll synthesis as amytal. Atebrin inhibits photophosphorylation in /?.rM&rMW chromatophores (7) and DPNH oxidation in mitochondria (8), In each case it is likely that the inhibition is on a flavoprotein. It would be expected that atebrin should have the same effect as amytal. As can be seen from Table 1, this is not the case: atebrin inhibits growth and reduces the specific bacteriochlorophyll content. However, it should be noted that the pig- ment content is reduced by only 15%, although growth is inhibited by almost 50%. This is in contrast to the results with DNP, dicoumarol and 5-methyl-tryptophan. It is possible that atebrin inhibits growth not only by interfering with oxidation of DPNH but also in some other fashion unrelated to electron transport. The small reduction in specific bacteriochlorophyll content which is observed may represent the net result of these contrary tendencies. ACKNOWL EDGME NTS This work has been supported by grants from the National Science Foundation and the National Institutes of Health. REFERENCES 1. Cohen-Bazire, G.,W. R. Sistrom.and R.Y. Stonier, J. Cell. Co mp. Physiol., 49, 25 (1957). 2. Sistrom, W. R.,J. Gen. Microbiol., 28,607 (1962). 3. Sistrom, W. R., J. Gen. Microbiol., 22, 778 (1960). 4. Cohen-Bazire, G., and R. Kunisawa, personal communication. 5. Baltscheffsky, H., and M. Baltscheffsky, Acta Chem. Scand., 11,2^1 (1960). 6. Chance, B., In G3.eh\eY , Enzymes: Units of Biological Structure and Function, p. 447. Academic Press, New York (1956). 7. Baltscheffsky, H., Biochim. Biophys. Acta, 10, 1 (1960). 8. Low, H., Biochim. Biophys. Acta, 32, 1 (1959). THE HEME PROTEINS OF PHOTOSYNTHETIC BACTERIA* MARTIN D. KAMEN University of California, San Diego La Jolla, California INTRODUCTION The area of research I have been asked to discuss was inaugurated by an observation of Dr. L. P, Vernon when he and I were engaged in collaboration (1,2) on energy- storage reactions in chromatophore preparations in 1952. It is a source of particular satisfaction a decade later to be here at this session sponsored by the C. F. Kettering Foun- dation and its new research director, Dr. Vernon. The heme proteins of the photosynthetic bacteria have been con- sidered in the past primarily from the standpoint of function. This is so well documented now (3), even if still incompletely understood, that I propose in this paper to adopt another approach— that is, to consider the heme proteins as objects of intrinsic biochemical interest. To paraphrase President Kennedy's famous exhortation— "Let us not ask what heme proteins have done for photosynthesis, but what photosyn- thesis has done for heme proteins!" CLASSIFICATION The index of progress in any science can be indicated by the status of classification schemes. In the case of heme proteins, and natural tetrapyrrole proteins in general, classifications based on function have maintained a certain priority over those based on structure. This follows, obviously, because it is usually easier to make functional assignments than to solve structural problems, especially for im- portant macromolecules of biological importance, such as proteins. Thus, in the tetrapyrrole- conjugated proteins, which include the mag- Researches on which this paper is based were performed with the aid of sub- sidies from the National Institutes of Health, National Science Foundation, and the Charles F. Kettering Foundation, and have extended over a period of ten years, beginning in 1953. My associates in chronological order have been L. P. Vernon, J. W. Newton, R. G. Bartsch, T. Horio,J. A. Orlando, S. Taniguchi, and K. Dus. 61 62 THE BACTERIAL PHOTOCHEMICAL APPARATUS nesium and iron chelates, functions which are readily apparent, such as photosensitization for the chlorophylls (magnesium chelates) and oxidation catalysis for heme proteins (iron chelates), are familiar to all of us. The structural classification of heme proteins must be made on the basis of imperfectly understood chemical properties and behavior, as deduced from magneto- chemical, spectrochemical, and physical or- ganic studies on the proteins and on model compounds. A simple summary of present classification schemes can be presented, based on four generally accepted criteria: (1) The basic chelate structure. This is the beta- substituted tetra- pyrrole structure, associated with type III porphyrins. Three possible variations occur. The first, and most frequent, is based on the por- phyrin skeleton, in which two of the pyrrole moieties possess double bonds which do not participate in the conjugated macrocyclic resonance system. This we can call the "P" class. The second is the dihydro- porphyrin, or chlorin, in which one of these double bonds is removed by hydrogenation. This occurs primarily among the magnesium chel- ates (chlorophylls), but one example is known in the heme proteins— namely, the D-type cytochrome (4) (5), formerly called "cytochrome a^-" This we can call the "D" class. Thirdly, we must entertain the possibility that the tetrahydroporphyrin structure, in which no non- participating double bonds are left, may also be found to occur for heme chelates, as it does for the magnesium chelates (bacteriochloro- phyll). This gives a third class, which we can term "T." For the present, we can assume that general structures, other than type III, will not occur. One contribution from the study of bacterial cytochromes is already apparent in that we are prepared to accept the possibilities of classes "D" and "T." (2) Ligand interactions. The special effects created by the intro- duction of ligand groups from the protein into the extraplanar positions of the tetrapyrrole ring cannot be defined completely on the basis of any single chemical criterion, but the magnetochemical behavior is as adequate as any. This criterion leads to specification of three sub- classes for each of the main classes established by criterion 1, These are "high spin" (H), "low spin" (L), and "mixed spin" (M). The M subclass is actually derivative from the other two because it is probable that no absolutely pure H or L state exists. However, as a practical matter, the totality of heme proteins known at present falls easily into three such classes. (3) Binding type. The chelate may be bound to the protein by simple acid-base linkages which may be split easily by the usual acid- acetone treatment; or it may be bound by covalent linkages (thioether, ester, carbon-carbon bonds, etc.), which resist such treatment. These two cases can be distinguished by the letters "F" ("free") and "C" ("cova- lent"). HEME PROTEINS 63 (4) Degree of saturation of side chains. The presence of electron- withdrawing side- chains (unsaturated groups) provides one further sub- division, which may be termed "U," whereas their absence gives a subdivision, termed "S" (saturated). It follows thatthirty-sixpossible combinations exist, twelve for each of the three main classes; i.e., for the P class there are the following subclasses: HFU, HFS, HCU, HCS, LFU, LFS, LCU, LCS, MFU, MFS, MCU, MCS. The conventional classes of heme proteins, derived from studies of aerobic tissues (beef, yeast, etc.) are bunched into a relatively few of the above subclasses. Thus, the oxygen storage and transport pro- teins (myoglobins, hemoglobins, chlorocruorins) fall into the PHFU class. So do the catalases and some peroxidases. Myeloperoxidase probably belongs in the class PHCU, as does lactoperoxidase. The oxidation catalysts, cytochromes b and c, appear in the classes PLFU and PLCS, respectively. The oxidases, or "a" type cytochromes, appear to belong to the class, PMFU. Thus, just five combinations appear to include all the conventional heme proteins. The bacterial heme proteins supply a few more. Thus, the so-called "fl^" ^^^ "^4" cytochromes, from E. coll and Acetobacter s^. (4) (5), belong somewhere in the "D" class. In addition, a diheme protein which occurs in Pseiidomonas sp. (6) appears to fall into both classes «D. . ." and "PLCS". The photosynthetic bacteria, together with the plant systems, pro- vide representatives of the "b" and "c" type cytochromes, as well as peroxidases and catalases. In addition, the purple photosynthetic bac- teria have a type of oxidase— called "cytochrome o" (7) (8)— which appears to be a particle-bound member of the same class as the per- oxidases. It is uncertain whether it is similar in spin type, so it may conceivably fall into a "PM" category, like the "a" type cytochromes. Finally, there is the authentic new class of heme proteins— provision- ally called "cytochromoids" and known previously as "RHP"-type proteins (3e)— which falls into the class PMCS. Thus, we see that the bacteria have added at least three new combinations, on the basis of structure alone. More than this, they have alerted biochemists to the possibility that new combinations may exist, not only in bacteria and in plants, but also in animal tissues. PROPERTIES AND FUNCTIONS I wish now to summarize briefly present notions about the structures and functions of heme proteins in photosynthetic bacteria. I shall deal with these in terms of the various general classes known. 64 THE BACTERIAL PHOTOCHEMICAL APPARATUS (a) General considerations. First, I should bring to your attention certain general considerations which apply to all the work which has been done. In researches with bacteria as source material, amounts of protein available rarely exceed a few milligrams. Even this minute amount is obtained only after rather arduous culture and isolation procedures. For instance, to obtain 10 milligrams of the pure c-type cytochrome ironx Rho do spirillum rubrum — a. microorganism very rich in this heme protein— 10 grams dry weight of bacteria are required. Most work on protein fine structure— such as sequence determination, x-ray analyses— demands the sacrifice of literally kilograms of ma- terial. Another uncertainty, less often encountered in conventional aerobic systems where function is rather obvious, is the determination of function in bacterial metabolism. Bacterial systems employ heme proteins in a variety of ways which are quite different from those commonly associated with the aerobic processes of mammals. Hence, assays for activity cannot always be applied. Future researches will tend to the development of procedures for bulk culture so that greater amounts of material are available. In ad- diton, intensified enzymic analyses of particle, or insoluble, electron transport systems coupled to energy storage and reduction of substrates other than molecular oxygen will be likely, (b) c-Type cytochromes, Theseoccur most frequently in the greatest concentrations of all the bacterial heme proteins. Pure specimens have been obtained in varying amounts, ranging from a few micrograms up to hundreds of milligrams, from all species of the photosynthetic bacteria available. In general, they show no marked variance from cytochrome c, as indicated by the usual spectrochemical criteria, sequence, and other structural studies, but as a rule are wholly in- active in the classic cytochrome c oxidase system of mitochondrial tissues. Moreover, they usually exhibit acid isoelectric points, owing to a preponderance of aspartic and glutamic acid residues. Another in- teresting difference is that none of the bacterial proteins show an acetylated N-terminal group, as in cytochrome c. There are great variations in size, oxidation potentials, and associated properties. Some data (9,10,11) which relate to the composition of c-type cyto- chromes are exhibited in Table 1, Functionally, they differ completely from the conventional cyto- chrome c, in that they appear to be associated wholly with the photo- oxidase system of chromatophores. That is, they do not act as ter- minal catalysts in the reduction of molecular oxygen. The most in- formed guess, at present, based on the very extensive accumulation of data from enzymic studies, dynamic spectrophotometric and flash spectrophotometric observations (3,12-15), implicate the c-type cyto- chromes (including the analogous heme proteins of plant chloroplasts) as substrates for the primary photochemistry of the photosynthetic process. The reaction is not certain, but involves one of two alterna- HEME PROTEINS 65 TABLE 1 Composition ofc-Type Cytochromes* Amino Acid Composition Per Heme: R. rubrum - (LySj^^ HiS2 Asp-j^^ ThPg Ser^ Gly^ Pro^ Glyg Ala^g Valg lieu Leu Met Cys^ Tyr Phe Try^^N ; 9 amides Chromatium - (Lys^^ His^ Argg Asp^^ ^^^^12 ^^^15 ^^^3 ^^°18 ^1^29 ^^^35 ^^^9 "^^17 ^^V ^^% 'ry^l3 ^^^0 Rps. Pahistris- CLys^^ms^ ^^Pl5 "^^^8 ^^^2 ^^^8 ^^^3 ^^^12 Ala^g Valg Ileu^ LeUg Cys^ Tyr^ Phe^ Try^) Heme Peptide Sequence: R. nibnim - H^N-Ser-Lys-Cys-Leu-Ala-Cys-His-Thr-Phe-Asp-Glu-Gly- Ala-Asp NH2 -Lys-COOH (Residues 14-28) End-Group Sequences ; Glu -Asp- (NH,) R. rubrum - H N-Glu-Gly-Asp-Ala-Gly-Ala Lys-COOH Chromatium - H^N-Glu * These data have been taken both from published articles (9,10,11) and unpub- lished observations by K. Dus, H. de Klerk, and M. D. Kamen. tives— either the excited chlorophyll oxidizes the cytochrome, after loss of an electron, or it oxidizes the cytochrome before loss of an electron. Either possibility is consistent with the data now at hand, (c) &-Type cytochromes. Very little is known about these heme pro- teins, except that they exist in amounts which may approach those of the c-type cytochromes. Only one instance of a solubilized specimen has been reported (16). Attempts to link the6-type cytochromes of the purple photosynthetic bacteria with a conventional function in the elec- tron transport chain, coupled to phosphorylation either in dark or light, have been frustrated by the simultaneous occurrence of the o-type cytochromes and cytochromoids, spectrochemical characteristics of which mask expected spectral shifts which might be ascribed to &-type cytochromes, (d) Catalases and peroxidases. Practically nothing is known about these types of heme proteins, except that their presence is evident 66 THE BACTERIAL PHOTOCHEMICAL APPARATUS in chromatophore preparations, even those from strict anaerobes. Some qualitative observations indicate the catalase of Chromatium to be a heme protein (17), Proto-heme, the usual prosthetic group, can be extracted by acid- acetone (18), but this may arise wholly from the 6-type cytochrome which is present, (e) "o-Type" cytochromes. The dark oxidase activity of R. rubrum seems to require the presence of a particle-bound heme protein with spectrochemical characteristics like those of a cytochromoid (see below), in the visible range. No isolation of these proteins has been reported. Researches, based on analysis of spheroplasts, obtained by lysis of dark-grown cells, and other systems enriched in the aerobic phosphorylation system, are needed. It may be remarked here that not a single bacterial oxidase, let alone the "o-type" proteins, has been characterized as yet. (f) Cytochromoids (3e). This class illustrates best the surprises which may be in store for biochemists, when they begin to take bac- terial heme proteins more seriously. Again, we owe to Leo Vernon the original observation which led to the discovery of these proteins. He noted in 1953 that trichloracetic acid extracts of R. rubrum con- tained a heme protein, other than a cytochrome of the "c" type, with an absorption spectrum like that of myoglobin, or hemoglobin (2), Since then, our laboratory has been engaged in continuous research in an effort to elaborate the nature and function of what we now call "cytochromoids," Parenthetically, I may add that the existence of cyto- chromoids is a surprise only to those who persist in ignoring possibili- ties of combinations such as those given in the classification scheme I have described above. The official definition, as proposed by the Commission on Enzymes of the International Union of Biochemistry (5), is "heme proteins with a hemoglobin- like spectrum and a reactivity with ligands which do not react with cytochrome c." Cytochromoids are essentially heme pro- teins in which the normal heme prosthetic group retains the high- spin or mixed-spin character of "open" type heme proteins (19), while being bound covalently as in cytochrome c. The mixed- spin character used in classification emerges only in the oxidized forms of cyto- chromoids so far studied. There are just two specimens— one from R. rubrum (I), the other from Chromatium (II). Another specimen, recently isolated from Rps. palustris (2), shows the same spectral characteristics and ligand behavior as I and II, but differs in that it has a high oxidation potential (E^i 7 250 mv) and is not autoxidizable, whereas I and n have low oxidation potentials (Ej^ 7 ~-8 to -5 mv) and are rapidly autoxidizable (3e). Thus, with just three specimens ex- amined thoroughly, the cytochromoids appear already to exhibit a wide range of physicochemical character, just as do the "c" cytochromes. Both I and II have molecular weights close to 28,000, and contain two heme groups per mole. Sequence analysis for the diheme- HEME PROTEINS 67 containing moiety of n has been accomplished and results are consistent with the assumption that one heme is attached by thioether bonds to two cysteinyl residues, separated by two residues, as in cytochrome c. The placement of the other heme is uncertain, as only one more cysteine is available for linkage of thecovalenttype. Larger quantities of protein will be needed to permit further elaboration of the peptide structure as well as that for the whole protein. The nature of the second heme group is still in question, also. It is certain that it does not differ in oxidation potential appreciably from the value found by titration of the protein. The primary sequences for I and II, beginning at the terminal end, show interesting correlations (see Table 2), Thus, for II, the sequence TABLE 2 Composition of Cytochromoids Amino Acid Composition Per Heme: R. ruhrum - (hys^^ HiSg Arg^ ^ ^^Pg 5 '^^^7 ^^^10 ^^'^12.5 ^^°4 Gly^ Ala^g Val^ ^ IleUg ^ LeUg Tyrg Phe^ Met^ Cys^ Try,) Chromatium - A.ySg His, Arg^ ^^^u ^^P(^"2)2 '^^^'g ^^^3 ^^"^17 Glu NH^ 2 PrOa Try, Gly,^ Ala^^ Val,^ Leu^ Ileu^ Met^ Cys^ Tyr^ Phe^^) Rps. PalJistris - (Lys,^ His, Arg^ Asp,, Thr^ Ser^ GlUg Pro^^ Glyg Ala,^ Val^ Ileu^ LeUg Met^^^^^ Cys. Heme Peptide Sequence Val-Ala-Asp-Glu-Gly-Ser-Ala-Lys-Cys-His-Thr-Phe- A sp-Glu-Gly-Ser-COOH End-Group Sequences R. nibrum- H^N-Ala-AspNH^-Val-Ala-Gly Glu-COOH Chromatium - H N-Ala-Gly-Leu/ -Ser-AspNH^ — Ala-COOH lieu 68 THE BACTERIAL PHOTOCHEMICAL APPARATUS is: H2N-AIa-Gly- Leu (lieu) •Ser-AspNH2. . .; for litis: H2N- Ala- AspNHg •Val-Ala-Gly. . . Thus, both end groups are identical; there are simple cross-overs between the asparagine residues andthe glycine residues, while leucine-valine, and serine-alanine are comparable pairs. The carboxyl end group for I is glutamic acid, and for II is alanine. It is of interest that the two residues which separate the cysteinyl groups, presumably holding one of the hemes in II, are serine and glutamine, just as in chicken cytochrome c. In all other mammalian and animal c-type cytochromes, these two residues are alanine and glutamine. These are only a few of the results at hand; space prohibits pre- sentation of others, which in any case are still very preliminary, I have included the results shown only to provide examples of the sort of information coming out of our present studies. The most remarkable properties of cytochromoids are exhibited in their behavior with ligands (2,21), with which they would normally be expected to react; such reagents as azide, cyanide, hydrosulfide, methylimidazole, nitrosobenzene, etc., fail to attach to the central iron atom even at extremes of pH, The proteins in the oxidized state appear to be accessible only to protons, CO and NO, and not even to protons when in the reduced state (21,22), Reactions with NO occur with both reduced and oxidized forms, but only marginally. Affinities are many orders of magnitude less than for usual NO-heme interactions, CO reacts more strongly than NO and, of course, only with the re- duced form, with which it forms an easily photodissociable complex. The fact that cytochromoids attach NO less firmly than CO is anomal- ous, and so is another finding— that the photodissociability is pH- dependent. These ligand interactions are sufficiently weak so that taken with the absolute lack of reaction with ligands in general they weight the similarity to "c" cytochromes greater than to myoglobins or per- oxidases, as based on the spectral and magnetic susceptibility data. The sequence studies provide more firm evidence for the close rela- tion to the "c" cytochromes. Finally, the testimony of data on functional involvement of cytochro- moids as intermediates in the photoactivated electron-transport chain coupled to photophosphorylation (23,24,25), rather than as oxidases or peroxidases, completes the similarity to the "c" cytochromes. The relevance of the term "cytochromoids" is obvious. Other counts against an oxidase function for cytochromoids are: (1) One of them occurs, as in (H), in a strict anaerobe; (2) R. riibrum oxidase from dark-grown aerobic cultures contains no extractable cytochromoids and shows little (26) or no (27) detectable bound cyto- chromoids; (3) The oxidase activity of R. ruhrum is inhibited by approximately 10"5 M cyanide (2), whereas cytochromoid I does not combine with cyanide even at a cyanide concentration of 10-2 m (2,21). There have been data on action spectra for the relief of the CO inhibi- HEME PROTEINS 69 tion of oxidase activity in dark-grown aerobic R. rubrum cultures (7, 28) which show very good correlations between such spectra and the CO, reduced-minus- reduced difference spectra of cytochromoidI,butthese apparently only prove the identity of these spectra for "cytochrome o" and cytochromoid I. CONCLUDING REMARKS I have not burdened you with many data in this presentation because it is certain that neither space nor time allotted permitted a detailed discourse, and because I feared the general outlines which have emerged in this area of research might not be visible through the flood of tables and figures which would have resulted from any attempt to document my remarks. I have included references which I hope will aid those of you who wish to inquire further into the subject matter I have presented. REFERENCES 1. Elsden, S. R., Kamen, M. D., and Vernon, L. P., A new soluble cyto- chrome. J.Am. Chem. Soc, 75, 6347 (1953). 2. Vernon, L. P., and Kamen, M. D., Hematin compounds in photosynthetic bacteria. J. Biol. Chew., 211, 643 (1954). 3. For reviews of structure and function of heme proteins, the following arti- cles by the author may be consulted: (a) New problems in the biochemistry and metabolism of heme proteins, p. 245 in Proc. Inter)}. Syiiip. on Enzyme Chemistry, Toliyo and Kyoto. 1957 {K. Ichihara, ed.), Vol. 2. Maruzen, Tokyo, 1958. (b) Hematin compounds in photos joithesis. p. 323 in Comparative Bio- cliemistry of Pholoreaetive Systems (M. B. Allen, ed.). Academic Press, New York, 1960. (c) Comments on function of haem proteins as related to primary photo- chemical processes in photosynthesis, p. 483 in A Symposium on Liglit and Life (W. D. McElroy and B. Glass, eds.). Johns Hopkins Press, Baltimore, 1961. (d) Cytochrome systems in anaerobic electron transport (with J. W. Newton), p. 397 in Tlie Bacteria (I. C. Gunsalus and R. Y. Stanier, eds.). Vol. 2. Academic Press, New York, 1961. (e) The atj^Dical haemoprotein of purple photosynthetic bacteria (with R. G. Bartsch). p. 419 in Haematin Enzymes (J. E. Falk, R. Lemberg, and R. K. Morton, eds.). Part 2. Pergamon Press, New York, 1961. (f ) Haem protein content and function in relation to structure and early photochemical processes in bacterial chromatophores. p. 277 in Biologieal Structure and Function (T. W. Goodwin and O. Lindberg, eds.), Vol. 2. Academic Press, New York, 1961. 4. Lemberg, R., Clezy, P., and Barrett, J., Haem A2 and chlorin A2. p. 354 in Haematin Enzy)}ies (J. E. Falk, R. Lemberg, and R. K. Morton, eds.), Part 1. Pergamon Press, New York, 1961. 70 THE BACTERIAL PHOTOCHEMICAL APPARATUS 5. The nomenclature of tetrapyrrolic chelated proteins is discussed in Report of the Commission on Enzymesof the International Union of Biochemistry, 1961. Pergamon Press, New York, 1961. See p. 24 et seq. and 57 et seq. 6. Horio, T., Higachi, T., Yamanaka, T., Matsubara, H., and Okunukl, K., Purification and properties of cytochrome oxidase from Pseudomonas aeruginosa. J. Biol. Chem., 236, 944 (1961). 7. Chance, B., Cytochrome o. p. 433 in Hae matin Enzynies (J. E. Falk, R. Lemberg, and R. K. Morton, eds.). Part 2. Pergamon Press, New York, 1961. 8. Castor, L. N., and Chance, B., Photochemical determinations of the oxidase of bacteria. J. Biol. Chem., 234, 1587 (1959). 9. Bartsch, R. G., Coval, M. L., and Kamen, M. D., The amino acid compo- sition of the soluble chroma tium haem proteins. Biochim. Biophys. Acta, 51, 241 (1961). 10. Coval, M. L., Horio, T., and Kamen, M. D., The amino acid composition of some bacterial haem proteins, ibid., 51, 246 (1962). 11. Dus, K., Bartsch, R. G., and Kamen, M. D., The diheme peptide of Chro- matinm RHP. J. Biol. Chem., 237, 3083 (1962). 12. Smith, L., and Ramirez, J., Reactions of pigments of photosynthetic bacteria following illumination or oxygenation. Brookhaven Symp. Biol., 11. 310 (1959). 13. Olson, J. M., and Chance, B., Oxidation-reduction reactions in the photo- synthetic hacterium, Chromatinm. Arch. Biochem. Biophys., S8, 26 (1961). 14. Duysens, L. N. M., and Amesz, J., Function and identification of two photo- chemical systems in photosynthesis. Biochim. Biophys. Acta, 64, 243 (1962). 15. Rumberg, B., Muller, A., and Witt, H. T., New results about the mechanism of photosynthesis. Nature, 194, 854 (1962). 16. Orlando, J. A., and Horio, T., Observations on a 6 -type cytochrome from Rhodopseudomonas spheroides. Biochim. Biophys. Acta, 50, 367 (1960). 17. Newton, J. W., Unpublished observations. 18. Kamen, M. D., Unpublished observations. 19. Ehrenberg, A., and Kamen, M. D., Unpublished observations. 20. de Klerk, H., and Kamen, M. D., Unpublished observations. 21. Taniguchi, S., and Kamen, M. D., On the anomalous interactions of ligands with Rliodospirillum haemprotein. Biochim. Biopliys .Acta, 74, 438(1963). 22. Horio, T., and Kamen, M. D., Preparation and properties of three pure crystalline bacterial haem proteins, ibid., 48, 266 (1961). 23. Horio, T., and Kamen, M. D., Some haem protein-linked pyridine nucleotide oxidation systems in Rhodospirillum rubrum. ibid., 43, 382 (1960). 24. Horio, T., and Kamen, M. D., Optimal oxidation-reduction potentials i— J endogenous co-factors in bacterial photophosphorylation. Biochemistry, 1, 144 (1962). 25. Horio, T., and Kamen, M. D., Observations on the respiratory system of Rhodospirillu)ii. ibid., 1, 1141 (1962). 26. Geller, D. M., Oxidative phosphorylation in extracts of Rhodospirillum rubrum. J. Biol. Chem., 237, 2947 (1962). 27. Taniguchi, S., Unpublished observations. 28. Horio, T., and Taylor, C. P. S., Unpublished observations. THE STRUCTURE OF THE PHOTOSYNTHETIC APPARATUS IN THE GREEN AND PURPLE SULFUR bacteria! R. C. FULLER, S. F. CONTI and D. B. MELLIN Department of Microbiology , Dartmouth Medical School Hanover, Neiv Hampshire INTRODUCTION The concept of the chromatophore as the basic structural unit of bacterial photosynthesis has grown steadily since this structure was first described by Schachman, Pardee, and Stanier (1) in 1952 and subsequently shown to be functional in photosynthetic phosphorylation by Frenkel (2). However, obvious exceptions to this rule have been observed in recent years. Under most conditions of growth the cyto- plasm of the photosynthetic bacterium Rhodomicrobium vannielii con- tains a lamellar system which in all probability is the photosynthetic apparatus of the cell (3). It also has been reported that lamellar structures are characteristic of Rhodospirillum molischiamim (4,5) and in some cases can be observed in R. rubrum. Recently Cohen- Bazire and Kunisawa (6) have shown that in cells of R. rubrum grown at high light intensity chromatophore synthesis is suppressed and that the chromatophores, if present at all, are localized in the peripheral areas of the cell. They suggest further that the photochemical apparatus of this organism has its origin in the bacterial cytoplasmic membrane. Boatman and Douglas have indications of similar relationships in subaerobically grown cultures of R. rubrum (7). Electron microscopy of thin sections of cells of the purple bacterium Chromatium strain D reveals the presence of circular vesicular chromatophores throughout the cell. Disruption of the cells by various means yields a preparation containing stable, uniform chromatophores approximately 300 A in diameter, with a molecular weight of approxi- mately 15,000,000; these are capable of catalyzing photosynthetic phosphorylation without the addition of artificial electron transport carriers (8), It has not been possible to obtain afunctional subunit of this structure. This organism has therefore served well as a model 1 This work was supported in part by the Charles F. Kettering Foundation; Grant No. GB-76 from the National Science Foundation; and Contract No, AT (30-1) 2801 from the U. S. Atomic Energy Commission, 71 72 THE BACTERIAL PHOTOCHEMICAL APPARATUS system for investigations dealing with the nature of the bacterial chromatophore, A much simpler unit of photochemical activity has been shown to be present in the green sulfur hsLCteriumChlorobiumthiosulfatophilu})! (9), The cytoplasm appears granular and contains small (100 A) parti- cles. After cell disruption, similar particles can be isolated and are functional in carrying out photosynthetic phosphorylation. These parti- cles have amolecularweightof about 1,500,000 and contain chlorophyll, cytochromes, quinones, carotenoids, and lipids in amounts that seem to indicate that this isolated structure may indeed represent the minimal structural unit necessary for photosynthetic phosphorylation (10,11). Thus, it now seems clear that there really are significant differ- ences in the architecture of the photochemical apparatus of the various photosynthetic bacteria. THE STRUCTURE OF THE GREEN SULFUR PHOTOSYNTHETIC BACTERIUM CHLOROBIUM THIOSULFATOPHILUM STRAIN L As previously stated, Chlorobium appears to be an exception among the photosynthetic organisms studied to date in that thin sections of the cells observed by electron microscopy indicate the absence of any form of visicular or lamellar structure known to be associated with a photochemical apparatus. However, the authors have not yet made a detailed analysis of the cell with electron microscopy. 2 Following disruption of the cell, either by ultrasonic treatment or with the Hughes press (which is not subject to the criticism of breaking up internal membranes as is ultrasonic treatment), a sedimentable fraction containing most of the pigment of the cell could be obtained. Examination of the pellet by electron microscopy did not disclose any vesicular or membrane associations, in agreement with our observa- tions on cells of Chlorobium thiosiilfatophilum (strain L), These photochemical macromolecules behave differently from the chromatophores obtained from other photosynthetic bacteria in several ways. The pigments are more readily dissociated from them than from the chromatophores of the purple bacteria and, indeed, the particles appear to be perhaps the equivalent of subunits or chromatophore fragments of a more highly organized system. Therefore, for this and many other reasons, we prefer not to regard this pigmented component of the cell as a chromatophore. Fig, 2 illustrates some of the physico- chemical properties of this macromolecular system and the degree of 2 Cohen-Bazire, in this volume, shows thin sections of other strains of Chloro- hium which show ultrastructural differentiation which may or may not be as- sociated with the pigment-bearing structure. STRUCTURE OF PHOTOS YNTHE TIC SULFUR BACTERIA 73 , ti • •^-f-*''^ #; f^t.\d\^''^'''T ^^^.^"gr^Ph of Chlorobium thiosulfatophilum strain L The C^!o2! ''"" intracellular inclusions are polymetaphosphate granules ^P oo.uuu X (±iar - 1 //). Bottom - enlargement to 140,000 x. 74 THE BACTERIAL PHOTOCHEMICAL APPARATUS G) 00 r-- (X) in sn ro CO o O o o o o o o AIISNBQ nVOIldO ■^ — ( -i_> T3 i-H C S tH •2 b c 0) ■:2 rt s > 0) ^s CO 2 1^ r^ ^^ o o -^ d ^ s •t: 0) M S o .i; 0) bD-T! -tf ^ a G Sh _ C (J -!-> STRUCTURE OF PHOTOS YNTHE TIC SULFUR BACTERIA 75 homogeneity of fractions after isolation and purification. With minimal purification, the pigmented fraction shows a strong absorption maximum at 260 m/2, indicating the presence of nucleic acids, but upon repeated sedimentation and purification in buffered systems a single ultracentri- fugal peak can be obtained. It is clear that the photochemical pigment system is in an unaltered form and is free of both nucleic acids and slower sedimenting components. The estimated molecular weight of this particle from the above data is one and one-half million. This is by far the simplest defined photochemical system described to date which occurs in a natural state. Chemical analysis shows that the particulates isolated from Chlorobium contain carotenoids, chlorophyll, cytochromes, quinones, and phospholipids; the data of Hulcher and Conti (12), as shown in Table 1, give a reasonable analysis of the particles. It is interesting to note that the quinone associated with these particles is a new one and very similar to plastoquinone (13). Olson and Romano (14) have TABLE 1 Estinmtion of Cytochromes and Other Components in Chlorophyll -containing Particles of Chromatium and Chlorobium Constituent Chromatium C. thiosidfatophilum myUmoles/1.0 ml suspension Cytochromes: type 555, 552 0.91 X 10-5 0.85 X 10-5 type 612 none 1.66 X 10-5 type 630-640 0.78 X 10-5 none Total Cytochrome (ext. coef.) 1.69 X 10-5 2.51 X 10-5 Total Cytochrome (pyridine hemochromogen) 3.18 X 10-5 1.59 X 10-5 Chlorophyll 7.10 X 10-4 2.34 X 10-3 Carotenoids 4.00 X 10-4 4.48 X 10-4 Proteinl 4.56 mg/ml 24.0 mg/ml Molar ratios: Chhcarotenoid: cytochromes 40:20:1 100:20:1 Chhcarotenoid: cytochrome 20:10:1 150:30:1 1 Protein determined by the spectrophotometric method of O. Warburg and W. Chv\si\a.n,Biochem. Z., 310, 384 (1941). 2 Calculated from total cytochrome, determined from extinction coefficients. From Hulcher and Conti, Biochem. Biophys. Res. Commun., 3, 497 (1960). 76 THE BACTERIAL PHOTOCHEMICAL APPARATUS Fig. 3. The effect of serial transfer ot Chlorobium thiosidfatophilum in the absence of inorganic phosphate in the growth medium, on the content of polyphosphate in the cells. (Tq T4 = 4 transfers). Growth was relatively unaffected through three such transfers (To) but polyphosphate was essentially depleted. STRUCTURE OF PHOTOS YNTHE TIC SULFUR BACTERIA 77 PO4 RELEASE 5 10 TIME 15 20 25 IN MINUTES 35 Fig. 4. Phosphorus uptake by a particulate fraction from Chlorobium. The fraction consisting of pigmented particles was incubated in light and dark in Warburg vessels contain- ing 0.8 ml extract; 10 //moles MgCl2; lO^umoles Kh2P04; 100 yUmoles Tris pH 7.0; 7.5 yUmoles ADP; 0.5 ml Hexokinase (Sigma), and water to a total vol. of 2.5 ml. 0.5 ml of TCA (10%) was added to stop the reaction and the precipitate re- moved by centrifugation. Rates of phosphate uptake are ex- pressed as/imolesorthophosphateesterifiedAr/mg chloro- phyll. recently shown that bacteriochlorophyll is also associated with the particle from Chlorobium and have suggested that Chlorobium chloro- phyll may act as an accessory pigment to the photochemically active bacteriochlorophyll. It is only recently that we have been able to show that the small particle is indeed an active photochemical system. During experi- ments designed to measure light-dependent phosphate uptake, inorganic phosphate was always released into the medium, making measurements of photophosphorylation difficult. It has recently been shown by Hughes et al. (11) that Chlorobium thiosulfatophilum accumulates large amounts 78 THE BACTERIAL PHOTOCHEMICAL APPARATUS of polymetaphosphate in discrete intracellular granules, from which inorganic phosphate is released by an ADP- dependent and light- independent reaction. The polymetaphosphate granules represent a large proportion of the intracellular material. Fig. 3 illustrates the effect of continued serial transfers on growing cultures of Chlorobium in a phosphate deficient medium. It can be readily seen that the size of the polymetaphosphate granules decreases with each transfer. Cells essentially free of these polymetaphosphate granules were used to study photosynthetic phosphorylation. The results shown in Fig. 4 indicate a rapid (and reproducible) rate of photosynthetic phosphorylation over short periods of time. There is still an indication of phosphate release which can mask the photo- phosphorylation during longer periods of incubation. However, it now seems clear that these functional macromolecules can catalyze the light- dependent esterification of inorganic phosphate into ATP in the absence of any artificial electron transport carriers. It is interesting to note that Levine (15) recently has been able to demonstrate photo- synthetic phosphorylation by Chlamydomonas chloroplasts for the first time by growing cells deficient in polymetaphosphate. THE STRUCTURE OF THE PURPLE SULFUR PHOTOSYNTHETIC BACTERLA. As stated previously, electron microscopy of thin sections of cells of the purple bacterium Chromatiiim strain D has always revealed the presence of circular vesicular chromatophores throughout the cell. Fig. 5 is a thin section of the bacterium indicating the classical picture of the internal contents of the cell packed with spherical chromatophores. The relationship of these chromatophores to the cytoplasmic mem- brane in the photosynthetic bacteria has been a subject of speculation for some time. In the succeeding article in this volume by Cohen- Bazire, this discussion is extended in great detail in regard to the nonsulfur photosynthetic bacterium R. ndrnim. It has been possible to obtain a separation of the photosynthetic pigments and enzymatic properties ordinarily associated with the bacterial cell membrane, such as succinic dehydrogenase, in Chlorobium thiosuIfatopJiihn)! (IG). However, such a separation has not been successfully achieved in the purple bacteria. Recent experiments in our own laboratory in associa- tion with Dr. R. Bennett have indicated that succinic dehydrogenase activity is always associated with purified chromatophores, but can be separated from the major fraction containing hexosamine which is a constituent of the cell wall. Therefore, chromatophores appear bio- chemically related to the cell membrane and not to the cell wall. However, the difficulties involved in electron microscopy of such STRUCTURE OF PHOTOSYNTHETIC SULFUR BACTERIA 79 Fig. 5. Electron micrograph of thin section of Chromntiiou strain D grown at low light intensity (see text). small chromatophores as appear in Chromatium have limited us in further speculation concerning a structural association. Variations in the far- red region of the absorption spectrum of bacteriochlorophyll have also been a subject of considerable specula- tion. The wide diversity of the in vivo spectrum of this pigment in a variety of photosynthetic bacteria is shown in Fig, 6. The divergence in spectra of Chromatium is caused by different nutritional conditions of growth (see below). Wassink et al . (17) have reported variations in the far-red chlorophyll spectrum of Chromatium, anditwas suggested that the multiplicity of absorption maxima represented varieties of bacteriochlorophyll-protein complexes produced under different condi- tions of growth. Cohen-Bazire, et al. (18) and Bergeron and Fuller (19) demonstrated changes in the in vivo spectrum of bacteriochloro- phyll that were associated with carotenoid deficiency. Wassink and Kronenberg (20), however, were able to grow carotenoid deficient cells of Chromatium with a relatively normal spectrum. Bril (21) has re- cently confirmed Wassink' s observations and points out that these divergent experimental results are not readily explained. One point is clear, viz., that alteration of the fine structure in the far-red spec- trum is not due directly to carotenoid deficiency. Frenkel and Hick- 80 THE BACTERIAL PHOTOCHEMICAL APPARATUS 700 800 900 1,000 mjj 0.6 700 800 900 1,000 mjj 0.6 700 800 900 1,000 mjj 700 800 900 1,000 m/j Fig. 6. Diagramatic representation of various far-red presentations of the spectrum of bacteriochlorophyll associated with the photochemical apparatus in a variety of bacteria. The 800 m^ peak is always present and the variation occurs in the region from 800-900 m/i. The simplest spectrum, with a single peak at 800 m/z, occurs in a pigmented lipoprotein from CJilorobiiiDi (14). The divergent spectra in Chro»ialin»i D arc dependent on nutritional conditions and light intensity during growth, which may well affect the photochemical struc- tures (see text). STRUCTURE OF PHOTOS YNTHE TIC SULFUR BACTERIA 81 OPTICAL DENSIT\ 82 THE BACTERIAL PHOTOCHEMICAL APPARATUS man (22), Bril (21), and Fuller (10) have noted that chromatophore structure in a single species might be dependent on such factors as groAvth conditions and age of cells and that the far- red fine structure of the spectrum of bacteriochlorophyllmay be related to altered struc- tures. These conflicting observations can now be reconciled. Fig. 7 illustrates the dramatic changes that occur in the far- red region of the in vivo absorption spectra; these appear to be dependent upon the light intensity to which the cells are exposed during growth. High light intensity seems to enhance the 850 m/i and especially the 890 m// absorption maxima. Heterotrophic conditions of growth seem to depress this effect to some extent. Other carbon sources such as acetate or succinate yield the same spectrum as malate. In previous experiments, where diphenylamine was used to suppress carotenoid synthesis, red cellophane was placed between the light source and the culture bottles to depress the light- catalyzed breakdown of diphenyl- amine. Inadvertently this reduced the effective far- red incident light. Although growth did not appear to be affected, the spectrum was altered in a manner similar to that shown for cells grown at low light in- tensities as seen in Fig. 7. STRUCTURAL ALTERATIONS Chemical analysis of the cells grown at 7000 and 100 foot-candles of incident light was undertaken, and the results are shown in Table 2. TABLE 2 Cheuiical Analysis o/Chromatium Cells Groirii at High and Low Light Intensities 7000 Ft.-Candles 100 Ft.-Candles CONSTITUENTS % of Dry Wt. % of Dry Wt. CHLOROPHYLL 0.3 1,0 PROTEIN 38.0 35.0 LIPIDS (Including Carotenoid Pigments) 19.0 29.6 CARBOHYDRATES 26.2 20.0 NUCLEIC ACIDS 6.0 6,5 INORGANIC ASH 11.7 9.0 TOTALS 101.2 101.3 CHL/PROTEIN RATIO .008 .02 STRUCTURE OF PHOTOSYNTHETIC SULFUR BACTERIA 83 Fig. 8. Electron micrograph of a thin section of Chromatiiini strain D grown at 7000 foot candles of incident light. Bar equals l/u. (See text for details). Fig. 9. Electron micrograph of a thin section of Chromatiiim strain D grown at 7000 foot candles of incident light. Bar equals 1^. (See text for details). 84 THE BACTERIAL PHOTOCHEMICAL APPARATUS Fig. 10. Electron micrograph of a thin section of Chromatiuni strain D grown at 7000 foot candles of incident light. Bar equals Ifl. (See text for details). A Striking change in the chlorophyll to protein ratio of the cells can be seen. At low light intensities the amount of chlorophyll per mg pro- tein doubles. These data are similar to those obtained by Cohen- Bazire et al. (18), Sistrom (23), and Lascelles (24), who noted an in- crease in the chlorophyll content of cells of Rhodopseudomonas spheroides when grown at low light intensities. Of equal interest is the rather striking increase in total lipids in cells grown at low light intensity. This increase may represent structural lipids associated with the chlorophyll-bearing structure. Although Chromatium strain D has always appeared to contain chro- matophores at all stages of growth, electron microscopy of cells exhibiting spectral and chemical differences was undertaken to ascer- tain if the changes in the spectrum and chlorophyll-protein ratios might be reflected in the structure of the photochemical apparatus. Fig, 5 illustrates the appearance in electron micrographs of Chro- matium cells grown under conditions of low light intensity. The pres- ence of chromatophores (-300 A in diameter) throughout the cell is in conformity with the observations of other investigators. In contrast to this, cells grown under conditions of high light intensity (7000 foot- candles incident light) show marked structural alterations (Figs, 8 and STRUCTURE OF PHOTOSYNTHETIC SULFUR BACTERIA 85 Fig. 11. Electron micrograph of a thin section of Chromatiutn strain D grown at 7000 foot candles of incident light. Bar equals Ifl. (See text for details). 9). Although chromatophores are still present, an intracytoplasmic membrane system can also be observed. These intracytoplasmic mem- branes are observed most frequently in the peripheral areas of the cells. Close inspection of the plates reveal that these membranes are paired. Each membrane (~ 100 A thick) corresponds to a unit membrane, but this has not yet been clearly demonstrated by high resolution studies. Fig. 10 is another micrograph of a cell grown under conditions of high light intensity. It appears that chromatophore-like structures are formed at the terminal ends of the invaginated paired membranes (see arrow). Occasionally, cells groAvn under conditions of high light intensity appear to have the intracellular membranes organized in a lamellar system (Fig. 11). It was observed that when cells are grown under conditions which induce the formation of intracellular membranes and cell-free ex-- tracts prepared, the majority of the pigment-bearing particulate material sediments at relatively low centrifugal forces as compared to the situation encountered with "normal" cells. 86 THE BACTERIAL PHOTOCHEMICAL APPARATUS SUMMARY Thin sections of the green photosynthetic bacterium Chlorobium thiosulfatophilum strain L are unlike other photosynthetic organisms in that no lamellar or chromatophore-like structures appear in the cell. A homogeneous fraction containing the photosynthetic pigments can be isolated and shows the capacity to catalyze photosynthetic phosphorylation. This particulate fraction is by far the simplest photo- chemical system thus far described. The intracellular origin of this fraction, particularly in relation to the bacterial cytoplasmic mem- brane, is not clearly understood at this time. In the purple sulfur bacterium Chromatium strain D, where the classical chromatophore structure has always been observed in sec- tions of cells, a complex intracytoplasmic membrane system is pro- duced under certain environmental conditions. The pigment and lipid contents of the cell increase sharply when low light intensities are used for growth. The structure of the photochemical apparatus and the physical and chemical environmentof the bacteriochlorophy 11— reflect- ed by the fine structure of its far red spectrum— are under control of both incident light intensity and metabolic conditions of growth, REFERENCES 1. Schachman, H. K., Pardee, A. B., and Stanier, R. Y., Studies on macro- molecular organization in microbial cells. Arch. Biochem. Biophys., 38, 245 (1952). 2. Frenkel, A. W., Light induced phosphorylation by cell-free preparations of photosynthetic bacteria. J. A;;z. Chem. Soc, 76, 5568 (1954). 3. Vatter.A. E., Douglas, H. C, and Wolfe, R.S., Structure of Rhodomicrobium uaniiielii. J. Bacteriol., 77, 812 (1959). 4. Drews, G., Untersuchungen zur Substruktur der "Chromatophoren" von Rhodospirillum rubrum und Rhodospirilluni molischianum. Arch. Micro- biol., 36, 99 (1960). 5. FrenkeL A. W., This volume, p. 89. 6. Cohen-Bazire, G., and Kunisawa, R., The fine structure of Rhodospirillum mbrum. J. Cell. Biol., i6,401 (1963). 7. Boatman, E. S., and Douglas, H. C., Protoplast morphology and chromato- phore formation in Rhodospirillum rubrum. In Electron Microscopy, vol.2. Academic Press, New York and London, 1962. 8. Anderson, I. C., and Fuller, R. C., Photophosphorylation by isolated chro- matophores of the purple sulfur bacteria. Arch. Biochem. Biophys., 76, 168 (1958). 9. Bergeron, J. A., and Fuller, R. C, The photosynthetic macromolecules of Chlorobium thiosulfatophilum . p. 307 in Biological Stnicture and Function (T. W. Goodwin and O. Lindberg, eds.), vol. II. Academic Press, New York and London, 1961. 10. Fuller, R. C., The comparative structure and activities of the microbial photosynthetic apparatus, In General Physiology of Cell Specialization (D. Mazia and A. Tyler, eds.). McGraw-Hill, New York, 1963. STRUCTURE OF PHOTOSYNTHETIC SULFUR BACTERIA 87 11. Hughes, D. E., Conti, S. F., and Fuller, R. C, Inorganic polyphosphate metabolism in Chlorobiuni thiosidfatophiliou . J. Bacteriol., S5, 577 (1963). 12. Hulcher, F. H., and Conti, S. F., Cytochromes in chlorophyll-containing particles of C//ro;;z(7^/?i;>z and Chlorobium thiosulfatophilum. Biochem. Bio- phys. Res. Conimim., 3, 497 (1960). 13. Fuller, R. C, Smillie, R. M., Rigopolous, N., and Yount, V., Comparative studies of some quinones in photo synthetic systems. Arch. Biochem. Bio- phys., 95, 197 (1961). 14. Olson, J., and Romano, C. A., A new chlorophyll from green bacteria. Bio- chim. Biophys.Acta, 59, 726 (1962). 15. Levine, R. P., Personal communication. 16. Smillie, R. M., and Bergeron, J. A., Succinic dehydrogenase of the photo- synthetic sulfur bacteria. Plant Physiol., 36, xlix, (1961). 17. Wassink, E. C, Katz, R., and Dorrestein, R., Infrared absorption spectra of various strains of purple bacteria. Enzymologia, 7 , 113 (1939). 18. Cohen-Bazire, G., Sistrom, W. R., and Stanier, R. Y., Kinetic studies of pigment synthesis by non-sulfur purple bacteria. J. Cell. Comp. Physiol., 49, 25 (1957). 19. Bergeron, J. A., and Fuller, R, C, Influence of carotenoids on the infra- red spectrum of bacteriochlorophyll in Chromatiurn . Nature, 184, 1340 (1959). 20. Wassink, E. C, and Kronenberg, H., Strongly carotenoid-deficient Chro- matiuDi strain D cells with 'normal' bacteriochlorophyll absorption peaks in the 800-850 mju region. Nature, 194, 553 (1962). 21. Bril, C, Studies on bacterial chroraatophores. II. Energy transfer and photooxidative bleaching of bacteriochlorophyll in relation to structure in normal and carotenoid-depleted Chromatiurn. Biochim. Biophys. Acta, 66, 50 (1963). 22. Frenkel, A. W., and Hickman, D. D., Structure and photochemical activity of chlorophyll-containing particles from Rho do spirillum rubrum. J. Bio- chem. Biophys. Cytol., 6, 284 (1959). 23. Sistrom, W. R., Observations on the relationship between the formation of photopigment and the synthesis of protein in Rhodopseudomonas spheroides. J. Gen. Microbiol., 28, 599 (1962). 24. Lascelles, J., Adaptation to form bacteriochlorophyll in Rhodopseudomonas spheroides: Changes in activity in enzymes concerned in pyrrole synthesis. Biochem. J., 72, 508 (1959). SOME OBSERVATIONS ON THE ORGANIZATION OF THE PHOTOS YN THE TIC APPARATUS IN PURPLE AND GREEN BACTERIA^ GERMAINE COHEN- BAZIRE Department of Bacteriology and Electron Microscope Laboratory , University of California, Berkeley, California The light microscope does not have the resolving power necessary to reveal the structure and organization of the photosynthetic apparatus in cells of procaryotic organisms. Only one conclusion could be drawn from the observations on photosynthetic bacteria and blue-green algae by light microscopy: namely, that even the largest members of these groups were devoid of organelles having the typical structure of chloro- plasts. Our first positive information about the structure of the photo- synthetic apparatus in these groups was accordingly derived from physico-chemical studies on the pigment- bearing elements that could be isolated from extracts of broken cells. As early as 1938, French (1) showed that the native pigment complex can be extracted from purple bacteria in water-soluble form and is bound to protein, as shown by its precipitability with ammonium sulfate. In 1952, Schachman, Pardee, and Stanier (2) demonstrated that the pigment complex in extracts of R. rubrum prepared by mechanical abrasion or sonic oscillation can be readily sedimented by high-speed centrifugation. By applying a series of differential centrifugations, they succeeded in isolating a physically homogeneous particulate fraction which had the typical in vivo absorption spectrum of this bacterium. The particles were isodiametric, and about 600 A in diameter. They were designated as chromatophores. No fraction of corresponding size and uniformity could be isolated from aerobically grown cells of /?. rubrum, which are essentially devoid of photosynthetic pigments. Soon after, Frenkel (3) made the important discovery that chromato- phores could perform photophosphorylation, and the investigation of their biochemical properties was undertaken in several laboratories. The finding that chromatophores possess significant photochemical functions lent support to the belief that they are the structures re- sponsible for the performance of photosynthesis in the bacterial cell. 1 This work was supported by a grant from the National Science Foundation to Professor Michael Doudoroff. 89 90 THE BACTERIAL PHOTOCHEMICAL APPARATUS It is, therefore, not surprising that when ultra-thin sections of photosynthetic bacteria were eventually examined in the electron micro- scope the interpretation of the observed intracellular structures was strongly influenced by earlier work on isolated chromatophores. For example, Vatter and Wolfe (4), who performed the first extensive study of fine structure in purple bacteria, stated that: "... the cyto- plasm [of Rho do spirillum ruhrum ] is organized into discrete units which appear less dense than the cytoplasm and are surrounded by a membrane. Since these structures are of the size of the isolated chro- matophores of Schachman et al., they are believed to be identical with them and to represent the chromatophore arrangement which exists within the cell," The studies of Vatter and Wolfe (4), Bergeron (5), and Drews (6) established the presence of abundant membrane-bounded vesicles of low electron opacity in the cytoplasmic region of photosynthetically grown cells of several purple bacteria: i?.« rubrum,Rhodopseudomonas spheroides , and Chromatium strain D. However, this is not a universal feature of internal structure in purple bacteria. As shown by Drews (6), Vatter, Douglas and Wolfe (7), and Boatman and Douglas (8), Rhodo- spirillum molis chianum and Rhodomicrobium vannielii do not contain such elements; instead, thin sections reveal the presence of regularly disposed lamellae, the intracellular arrangement of which is character- istic for each species. These structures resemble the lamellar sys- tems found in the cytoplasm of blue-green algae, which likewise assume special patterns characteristic of the different species, as shown by the work of Ris and Singh (9), When Schachman et al. (2) first isolated and described bacterial chromatophores, it was not recognized that ultrasound and mechanical abrasion do not simply break bacterial cells open; these treatments also comminute, to a greater or lesser extent, the internal structures. The comminutive effects of these methods of cellular breakage could be clearly appreciated only after the effects of the much gentler pro- cedure of osmotic lysis on cell structure had been explored, a field that was opened by the work of Weibull in 1954 (10), In the light of subsequent knowledge about bacterial cell structure, it is by no means evident that the chromatophores as defined by Schachman et al. do in fact exist as discrete structural entities in the intact cell; they could be physical artifacts, produced by the fragmentation of larger struc- tural elements during the treatment employed for breakage of the cells. This possibility was first recognized by Karunairatnam et al. (11), as a result of their observation that lysed spheroplasts of /?. ruhrum retain essentially all of the photosynthetic pigment system. At about the same time, we independently performed similar experi- ments with R. ruhrum and Rhodopseudo))2onas splieroides ,'mthsimila.r findings. In the case of Rhodopseudomonas spheroides, we observed that the cellular inclusions of poly-/3-hydroxybutyric acid, which are STRUCTURE OF PHOTOSYNTHETIC BACTERIA 91 easily visible by phase contrast microscopy, are liberated from lysed spheroplasts under conditions where the release of pigment is neglig- ible. Clearly, therefore, the lysed spheroplasts could not physically retain individual structural elements of the dimensions of chromato- phores. The gaps and apparent contradictions in our present knowledge about the internal organization of the cell in photo synthetic bacteria have led us to undertake an extensive study of fine structure in this group. The present communication constitutes a progress report. EXPERIMENTAL PROCEDURES These have been described in a recent paper (12). Here, it is suf- ficient to say that we have used, throughout, the preparative technique for electron microscopy of bacteria developed by Ryter and Kellen- berger (13), with two minor modifications. The period of main fixation has been reduced in recent work from 16 to 2 hours, since this appears to give equally good, if not superior, preparations. Secondly, most of the sections shown have been post-stained with lead hydroxide (14), a treatment which greatly sharpens the definition of wall and membrane structure. Negative staining with phosphotungstic acid was performed as described by Huxley and Zubay (15). All electron micrographs were taken with a Siemens Elmiskop I, operating at 80 KV. THE FINE STRUCTURE OF RHODOSPIRILLUM RUBRUM AND RHODOPSEUDOMONAS SPHEROIDES Physiological studies in our laboratory have shown that environ- mental conditions profoundly affect the chemical composition of the cell in nonsulfur purple bacteria (16,17). In particular, the cellular content of photosynthetic pigments can vary over a wide range, even under conditions of strictly photosynthetic growth (anaerobiosis), in response to such factors as light intensity and temperature. In faculta- tively aerobic strains, pigment synthesis can be completely suppressed by growth in the presence of air; if oxygen access becomes limited, photosynthetic pigment synthesis resumes, even in the absence of light (18). The magnitude of these environmental effects on the pigment content of R. nibnan and Rhodopseudomonas spheroides is shown by the representative data assembled in Table 1, Accordingly, our first goal in the cytological study of these bacteria was to compare the fine structure of cells with different pigment con- tents. Typical results are shown in Figs, 1 to 7, It can be seen that the 92 THE BACTERIAL PHOTOCHEMICAL APPARATUS TABLE 1. Bacteria chlorophyll content of no nsul fur purple bacteria growing under different environmental conditions Rhodospirillmn rubrum Photosynthetic growth at Chlorophyll content of cells, /Ug/mg light intensity of: cellular protein. 50 foot-candles 25.0 2000 foot-candles 10.2 6000 foot-candles 5.6 Respiratory growth conditions: Full aeration 0.2 Full aeration, then limiting oxygen for 3 hr. 3.3 Rhodopseudomonas spheroides Photosynthetic growth at light intensity of: 50 foot-candles 66 9500 foot-candles 5.9 Aerobic growth 0.14 classical profile of a photosynthetically grown cell with a vesicle- filled cytoplasmic region, as first described for these two species by Vatter and Wolfe (4), is characteristic only for cells which have been grown at very low light intensities, and therefore have a high specific chlorophyll content (Figs. 1 and 5), In sections of cells grown at pro- gressively higher light intensities, the abundance of vesicles declines systematically, and their location in the cell becomes increasingly peripheral. In cells grown anaerobically at light intensities of 5000 foot-candles or higher, most of the cytoplasmic region in sections is filled with a dense array of ribosomes, and the relatively rare membrane-bounded vesicles occur exclusively in the neighborhood of the cytoplasmic membrane (Figs, 2 and 6), Two other important struc- tural features become evident in such cells. Firstly, the typical unit membrane which bounds each vesicle is identical in thickness and fine structure to the cytoplasmic membrane. Secondly, the membrane that encloses a vesicle is sometimes sectioned in a plane that reveals its continuity with adjacent regions of the cytoplasmic membrane. The central, transparent area of such vesicles opens through a narrow aperture into the space that lies between wall and membrane, and is consequently external to the cytoplasm proper. Contrary to the report Fig. 1. Section of R. ruhrum growing exponentially at a light intens- ity of 50 foot-candles. The membrane bounded vesicles (v) extend deep into the cytoplasm, but are most abundent at the periphery of the cell. Main fixation 18 hours, no post staining. X 84,000. Fig. 2. Section of a dividing cell of R. rubriim growing exponentially at a light intensity of 6000 foot-candles. Membrane-bounded vesicles are sparsely and irregularly distributed around the periphery of the cell. At the point indicated by an arrow, a portion of the vesicular membrane is continuous with the cytoplasmic membrane and the light central area of the vesicle opens into the space that lies between the cell wall (w) and the cell membrane. Nucleoplasm (n) and ribosomes (r) are particularly well preserved in this section. Main fixation 2 hours, post staining with lead hydroxide. X 120,000. .w. 'a :..-, . .. iS.#,' Fig. 3. Section of R. rubrion growing exponentially under strictly aerobic conditions in the dark and containing only traces of photosyn- thetic pigments. The membrane-bounded vesicles are extremely rare in this type of cell. Arrows indicate two such vesicles. Main fixation 2 hours, followed by a short treatment with ribonucleasc; post stain- ing with lead hydroxide. X 120,000. Fig. 4. Section of R. rubruniirom a culture grown aerobically in the dark, and then allowed to synthesize photosynthetic pigments for 3 hours under semiaerobic conditions in the dark. Peripheral membrane- bounded vesicles are more numerous than in cells grown strictly aerobically. Compare with Fig. 3. Treatment as described for Fig. 3. X 120,000. ^^^N^^r Fig. 5. Section of Rhodopseudomouas spheroides growing photosyntheti- cally at a light intensity of 50 foot-candles. The cytoplasm is packed '^ith membrane-bounded vesicles (v) of relatively uniform dimensions, 500 A in diametet-. Main fixation 2 hours, post staining with lead hydroxide. X 120,000. Fig. 6. Section of Rhodopseudomouas spheroides growing exponentially under photosynthetic conditions at a light intensity of 9500 foot-candles. Internal membranes (arrows) can be distinguished at the periphery of the cell; in this section, only one irregularly shaped, membrane-bounded vesicle (v) is visible. The very electron-dense structure in the middle of the cell is polymetaphosphate. indicates areas originally occupied by poly-y8-hydroxbutyric acid. Several deposits of glycogen (g) are also vis- ible in this section. Main fixation 2 hours, post staining with lead hydrox- ide. X 120,000. 96 THE BACTERIAL PHOTOCHEMICAL APPARATUS i. ^ ^1^ < % •rj (%.. ^^BSS^SSSBBSB^^^- STRUCTURE OF PHOTOSYNTHETIC BACTERIA 97 of Vatter and Wolfe (4), vesicles are not completely absent from aero- bically grown depigmented cells (Figs. 3 and?), although they are rare, particularly in R. rubrum. These cytological studies lead us to the following interpretation of the organization of the photosynthetic apparatus in the two species so far discussed. The photosynthetic pigment system is incorporated into a continuous unit membrane, which arises from the cytoplasmic mem- brane and can intrude into the cytoplasm to a greater or lesser extent, depending on the specific pigment content of the cell. Since structurally indistinguishable membranous intrusions occur in aerobically grown cells with a negligible pigment content, these intrusions are not nec- essarily always associated with the presence of a functional photo- synthetic apparatus. However, recent fine structure studies have shown that a variety of membranous intrusions derived from the cytoplasmic membrane may occur in nonphotosynthetic, aerobic bacteria (19). The "simple intrusives" discovered by Murray (19) in the nonphotosynthet- ic organism Spirillum serpens are quite similar in appearance and intracellular position to the sparse, peripheral vesicles characteristic of R. rubrum and Rkodopseudomonas spheroides grown aerobically in the dark or anaerobically at high light intensities. A general functional interpretation of these intrusions in aerobic and photosynthetic bac- teria has been offered by Stanier (20). On the basisof a recent study of the fine structure of Rho do spirillum molischianum, Giesbrecht and Drews (21) have proposed a very similar interpretation of the organization of the photosynthetic apparatus in this species. In cells sectioned after osmotic lysis, the physical con- nection between the characteristic lamellar bundles and the cytoplasmic membrane was clearly evident. THE FINE STRUCTURE OF SOME NEWLY ISOLATED PURPLE SULFUR BACTERIA Thus far, fine structure studies on the sulfur purple bacteria have been confined to a single strain, Chromatium strain D, the only representative of this group which has been generally available in pure culture. Recently, Dr. Norbert Pfennig has succeeded for the first time in isolating and growing in pure culture a number of other types, including Chromatium okenii and Thio spirillum jenense (22). Through his kindness, we have been able to examine the fine structure of some of the strains in his collection. C. okenii and T. jenense were of particular interest, since their cells are much larger than those of any other photosynthetic bacteria, with a volume about a thousand times that of the cell of Rkodopseudomonas spheroides or R. rubrum. Figs. 8 and 9 show typical thin sections of these two large purple sul- NC ">.«. .'5-^;.;, :s: -■.-«£. :-fl^./ «^-- -'■ Fig. 8. Part of a section of a cell of Chromatium okeiiii grown at low light intensity (less than 40 foot-candles) and depleted of sulfur inclusions. The cyto- plasm is filled with membrane-bounded vesicles. The cytoplasm has retracted from the cell wall at the pole of the cell. Several slime layers (SI) are present around the cell wall (w). The figure includes also a longitudinal section of the flagellar tuft (F) of this bacterium. Main fixation, 2 hours, post stammg with lead hydroxide. X 100,000. STRUCTURE OF PHOTOSYNTHETIC BACTERIA 99 t: £: ?^l^» . ■*'■*- ■=> . si T-;%^ ^^ /^> Fig 9 Part of a longitudinal section of Thiospirillum jenense grown at low light intensity. As in Chromatiuni okenii (Fig. 8), the cytoplasm is almost filled with membrane-bounded vesicles. The cell wall is surrounded by a thm slime layer. Main fixation 2 hours, post stained with lead hydroxide. X 120,000. 10 Fig. 10. Relatively thick section of a small sulfur purple bacterium, Thiocapsa, grown at low light intensity. The cytoplasm contains typical membrane-bounded vesicles (v), as well as a large inclusion of paired lamellae (L). The very electron-dense areas represent polymetaphosphate deposits. The clear areas are probably deposits of glycogen. Main fixation, 18 hours, post staining with lead hydroxide. X 120,000. STRUCTURE OF PHOTOSYNTHETIC BACTERIA 101 fur bacteria; the sections were prepared from cells grown in very dim light (<40 foot- candles) and depleted of internal sulfur deposits prior to fixation. The cytoplasmic region is in each case densely filled with vesicles bounded by unit membranes. The individual vesicles are approximately 500A in diameter and are thus structurally indistinguish- able from the vesicles of R.rubrum andRhodopseudomonas spheroides. A markedly different internal membrane structure occurs in the small, spherical cells of Thiocapsa (Fig, 10). These cells were also taken from a culture grown in very dim light. Just as in the case of the larger purple sulfur bacteria, much of the cytoplasm is filled with typical 500A vesicles surroundedbyunit membranes; however, in many sections, a large area is occupied by an extensive system of parallel paired lamellae. THE STRUCTURE OF ISOLATED MEMBRANE FRACTIONS The interpretation which has been offered for the structure of the photosynthetic apparatus in purple bacteria implies that the chromato- phores which can be isolated from broken cells constitute fragments derived from an initially continuous membrane system, no doubt fre- quently contaminated with associated wall material. In the case of R. ruhrum, the presence of numerous wall fragments in crude chromato- phore preparations can be readily established by the examination of material negatively stained with phosphotungstic acid, since the wall of this species has the distinctive fine structure characteristic of the wall of spirilla. This probably explains the observation of Newton (23) that antisera prepared against chromatophore material are capable of agglutinating intact cells, K one avoids methods of cell breakage that cause considerable com- minution, and fractionates the extract by sucrose gradient centrifuga- tion, it is possible to obtain from R. rubrum, though in small yields, membrane fractions that are essentially devoid of wall material. Sec- tions show that the unit membrane structure is well preserved in such material (Fig. 12), The geometric form of these membrane fragments is revealed by negative staining with phosphotungstate. Fig. 11 shows our best preparation of this kind to date, made with membrane material prepared from a large Chromatium species provided by Dr. Pfennig. Similar preparations from R. ruhrum give a virtually indistinguishable picture. To judge from such preparations, most of the membrane frag- ments have the form of cups or hemispheres, flattened to a greater or lesser extent by drying. We have thus far been unable to detect any fine structure in them. 102 THE BACTERIAL PHOTOCHEMICAL APPARATUS Fig. 11. Preparation of pigmented membrane fraction isolated from a large Chromatium (strain Tassajara) negatively stained with phos- photungstate. X 180,000. Fig. 12. Thin section of pigmented membrane fraction isolated from R. ruhrum grown photosynthetically at 50 foot-candles. Main fixation 2 hours, post staining with lead hydroxide. X 180,000. STRUCTURE OF PHOTOSYNTHETIC BACTERIA 103 THE INTERNAL ORGANIZATION OF THE CHLOROBIUM CELL Fine structure studies on green bacteria have so far been very lim- ited, Vatter and Wolfe (4) published one micrograph of a thin section of Chlorobium limicola. The cytoplasm contained numerous very electron- dense bodies, which the authors equated with chromatophores, despite the fact that these bodies showed no resemblances to the vesicular elements of purple bacteria. They could perhaps be more reasonably interpreted as deposits of polyphosphate. Bergeron and Fuller (24) have published a micrograph of a thin section of Chlorobium thiosul- fatophilum, which shows, according to the authors, that the fine struc- ture of this bacterium is essentially indistinguishable from that of an ordinary nonphotosynthetic true bacterium such as £s c/zencMa coZ/. Bergeron and Fuller (24) also made the first detailed study of the lo- cation of the pigment system in cell-free extracts of a green bacterium. They broke the cells of C. thiosulfatophilum by a highly comminutive treatment: sonic oscillation of cells suspended with a fine synthetic sapphire abrasive. The bulk of the pigment system in such extracts was associated with particles about 150 A in diameter, difficultly separable from ribosomes, and having a molecular weight of approxi- mately 1,5 million. Since the most conspicuous cytoplasmic elements observable in their thin sections were also particles with an approxi- mate diameter of 150 A, Bergeron and Fuller assumed the identity of their isolated "holochrome" with these cellular elements. We have recently collaborated with Dr. Norbert Pfennig on a study of the fine structure of five strains of Chlorobium. The strains in- cluded representatives of the limicola 3nA thiosulfatophilum physiologi- cal types, and also of the two sub-groups which can be distinguished on the basis of the chlorobium chlorophyll that they contain. They are thus representative of the group as a whole. Although our work on these organisms is still in progress, the find- ings to date, with respect both to the structure of the intact cell and to the properties of the pigment system in extracts, differ in major re- spects from the findings of Bergeron and Fuller (24). The green bac- teria that we have examined all share a highly distinctive and complex fine structure, quite unlike that found in purple bacteria— or, indeed, in any other type of bacterium so far studied by modern techniques of electron microscopy. Thin sectionsof three different strains are shown in Figs, 13-15. The cell wall has the double- layered structure characteristic of Gram-negative bacteria, with a thin, sharply defined inner layer, and a thicker, less dense outer layer. In some strains, it is ornamented by rod-shaped extensions about 300 A wide, with a helically patterned surface structure (Figs. 13 and 14); the presence of these elements on the wall seems to be characteristic of strains that give a slimy type of growth in liquid media. Underlying the wall is a complex membrane Fig. 13. Sections of Chlorobium limicola strain R grown at low light intensity. The median section at the lower right shows clearly the com- plex wall (w), with its rod-shaped extensions and the cell membrane (m) of this organism. The large oblong chlorobium vesicles (Cv) sur- rounded by an electron-dense membrane are clearly seen in the me- dian section. Their general distribution through the cortex of the cell is well shown by the tangential section at the upper right. At upper left, a tangential section has passed transversely through some of the rod- shaped extensions from the cell wall. Main fixation 2 hours, post staining with lead hydroxide. X 135,000. . /*^M^ '.^r / 14 r^ to thatof C. i»,„co/a, strain R Stp ' *"'■"';•"'■<= is very similar Main «.atio„ 3 „„„rs, Poirs^r^w^ reTa^n'^.T-fo-otSf ' 106 THE BACTERIAL PHOTOCHEMICAL APPARATUS system, the detailed structure of which is not yet fully interpretable; we have the impression that it may not be organized in the same fashion in every strain. Immediately within the inner layer of the cell wall, there are either one or two unit membranes. Parallel and adjacent to this unit membrane system is a thinner, electron-dense membrane, some 50 A thick. This thin membrane appears to surround and enclose the series of large, clear oblong areas which line the cortex of the cytoplasm between the surface membrane system and the ribosomal region. Every strain that we have examined contains these character- istic oblong structures, the shape and disposition of which are particu- larly well shown in tangential sections (Figs. 13 and 14). We shall term them chlorobium vesicles. They are relatively large, and of somewhat variable dimensions, 1000-1500 A long and 300-400 A wide. In vestopal- embedded cells (Figs. 13-15), the chlorobium vesicles are transparent to the electron beam, and hence might be interpreted as empty. However, in epon-embedded cells, they appear much denser than the adjacent cytoplasm. All the strains of green bacteria that we have examined also contain large, conspicuous, and complex membranous intrusions similar in structure and derivation to the so-called "mesosomes" found in a variety of nonphotosynthetic true bacteria. There are relatively few (1 to 3) in the cell; they commonly lie deep in the cytoplasm, some- times intruding into the nuclear region. In favorable sections, the mesosomal membranes can be seen to connect with the inner layer of the cytoplasmic membrane (Fig. 14), In this anatomical labyrinth, where is the photosynthetic apparatus located? We are not yet prepared to give a categorical answer to this question, and will simply describe a few observations which bear on it. The first point to be emphasized is that all the strains of green bac- teria we have examined have a very high specific chlorophyll content- far higher than that of purple bacteria growing at the same light in- tensity. This is shown by the data in Table 2, In view of our cytological experience with purple bacteria, it seems reasonable to assume that the photosynthetic apparatus of a procaryotic organism with such a high chlorophyll content should occupy a substantial volume of the cytoplasmic region of the cell. For this reason, it is unlikely that the mesosomal elements constitute the sole, or even the major, site of the photosynthetic pigment system. The cells of green bacteria can be readily broken in the French pressure cell, or (in the case of cultures in the stationary phase) by osmotic lysis, following treatment with lysozyme and versene. After osmotic lysis, followed by DNAse treatment to reduce the viscosity of the lysate, all the chlorophyll and carotenoid in the extract is sedi- mented at low gravitational fields, in association with the lysed spheroplasts. Such behavior would not be expected of a holochrome with a molecular weight of 1,5 million. In another experiment, cells were STRUCTURE OF PHOTOSYNTHETIC BACTERIA 107 TABLE 2. Chlorophyll content of green bacteria grown at low light intensity (<40 foot-candles) . 2 Chlorophyll^ content of cells , ^^^'^^ l^g/mg cellular protein Chlorobium limicola R (660) 142 Chlorobium limicola ML (650) 120 Chlorobium limiCola 17 CR (650) 118 Chlorobium thiosulfatophilum , B (660) 115 Chlorobium thiosulfatophilum Tassajara (660) 190 Chlorobium thiosulfatophilum 6 CR (660) 100 1 The absorption coefficients for chlorobium chlorophylls 650 and 660 in meth- anol were taken from the publication of R. Y. Stanier and J. H. C. Smith, The chlorophylls of green bacteria, Biochim. Biophys. Acta, 41, 478 (1960). 2 The values in parentheses represent the type of chlorobium chlorophyll pres- ent in each particular strain. broken in the French pressure cell, and after a low-speed centrifuga- tion to remove residual intact cells and large fragments, the extract was centrifuged at 100,000 x g for 2 hours. The pigment system was completely sedimented. A very small fraction of the total pigment was contained in a loose layer overlying the pellet. After removal of this layer, the pellet was resuspended and subjected to centrifugation through a linear sucrose gradient (0,5-2,0 M) for 2 hours at 25,000 rpm. At the end of this period, most of the pigment was contained in a broad, deep-green band. This material had a considerably higher spe- cific chlorophyll content (325 /Ug/mg protein) than the original cells (190 /ig/mg protein). It proved to be structurally heterogeneous upon examination in the electron microscope after negative staining with phosphotungstate (Fig, 16), However, the bulk of the material consisted of vesicles 1000-1800 A long and 500-750 A wide. They are accordingly similar in form and dimensions to the peripheral chlorobium vesicles which are such a constant feature of thin sections of intact cells. Al- though this evidence certainly cannot be considered conclusive, it suggests that the chlorobium vesicles could be the major site in the cell of the photosynthetic pigment system. If this supposition is con- firmed by subsequent work, the 150 A holochrome particles isolated by Bergeron and Fuller must either be contained within the vesicles or else derived from them by comminution. 108 THE BACTERIAL PHOTOCHEMICAL APPARATUS Cv i ^^-^ \^ \ ^ ■'1 r%. 't 1 is 16 Fig. 15. Section of Chlorobium thiosulfatophilum , strain 6CR. The cell wall of this strain does not show the rod-shaped extensions char- acteristic of the strains illustrated in Figs. 12 and 13. The general structure of the cell is very similar to that of C. thiosidfatophilio)/, strain Tassajara. Chlorobium vesicles (Cv), mesosomal element (M). Main fixation, 2 hours, post staining with lead hydroxide. X 120,000. Fig. 16. Negatively stained preparation of the main pigmented frac- tion isolated from extracts of C, thiosulfatophilum , strain Tassajara (see text). STRUCTURE OF PHOTOSYNTHETIC BACTERIA 109 ACKNOWLEDGMENTS This work owes much to the generous help and advice given by Dr. J. H. McAlear and his associates on the staff of the Electron Microscope Laboratory. We are particularly indebted to Mr. Lloyd Thibodeau and Mr. Philip Spencer for their technical assistance. Much valuable help in the performance of ex- periments was given by Miss Riyo Kunisawa. REFERENCES 1. French, C. S., The chromoproteins of photosynthetic purple bacteria. Science, 88, 60 (1938). 2. Schachman, H. K., Pardee, A.B., and Stanier, R.Y., Studies on the macro- molecular organization of microbial cells. Arch. Biochem. Biophys., 38, 245 (1952). 3. Frenkel, A. W., Light induced phosphorylation by cell-free extracts of photosynthetic bacteria. J. Am. Chem. Soc, 76, 5568 (1954). 4. Vatter, A. E., and Wolfe, R. S., The structure of photosynthetic bacteria. J. BacterioL, 75, 480 (1958). 5. Bergeron, J. A., The bacterial chromatophore. Brookhaven Symp. Biol., 11, 118 (1958). 6. Drews, G., Untersuchungen zur Substruktur der "Chromatophoren" von Rhodospirilliim rubnim und Rhodospirillum molischianum. Arch. Mikro- biol., 36, 99 (1960). 7. Vatter, A. E., Douglas, H. C, and Wolfe, R.S., Structure of Rhodomicrobiurn vannielii. J. Bacterial., 77, 812 (1959). 8. Boatman, E. S., and Douglas, H. C, Fine structure of the photosynthetic bacterium Rliodomicrobium vannielii. J. Biophys. Biochem. Cytol., ii, 469 (1961). 9. Ris,H.,and Singh, R.N. , Electron microscope studies on blue-green algae. J. Biophys. Biochem. Cytol. , 9, 63 (1961). 10. Weibull, C, Characterization of the protoplasmic constituents of Bacillus megaterium. J. BacterioL, 66, 696 (1953). 11. Karunairatnam, M.C., Spizizen, J., and Gest, H., Preparation and proper- ties of protoplasts of Rhodospirillum nibrum. Biochim. Biophys. Acta, 29, 649 (1958). 12. Cohen-Bazire, G., and Kunisawa, R., The fine structure of Rhodospirillum rubrum. J. Cell. Biol., 16, 401 (1963). 13. Ryter, A., and Kellenberger, E., Etude au microscope electronique de plasmas contenantdel'acidedesoxyribonucleique. Z. Naturforsch., 136,597 (1958). 14. Millonig, G., A modified procedure for lead staining of thin sections. J. Biophys. Biochem. Cytol., 11, 736 (1961). 15. Huxley, H. E., and Zubay, G., Electron microscope observations on the structure of microsomal particles from Escherichia coli. J. Mol. Biol., 2, 10 (1960). 16. Cohen-Bazire, G., Sistrom, W. R., and Stanier, R. Y., Kinetic studies of pigment synthesis by non-sulfur purple bacteria. J. Cell. Comp. Physic'., 49, 25 (1957). 17. Stanier, R.Y., Doudoroff, M., Kunisawa, R., and Contopoulou, R., The role of organic substrates in bacterial photo sjmthe sis. Proc. Natl. Acad. Sci. U. S., 45, 1246 (1959). 110 THE BACTERIAL PHOTOCHEMICAL APPARATUS 18. Cohen-Bazire, G., and Kunisawa, R., Some observations on the synthesis and function of the photosynthetic apparatus in Rhodospirillum rubrum. Proc. Natl. Acad. Sci. U. S., 46, 1543 (1960). 19. Murray, R. G. E., The structure of microbial cytoplasm and the develop- ment of membrane systems. Lectures on theoretical and applied aspects of modern microbiology . University of Maryland, 1960-61. 20. Stanier, R. Y., The organization of the photosynthetic apparatus in purple bacteria. In General Physiology of Cell Specialization (D. Mazia and A. Tyler, eds.). McGraw-Hill, New York, 1963. 21. Giesbrecht, P., and Drews, G., Elektronenmikroskopische Untersuchungen uber die Entwicklung der "Chromatophoren" von Rhodospirillum molisch- ianum Giesher ger. Arch. Mikrobiol., 43, 152 (1962). 22. Schlegel, H. G., and Pfennig, N., Die Anreicherungskultur einiger Schwefel- purpurbakterien. Arch. Mikrobiol., 38, 1 (1961). 23. Newton, J. W., Macromolecular variation in the chromatophores of the photosynthetic bacterium Rhodospirillum rubrum. Biochim. Biophys. Acta, 42,34 (1960). 24. Bergeron, J. A., and Fuller, R. C., The photosynthetic macromolecules of Clilorobium thiosidfatophilum . p. 307 in Biological Structure and Fimction (T. W. Goodwin and O. Lindberg, eds.), Vol. II. Academic Press, New York, 1961. isolation of photochemically active chroma tophores from rhodospirillum molischianum'^ DONALD D. HICKMAN, ALBERT W. FRENKEL and KONSTANTINE COST^ Department of Botany University of Minnesota Minneapolis 14, Minnesota A bacteriochlorophy 11- containing particulate fraction capable of carrying out light-induced phosphorylation has been isolated from cells of Rhodospirillum molischianum.^ The cells from which these particles are derived contain lamellar structures in the cytoplasm (Fig. 1) showing the same characteristics as those described by Drews (1) and by Giesbrecht and Drews (2). When such cells are sub- jected to sonic disintegration followed by fractionation in the ultra- centrifuge according to the procedures described earlier fori?, rubrum (3), a fraction is obtained which contains well-defined particles as indicated in Fig. 2 A, The chromatophores oiR. molischianum, when compared with those isolated from R. nibnim (Fig, 2 B), usually are of greater diameter and also reveal more surface detail in phospho- tungstate- treated preparations. Judging from the appearance of sec- tioned cells, the chromatophores oiR. molischianum are disc shaped, while those of R. rubrum appear to be nearly spherical, Giesbrecht and Drews (2) have described chromatophores in R. molischiamim as consisting of stacks of 5-15 lamellae. The chromatophores described here (Fig, 2 A) appear to be subunits of such stacks. Rates of light-induced phosphorylation observed with isolated chromatophores from/?, molischiamim (150-200/^molesorthophosphate esterified as ATP per hour per jumole bacteriochlorophy 11) are com- parable to those reported by us for R. rub mm (3), Thus far we have not been able to elicit photoreduction of pyridine nucleotides by these particles. 1 This study was supported by grants E-2218 and E-3989 from the National Institutes of Health. 2 We wish to thank Dr.C.B. van Niel for the culture of Rhodospirillum molisch- ianum . 3 Holder of Graduate Fellowship from the Charles F. Kettering Foundation. Ill 112 THE BACTERIAL PHOTOCHEMICAL APPARATUS 4 1 1 Fig. 1. Section of RhodospiriUum molischianmn from a 3-day-old anaerobic light grown culture. Fig. 2. Isolated chromatophores. Negative staining with neutral phosphotung- state. A. From cells of 3-day-old culture of RJiodospirillum molischianum comparable to those in Fig. 1. B. From cells of 2-day-old culture of Rhodo- spirillum rubrum grown anaerobically in the light. Length of black bar rep- resents 0.1^. 114 THE BACTERIAL PHOTOCHEMICAL APPARATUS REFERENCES 1. Drews, G., Untersuchungen zur Substruktur der "Chromatophoren" von Rliodo spirillum rubrum und Rhodospirillum molischianum. Arch. Mikro- biol., 36, 99 (1960). 2. Giesbrecht, P., and Drews, G., Elektronenmikroskopische Untersuchungen fiber die Entwicklung der "Chromatophoren" voti Rhodospirillum molisch- ianum Giesberger. ibid., 43, 152 (1962). 3. Frenkel, Albert W., and Hickman, Donald D., Structure and photochemical activity of chlorophyll containing particles irom Rhodospirillum rubrum. J. Biophys. Biochem. Cytol., 6, 285 (1959). ISOLATION OF BACTERIOPHEOPHYTIN-CONTAINING PARTICLES FROM RHODOSPIRILLUM RUBRUM'^ TORU KIHARA and ALBERT W. FRENKEL Department of Botany University of Minnesota Minneapolis 14, Minn. The existence of two or three red peaks in the absorption spectra of members of the Athiorhodaceae and Thiorhodaceae has given rise to much speculation as to their identity. A partial fractionation of detergent-treated chromatophores has been achieved by Bril (1); the resulting material, however, was unstable and was not further identi- fied. When two-day-old, actively growing cultures ofR. rubrum are ex- tracted with distilled water or with phosphate buffer (pH 7.0) at 5°C, dilute colloidal extracts are obtained whose ratios of extinction at 800 m/i to that at 880 mfi are relatively high as compared with the ab- sorption characteristics of whole cells or of isolated chromatophores. By differential centrifugation it is possible to obtain a preparation which, in the red portion of the spectrum, shows only one peak at 794 m/j and none at 880 m/y (Fig. 1). This material can be sedimented by centrifuging at 60,000 x g for 1 hour and is free of cytochrome- 5 50. The spectrum of the resuspended B794 fraction (Fig. 1) also shows a pronounced peak at 364 m/i, but little if any carotenoid absorption. When this material is extracted with methanol and the pigment transferred to ethyl ether, it becomes apparent that the spectrum of the extract corresponds to that of bacteriopheophytin obtained through the acidification of chromatographically purified bacteriochlorophyll. This spectrum also is in agreement with published spectra (Fig. 2, 3) of bacteriopheophytin (2). It does not appear likely that the B794 ma- terial represents a decomposition product as it can be obtained most abundantly from cultures of rapidly growing cells and is more difficult to obtain in any quantity from older cultures. The origin and possible role of these particles remains to be determined. Methanol extracts of intact cells or of isolated chromatophores from R. nibnim, when prepared in the presence of CaCOs and separated chromatographically, also yield a trace of bacteriopheophytin. We do not 1 A preliminary communication. This investigation has been supported by grants from the National Institutes of Allergy and Infectious Diseases (E-2218) and the U. S. National Science Foundation (G-9888). 115 116 THE BACTERIAL PHOTOCHEMICAL APPARATUS 364 0.5 794 880 -Absorption spectrum of "normal", washed chromatophores Fig. 1. from R. ruhrum in 0.1 M K giycylglycine pH 7.7. Absorption spectrum of B794 particles (obtained from 2-day-old cultures) after removal of 880 m/i absorbing material by differential centrifu- gation; in 0.1 M K giycylglycine pH 7.7. believe this to be an extraction artifact since bacteriopheophytin does not form readily from bacteriochlorophyll even in the presence of dilute acids, and the isolation procedures have been carried out under con- ditions which should prevent the accumulation of acids in the extrac- tion mixtures. The cause for the discrepancy in the position of the in situ absorp- tion peaks near 800 m^ in the two types of particles may be ascribed to the presence or absence of the BgSO bacteriochlorophyll. The BssQ material has a higher extinction at 800 vafj. than at 794 m/i; thus, in the presence of excess B88O the absorption peak of the presumed bacteriopheophytin would tend to be shifted toward longer wavelengths {i.e., from 794 m^ toward 800 m/i). A more quantitative evaluation of this spectral shift will be attempted. BACTERIOPHEOPHYTIN-CONTAINING PARTICLES 117 400 600 800 m/j Fig. 2. Upper curve: Absorption spectrum of bacteriopheophytin pre- pared from purified bacteriochlorophyll by acidification with aqueous acetic acid in acetone. Resulting bacteriopheophytin extracted with ethyl ether. Spectrum measured in dry ethyl ether after removal of acetic acid and acetone. Lower curve: Spectrum of B794 extract in dry ethyl ether. An additional observation points to the possible identity of the material giving rise to the 800 niju peak in "normal" chromatophores and the 794 m/i peak in the "abnormal" particles: the treatment of both types of preparations with dilute KOH (pH 10) leads to a shift of the 800 mju peak or of the 794 m/i peak to 754 m/i; the position of the latter peak is identical in the two preparations. Work is in progress to establish a difference spectrum between "normal" chromatophores and B794 particles in an effort to obtain an approximation of the in situ absorption spectrum of Bsso alone. Also, we hope to be able to separate particles containing only the 880 mju peak and thereby obtain the in situ spectrum of Bgso directly. 118 THE BACTERIAL PHOTOCHEMICAL APPARATUS e 1.0 ' 358 1 M »• • • •* 770 ( • A « • A • • /I t • 11 • « II • • 1 t • 750l ;. » • 'n* * 0.5 h 522 568 1 ; ; A 680 •/ • I v. ; < /\ 7 *. I I*, 1 1 y .1 v:< O- /.\ 4. 400 600 800 m/j Fig. 3. -Absorption spectrum of bacteriochlorophyll in dry ethyl ether. Absorption spectrum of bacteriopheophytin in dry ethyl ether. We have tested the B794 particles for their capacity to photoreduce pyridine nucleotides but with negative results thus far. It would, how- ever, be most surprising if these particles did show any capacity to carry out partial photosynthetic reactions in the absence of true bacteriochlorophyll. CONCLUSIONS A particulate material which has only one red absorption peak at 794 Ytiju has been isolated from actively growing cultures of R. rnbnim. The pigment responsible for the absorption peak has been character- ized as bacteriopheophytin on the basis of its spectral characteristics. BACTERIOPHEOPHYTIN-CONTAINING PARTICLES 119 REFERENCES Bril, C, Action of a non-ionic detergent on chromatophores of Rhodopseudo- monas spheroides. Biochim. Biophys. Acta, 29,458 (1958). , Studies on bacterial chromatophores. I. Reversible disturbance of transfer of electronic excitation energy between bacteriochlorophyll- types in Chrotnatiutn. Biochim. Biophys. Acta, 39, 296-303 (1960). Goedheer, J. C, Optical properties and in vivo orientation of photo synthetic pigments. Thesis, Utrecht, 1957. Holt, A. S., and E.E.Jacobs, Spectroscopy of plant pigments. II. Methyl bac- teriochlorophyllide and bacteriochlorophyll. Am. Jour. Botany, 41, 718-722 (1954). STRUCTURE AND FUNCTION IN BACTERIAL PHOTOSYNTHESIS 1 HOWARD GEST and SUBIR K. BOSE The Henry Shaiv School of Botany and the Adolphus Busch III Laboratory of Molecular Biology, Washington University, St. Louis, Missouri In the concluding section of his recent monograph on Synthesis and Organisation in the Bacterial Cell, E. F. Gale (1) comments: "The title should also include 'disorganisation,' as most of the results dis- cussed in these pages have been concerned with the consequences of disorganisation— frequently on so drastic a scale that it is a miracle that anything of significance survives," During the past decade, our understanding of bacterial photosyn- thesis has been greatly increased through the study of biochemical activities which survive the violent cell-disruption procedures com- monly used (sonic oscillation, grinding with alumina, French pressure cell). In fact, the discovery (2) that pigmented subcellular particles ("chromatophores") from R. ruhrum and similar bacteria catalyze an anaerobic light- dependent phosphorylation of ADP has led to a pro- found revision of views on the fundamental features of the overall bacterial process. On the other hand, reflection on the recent history of analysis of other multi- component processes of comparable com- plexity {e.g., mitochondrial electron transfer and oxidative phosphory- lation) suggests that it would indeed be remarkable if particles ob- tained by the methods noted retained all of the biochemical capacities characteristic of the photochemical "apparatus" in its native state. The foregoing considerations provided the stimulus for earlier studies (3,4) on isolation of the particulate photochemical system by milder procedures. In particular, osmotic lysis of protoplasts appeared to represent a potentially ideal approach. The initial experiments (3) surprisingly revealed that upon osmotic lysis of R. nibriim proto- plasts2 (prepared by treatment with lysozyme + EDTA) the pigment system was not released but, rather, was retained in the membranous "ghost" structures. In sharp contrast with particles obtained by drastic methods of cell breakage, the pigmented "ghosts" are sedimentable 1 Supported by grants from the U. S. Public Health Service (E-2640) and the National Science Foundation (G-9877). 2 This term is used with the understanding that the cell-forms referred to are not entirely analogous to the protoplasts of Gram-positive bacteria. 121 122 THE BACTERIAL PHOTOCHEMICAL APPARATUS at low centrifugal force. As would be expected, washed "ghost" prep- arations photophosphorylate ADP when supplemented with appropriate cof actors (4). Pigmented membranous structures which catalyze photophosphory- lation can also be obtained by disrupting cells of R. rubrum through the sequential action of lysozyme and polymyxin B, in the absence of an osmotic stabilizer (4), Polymyxin B is believed to react with lipid components of the cytoplasmic membrane of sensitive Gram-negative bacteria, causing disorganization of the membrane and an attendant loss of specific permeability properties (5), It is significant that the structures produced by the action of lysozyme + polymyxin Bon/?. rubrum cells do not contain the large opaque granules characteristi- cally seen in this organism when it is grown in certain media. This observation reinforced the conclusion (4) that retention of the pigment system in the "ghosts" was not due to physical entrapment of "chro- matophores" by a limiting (damaged) membrane, "Chromatophores," however, are readily released when the relatively fragile protoplasts or "ghosts" are further disintegrated, e.g., by shaking with Ballotini beads in the Mickle apparatus or by exposure to surface- active agents such as sodium lauryl sulfate (4), The results briefly summarized above were the first indications^ that the system of pigments and electron carriers responsible for photochemical generation of ATP in R. rubrum is normally integrated with the cytoplasmic membrane or intracytoplasmic extensions of the membrane. This conception (3,4) received additional support from a cytological investigation by Giesbrecht and Drews (7) with another species oi Rhodospirillmn (molischianum) ; their electron micrographs indicate that "chromatophores" ariseby invagination of the cytoplasmic membrane and may, in fact, remain attached to the latter by tubular stalks, A recent study (8) with R. nibnim has provided similar cyto- logical evidence for an association of the photochemical apparatus with the cytoplasmic membrane. It is well known that the properties of "particulate" enzymes can be greatly influenced by the state of the structure with which they are combined. Accordingly, in a photosynthetic bacterium such as R. rubrum, disengagement of the pigment system and its associated catalysts from the complex in vivo matrix might be expected to result in significant alteration— or even total loss— of certain biochemical properties (4). It appears that the normal redox balances of electron 3 Immunochemical studies conducted with ChrojuatiH in at about the same time by J. W. Newton (6) led him to conclude that "the photosynthetic apparatus of purple bacteria is not necessarily a unique, discrete intracellular entity, since it contains macromolecular configurations common to the cell surface on one hand and certain 'intracellular' proteins on the other. It seems advisable, therefore, provisionally to consider it a part of a more complex organizational state within the cell." 240 BV + DOC 180 120 O BV MINUTES Fig. 1. Effect of lipid-dispersing agents on hydrogenase activity oiR. rubrinii particles. Washed R. nibriDU cells, grown in a malate + glutamate medium, were sus- pended in 0.05 M potassium phosphate buffer pH 7.1 and sonicated (Raytheon 10 KG) under hydrogen for 3 minutes. Pigmented particles werecoUected bycentri- fugation at 144,000 x g for 60 minutes and resuspended in 0.05 M phosphate buffer pH 7.6. The Warburg vessels contained, in a final fluid volume of 2.0 ml: particles equivalent to 0.75 mg bacteriochlorophyll; potassium phosphate pH 7.6, 275/ymoles; benzyl viologen (BV), 100 ^moles (added at zero time); where indicated, 10 mg sodium deoxycholate (DOC) or 10 mg sodium lauryl sulfate (LS). Gas phase, H2 (KOH in center well); temperature 30°C. 124 THE BACTERIAL PHOTOCHEMICAL APPARATUS transfer carriers involved in photophosphorylation are significantly disturbed by disruption of the native structure. Thus, optimal phos- phorylation by isolated pigmented particles is observed only if the overall redox potential is experimentally adjusted to a favorable region by addition of suitable reductants or "redox buffers" (9,10). The process of photochemical H2 evolution apparently is even more dependent on maintenance of in vivo structural integrity. Despite many attempts, this reaction has not yet been definitively demonstrated in a cell-free system using components exclusively derived from active cells of photosynthetic bacteria. In this connection, it is noteworthy that conversion of intact cells of R. rubrum to protoplasts using the lysozyme technique causes almost complete loss of H2- evolution activity (3). This could evidently be due to a direct inhibitory effect of lysozyme or to loss of a required enzyme, or cofactor, from the protoplasts. It is also conceivable that a relatively minor disturbance of a delicate organizational state may be responsible for the disappear- ance of activity. Although pigmented particles from R. ruhnim catalyze light- dependent phosphorylation and certain oxidation- reduction reactions, it seems possible that the properties of such particulate fragments may be substantially modified by spontaneous architectural changes which occur during their release from the cell. For example, struc- tural lipids or lipoproteins organized in some sort of membranous fabric may envelop, or partially overlay, normally accessible enzy- matic components. This possibility is suggested by the effects of de- oxycholate and other surface active agents on hydrogenase activities of pigmented particle preparations. As shown in Fig. 1, lipid-dispersing agents can cause marked acceleration of hydrogenase activity with an artificial electron acceptor (activity with ferricyanide is also stimu- lated). Similarly, hydrogenase activity of R. rubrum particles with FMN or FAD is completely dependent on the presence of deoxycholate or sodium lauryl sulfate (11). CONCLUSIONS The data now available from experiments with photosynthetic bac- teria and previous experience with other complex particulate systems indicate the desirability of exploring the photo- and biochemistry of bacterial photosynthesis in subcellular preparations produced by milder methods of cell rupture than are usually employed. Ideally, one could hope to develop procedures which preserve the native sur- face structures (cell "envelope" or "hull") and intracytoplasmic mem- branes essentially intact but which effectively eliminate the perme- STRUCTURE AND FUNCTION 125 ability barriers of the cell.^ Studies with such preparations possibly will provide evidence that the cellular organization of photosynthetic bacteria resembles that of other types of cells in the sense that "... electron transport- coupled generation of ATP and ATP-dependent energy andiontransportmay be specialized manifestations of a common enzymic function, inherent in cellular membrane structures, . ," (15), REFERENCES 1. Gale, E. F., Synthesis and Organisation in the Bacterial CeZZ. John Wiley and Sons, Inc., New York, 1959. 2. Frenkel, A. W., Light- induced phosphorylation by cell-free preparations of photosynthetic bacteria. J. Am. Chein. Soc., 76, 5568 (1954). 3. Karunairatnam, M. C, Spizizen, J., andGest, H., Preparation and properties of protoplasts of Rhodospirilluni rubrnm. Biochim. Biophys. Acta, 29,649 (1958). 4. Tuttle, A. L., and Gest, H., Subcellular particulate systems and the photo- chemical apparatus of /?//or/os/j/n7/»;» mbnim. Proc. Nat. Acad. Set. U. S., 45, 1261 (1959). 5. Newton, B. A., The properties and mode of action of the polymyxins. Bac- terial . Revs., 20, 14 (1956). 6. Newton, J. VV., Immunochemical reactions of the photosynthetic apparatus in purple bacteria. Brookhaven Synip. Biol., 11. 289 (1959). 7. Giesbrecht, P., and Drews, G., Elektronenmikroskopische Untersuchungen uber die Entwicklung der "Chromatophoren" won RhodospirilliDn nionsch- ianiim Giesberger. Arc/?. Mikrobiol., 43, 152 (1962). 8. Cohen-Bazire, G., and Kunisawa, R., The fine structure oi Rhodospirillitm rubrnm. J. Cell Biol., 16, 401 (1963). 9. Horio, T., and Kamen, M. D., Optimal oxidation-reduction potentials and endogenous cofactors in bacterial photophosphorylation. Biocheniistry, 1, 144 (1962). 10. Bose, S. K., and Gest, H., Bacterial photophosphorylation: regulation by re- dox balance. Proc. Nat. Acad. Sci. U. S., 49, 337 (1963). 11. Bose, S. K., and Gest, H., Hydrogenase and light-stimulated electron trans- fer reactions in photosynthetic bacteria. Nature, 195. 1168 (1962). 12. Bose, S. K., Gest, H., and Ormerod, J. G., Light-activated hydrogenase activity in a photosynthetic bacterium: a permeability phenomenon. J. Biol. Chem., 236, PC 13 (1961). (Please note erratum: for the concentrations of ferricyanide in figures and text, read molarity rather than millimolarity.) 13. Robrish, S.A.,and Marr, A.G., Location of enzymes in Azotobacter agilis. J. Bacteriol., S3, 158 (1962). 14. Hughes, D. E., The bacterial cytoplasmic membrane. J. Gen. Microbiol., 29, 39 (1962). 15. Ernster, L., Siekevitz, P., and Palade, G. E., Enzyme-structure relation- ships in the endoplasmic reticulum of rat liver, a morphological and bio- chemical study../. Ceil Biol., 15, 541 (1962). 4 Possible approaches are described in refs. 12-14. Top: W. Wiesner, D. I. Arnon; bottom: B. F. Burnham, J. Lascelles. II METABOLISM and PHYSIOLOGY METABOLIC ASPECTS OF BACTERIAL PHOTOS YNTHESISl HOWARD GEST The Henry Sliaiv School of Botany and the Adolphus Busch III Laboratory of Molecular Biology Washinsion Universitv, St. Louis, Missouri The overall metabolism of photosynthetic bacteria is clearly dis- tinguished from that of green plants in at least two particular respects (Table 1). These properties have been known for many decades and much effort has been expended in obtaining evidence for hypotheses designed to reconcile the "apparent" differences in bacterial and green plant photosynthesis. With increasing awareness of comparative bio- chemical correlations, it became natural to suppose that the light- TABLE 1 Distinctive features of autotrophic )uetaboUs))i in photosynthetic bacteria and green plants Photosynthetic bacteria Green plants Requirement for an "accessory" hydrogen (electron) donor Production of oxygen dependent metabolism in all photosynthetic organisms must be "basi- cally similar," Thus, the discordant properties of the bacteria as com- pared with higher plants could be viewed as the result of relatively minor variations on a basic theme (1,2). Since the time of Pasteur, the power of comparative biochemistry in rationalizing common principles in the framework of metabolism has been amply demonstrated. It is undoubtedly safe to say that at present the comparative biochemical approach is so ingrained that it has become a stock-in-trade aspect of our expanding methodology for 1 Research of the author is supported by grants from the National Science Foundation (G-9877) and the U. S. Public Health Service (E-2640). 129 130 METABOLISM AND PHYSIOLOGY exploring metabolic phenomena. This is illustrated by the fact that we routinely mix subcellular components isolated from very different organisms in order to reconstruct certain complex metabolic proces- ses; at least, to serve as models. At the same time, a number of in- vestigators working in different fields of biology have expressed dis- quieting attitudes on the general aims of comparative biochemistry and have also criticized current tendencies in its usage. This has resulted in some debate as to whether comparative biochemistry should now be concerned primarily with the study of common chemical principles shown by different forms of life, or with the origin and nature of biochemical variability from patterns common to many organisms. It can be argued that this difference in viewpoint represents a trivial philosophic confusion concerning two sides of the same coin, but I am inclined to believe that the debate is ultimately inspired by important questions which have not as yet received the scrutiny they deserve. One can legitimately ask whether undue emphasis on a particular apparent correlation may not have the effect of obscuring (or delaying the recognition of) a more profound relationship or a significant in- stance of biochemical "disunity." This possibility has led Ernest Bueding (3) to caution that excessive attention to biochemical unity, at the expense of adequate consideration of biochemical diversity, may lead to a distorted picture. In a very perceptive recent essay (4) relating to this topic, Seymour Cohen concludes that "the notion of the 'unity of biochemistry' has been advanced in an overly simplified form and reflects a primitive stage in the development of the discipline," Furthermore, he cites the area of photosynthesis as providing an ex- ample "in which the predilection for simplicity has impeded the de- velopment of understanding." Research advances during the past decade or so have sharpened the focus on the comparative biochemistry of photosynthesis consider- ably, but it is my opinion that a really satisfactory conception has not yet been achieved. The remainder of this paper is concerned with general metabolic properties of photosynthetic bacteria and with the discussion of several questions which are of particular significance for reevaluation of the comparative problem. Carbon metabolism and its ramifications From the standpoint of carbon nutrition, photosynthetic bacteria can grow anaerobically under two markedly different sets of conditions (2,5). On the one hand, CO2 can serve as the sole or primary source of carbon, provided an inorganic hydrogen donor is present. On the other hand, many types can grow luxuriantly in a completely synthetic medium containing a single organic compound, such as malate, in place of CO2 and the inorganic "accessory" hydrogen donor. Certain METABOLIC ASPECTS 131 photosynthetic bacteria (e.g., Chromatium) can grow under both sets of circumstances (see Table 2). The term "autoheterotrophic" (6) is suggested as a convenience to designate the growth pattern when an organic compound constitutes the carbon source, in cognizance of an autotrophic mechanism for the energy supply and a heterotrophic carbon metabolism. Fixation of CO2, liberated during the metabolic conversions of added organic com- pounds, also occurs during autoheterotrophic metabolism, but with most substrates the extent of this process is unknown and undoubtedly varies considerably depending on the nutritional conditions. In this connection, it is noteworthy that certain anaerobic heterotrophs re- quire surprisingly large quantities of CO2 for optimal growth (9,10). An example is given by the myxobacterium Cytophaga succinicans , whose CO2 requirement has been shown (11) to be related to the energy metabolism of the organism. In this instance, CO2 is essential in sub- TABLE 2 Sources of carbon, reducing poiver, and energy for anaerobic growth of photosynthetic bacteria Mode of growth: "autotrophic" "autoheterotrophic" Carbon: CO^, organic compounds Reducing power: H , S" etc. organic compounds Energy: light light strate amounts because it is a precursor of oxaloacetate (through con- densation with phosphoenolpyruvate), which in turn functions as a major oxidant for NADH2 generated during the fermentative breakdown of glucose. Accordingly, the CO2 is eventually converted to the carboxyl group of succinate and in essence has served as an electron acceptor. It seems quite possible (12,13) that CO2 may be used, to some ex- tent, in a similar way during autoheterotrophic metabolism of photo- synthetic bacteria (i.e., simply as an "accessory electron acceptor"). In fact, some thirty years ago F. M. Muller (14) concluded that this probably was the case during growth of purple sulfur bacteria on or- ganic substrates such as succinate, acetate, and butyrate. Muller also appreciated the autoheterotrophic character of the purple bacteria, which is clearly evident from the following statements made in his well-known paper published in 1933: 132 METABOLISM AND PHYSIOLOGY In the meantime it has to be kept in mind that the experiments referred to above yielded another fact of primary importance, viz. that also in this heterotrophic metabolism the cooperation of radiant energy is an essential factor. We must conclude from this fact that photochemical processes play a part in the conversion of the organic substrates, which is a most remarkable phenomenon. Besides the purple sulphur bacteria and the Athiorhodaceae , no heterotrophic organisms are known in whose metabo- lism radiant energy plays such an important part. The potential significance of heterotrophic carbon metabolism in bacterial photosynthesis was also recognized by others (e.g., see 15, 16,17), leading to the suggestion (16,18) that CO2 fixation may be by- passed or suppressed during photometabolism of certain organic sub- strates. This point of view, however, was not generally considered or discussed for some time, apparently because of preoccupation with the concept that CO2 reduction was a central feature of all bacterial photosyntheses. Combined autotrophy and heterotrophy in the sense used here is not confined to photosynthetic bacteria and may actually be more widespread in nature than is commonly supposed. An inter- esting example is found in the organism Desulfovibrio desidfuricans (Hildenborough strain). Mechalas and Rittenberg (19) have shown quite conclusively that this bacterium can use the "autotrophic" oxidation of H2 with sulfate as the energy source for "heterotrophic" growth on assimilable organic compounds. They suggest that other "heterotrophs" which contain hydrogenase, suchsis Escherichia coli,Azotobacter, etc., may also be capable of using "autotrophic" oxidation of H2 (e.g., with O2) as an energy supply for "heterotrophic synthesis." There has been considerable discussion in the literature on the fundamental differences between autotrophs and heterotrophs, and it is becoming increasingly difficult to make satisfactory definitions. Since the discovery of CO2 fixation in heterotrophs, many other distinctions which were once thought to be decisive have diminished in significance. In addition, the list of facultative autotrophs and organ- isms in the "grey zone" between the extremes constantly increases in length. This is really another way of saying that we are gradually seeing an enlarged spectrum of similarities and this, in itself, poses problems for comparative biochemical interpretations. It is interest- ing that this development was clearly anticipated by B. C. J. G. Knight in his early classic monograph (20) on bacterial nutrition. Knight not only emphasized the arbitrariness of the classification of organisms into the two categories of autotrophic and heterotrophic, but also held the view that this separation had "serious disadvantages from the point of view of the use of comparative physiology as a guide to further investigation." It was noted earlier that certain photosynthetic bacteria (e.g., ChroDiatiiou and R. nihriim) can grow using either CO2 or a single organic compound as the sole, or primary, carbon source. There is METABOLIC ASPECTS 133 little doubt that there must be numerous qualitative and quantitative metabolic differences between cells (of the same strain) growing under the two sets of conditions. First of all, we can expect that different pathways of carbon metabolism will predominate. The types of changes that might be anticipated are illustrated by the studies of H, L. Rom- berg and his colleagues (21) with the facultative chemosynthetic auto- troph Micrococcus deuitrificans . Their results indicate that when the micrococcus grows autotrophically the reductive pentose cycle oper- ates as a primary mechanism. On the other hand, cells cultivated as heterotrophs on acetate do not contain significant quantities of ribulose diphosphate carboxylase or other enzymes of the cycle for which tests were made. In place of the pentose cycle, cells growing on acetate apparently utilize the glyoxylate cycle as an important means of ob- taining carbon skeletons for biosynthesis of cellular constituents. One of the key enzymes of the glyoxylate cycle is isocitratase, and this enzyme is found in high concentration in acetate-grown cells of M. denitrificans and other organisms which use this particular cycle. Formation of isocitratase in such bacteria is ordinarily greatly sup- pressed when the carbon source is a C4 dicarboxylic acid, and this effect is believed to involve repression or related mechanisms of regulatory control. We can confidently predict that in the photosynthetic bacteria the transition from autotrophic to autoheterotrophic growth, or y/ce versa, will also evoke a number of enzymatic changes governed by repres- sion, derepression, or induction mechanisms. Although the data available are relatively limited, it appears that appreciably different patterns of carbon metabolism are found in different photosynthetic bacteria and, consequently, alteration of the carbon source used for growth leads to varied responses. Autotrophically grown cells of Chromatium contain a high level of ribulose diphosphate carboxylase, but the isocitratase content is low (22), Growth on acetate is charac- terized by a striking increase in isocitratase and a significant de- crease in the carboxylase level. It seems that when acetate is the carbon source for autoheterotrophic growth a modified type of glyoxy- late cycle becomes a prominent pathway inChromatium (22). The non- sulfur purple bacteria Rhodopseudo monas palustris and Rhodopseudo- monas capsidatiis resemble Chromatium (and M. denitrificans) in that they also contain isocitratase in large amount when grown on acetate (23). On the other hand, only traces of this enzyme are found in acetate- grown cells of Rhodopseudo monas sphcroides and /?. rubrum (23); in these particular organisms, the glyoxylate cycle evidently does not function to a quantitatively significant extent. The general picture emerging from studies on carbon pathways is that the mechanisms used by photosynthetic bacteria may vary signifi- cantly, depending on the nutritional conditions, and may range from 134 METABOLISM AND PHYSIOLOGY predominance of one particular sequence to a balanced mixture of al- ternative pathways. These appear to include, in whole or in part, most of the major pathways now recognized, such as the reductive pentose cycle, the citric acid cycle (13,24), and the glyoxylate cycle. It is relevant to note that the reductive pentose cycle is only one of several mechanisms by which photosynthetic bacteria can reduce substantial quantities of CO2 to cell material. For example, there is some evi- dence (25,26) that, in Rho do spirillum , CO2 may condense with a C2 fragment to form a C3 compound such as pyruvate, and it is possible that under certain conditions this type of reaction is of importance in net synthesis (see also ref. 27), Another type of CO2 assimilation is encountered in the metabolism of Rhodopseudomonas gelatinosa. Ace- tone serves as a carbon source for photosynthetic growth of this or- ganism, but only if CO2 is also provided, Siegel's studies (28-32) on this system suggested the following sequence of carbon conversions: CO + acetone ^ acetoacetate ^2 acetate ^ cell materials The initial step is an endergonic carboxylation and Siegel's investiga- tions were particularly instructive in that they clearly indicated the required energy can be supplied alternatively by: light-induced phos- phorylation, dark oxidative phosphorylation, or substrate-level phos- phorylation coupled with the dark anaerobic fermentation of added acetoacetate. An important question that arises in connection with enzymatic al- terations caused by variation of the carbon source is whether such changes will be primarily limited to enzymes specifically concerned with carbon transformations. It seems very unlikely that this would be the case, since the carbon conversions which occur during growth are obviously intermeshed or connected in some way with a multitude of other types of reactions. These interconnections provide a potential basis for an amplification of alterations, which may eventually attain relatively major proportions. In growing cells, particularly close relationships exist between carbon and nitrogen metabolism and, in the purple bacteria, the nature of the nitrogen growth source exerts a profound influence on the ulti- mate fate of organic compounds supplied in the environment. This has been studied particularly in R. nihrnm, but the available data suggest that many other photosynthetic bacteria behave similarly. Let us com- pare gross metabolic events in cultures oiR. rubrum growing anaero- bically in the light on malate, with either an ammonium salt or an amino acid such as glutamate as the nitrogen source. With the am- monium salt, the bacterium grows rapidly and, considering that it is an anaerobe, the cell yield is remarkably high (33): aside from CO2, metabolic byproducts are not found in the medium in appreciable METABOLIC ASPECTS 135 quantity. With glutamate serving as the nitrogen source, the growth rate is somewhat reduced and an additional metabolic product is now observed, viz., molecular hydrogen (34), The source of this hydrogen and the mechanism of its formation are problems of great interest because of their relationships with the energy metabolism of photo- synthetic bacteria. Hydrogen production by R. rubrum is dependent on light and, except for cells which contain readily expendable endogenous reserves, on the addition of oxidizable organic substrates, e,g., citric acidcycle inter- mediates (16,34), Maximal yields of gas are observed with truly resting cells and the quantities produced closely approximate those predicted on the basis of complete conversion of the organic com- pound to H2 and CO2 (24), i,e.: Acetate: Succinate: Fumarate: L-Malate: '=2«4°2 * 2H2O 2CO2 + 4H2 '=4«604 " ""2° "- *'=°2 * ™2 C^H^O^ + 4H2O 4CO2 + 6H2 C4H6°5 * '"2° ^ ^^°2 * ««2 Studies with intact cells have led to the conclusion that these re- markable conversions occur through the reactions of the well-known citric acid cycle coupled with an additional light-dependent process which, either directly or indirectly, effects the oxidation of reduced pyridine nucleotide by liberation of H2 (13,24), If certain utilizable nitrogen sources such as N2 or NH4+ are added to cells metabolizing in this fashion, the evolution of H2 is completely inhibited (34) and the quantity of CO2 produced is greatly diminished (35), These effects are especially clear-cut with N2, which establishes inhibition very rapidly. In our earliest publications (16,17,36; see also ref. 37) on this phenomenon, we made the obvious suggestion that the inhibition was due to diversion of a reductant, created by light, from the hydrogen- evolving system to reactions of reductive amination of keto acids derived from the organic substrate. 136 METABOLISM AND PHYSIOLOGY Although a competition of this sort may still be entertained as a first approximation, the effect is evidently more complex since the amount of N2 actually utilized in short-term experiments is consider- ably less than the predicted quantity (38), In other words, N2 appears to inhibit by two mechanisms, one of which is essentially "catalytic," It seems quite possible that the "catalytic" effect of N2 on H2 produc- tion may be a regulatory cut-off device similar to the so-called "allosteric inhibition" (39) observed in other enzyme systems. Further investigations will be required to unravel this interesting puzzle. The overall effect of N2, or ammonia, on the photometabolism of organic substrates by hydrogen-producing cells of R. nihrum is to divert carbon from the dissimilatory anaerobic citric acid cycle, and toward assimilatory pathways (5,24), This is hardly unexpected be- havior for a cell with an abundant energy supply suddenly faced with a utilizable nitrogen source. The absence of H2 production during active growth in media containing excess NH4"'" is partly understandable on the basis of the fact that reduction is required to convert this inorganic nitrogen source to the level of the amino group. There is, however, an additional mechanism which ensures the prevention of hydrogen evolution during growth in ammonium salt media. Substantial evidence has been obtained (33) indicating that NH4+, or a metabolic derivative, effectively represses the formation of one or more com- ponents of the hydrogen- evolving system. Accordingly, cells harvested from a malate + NH4+ medium before the nitrogen source is exhausted are initially incapable of producing hydrogen. After a period of con- tinued illumination in the presence of suitable organic substrates, however, H2 evolution begins and gradually increases in rate (see Fig. 1). Experiments (33) with chloramphenicol and other inhibitors indicate that the derepression of hydrogen production involves protein synthe- sis, which presumably can occur at the expense of amino acids in the endogenous pool. In agreement with expectations, cells harvested sev- eral hours after exhaustion of the nitrogen source from cultures grown with limiting amounts of ammonia are immediately capable of photoproducing hydrogen. This brief review of the carbon and nitrogen metabolism of photo- synthetic bacteria should serve, in part, to underline the dynamic and variable aspects of their metabolic behavior. In this respect, the photosynthetic bacteria obviously do not differ materially from other microorganisms. Their metabolic variability is emphasized here be- cause of a tendency to regard photosynthetic bacteria as "laboratory reagents" to be used primarily as a convenient source of photoactive subcellular particles. It is probable that the composition and proper- ties of such particles will differ considerably depending on the organism and its nutritional history, and it is prudent not to underestimate the possibility (5,40,41) that violent disengagement of the photochemical METABOLIC ASPECTS 137 200 B, "GLUTAMATE CELLS" 100 200 MINUTES 20 40 60 80 100 MINUTES 'resting' cells of R. rtibrum de- I- glutamate (B.) medium. After Fig. 1. Kinetics of photoproduction of H2 by rived from a malate + NH4''" (A.) or malate Ormerod and Gest (13). In A., the cells were harvested while ammonia was still present in the cul- ture medium. Except for the endogenous control in B., L-malate was added at zero time. For other experimental details, see ref. 13. apparatus from the complex cellular matrix may result in significant alteration or even total loss of certain biochemical characteristics. Remarks on the absence of oxygen production in bacterial photosyn- thesis Before the discovery of light-inducedphosphorylation, the similari- ties and differences in the autotrophic metabolism of green plants and photosynthetic bacteria could be visualized mainly in terms of forma- tion and subsequent fate of a photoreductant and a photooxidant. Ac- cording to the concept developed by van Niel (2), the photoreductant would be used for reducing CO2 in both types of organisms, while dis- posal of the photooxidant presumably occurs through alternative routes, i.e., conversion to O2 in green plants and reduction of the po- tential 02-precursor by the accessory electron donor in the bacteria. An appealing experimental system for investigating this hypothesis is furnished by green algae which can switch by '^anaerobic adapta- tion" from typical green plant photosynthesis to a bacterial- type of photoreduction of CO2 with H2 (42). The change is reversible in that O2 formation is resumed if the anaerobic algae are exposed to high light intensities. In a recent extension of studies with such organisms. Bishop (43) has isolated a mutant of Scenedesmus obliquus which 138 METABOLISM AND PHYSIOLOGY apparently has a genetic block in the oxygen-producing mechanism. This organism, consequently, isincapableof producing O2 when illumi- nated, but it retains the capacity for photoreducing CO2 with molecular hydrogen. It seems reasonable that further investigation of algae of the kind under discussion may provide valuable insights into the biochem- ical evolution of photosynthesis. On the basis of the data at hand, Gaffron (44) has concluded that it is: . . . extremely plausible that whatever else had to change to make the release of free oxygen possible— an indispensable part of this last evolu- tionary step must have been the addition of certain catalysts to an already existing complete mechanism for the fixation of carbon dioxide, for photo- phosphorylation, and for photo reduction with hydrogen. In place of the reaction leading to molecular oxygen there has been originally a coupling to hydrogenases or dehydrogenases. According to this point of view, O2 formation represented the acquisi- tion of a new and singular property. Perhaps there is really no need to ask why the bacteria do not produce O2; the significant question may be why green plants do, Gaffron (45) subscribes to the view that the function of the accessory hydrogen donor in autotrophic photosynthesis of the bacterial type is to reduce the potential 02-precursor (photooxidant) and has marshaled arguments against an alternative hypothesis, viz., that the accessory donor is, in fact, used as a source of hydrogen for CO2 reduction. In his terms, the alternative stipulates that water is no longer oxidized, but instead NADP is reduced"by a more direct photocatalytic hydrogen (or electron) transport," If we use the formalism of light- dependent "water cleavage," this seems to imply that the only alternative is one in which the "water cleavage" system is entirely eliminated. We can formulate schemes, however, which retain "water cleavage" for the purpose of producing the reductants and oxidants required for cyclic photophosphorylation, or generation of energy- rich precursors of ATP, and postulate that the hydrogen necessary for net reduction is derived from the accessory donor. This type of conception would relegate O2 production to the category of a unique feature of green plant photosyn- thesis which is not readily rationalized as a minor comparative bio- chemical variation. The notion that the accessory donor serves as the source of hydro- gen for CO2 reduction is an old idea which can be found in some of the earliest discussions (e,g,, see ref, 1) on the mechanism of bacterial photosynthesis. Similarly, the concept of a hydrogen transfer, between the added hydrogen donor and CO2, which is "driven" in some way by light- energy also was proposed in relatively modern terms many years ago (46). The knowledge accumulated during the past fifteen years on electron transfer processes and phosphorylation in a variety METABOLIC ASPECTS 139 of biological systems makes it possible to reexamine these ideas in a more sophisticated and meaningful way. Net reducing power in bacterial photosynthesis There is still no unambiguous supporting evidence for the proposal (see ref. 13) that a photoreductant provides the hydrogen atoms, or electrons, required for net formation of reduced pyridine nucleotide or molecular hydrogen. To my mind, it has become increasingly dif- ficult to account for the available biochemical and physiological facts on the basis of this kind of postulated mechanism, which has come to be called "noncyclic electron flow," In this connection, Bose and I (47) have recently examined the experimental basis of the claim (48) that an antimycin- resistant "noncyclic photophosphorylation" (i,e,, aphos- phorylation presumably dependent on light-stimulated"noncyclic elec- tron flow") system operates in the normal metabolism of purple bac- teria. The evidence cited in favor of this conclusion consists essen- tially of the demonstration, using pigmented particles from /2. ruhnim, of photophosphorylation ostensibly dependent on the presence of both an added reductant (ascorbate + DPIP) and oxidant (NAD). Our studies indicate that the phosphorylation observed in such experiments is in reality cyclic photophosphorylation, which is antimycin- resistant be- cause of the ability of DPIP to effect a by-pass in electron transfer around the antibiotic- sensitive region. The effects of added reductants or oxidants, or both, can be readily explained by the fact that optimal photophosphorylation by isolated particles requires maintenance of a suitable redox potential (49). The particles show very little inherent poise and, consequently, the rate of phosphorylation is quite sensitive to changes in the redox potential caused by electron donors or acceptors which interact with the electron transfer system. It is easy to show, in fact, that in the presence of excess reductant, phosphorylation will not proceed at a significant rate unless a suitable oxidant is added, even though there is no net electron transport (Table 3). The gas phase in both experiments was 100% H2. At 30 °C. (Exp. I), the hydrogenase in the R. rubnim particle preparation is active and when ascorbate + DPIP are also present the particles become "over- reduced," leading to an inhibition of photophosphorylation. It is ap- parent that the over- reduction effect can be completely prevented, or reversed, by addition of fumarate. At 20°C. (Exp, II), the hydrogenase is practically inactive and inhibition of light-induced phosphorylation by over- reduction is, accordingly, not observed. It is of importance to note that under the conditions employed there was no detectable consumption of H2 and as much as 8.5 /imoles of Pi could be este ri- fled in the presence of only 0,2 //mole of ascorbate. The results in Table 3 also indicate that DPIP can catalyze an antimycin- resistant bypass in electron transfer associated with photophosphorylation when the overall redox potential is suitably adjusted. 140 METABOLISM AND PHYSIOLOGY TABLE 3 Activation of antiiuycin-resistant cyclic photophosphorylation by DP IP and "redox buffers"* Experiment Temperature 30°C. 20°C. Additions Pi utilized (//moles) None 7.5 Antimycin A (10 ^g) 0.7 Ascorbate (0.2 //mole) + DPIP (0.2 /imole) 0.7 Ascorbate + DPIP + fumarate (10 //moles) 8.5 Ascorbate + DPIP + fumarate + antimycin A 4.4 None 4.4 Antimycin A (10//g) Ascorbate (0.2//mole) + DPIP (0.2 //mole) (0.2 //mole) Ascorbate + DPIP + antimycin A 5.0 Ascorbate + DPIP + NAD (1 //mole) 7.5 Ascrobate + DPIP + NAD + antimycin A 5.1 * After Bose and Gest (47). In experiment I, the R. riibnDn particles contained 0.11 mg bacteriochlorophyll; in experiment II, 0.20 mg. The reaction mix- tures contained in a final volume of 3 ml: particles; Tris-HCl pH 8.0, 100 //moles; ADP, 0.5 //mole; K2HPO4, 10.2 //moles; MgS04, 5 //moles; hexo- kinasc, 1 mg; glucose, 30 //moles. Gas phase, H2; incubation time, SOminutes in red light. As an alternative to "noncyclic electron flow," studies by a number of investigators, notably Chance and his colleagues (50,51), have pro- vided the basisfor a completely different, but at least equally plausible, mechanism for light- stimulated electron transfer between added donors and acceptors. These researches, with mammalian mitochon- dria, indicate that electron flow against an apparent thermochemical gradient can be driven by energy-generating systems. The reaction which has been studied in most detail is the reduction of NAD by suc- cinate, i.e., X "^ or X "^ P succinate + NAD ^fumarate + NADH METABOLIC ASPECTS 141 Considering the relative redox potentials of the succinate-fumarate (Eq = +0.031V.) and NADH2 - NAD (Eq = -0.32v.) couples, it canbe concluded that an input of energy is necessary to drive the reaction from left to right. This can be provided either by ATP or energy- rich intermediates (precursors of ATP) associated with oxidative phos- phorylation. Recent experiments by Griffiths and Chaplain (52) suggest the intriguing possibility that a labile phosphorylated form of NAD may be involved in this and analogous reactions. An energy- dependent reduction of NAD by succinate has been ob- served with R. rubrum particles, in the sense that a net reaction occurs only when the system is illuminated. The reaction catalyzed by the Rho do spirillum system was found by Frenkel (53) to be sig- nificantly inhibited when optimal amounts of ADP, Pi, and Mg++ were added. This observation suggests, among other possibilities, that there is a competition between the oxidation- reduction reaction and the phosphorylating system for a common intermediate. The oxidation of succinate to fumarate is a particularly interesting reaction in the metabolism of both mitochondria and photosynthetic bacteria. In both instances, there appears to be a close structural association between succinic dehydrogenase and the phosphorylating electron transfer system. In mitochondria, this physical integration presumably is one of the factors responsible for the unusually high "electron pressure" exerted by succinate on the electron transfer chain. Another property worthy of special attention is the relatively high redox potential of the succinate-fumarate couple. It seems un- likely that the presence of one oxidation- reduction step in the citric acid cycle with this redox character is an accident of nature, and I am inclined to believe that thisparticular aspect of the succinate-fumarate conversion may be of great importance in the regulation of electron transfer in both aerobic and anaerobic systems. Purple bacteria such as R. rubrum can grow anaerobically in the light with succinate as the primary carbon source and evidently must be able to oxidize succinate to fumarate under these conditions. What, then, is the electron acceptor? There is some evidence (54) from in- tact cell experiments that the anaerobic oxidation of succinate can be coupled with CO2 fixation, i.e., the "direct" acceptor in this case is presumably NAD, as in Frenkel's in vitro system. There is also evi- dence (13,24) that a coupling with CO2 reduction is not obligatory, in that glutamate- grown cells of R. rubrum can oxidize succinate with the liberation of H2; the gas yield (~7 moles H2/mole succinate) under optimal conditions indicates that the two hydrogen atoms removed in the initial oxidation step must be convertible to molecular hydrogen. To my knowledge, photosynthetic bacteria are the only known organisms which have the apparent potentiality of producing H2 from the conver- sion of succinate to fumarate. 142 METABOLISM AND PHYSIOLOGY Reasoning on the basis of redox potentials, the formation of H2 from succinate by illuminated cells of purple bacteria would be expected to require an even greater input of energy than the reduction of NAD. Although we could calculate the probable magnitude of the energy re- quirements for such reactions, or for formation of H2 from NADH2, it seems wise to heed Mansfield Clark's admonition (55) in this regard. He advises us to: "Beware when the attempt is made to apply such [thermodynamic] data too casually to the dynamic affairs of living cells and to such heterogeneous systems as are living cells." In any event, the contribution of the photochemical system to light- stimulated net electron flow in bacterial photosynthesis can be reasonably visualized in terms of energy-dependent "reverse" electron transfer (see Fig. 2). Hydrogen (electron) donor; inorganic or organic Fig. 2. Scheme for hydrogen (electron) flow from donors to acceptors and photoproduction of H2 in bacterial photosynthesis. The pyridine nucleotide may be either NAD or NADP; for convenience, only the former is shown. This hypothesis assumes that electrons (or hydrogen) required for net generation of reduced pyridine nucleotide are derived from an ac- cessory inorganic electron donor or organic compounds. Depending on the redox potential of the donor, or on the steady- state concentra- tions of the reduced and oxidized forms of donor and pyridine nucleo- tide, the formation of NADH2 may be promoted by energy- rich inter- mediates created by the action of light on the photochemical apparatus. Parenthetically, it may be noted that a similar promotion of "reverse" electron transfer by intermediates of oxidative phosphorylation could METABOLIC ASPECTS 143 explain the unsolved problem (56) of how certain chemosynthetic auto- trophs generate reduced pyridine nucleotide from the oxidation of in- organic electron donors of relatively high redox potential. Reduced pyridine nucleotide would be utilized for a variety of re- ductive biosyntheses, e.g., conversion of CO2 to organic compounds, transformation of C2 and C3 intermediates to reserve materials (5, 13,27,57), or reductive aminations to provide amino acids for protein synthesis. If, however, the illuminated cell is producing "excess" ATP and NADH2 relative to the demands of the biosynthetic machinery, the reduced electron carrier would be reoxidized, through liberation of molecular hydrogen, by an energy-dependent process. Conditions of this kind apparently obtain during photosynthetic growth with certain amino acids serving as the nitrogen source, and in illuminated resting cell suspensions which are rapidly metabolizing oxidizable compounds in the absence of utilizable nitrogen sources. According to the fore- going conception, H2 evolution is interpreted as the reflection of a kind of regulatory device which maintains ATP and reduced pyridine nucleotide at levels consistent with the overall rate of biosynthetic activity (13,24,33), Chance and Hollunger (58) have recently suggested that a similar type of control mechanism may operate in mitochondria, i.e., the succinate-dependent reduction of NAD shows great sensitivity to "uncouplers" of oxidative phosphorylation and this could provide a mechanism for delicate regulation of the concentration of NADH2 with- in the mitochondrion. The relatively simple scheme shown in Fig, 2 provides a working hypothesis which appears to be compatible with the salient oxidation- reduction features of bacterial photosynthesis. With particular refer- ence to light-dependent H2 formation, the scheme predicts that this process should be inhibited by compounds which "discharge" or "de- energize" the postulated ATP-precursors, Certain inhibitors of light- induced phosphorylation (such as antimycin A and redox dyes) are, in fact, potent inhibitors of H2 evolution (24). Addition of such compounds to illuminated intact cells of R. rubrum not only abolishes H2 forma- tion, but usually also causes the cells to resort to an anaerobic fer- mentation of endogenous reserves (to fatty acids). Such fermentation, which is characteristic of dark anaerobic metabolism, normally does not occur to an appreciable extent during illumination (59), In other words, it appears as if inhibition of the phosphorylating system can have the effect of suppressing overall photometabolism, which in turn results in the appearance of a fermentative pattern frequently seen in heterotrophic anaerobes. This striking phenomenon suggests that photophosphorylation activity inhibits fermentation and, accordingly, the induced transition could be characterized as a "photosynthetic Pasteur effect" (47), It is encouraging that other types of experiments have given results consistent with the occurrence of energy-linked "reverse" electron 144 METABOUSM AND PHYSIOLOGY flow in bacterial photosynthesis. Chance and Olson (60) have proposed that a process of this kind can account for the kinetics, obtained by dynamic spectrophotometry, of light- stimulated NAD reduction in intact cells of purple bacteria. Relevant evidence has also been obtained with a model cell-free "photohydrogenase" system in which light stimulates the oxidation of H2 with fumarate (61). Pigmented particles isolated from R. ruhrum do not reduce fumarate with H2 unless a suitable electron carrier is added. When a dye of relatively low redox potential such as brilliant cresyl blue (BCB; Eq = +0.047v.) is sup- plied, a rapid reaction occurs in darkness as shown in Fig, 3. Illumination has no effect on the rate of the reaction with BCB serving as the mediator. On the other hand, with DPIP( Eq = +0.22v,) negligible fumarate- reducing activity is observed in the dark and the reaction is now dependent on, or stimulated by, light. Although ferri- cyanide is rapidly reduced by the particulate hydrogenase, this com- pound does not catalyze fumarate reduction in dark or light. According to our interpretation, when DPIP is the mediator, light provides energy for driving hydrogen transfer against an unfavorable thermodynamic gradient, i,e,, from DPIPH2 to fumarate. Further evi- dence for this view has been obtained by Bose (62) using N,N,N',N'- tetramethyl-p-phenylenediamine (^TMPD; Eq = ~ +0,26v,) as the cata- lyst for light- dependent reduction of fumarate by H2. Table 4 sum- marizes some of his results showing the effects of various uncouplers of phosphorylation on the light- stimulated oxidation- reduction reac- tion. It can be seen that with the first three compounds both the oxidation of H2 and light-induced phosphorylation were inhibited. The results suggest that of the two processes, photophosphorylation is perhaps somewhat more sensitive to inhibition. With appropriate concentra- tions of oligomycin, on the other hand, photophosphorylation was in- hibited while the oxidation- reduction reaction was consistently stimu- lated; separate experiments disclosed similar effects with atebrin and gramicidin D. These data show a striking parallelism with the re- ported (58,63,64) effects of these inhibitors on oxidative phosphoryla- tion and energy-linked "reverse" electron transfer in mitochondria and, consequently, lend further support to the suggested interpretation. There is good reason to expect that the availability of active cell- free systems and new experimental techniques will lead to an intensi- fication in study of the basic features of light- activated electron trans- fer pathways in bacterial photosynthesis. Many significant questions remain unanswered, which is not surprising in view of the fact that our understanding of electron transfer in nonphotosynthetic systems is still far from complete in spite of decades of active investigation. One of the prominent unresolved problems in bacterial photosynthesis concerns the nature and extent of cross-connections between the elec- tron carrier systems associated with net electron transfer (i.e., from METABOLIC ASPECTS 145 60 80 100 MINUTES Fig. 3. Oxidation of H2 with fumarate by R. rubrum particles in the presence of redox carriers; light- stimulated catalysis by DPIP. After Bose and Gest (61). Conditions: Tris-HCl pH 8.0, 150 ^umoles; potas- sium phosphate pH 7.6, 3.5 //moles; sodium fumarate, 40^moles; washed R. rubrum particles (fromalurama- ground cells) containing 0.98 mg bacteriochlorophyll; water to make 2.0 ml final volume; 0.2 ml of 20% KOH in center well; gas phase, H2; temperature, 30 C; light intensity, 1000 foot-candles. Redox carriers (approximately 4 //moles) were added at zero time, and fumarate at 55 minutes. Brilliant cresyl blue, BCB; potassium ferricyanide, FeCy. 146 METABOLISM AND PHYSIOLOGY TABLE 4 Effects of 'hmconplers" on light-stimulated oxidation of H2 ivith fiimarate and photophosphorylation by R. ruhrum. particles* Inhibitor (M) Particle Cone. (mg bacterio- chl./3 ml) Activity ■Photohydro- Photophos- genase" phorylation (% of eontrol) Dicumaroi: 5 X 10-4 0.21 26 3 7 X 10-4 0.21 16 2.5 1 X 10"3 0.21 n-butyl-3,5-diiodo- 4-hydroxybenzoate: 1 X 10-4 0.18 70 78 5 X 10-4 0.18 16 4 1 X 10-3 0.18 m-chlorocarbonyl cyanide phenylhydrazone: 5 X 10-5 0.45 32 1 X 10"4 0.45 Oligomycin: 9 X 10-"^ 0.56 118 33 3 X 10-6 0.56 124 14 6x 10-6 0.56 124 9 x 10-6 0.56 113 After Bose (62). Particles were prepared as described in ref. 47 and illumi- nated with red light. Reaction mixtures (in Warburg vessels) contained, in a final volume of 3 ml: particles, as indicated; Tris-HCl pH 7.9, 200 /imoles; TMPD, 2 /Umoles; sodium fumarate, 40 fimoles; ADP, 2 /ymoles; K2HPO4, 40 //moles; MgCl2, 30 //moles; hexokinase, 1 mg; mannose, 60 //moles. Gas phase, H2 (KOH in center well); temperature, 30°C. The particles were preincubated with inhibitors for approximately 10 minutes and the extent of phosphorylation determined in the usual way after deproteinization of the suspensions used for the manometric assays. donors to acceptors) and cyclic photophosphorylation, A number of studies (47,49,65-68) indicate that electrons from substrates such as NADH2, succinate, and H2 can eventually interact with carriers of the photophosphorylating chain. It is possible that such interconnections are merely concerned with maintaining the latter carriers in states of optimal redox balance, but further investigation is required to estab- lish their true import. METABOLIC ASPECTS 147 REFERENCES 1. van Niel, C. B., On the morphology and physiology of the purple and green sulphur bacteria. Arch. Mikrobiol., 3, 1 (1931). 2. van Niel, C. B., The bacterial photosyntheses and their importance for the general problem of photosynthesis. Advan. Euzymol., 1, 263 (1941). 3. Bueding, E., Comparative biochemistry of parasitic helminths. Conip. Biochem. Physiol., 4, 343 (1962). 4. Cohen, S. S., On biochemical variability and innovation. Science, 139. 1017 (1963). 5. Gest, H., and Kamen, M. D., The photosynthetic bacteria. Handbuch der Pflanzenphysiologie, 5, Pt. 2, 568 (1960). 6. Autoheterotrophy may also take the form of a "heterotrophic" energy sup- ply combined with "autotrophic" carbon metabolism; e.g., in Pseudo)}ionas oxalaticus growing with formate as sole carbon and energy source, oxida- tion of an organic compound (formate) provides the energy and reducing power required for assimilation of CO2 (derived from formate) via the reductive pentose cycle (see refs. 7 and 8). 7. Quayle, J. R., and Keech, D. B., Carbon assimilation by Pseiidomonas oxalaticus (OX 1), 1. Formate and carbon dioxide utilization during growth on formate. Biochem. J., 72, 623 (1959). 8. Quayle, J. R., and Keech, D. B., Carbon assimilation by Pseiidomonas oxalaticus (OX 1), 2. Formate and carbon dioxide utilization by cell-free extracts of the organism grown on formate. Biochem. J., 72, 631 (1959). 9. Anderson, R. L.,and Ordal, E. J., Cytophaga succinicans sp. n,, a faculta- tively anaerobic, aquatic myxobacterium. J. Bacteriol., 81, 130 (1961). 10. Bryant, M. P., Small, N., Bouma, C, and Chu, H., Bacteroides ruminicola n. sp. and Succinimonas amylolytica, the new genus and species. Species of succinic acid-producing anaerobic bacteria of the bovine rumen. J. Bac- terial., 76, 15 (1958). 11. Anderson, R. L., and Ordal, E. J., C02-dependent fermentation of glucose by Cytophaga succii/icaiis.J. Bacteriol., SI, 139 (1961). 12. Lascelles, J., The formation of ribulose l:5-diphosphate carboxylase by growing cultures oi Athiorhodaceae . J. Gen. Microbiol., 23, 499 (1960). 13. Ormerod, J. G., and Gest, H., Hydrogen photosynthesis and alternative metabolic pathways in photosynthetic bacteria. Bacteriol. Revs., 26, 51 (1962). 14. MuUer, F. M., On the metabolism of the purple sulphur bacteria in organic media. Arch. Mikrobiol., 4, 131 (1933). 15. Gaffron, H., tJber den Stoffwechsel der Purpurbakterien, II, Biochem. Z., 275, 301 (1935). 16. Gest, H., Kamen, M. D., and Bregoff, H. M., Studies on the metabolism of photosynthetic bacteria, V. Photoproduction of hydrogen and nitrogen fixa- tion by Rhodospirillum rubrum. J. Biol. Chem., 182, 153 (1950). 17. Gest, H., Metabolic patterns in photosynthetic bacteria. Bacteriol. Revs., 15, 183 (1951). 18. Kamen, M. D., Hydrogenase activity and photoassimilation. Fed. Proc, 9, 543 (1950). 19. Mechalas, B. J., and Rittenberg, S. C, Energy coupling in Desulfovibrio desidfuricans . J. Bacteriol., 80, 501 (1960). 20. Knight, B.C.J.G., Bacterial Nutrition, Material for a Comparative Physi- ology of Bacteria, Medical Res. Council Special Report Series, No. 210, His Majesty's Stationery Office, London, 1938, see p. 27. 148 METABOLISM AND PHYSIOLOGY 21. Kornberg, H. L., Collins, J. F., and Bigley, D., The influence of growth substrates on metabolic pathways in Micrococcus denitrificans . Biochim. Biophys. Acta, 39, 9, (1960). 22. Fuller, R. C, Smillie, R. M., Sisler, E. C, and Kornberg, H. L., Carbon metabolism in Chromatium. J. Biol. Chem., 236, 2140 (1961). 23. Kornberg, H. L., and Lascelles, J., The formation of isocitratase by the Athiorhodaceae . J. Gen. Microbiol., 23, 511 (1960). 24. Gest, H., Ormerod, J. G.,and Ormerod, K. S., Photometabolism of Rhodo- spirillum nibnim: light-dependent dissimilation of organic compounds to carbon dioxide and molecular hydrogen by an anaerobic citric acid cycle. Arch. Biochem. Biophys., 97, 21 (1962). 25. Cutinelli, C, Ehrensvard, G., Reio, L., Saluste, E., and Stjernholm, R., Acetic acid metabolism in Rhodospirillum rubrumunder anaerobic condi- tion, n. Ark. Kemi, 5,315 (1951). 26. Cutinelli, C, EhrensvSrd, G., HSgstrom, G., Reio, L., Saluste, E., and Stjernholm, R., Acetic acid metabolism in Rhodospirilluiu nibnon under anaerobic condition. III. Ark. Kemi, 3, 501 (1951). 27. Stanier, R. Y., Doudoroff, M., Kunisawa, R.,and Contopoulou, R., The role of organic substrates in bacterial photosynthesis. Proc. Natl. Acad. Sci. U. S., 45, 1246 (1959), 28. Siegel, J. M., The metabolism of acetone by the photo synthetic bacterium Rhodopseudomonas gelatinosa. J. Bacteriol . , 60, 595 (1950). 29. Siegel, J. M., The photo synthetic metabolism of acetone by Rhodopseudo- monas gelatinosa. J. Biol. Chem., 208, 205 (1954). 30. Siegel, J. M., and Smith, A. A., The dark aerobic metabolism of acetone by the photosynthetic bacterium Rhodopseudomonas gelatinosa. J. Biol. Chem., 214, 475 (1955). 31. Siegel, J. M., The application of carbon-14 to studies on bacterial photo- synthesis, p. 334 in Proc. Intern. Conf. Peacefid Uses Atomic Energy, Geneva, 1955, vol. 12. United Nations, New York, 1956, 32. Siegel, J. M.,The dark anaerobic metabolism of acetone and acetate by the photosynthetic bacterium Rhodopseudomonas gelatinosa. J. Biol. Chem., 228, 41 (1957). 33. Ormerod, J. G., Ormerod, K. S., and Gest, H., Light-dependent utilization of organic compounds and photoproduction of molecular hydrogen by photo- synthetic bacteria; relationships with nitrogen metabolism. Arch. Biochem. Biophys., 94, 449 (1961). 34. Gest, H., and Kamen, M. D. , Photoproduction of molecular hydrogen by Rhodospirillum nibrum. Science, 109, 558 (1949). 35. Siegel, J. M.,and Kamcn, M. D., Studies on the metabolism of photosynthe- tic bacteria, VII. Comparative studies on the photoproduction of H2 by Rhodopseuflonionas gelatinosa and Rhodospirillum rubrum. J. Bacteriol., 61, 215 (1951). 36. Gest, H., and Kamen, M. D., Studies on the metabolism of photosynthetic bacteria, IV. Photochemical production of molecular hydrogen by growing cultures of photosynthetic bacteria. J. Bacteriol., 58, 239 (1949). 37. Gest, H., Judis, J., and Peck, H. D., Jr., Reduction of molecular nitrogen and relationships with photosynthesis and hydrogen metabolism, p. 298 in Inorganic Nitrogen Metabolism (W. D. McElroy and B. Glass, eds.). Johns Hopkins Press, Baltimore, 1956. 38. Bregoff, H. M., and Kamen, M. D., Studies on the metabolism of photosyn- thetic bacteria, XIV. Quantitative relations between malate dissimilation, photoproduction of hydrogen, and nitrogen metabolism in Rhodospinlluni rubrum. Arch. Biochem. Biophys., 36, 202 (1952). METABOLIC ASPECTS 149 39. Monod, J., and Jacob, F., General conclusions: teleonomic mechanisms in cellular metabolism, growth, and differentiation, p. 389 in Cellular Regu- latory MechanisDis, Cold Spring Harbor Syiiip. Quant. Biol., vol. XXVI. Waverly Press, Inc., Baltimore, 1961. 40. Newton, J. W., and Kamen, M. D., Photophosphorylation by subcellular particles from Chromatium. Biochim. Biophys . Acta, 25,462 (1957). 41. Tuttle, A. L., and Gest, H., Subcellular particulate systems and the photo- chemical apparatus of Rhodospirillum rubrum. Proc. Natl. Acad. Set. U.S., 45, 1261 (1959). 42. Gaffron, H., Photosynthesis, photoreduction and dark reduction of carbon dioxide in certain algae. Biol. Rev. Cambridge Phil. Soc, 19„1 (1944). 43. Bishop, N., Separation of the oxygen evolving system of photosynthesis from the photochemistry in a mutant oi Scenedesnnis . Nature, 195, 55 (1962). 44. Gaffron, H., Evolution of photosynthesis. Comp. Biocliein. Physiol., 4, 205 (1962). 45. Gaffron, H., On dating stages in photochemical evolution, p. 59 in Horizons in Biochemistry {M. Kasha and B. Pullman, eds.). Academic Press, Inc., New York, 1962. 46. Wassink, E. C., Photosynthesis as a light-sensitized transfer of hydrogen. Anionic van Leeuwenhoek, J. Microbiol. Serol., 12,281 (1947). 47. Bose, S. K., and Gest, H., Bacterial photophosphorylation: regulation by redox balance. Proc. Natl. Acad. Sci. U. S., 49,337 (1963). 48. Nozaki, M.,Tagawa,K., and Arnon, D. I., Noncyclic photophosphorylation in photosynthetic bacteria. Proc. Natl. Acad. Sci. U. S., 47,1334 (1961). 49. Horio, T., and Kamen, M. D., Optimal oxidation-reduction potentials and endogenous co-factors in bacterial photophosphorylation. Biochemistry, 1, 144 (1962). 50. Chance, B., and Hollunger, G., The interaction of energy and electron trans- fer reactions in mitochondria, VI. The efficiency of the reaction, J. Biol. Chem., 236, 1577 (1961). 51. Chance, B., Reversal of electron transfer in the respiratory chain, p. 119 in Biological Strricture and Function {T .W . Goodwin and O. Lindberg, eds.), vol. II. Academic Press, Inc., New York, 1961. 52. Griffiths, D. E., and Chaplain, R. A., Some properties of a new phosphory- lated derivative of NAD, an intermediate in oxidative phosphorylation, Bio- chem. Biophys. Res. Comnmn., 6\ 501 (1962). 53. Frenkel, A. W., Light-induced reactions of chromatophoresof Rhodospiril- lum rubrum. Brookhaven Symp. Biol., 11, 276 (1959). 54. Ormerod, J. G., The use of radioactive carbon dioxide in the measurement of carbon dioxide fixation in Rhodospirillum rubrum. Biochem. J., 64,373 (1956). 55. Clark, W. M., Introduction to "Symposium on electron transport in the metabolism of microorganisms." Bacteriol. Revs., 19, 234 (1955). 56. Lees, H., Some thoughts on the energetics of chemosynthesis. Bacteriol. Revs., 26, 165 (1962). 57. Elsden, S. R., and Ormerod, J. G., The effect of monofluoroacetate on the metabolism of Rhodospirillum rubrum. Biochem. J., 65,691 (1956). 58. Chance, B., and Hollunger, G., Inhibition of electron and energy transfer in mitochondria, IV. Inhibition of energy-linked diphosphopyridine nucleo- tide reduction by uncoupling agents. J. Biol. Chem., 238, 445 (1963). 59. Kohlmiller, E. F., Jr., and Gest, H., A comparative study of the light and dark fermentations of organic acids by Rhodospirillum rubrum. J. Bac- teriol., 6i,269 (1951). 60. Chance, B., and Olson, J. M., Primary metabolic events associated with photosynthesis. Arc/?. Biochem. Biophys., SS, 54 (1960). 150 METABOLISM AND PHYSIOLOGY 61. Bose.S. K.,and Gest.H., Hydrogenase and light-stimulated electron trans- fer reactions in photosynthetic bacteria. Nature, 195, 1168 (1962). 62. Bose, S. K., Studies on electron transfer and energy metabolism in photo- synthetic bacteria. Ph.D. thesis, Washington University, St, Louis, Mis- souri, 1963. 63. Slater, E. C, Tager, J. M., and Snoswell, A. M., The mechanism of the reduction of mitochondrial DPN^ coupled with the oxidation of succinate. Biochim. Biophys. Acta, 56, 111 (1962). 64. Sanadi, D. R., Fluharty, A. L., and Andreoli, T. E., Partial resolution and reconstitution of the adenosine triphosphate-dependent reduction of di- phosphopyridine nucleotide by succinate. -Bzoc/^ew. Biophys. Res. Cominun., 8, 200 (1962). 65. Frenkel, A. W., Photophosphorylation of adenine nucleotides by cell-free preparations of purple bacteria. J. Biol. Chem., 222, 823 (1956). 66. Horio, T., and Kamen, M. D., Some haem protein-linked pyridine nucleo- tide oxidation systems in RJio do spirillum rubrum. Biochim. Biophys. Acta, 43, 382 (1960). 67. Geller, D. M., and Lipmann,F., Photophosphorylation in extracts of Rhodo- spirillum rubrum. J. Biol. Chem., 235, 2478 (1960). 68. Frenkel, A. W., and Cost, K., Reduction of mammalian cytochrome c by Rhodo spirillum hydrogenase. Nature, 195, 1171 (1962). BIOCHEMICAL BASIS FOR THE OBLIGATE PHOTO- AUTOTROPHY OF GREEN BACTERIA OF THE GENUS CHLOROBIUM ROBERT M. SMILLIE and W. R. EVANS Biology Department, Brookhaven National Laboratory Upton, N. Y. Acetate and a variety of other organic compounds can serve as sources of carbon and reducing power for the light- dependent growth under anaerobic conditions of most photosynthetic bacteria. This is not true for certain species of the genus Chlorobium (C. limicola and C. thiosulfatophilum). These green sulfur bacteria grow only anaero- bically in light in the presence of exogenous CO2 and reduced inorganic sulfur compounds (1,2), They may be classified as anaerobic, obligate photoautotrophs. Such restricted requirements for growth substrates are not characteristic of all green bacteria. Chlorobium chloro- chromatii can grow on peptone and malate (3). Chloropseudomonas ethylicum utilizes several carbon sources including ethanol, acetate, and glucose (4). The inability of C. limicola and C. thiosulfatophilum to utilize sim- ple organic compounds as the sole carbon source for growth was es- tablished by van Niel (1) and Larsen (2) and subsequently confirmed by others (3,5). Although these compounds cannot act as growth sub- strates in the absence of CO2, several are photometabolized in the presence of CO2. Larsen demonstrated a carboxylation of propionate to succinate by cells of C, thiosulfatophilum . Sadler and Stanier (5) showed that the growth yield of C. limicola was improved by the addi- tion of certain organic supplements. Of these, acetate was the most effective and doubled the cell yield. The gross intracellular distribu- tion of assimilated carbon from acetate was identical with that from CO2. The incorporation of acetate was strictly dependent on the pro- vision of both CO2 and sulfide. It was proposed that, whereas other photosynthetic bacteria oxidize acetate to provide the CO2 and reducing power necessary for the assimilation of other acetate molecules, an enzymic pathway for acetate oxidation is lacking in C. limicola. The utilization of acetate for cell growth by this organism would then be dependent upon an exogenous supply of CO2 and reducing power. Since the oxidation of acetate in photosynthetic bacteria such as R. rubrum can proceed via an anaerobic citric acid cycle (6), extracts of C. thiosulfatophilum were examined for citric acid cycle enzymes. 151 152 METABOLISM AND PHYSIOLOGY The results obtained support the hypothesis of Sadler and Stanier (5) and provide a possible biochemical basis for the inability of C. thiosulfatophilum to utilize acetate in the absence of CO2. EXPERIMENTAL PROCEDURE The green sulfur anaerobe Chlorobium thiosulfatophilum, strain L, was cultivated as described by Larsen (7). The growth temperature was 26 °C, Chloropseudomonas ethylicum was grown at 30°C in a medium containing ethanol and CO2 as the carbon sources (8). Chro- matium, strain D, was grown at 35°C in Hendley's medium (9). R. rubrum was cultivated as described by Kohlmiller and Gest (10). Rhodomicrohium vannielii was grown in a medium containing acetate as the carbon source (11), Since some of the growth media contained sulfide, an inhibitor of aconitate hydratase (12), the harvested cells were washed several times with 0,05 M Tris buffer, pH 7,8. Cells were broken by passage through a French pressure cell (13). The extracts were centrifuged at 144,000 X g for 60 minutes and the supernatant fluid assayed for enzymic activity. The protein content of the extracts was determined by the biuret reaction (14). All enzymic measurements were made with a Gary recording spec- trophotometer using a slide- wire showing full-scale deflection at an absorbancy of 0.1. The assay temperature was 23 °C. Two methods were employed to assay aconitate hydratase. In one, the enzyme was directly assayed by the change in absorbancy at 240 mfi (15). Gitrate, cis-aconitate, or isocitrate were used as substrates. In the other, an excess of purified pig heart isocitrate dehydrogenase and TPN were included in the reaction mixture and the reduction of TPN was meas- ured by the absorbancy change at 340 m/i (16). Gitrate or cis-aconitate were the substrates. Isocitrate dehydrogenase was measured by the reduction of TPN in the presence of isocitrate (16), Succinic dehy- drogenase was assayed by the reduction of cytochrome c in the pres- ence of catalytic amounts of iV-methylphenazonium sulfate (17). The absorbancy change at 240 m/y using malate as substrate was recorded to estimate fumarate hydratase activity (15). Malate dehydrogenase was determined by DPNH oxidation in the presence of oxalacetate (18), Aceto-GoA- kinase was assayed by hydroxamic acid formation (19). RESULTS The Conversion of Acetate to Isocitrate The incorporation of acetate into isocitrate is a prerequisite for any oxidation via the citric acid cycle. Incorporation of acetate into OBLIGATE PHOTOAUTOTROPHY OF GREEN BACTERIA 153 isocitrate is also required for the assimilation of acetate via the glyoxylate cycle. The condensing enzyme was not directly assayed in extracts of C. thiosulfatophilum , but its existence was indicated by experiments with labeled acetate, Chlorobium extract, acetate- l-C^"^, Coenzyme A, MgCl2, and oxalacetate were incubated for various time intervals and the products chromatographed on paper in two dimen- sions. Autoradiography revealed a radioactive compound which had migrated to the same region of the chromatogram as citrate. This compound cochromatographed with citrate in two different solvent systems. Its appearance was dependent upon the presence of oxalace- tate in the reaction mixture. These experiments indicated that C. thiosulfatophilum extracts contained condensing enzyme and aceto- CoA-kinase, The presence of the latter enzyme was verified by direct assay using the procedure of Jones and Lipmann (19), As shown in Table 1, aconitate hydratase activity could not be demonstrated in C. thiosulfatophilum extracts. An increase inabsorb- ancy at 240 m/i with either citrate or isocitrate or a decrease with aconitate was not detected. On the other hand, the extracts contained a very active TPN-linked isocitrate dehydrogenase, TPN reduction was not observed in this assay if isocitrate was replaced by either aconitate or citrate even in the presence of an excess of purified pig heart isocitrate dehydrogenase. The same results were obtained with cells grown in the presence of sodium acetate (0,005 M), In another experiment, a cell-free homogenate of C. thiosulfato- philum and the supernatant fluid and sediment from a centrifuged portion of the homogenate (144,000 x g for 60 minutes) were incubated separately with aconitate. After two hours, enzymic assay failed to TABLE 1 Enzymic assays ivith extracts of C. thiosulfatophilum Enzyme Assay Substrates Activity Aconitate hydratase Aconitate hydratase Isocitrate dehydrogenase Isocitrate dehydrogenase Absorbancy change at 240 m// Coupled to isocitrate dehydrogenase TPN reduction DPN reduction Citrate, aconitate or isocitrate Citrate or aconitate jumoles/min/g protein Isocitrate Isocitrate 1940 154 METABOLISM AND PHYSIOLOGY reveal the presence of isocitrate. Fractions obtained from the super- natant fluid by ammonium sulfate precipitation did not show aconitate hydratase activity. Possible activators such as /?-mercaptoethanol, EDTA, and various metal ions were tried without effect. Extracts were also inactive after preincubation with ferrous ions and cysteine under the conditions described for activation of pig heart aconitate hydratase (20), Aconitate Hydratase in Other Photosynthetic Bacteria Since aconitate hydratase was not found in C. thiosulfatophilum, several other photosynthetic bacteria were examined for this enzyme. The extraction and assay procedures were the same as used for C. thiosulfatophilum . The extracts were also assayed for isocitrate de- hydrogenase. Both enzymes were present in all the extracts examined (Table 2), For comparison, the aconitate hydratase activity of a pea leaf extract was determined, A value of 21 //moles aconitate utilized per minute per gram of protein was obtained. The activities of the bacterial extracts are either comparable or higher than this value. There is no reason to suspect that enzymes of C, thiosulfatophilum are easily inactivated during the extraction procedure. In a previous study, C. thiosulfatophilum extracts were assayed for thirteen enzymes of the photosynthetic carbon cycle and activity was demonstrated in every case (21), However, the possibility remains that these extracts contain an inhibitor highly specific for aconitate hydratase. According- ly, the effect oi C .thiosulfatophilum extrdLcts on the aconitate hydratase TABLE 2 Aconitate hydratase and isocitrate dehydrogenase activities in extracts of photosynthetic bacteria Type of bacterium Source of extract Aconitate* hydratase Isocitrate TPN dehydrogenase DPN yU moles/min/g protein Green sulfur Chloropseudo- uionas ethylicum 25 1080 2.0 Purple sulfur Chromatium 104 76 0.9 Purple non- sulfur Rhodospirillum ntbniii? 51 460 2.3 Purple non- sulfur Rhodo micro - bium vannielii 20 365 5.6 Assayed by TPN reduction in the presence of aconitate and excess isocitrate dehydrogenase. OBLIGATE PHOTOAUTOTROPHY OF GREEN BACTERIA 155 TABLE 3 The effect of C. thiosulfatophilum extract on aconitate hydratase activity Except where noted, the enzyme was assayed by following TPN reduction coupled to isocitrate dehydrogenase Source of enzyme Protein in assay (//g) Chlorobium Absorbancy Substrate extract change in (//g protein) 10 min Per cent inhibition Chromatium 59 Aconitate _ .397 59 Aconitate 570 .334 16 59 Aconitate 550* .340 14 59 Citrate - .091 59 Citrate 570 .070 23 118 Isocitratet _ .133 118 Isocitrate 380 .115 14 Pea leaf 280 Aconitate - .425 280 Aconitate 380 .430 * Extract was dialyzed * Assayed by absorbancy increase at 240 m// activities of extracts of other organisms was examined. Variable results were obtained. Some extracts of C. thiosulfatophilum had no effect; with other preparations some inhibition resulted. Table 3 shows part of an experiment in which inhibition was obtained. In the same experiment, the aconitate hydratase activities of Chloropseudomonas ethylicum, R. rubrum, and Rhodomicrobium vannielii were decreased from 10 to 30 per cent by the addition of C. thiosulfatophilum extract. The inhibition was not removed by dialyzing the extract. Since inhibi- tion was never severe, it appears unlikely that the inability to demon- TABLE 4 Enzymes involved in succinate oxidation in the green photo synthetic bacteria Bacterium Succinic dehydrogenase Fumarate hydratase DPN-malate dehydrogenase fxmoles/min/g chlorophyll Chlorobium 28 thiosulfatophilum Chloropseudomonas not assayed ethylicum Hmoles/min/g protein 285 22 233 2170 156 METABOLISM AND PHYSIOLOGY strate aconitate hydratase in C. thiosulfatophilum can be attributed to the existence of inhibitors in the extract. Other Enzymes of the Citric Acid Cycle Succinic dehydrogenase, fumarate hydratase, and malate dehydro- genase are present in C. thiosulfatophilum (Table 4), Results obtained with another green bacterium, Chloropseudomonas ethylicum, are in- cluded for comparison. It may be noted that the very active fumarate hydratase activity of C. thiosulfatophilum was measured by an assay procedure very similar to one of the assays used for aconitate hydra- tase, C, thiosulfatophilum was not examined for o-ketoglutarate de- hydrogenase. DISCUSSION The purple photosynthetic bacteria can photometabolize acetate by several alternative pathways. In some, acetate assimilation appears to be predominantly through a glyoxylate cycle (22,23), Others, including R. rubrum, lack isocitratase (22) and utilize acetate by an unknown pathway (24), In addition, acetate may be oxidized via an anaerobic citric acid cycle (6), condensed with pyruvate to yield citramalate (25,26) or reduced to a storage compound, poly y5-hydroxybutyric acid (27). As pointed out by Gest^^ al. (6), the direction of metabolism "is dictated by the nutritional history of the cells and the conditions of incubation." Regardless of the initial pathway of incorporation, the eventual utilization of acetate for cellular growth in bacteria which do not contain a glyoxylate cycle is C02-dependent (27,28,29), even though growth on acetate may result in a net production of CO2, This neces- sitates an enzymic pathway for the oxidation of acetate to CO2, In R. rubrum this can take place via a light- dependent anaerobic citric acid cycle (6,30,31). Among the green bacteria, Chloropseudomonas ethylicum can oxidize acetate to CO2. During growth on acetate there is a net production of CO2 amounting to 15 to 20 per cent of the acetate utilized. In contrast, C. limicola can quantitatively assimilate acetate but only with simultaneous assimilation of CO2. Experiments with acetate-Cl4 disclosed that only a very small fraction of the acetate taken up by these cells reappeared as labeled CO2 (5). Since acetate assimilation by C, limicola is CO2 dependent, the mechanism of assimilation may be similar to that in R. rubrum, except that whereas in the latter CO2 and reducing power are produced endogenously from acetate, these must come from an exogenous source in the case of C. limicola. Hence while C. limicola possesses an enzymic mechanism for acetate assimilation, it does not appear to contain a mechanism for acetate oxidation. OBUGATE PHOTOAUTOTROPHY OF GREEN BACTERIA 157 The experiments described in this paper indicate that a closely related bacterium C. thiosulfatophilunA is deficient in aconitate hy- dratase. With the exception of a-ketoglutarate dehydrogenase which was not assayed, all the remaining enzymes necessary for the oxida- tion of acetate via the citric acid cycle are present. A deficiency of aconitate hydratase would, firstly, preclude the possibility of acetate incorporation via a glyoxylate cycle such as occurs in Chromatium (23), and secondly, prevent oxidation of acetate via a citric acid cycle. While metabolism via the citric acid cycle cannot result in a net up- take of carbon, it can provide the CO2 and reducing power necessary for acetate assimilation by a pathway such as that occurring in R. rubrum (24), A deficiency of aconitate hydratase could thus explain why acetate alone cannot support the growth of C. thiosulfatophilum. Assuming that these cells, like C. limicola, can assimilate acetate provided CO2 and reduced sulfur compounds are present, a pathway for acetate assimilation exists which involves neither the glyoxylate nor citric acid cycles. In the presence of thiosulfate, the growth of C, thiosulfatophilum could theoretically be supported by any substance which is converted by the organism to CO2. The inability of dicarboxylic acids like suc- cinate and malate to function as growth substrates could be attributed in part to a deficiency of aconitate hydratase. However, in R. rubrum these acids can be diverted to form CO2 and Cs-fragments under con- ditions where the citric acid cycle is blocked by monofluoroacetate (32), This alternative to metabolism via the citric acid cycle does not appear to operate in C. thiosulfatophilum. ACKNOWLEDGMENTS The research at Brookhaven National Laboratory was carried out under the auspices of the U. S, Atomic Energy Commission. The assistance of Miss H. Kelly, Mrs, O. Ritter, and Mr. N. Rigopoulos in cultivating the photosynthetic bacteria was greatly appreciated. SUMMARY Experiments were carried out in an attempt to establish a bio- chemical basis for the obligate photoautotrophy of Chlorobium thio- sulfatophilum. It is suggested that this organism cannot utilize acetate in place of CO2 as the growth substrate because it lacks aconitate hydratase. This precludes assimilation of acetate by a glyoxylate cycle as well as oxidation via the citric acid cycle to provide the CO2 C. thiosulfatophilum is distinguished from C. limicola by the inability of the latter to oxidize thiosulfate (7). 158 METABOLISM AND PHYSIOLOGY and reducing power necessary for the assimilation of acetate by al- ternative pathways. REFERENCES 1. van Niel, C. B., On the morphology and physiology of the purple and green sulphur bacteria. Arch. MikrobioL, 3, 1 (1931). 2. Larsen, H., On the microbiology and biochemistry of the photosynthetic green sulphur bacteria. Kgl . Norske Videnskab. Selskabs. Skrifter, 1953, 1-205. 3. Mechsner,K., Physiolischeund Morphologische Untersuchungen an Chloro- bakterien. Arch. MikrobioL, 26,32 (1957). 4. Shaposhnikov, V. V., Kondrat'eva, E. N., Krasil'nikova, E. N.,and Ramen- skaya, A. A., Green bacteria utilizing organic compounds. Dokl. Akad. Naiik SSSR, 129, 1424 (1959). 5. Sadler, W. R., and Stanier, R. Y., The function of acetate in photosynthesis by green bacteria. Proc. Natl. Acad. Set. U. S., 46, 1328 (1960). 6. Gest, H., Ormerod, J. G., and Ormerod, K. S,, Photometabolism of Rhodo- spirillmn nibnan: Light-dependent dissimilation of organic compounds to carbon dioxide and molecular hydrogen by an anaerobic citric acid cycle. Arch. Biochem. Biophys., 97, 21 (1962). 7. Larsen, H., On the culture and general physiology of the green sulfur bac- teria. J. Bacteriol., 64, 187 (1952). 8. Shaposhnikov, V. V., Kondrat'eva, E. N., and Fedorov, V. D., A new species of green sulphur bacteria. Nature, 187, 167 (1960). 9. Hendley, D. D., Endogenous fermentation in Thiorhodaceae. J. Bacteriol ., 70, 625 (1955). 10. Kohlmiller, E. F., and Gest, H., A comparative study of the light and dark fermentations of organic acids by Rho do spirillum nibnim. J. Bacterial., 61, 269 (1951). 11. Duchow, E., and Douglas, H. C, Rhodomicrobium vannielii,a. new photo- heterotrophic bacterium. J. Bacteriol., 58, 409 (1949). 12. Krebs, H. A., and Eggleston, L. V., Micro-determination of zsocitric and czs-aconitic acids in biological material. Biochem. J., 38, 426 (1944). 13. Milner, H. W., Lawrence, N. S., and French, C. S., Colloidal dispersion of chloroplast material. Science, 111, 633 (1950). 14. Gornall.A. G., Bardawill,C. S., and David, M. M., Determination of serum proteins by means of the biuret reaction. J. Biol. Chem., 177, 751 (1949). 15. Racker, E., Spectrophotometric measurements of the enzymatic formation of fumaricand c/s -aconitic acids. Biochim. et Biophys. Acta, 4, 211 (1950). 16. Ochoa, S., Biosynthesis of tricarboxylic acids by carbon dioxide fixation. 111. Enzymatic mechanisms. J. Biol. Chem., 174, 133 (1948). 17. Massey, V., The microestimationof succinate and the extinction coefficient of cytochrome c. Biochim. et Biophys. Acta, 34, 255 (1959). 18. Ochoa, S., Malic dehydrogenase from pig heart, p. 735 inMelhods itiEuzym- ology (S. P. Colowick and N. O. Kaplan, eds.), Vol. 1. Academic Press, Inc., New York, 1955. 19. Jones, M. E., and Lipmann, F., Aceto-CoA -kinase, p. 585 in Mctliods in Enzyniology (S. P. Colowick and N. O. Kaplan, eds.). Vol. 1, Academic Press, Inc., New York, 1955. 20. Morrison, J. F., The purification of aconitase. Biochem. J., 56, 99 (1954). 21. Smillie, R. M., Rigopoulos, N., and Kelly, H., Enzymes of the reductive pentose phosphate cycle in the purple and in the green photosynthetic sul- fur bacteria. Biochim. et Biophys. Acta, 56, 612 (1962). OBLIGATE PHOTOAUTOTROPHY OF GREEN BACTERIA 159 22. Kornberg, H. L., and Lascelles, J., The formation of isocitratase by the Athiorhodaceae. J. Gen. Microbiol., 23, 511 (1960). 23. Fuller, R. C, Sraillie, R. M., Sisler, E. C, and Kornberg, H. L., Carbon metabolism in Chromatium. J. Biol. Chem., 236, 2140 (1961). 24. Elsden, S. R., Assimilation of organic compounds by photosynthetic bac- teria. Fed. Proc, 21, 1047 (1962). 25. Benedict, C. R., Early products of [l 4c] acetate incorporation in resting cells oi Rhodospirillum rubrum. Biochim. et Biophys. Acta, 56,620 (1962). 26. Losada, M., Trebst, A. V., Ogata, S.,and Arnon, D. I., Equivalence of light and adenosine triphosphate in bacterial photosynthesis. Nature, 186, 753 (1960). 27. Stanier, R. Y., Doudoroff, M., Kusisawa, M., and Contopoulou, R., The role of organic substrates in bacterial photosynthesis. Proc. Natl. Acad. Sci. U. S., 45, 1246 (1959). 28. Cutinelli, C, Ehrensvard, G., Reio, L., Saluste, E., and Stjernholm, R., Acetic acid metabolism in Rhodospirillum rubrum under anaerobic condi- tions. II. Arkiv Kemi, 3, 315 (1951). 29. Hoare, D. S., The photometabolism of [l-l^c] acetate and [2- 14c] acetate by washed-cell suspensions of Rhodospirillum rubrum. Biochim. Biophys. Acta, 59, 723 (1962). 30. Fuller, R. C, and Dykstra, A. P., The photometabolism of glutamic acid in Rhodospirillum rubrum. Plant Physiol., 37, Suppl., iv (1962). 31. Eisenberg, M. A., The tricarboxylic acid cycle in Rhodospirillum rubrum. J. Biol. Chem., 203, 815 (1953). 32. Elsden, S. R., and Ormerod, J. G., The effect of monofluoroacetate on the metabolism of Rhodospirillum rubrum. Biochem. J., 63, 691 (1956). SOME OBSERVATIONS CONCERNING THE PURIFICATION AND PROPERTIES OF THE AEROBIC PHOSPHORYLATION SYSTEM OF R. RUBRUM EXTRACTS 1 DAVID M. GELLER Department of Phaniiacology, Washington University School of Medicine St. Louis, Missouri INTRODUCTION Previous work (1) demonstrated that cell-free extracts of R. nib- rum carry out an aerobic phosphorylation of adenosine diphosphate in the absence of light. By means of differential centrifugation, the aerobic phosphorylation system of extracts of cells grown aerobically in darkness was found to consist of soluble dehydrogenases and a particulate complex which catalyzed phosphorylation of adenosine diphosphate associated with the oxidation of DPNH. The following is a report of results of a further analysis of /2. rubrum extracts by centrifugation in sucrose density gradients. This work indicates that the bulk of the DPNH oxidase activity of crude extracts of cells grown aerobically in darkness is associated with particles which are lighter than those which catalyze aerobic phos- phorylation with DPNH. The phosphorylation system appears to be closely associated with succinic dehydrogenase activity. This aerobic phosphorylation system, purified by centrifugation in sucrose density gradients, has been characterized by its response to various inhibitors of electron transport and phosphorylation. EXPERIMENTAL PROCEDURE Growth of bacteria and preparation of extracts Bacteria were grown aerobically in darkness or anaerobically in light, harvested, and washed as previously described (1). Extracts derived from cells grown aerobically in the dark are designated by the term "dark extracts" and extracts of cells grown anaerobically in the light, by "photosynthetic extracts." 1 This investigation was supported by Grant RG-7023 from The National Insti- tutes of Health, U. S. Public Health Service. 161 162 METABOLISM AND PHYSIOLOGY Extracts were made by disruption of cell suspensions (200-250 mg, wet weight, of washed cells per ml of 0,05 M glycylglycine buffer, pH 8) at 0° in a French Pressure Cell (American Instrument Com- pany, Silver Springs, Maryland) at 15,000 pounds (Carver Press, Fisher Scientific Company), Intact cells and large debris were removed by centrifugation at 37,000 X g for 30 minutes. The supernatant fluid was centrifuged again for 30 minutes at 37,000 x g. The final supernatant is referred to as the crude extract. Fractionation of crude extract by centrifugation in sucrose density gradients All operations were performed at 0-5 °C, Following the methods of Britten and Roberts (2), SW39 rotor tubes were loaded with a linear sucrose gradient made by emptying a Bock and Ling mixing device containing 2,3 ml 5% sucrose and 2,3 ml 20% sucrose. All sucrose solutions used also contained 0,05 M glycylglycine pH 8, This was followed by the addition of 0.2 ml of crude extract and 0,2 ml 4% sucrose to the gradient tubes with a smaller mixing device; accord- ingly, the extract concentration attained its highest value at the mini- mum sucrose concentration (at the top of the tube). The tubes were centrifuged in the SW39 rotor, following the procedure of Martin and Ames (3). The tubes then were punctured with a hollow needle and nine to ten fractions, each containing 40 drops, collected. Assay of sucrose gradient fractions The protein contents of sucrose gradient fractions were estimated routinely by measuring optical density at 280m/^with a Zeiss spectro- photometer. In other instances, protein was measured by the method of Lowry et al. (4), Chlorophyll was estimated on the basis of the optical density of sucrose gradient fractions at 880 m/i. Enzymatic assays were facilitated by fitting a Leeds and Northrup Speedomax Type G recorder to the spectrophotometer. With special chart paper (No, TCI 1180 Chart paper, Technical Recording Chart Division, Graphic Controls Corporation, Buffalo 10, N, Y.), the full span of the recorder was equivalent to 0, 1 optical density unit. About 15 seconds afte. mixing, the time course of enzymatic reactions was followed for one minute. All enzymatic assays were done at room temperature (23°-25°C), unless otherwise stated. The assay of DPNH oxidase activity consisted of measuring the rate of decrease in optical density at 340 m/i induced by the addition of enzyme preparation to 0,5 mlof 0,1 mM DPNH in 0,1 M K phosphate + 5 mM MgCl2 pH 7.0, In the succinic dehydrogenase assays, measurements were first made of the decrease in optical density at 600 rafi induced by addition of enzyme to 0,5 ml of 0.1 M K phosphate + 5 mM MgCl2 pH 7.0 AEROBIC PHOSPHORYLATION IN R. RUBRUM 163 containing 7 jug 2,6-dichlorophenolindophenol, The rate observed was deducted from the rate seen upon the subsequent addition of 10 jul of 1 M sodium succinate. Aerobic phosphorylation was measured by counting the radioactive mannose-6-P produced by incubation of an aliquot of enzyme prepara- tion with radioactive Pi. The standard incubation mixture consisted of the following, together with enzyme, in a total volume of 0.5 ml (pH 7.4): 0.1 fimole DPNH, 0.2//mole MgCl2, l.Oyumole K phosphate, lO'^ cpm p32_inorganic phosphate (Nuclear Consultants Corporation, St. Louis 19, Mo.), 0.1 yumole ATP, 10/t/moles mannose, 5.0/imoles glycylglycine, and 0.25 mg yeast hexokinase (Fraction IV, Sigma Chemical Corporation, St. Louis, Mo.), The reaction was started by addition of enzyme to the mixture contained in a 3 ml tube. After 10 minutes in a 30 °C water bath in darkness, the reaction was terminated by adding 10 jul 100% TCA.2 Zero time controls were prepared by addition of TCA prior to all of the components of the incubation mix- ture, Mannose-6-P was separated from Pi by electrophoresis. To each acidified incubation sample, 10 /ul aliquots of 1 M K2HPO4 and 1 M K mannose-6-P were added. A 10 /il aliquot of the mixture was electrophoresed on Whatman 3 MM paper in 0.04 M Na citrate pH 3.6 for 30 minutes at 80 V/cm. With the apparatus used, consisting of lucite tank No. LT-36 fitted with auxiliary steel cooling coils and a 5 K.V, power supply (Savant Instruments, Inc., Hicksville, N, Y.), 4500 volts were applied, with the current ranging from 100 to 160 mA. Thirty samples could be processed in one run. The paper then was thoroughly dried at 50 °C for one hour. The spots were visualized by several hours exposure to Kodak No-Screen Medical X-ray Film and by aniline hydrogen phthalate spray (5), In routine assays (e.g. inhibi- tion experiments) the X-ray film exposure was omitted. The sugar phosphate spots were cut out (3/4 in, x 1 in, rectangles), placed in vials containing 10 ml POPOP-PPO mixture (4 g PPO and 0.1 g POPOP per liter of toluene), and counted in a liquid scintillation spec- trometer (Packard Instrument Co., Inc., La Grange, Illinois). The mannose- 6- P count was compared with total phosphate count (obtained by counting a paper rectangle spotted with a 10 /il aliquot of one hundred fold-diluted incubation mixture). In this system, mannose-6-P migrated toward the anode at two-thirds the velocity of Pi. In studies of the distribution of the aerobic phosphorylation sys- tem in sucrose gradient fractions, a correction had to be made for activity which was presumably due to exchange reactions unrelated to DPNH oxidation. This was done by determination of phosphorylation in duplicate tubes containing PMS (0.1 mg/ml). The incorporation of 2 The abbreviations used are: TCA, trichloroacetic acid; PPO, 2,5-Diphenyl- oxazole; POPOP, l,4-bis-2-(5-phenyloxazolyl)-benzene; SDH, succinic dehy- drogenase; M-6-P, mannose-6-phosphate; P-F3COCCP, phenylhydrazone of p-trifluoromethoxyphenyl carbonyl cyanide. 164 METABOLISM AND PHYSIOLOGY Pi into mannose-6-P in the presence of PMS was subtracted from the corresponding value obtained in the absence of PMS. Previous experi- ments have established that aerobic phosphorylation is strongly in- hibited by PMS (1). The phosphorylation assay thus was an estimation of PMS-sensitive incorporation of Pi into mannose-6-P. The PMS control was not used in studies of the properties of the purified aerobic phosphorylation system (which was completely inhibited by PMS). Biochemicals were obtained from Sigma Chemical Corporation. PPO and POPOP were obtained from Packard Instrument Company. RESULTS Separation of enzymatic activity by sucrose density gradient centri- fugation Initial centrifugation experiments were carried out with crude extracts derived iromR. yuhnim cells grown in darkness under aero- bic conditions. Succinic dehydrogenase activity was associated with a collection of particles which sedimented more rapidly than the bulk of the DPNH oxidase (Fig, 1), This was even more evident when the centrifugation time was increased from 40 to 60 minutes (Fig. 2), In both examples, the incorporation of inorganic phosphate into mannose- 6-P was associated with succinic dehydrogenase activity. With crude extracts derived from cells grown anaerobically in the light (Fig. 3), succinic dehydrogenase activity paralleled chlorophyll content. The rate of sedimentation of the dark succinoxidase particles (Fig. 1) appeared to be similar to that of the "chromatophore fraction" (Fig. 3). However, the patterns of DPNH oxidase activity differed strikingly; most of the apparent DPNH oxidase in the photosynthetic extract remained in the top "soluble" fraction (Fig. 3), whereas much of the DPNH oxidase activity of the dark extract was found in particu- late form (Fig. 1), In other experiments it has been found that the specific DPNH oxidase activities of crude "photosynthetic" and "dark" extracts differ considerably, whereas the succinic dehydrogenase activities do not. For example, the specific activity of the DPNH oxidase (at 30°C) of a crude photosynthetic extract was 9.6 m/«moles DPNH oxidized/mg protein/min. This was one-fifth of the value shown by a crude dark extract, 49 m^wmoles. On the other hand, the specific activity of succinic dehydrogenase (at 3 0°C) of the crude photosynthetic extract was 26 m/i moles indophenol dye reduced/mg protein/min, only one and one-half times the value for the crude dark extract, 18 m/i moles. The aerobic phosphorylation system of the photosynthetic extract, however, appeared to sediment in the same zone as the chromatophore AEROBIC PHOSPHORYLATION IN R. RUBRUM 165 4 6 8 10 Fraction Fig. 1. Activity profile of Crude Dark Extract Centri- fuged Forty Minutes at 35,000 rpm-Crude dark ex- tract, 5.9 mg protein, was centrifuged in a sucrose density gradient; fractions were collected and analyzed as described in the text. The first fraction represented the bottom fraction of the tube. In order to translate arbitrary units into activity stated in terms of each undiluted fraction: one OD28O unit (open triangles) represents an optical density of 0.65 at 280 m/i, one succinic dehydrogenase unit (squares) represents 2.2 m/imoles indophenol reduced/ ml/min, one DPNH oxidase unit (triangles) represents 6.6 m/imoles DPNH oxidized/ml/min, and one M-6-P unit (circles) represents 2.1 m/ymoles M-6-P formed/ ml/min. 166 METABOLISM AND PHYSIOLOGY 12 ^)^ V M.DPNH 10 - / ^ Ov y \ ' ^ Co ■ V Ki \ 1 8 - V\ J V ^OD,go 6 ^,— /^ 1 1 f\ M-6-P •^ V ^ 4 - /^ 1 A \ A 2 £r'^- 1 -A-^' \J\sDH t 1 1 1 r 2 4 6 8 10 Fract ion Fig. 2. Activity profile of Crude Dark Extract Centri- fuged Sixty Minutes at 35,000 rpm— Crude extract, 8.2 mg protein, was treated in the manner outlined in Fig. 1 (except that the centrifugation time was extended to sixty minutes). The actual values of the arbitrary units were (in terms of undiluted fractions): forOD280 (open triangles), one unit represents an optical density of 0.865 at 280 m//, for succinic dehydrogenase (squares) 2.2 m^umoles indophenol reduced/ml/min, for DPNH oxidase (triangles) 3.7 m^t/moles DPNH ox- idized/ml/min, and for M-6-P (circles) 1.0 m//mole M-6-P formed/ml/min. AEROBIC PHOSPHORYLATION IN R. RUBRUM 167 14 12 - A DPNH ry Units 00 O / ' / / '\0Ds80 J 1\ / roD,,, <5 ^ 4 / ' / / ' / d 1 1 J ? If 1 1 % 2 ,0/ ^^'-'^X ^M-6-P 2 4 6 8 Frac tion Fig. 3. Activity profile of Crude Photo synthetic Ex- tract—Crude photo synthetic extract, 6.0 mg protein, was centrifuged and analyzed as in Fig. 1. One arbi- trary unit represents (in terms of each undiluted frac- tion): for OD28O (open triangles), an optical density of 1.0 at 280 m^; for ODgSO (chlorophyll, open circles), an optical density of 0.17; for succinic dehydrogenase (squares), 5.4 m^raoles indophenol reduced/ml /min; for DPNH oxidase (triangles), 4.5 m/imoles DPNH oxidized/ml /min; and for M-6-P (circles), 0.68 m^ moles M-6-P formed/ml/min. 168 METABOLISM AND PHYSIOLOGY TABLE 1 Substrate Requirement for Aerobic Phosphorylation by a Sucrose Gradient Fraction of Dark Extract The incubation volume was reduced to 200 |Ul, with proportionate reduction of all components (see experimental procedure). Each tube contained an aliquot of the sucrose gradient fraction used in Fig. 5 (29 /ig protein). The inorganic phosphate content was 0.37 //moles (107 cpm)/200 /il. All values given have been corrected for the zero time control (960 cpm). Additions Mannose-6-Phosphate formed (cpm) 1. None 1,600 2. DPNH (0.05 //mole) 163,000 3. DPN (0.05 //mole) 2,200 4. Succinate (0.5 //mole) 21,900 5. DPNH (Extract omitted) 80 fraction. It must be noted that the association of this phosphorylation system with either succinic dehydrogenase or chlorophyll does not appear to be as evident as the parallel distribution of succinic dehydro- genase and chlorophyll seen in Fig. 3, or succinic dehydrogenase and the phosphorylation system of dark extracts (Figs, 1 and 2), The phos- phorylation exhibited by the peak fraction in Fig, 3 (fraction 4) has been shown to have the same substrate requirements as the corre- sponding fraction from sucrose gradient experiments with crude dark extracts (see Table 1), Preliminary attempts to show a possible interaction between the "nonphosphorylating DPNH oxidase" fractions and the phosphorylation system have met with negative results (Fig. 4), The usual sucrose density gradient experiment (using a crude dark extract) was per- formed, the assays yielding the usual curves for DPNH oxidase, suc- cinic dehydrogenase, and phosphorylation (curve M-6-P-I). Phos- phorylation assays were also made using fraction 5 in combination with each sucrose fraction; the curve obtained (M-6-P-II) is identical within experimental error to the dotted curve calculated by a simple arithmetic addition of activities (expected if there were no interaction of fractions). Some properties of sucrose gradient fractions having maximum aerobic phosphorylation activity DPNH was required for aerobic phosphorylation by a sucrose gradient fraction of dark extract (Table 1), Little phosphorylation was seen with succinate. The same substrate requirement has been ob- served in similar experiments with the corresponding fraction from AEROBIC PHOSPHORYLATION IN R. RUBRUM 169 2- DP/VH M-6-PII 2 4 6 8 10 Fraction Fig. 4. Effect of Combining Fractions on Aerobic Phosphorylation— Crude dark extract, 8.0 mg protein, was centrifuged in a sucrose density gradient for forty five minutes at 35,000 rpm, fractionated and analyzed as in Fig. 1. Succinic dehydrogenase and DPNH oxidase activities were assayed at 30°C. One arbitrary unit represents: for succinic dehydrogenase (squares), 5.4 myU moles indophenol reduced /ml /min; for DPNH oxidase (triangles), 7.0 m/Umoles DPNH oxidized/ml/ min; and for M-6-P (curve M-6-P-I), 4.9 m^moles M-6-P formed/ml/min. Curve M-6-P-II represents the assay of PMS-sensitive aerobic phosphorylation carried out in the same manner as for M-6-P-I, ex- cept that equal volumes of fraction 5 were also added. The dashed curve was calculated by adding theM-6-P value for fraction 5 (in curve M-6-P-I) to the values obtained for each fraction in curve M-6-P-I. 170 METABOLISM AND PHYSIOLOGY 4 6 Minutes Fig. 5. Kinetics and Efficiency of Aerobic Phosphoryla- tion— An incubation mixture (see experimental procedure) was warmed for five minutes at 30°C (in the absence of enzyme). An aliquot was withdrawn and acidified with TCA (zero time), and the reaction was initiated by adding enzyme (26 jug of a sucrose fraction in a final incubation volume of 734 /il). Optical density at 340 m^ was followed (30°C,thermostated cuvette compartment) and at intervals aliquots were withdrawn and acidified with TCA. The su- crose gradient fraction used contained maximal succinic dehydrogenase activity (in a sucrose density gradient centrifugation of crude dark extract). All values given have been corrected for the zero time blank (0.25 m//moles/ ml). A control tube, lacking DPNH, incubated at 30°C for 12 minutes gave a value of 0.98 m//moles M-6-P/ml. AEROBIC PHOSPHORYLATION IN R. RUBRUM 171 TABLE 2 Effects of Antibiotie Inhibitors on Aerobie Phosphorylation The procedure followed and the enzyme preparation used were the same as in Table 1. Ethanolic solutions of the antibiotics were added to the incubation mixtures (final ethanol concentration, 0.5%). Inhibitor Cone. iUg/ml Mannose-6-Phosphate cpm Inhibition % Antimycin A 5 0.5 0.05 162,000 48,600 143,000 147,000 (0) 70 12 9 Oligomycin 5 0.5 0.05 13,000 84,000 140,000 92 48 14 Valinomycin 5 0.5 0.05 17,800 117,000 142,000 89 28 12 Dianemycin 5 0.5 92,500 139,000 43 14 Ethanol control 146,000 10 sucrose density gradient centrifugation of crude photosynthetic ex- tract. The rates of both oxidation of DPNH and phosphorylation were constant during the incubation period (Fig. 5), with a calculated P/2e ratio of 0.45. This ratio is of the same order as that obtained with washed crude 145,000 x g particles in earlier experiments (1), In separate experiments it has been observed that the rate of DPNH oxi- dation is independent of the presence of necessary components of phosphorylation, viz.. Pi, Mg, and ADP, In a study of inhibitors of aerobic phosphorylation (Table 2), in- hibitory concentrations of the antibiotics antimycin A, Valinomycin, and Dianemycin were one to two orders of magnitude above those effective in sensitive mitochondrial systems (6,7). On the other hand, Oligomycin exerted an effect at a concentration required for inhibition of mitochondrial phosphorylation (6) , It is of interest to note (Table 3) that aerobic phosphorylation was inhibited by low concentrations of a carbonyl cyanide phenylhydrazone derivative (lO-'^-lO'^ m), a potent uncoupling agent of mitochondrial oxidative phosphorylation (8); on the other hand, the system was only partially affected by dinitrophenol. In separate experiments, the DPNH 172 METABOLISM AND PHYSIOLOGY TABLE 3 Effects of Uncoupling Agents and Inhibitors of Electron Transport on Aerobic Phosphorylation The procedure and enzyme used were as described in Table 1. The solutions of inhibitors were made just before use; p-FgCOCCP and 2,4-DNP were dis- solved in excess NaOH, Inhibitor Cone. M Mannose-6-Phosphate cpm Inhibition % None 162,000 P-F3COCCP 1 1 1 1 X X X X 10-8 10-7 10-6 10-5 160,000 131,000 35,000 2,000 1 19 78 99 2,4-DNP 1 X 10-4 115,000 29 PMS 3 X 10-4 11,300 93 NaCN 1 X 10-4 100,500 38 Amytal 1 X 10-3 56,700 65 NaOH control 1 X 10-3 141,000 13 oxidase was found to be sensitive to amytal (and the phosphorylation correspondingly so, as shown here). Both DPNH oxidase and phos- phorylation were moderately affectedby high concentrations of cyanide, and aerobic phosphorylation was inhibited by phenazine methyl sulfate. DISCUSSION The fact that the bulk of the DPNH oxidase of crude dark extracts may be separated from the phosphorylation system does not neces- sarily imply that the two activities were not associated in the cell prior to disruption. Indeed, the nature of the particulates of such ex- tracts is most likely a function of the means used to break the cells. This is exemplified by the experiments of M. Baltscheffsky (9) which demonstrated that the type of abrasive used influenced the amount of "soluble" DPNH oxidase activity obtained in extracts of photosyntheti- cally grown R. rubnim. Furthermore, experience with "chromatophore fractions" in R. nibnim must be cited; all of the pigment of osmotic lysates of lysozyme-treated R. nihnun cells was found to sediment readily in very low centrifugal fields (10), AEROBIC PHOSPHORYLATION IN R. RUBRUM 173 It must be noted, however, that separation of most of the DPNH oxidase from the phosphorylating system in dark extracts was some- what unexpected. Previous work with similar extracts (1) repeatedly had shown that DPNH oxidase, succinic dehydrogenase and the phos- phorylation system sedimented together under the conditions of the usual differential centrifugation. Preliminary experiments with crude photosynthetic extracts indi- cate that the aerobic phosphorylation system sediments at a rate similar to that of the chromatophore fraction. The data are consistent with the concept that the aerobic phosphorylation system is associated with the chromatophore fraction (11), Further proof of association, however, must await results of purification of these particulate sys- tems using other means. The aerobic phosphorylation system obtained in sucrose density gradient experiments appears to be the same system observed pre- viously in crude washed 145,000 x g particles obtained by differential centrifugation (1), Both the latter particles and the active sucrose fractions required DPNH for activity, giving P/2e ratios below 1; the DPNH oxidase of both was amytal sensitive, but relatively unaffected by antimycin A, The most interesting observation made in the study of the inhibition of aerobic phosphorylation was that low concentrations of a potent uncoupling agent (F3COCCP) were strongly inhibitory. This fact and the sensitivity to oligomycin indicate that the mechanisms involved in phosphorylation by this preparation may be similar to those of other sensitive systems (6,8). SUMMARY 1, Centrifugation in sucrose density gradients of crude extracts of R. rubrum cells grown aerobically in darkness has shown that the aerobic phosphorylation system is associated withheavy particles rich in succinic dehydrogenase. The phosphorylation system is separable from the bulk of the DPNH oxidase activity, which is associated with light particles. 2. Similar centrifugation experiments with crude extracts of photo- synthetically grown R. rubrum indicate that the rate of sedimentation of the aerobic phosphorylation system is similar to that of the chroma- tophore fraction. Succinic dehydrogenase activity is also closely cor- related with the chlorophyll content of the chromatophore fraction. 3. The properties of the aerobic phosphorylation system in sucrose gradient fractions are similar to those of the cruder systems previ- ously studied. 4. The effects of a variety of inhibitors of electron transport and uncoupling agents of oxidative phosphorylation on the aerobic phos- 174 METABOLISM AND PHYSIOLOGY phorylation system purified by sucrose density gradient centrifugation are described. ACKNOWLEDGMENTS I am indebted to Dr. George Drysdale for his valuable advice and loan of chart paper for the recording equipment. REFERENCES 1. Geller, D. M., Oxidative phosphorylation in extracts of Rhodospirillum rubrum. J. Biol. Che?)}., 237, 2947 (1962). 2. Britten, R. J., and Roberts, R. B., High-resolution density gradient sedi- mentation analysis. Science, 131, 32 (1960). 3. Martin, R. G., and Ames, B. N., A method for determining the sedimenta- tion behavior of Enzymes: Application to protein mixtures. J. Biol. Chem., 236, 1372 (1961). 4. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J., Protein measurement with the folin phenol reagent. J. Biol. CIicdi., 193, 265 (1951). 5. Block, R. J.,Durrum,E. L.,and Zweig, G.,A Maiiuol of Paper Chromatog- raphy and Paper Electroplioresis . Academic Press, Inc., New York (1955), p. 133. 6. Lardy, H. A., Johnson, D. , and Mc Murray, W. C, Antibiotics as tools for metabolic studies. I. A survey of toxic antibiotics in respiratory, phosphory- lative and glycolytic systems. ^ re/?. Biochem. Biophys., 7S, 587 (1958). 7. McMurray, W. C, and Begg, R. W., Effect of valinomycin on oxidative phos- phorylation. A re/?. Biochem. Biophys., 84, 546 (1959). 8. Heytler, P. G., and Prichard, W. W., A new class of uncoupling agents— carbonyl cyanide phenylhydrazones. Biochon. Biophys. Res. Comnnai., 7, 272 (1962). 9. Baltscheffsky, M. , On light-induced phosphorylation and oxidation of reduced diphosphopyridine nucleotide in /?/?oc/os/)/r?7Z//;;? rubrum. Acta Chem.Scand., 15, 215 (1961). 10. Tuttle, A. L., and Gest, H., Subcellular particulate systems and the photo- chemical apparatus of /itorfos/^rnZZw;;^? rubrum. Proc. Natl. Acad. Sci. U. S., 45, 1261 (1959). 11. Horio, T., and Kamen, M. D., Observations on the respiratory system of Rhodospirillum rubrum. Biochemistry, 1, 1141 (1962). METABOLISM OF PHOTOS YN THE TIC BACTERIA. II. CERTAIN ASPECTS OF CYCLIC AND NONCYCLIC PHOTOPHOSPHORYLATION IN RHODOSPIRILLUM RUBRUM M. NOZAKI,! K. TAGAWA and DANIEL I. ARNON^ Department of Cell Physiology University of California, Berkeley Investigations with isolated chloroplasts led to the discovery of photosynthetic phosphorylation comprising two photochemical re- actions, now called cyclic (Eq. 1) and noncyclic (Eq, 2) photophos- phorylation (1-4), which produce the assimilatory power (ATP and re- duced pyridine nucleotides) to drive carbon assimilation during photo- synthesis. hv ADP + Pi >- ATP (1) chloroplasts 2TPN + 2ADP + 2Pi + 4H O *- 2TPNH ^ 1-1 1 X " hv chloroplasts + 2ATP + ©2 + 2H2O (2) The terms cyclic and noncyclic photophosphorylation were suggested to denote the difference between a "closed" and an "open" electron transport pathway (coupled with ATP formation) that are envisaged for reactions 1 and 2, respectively (4,5). In cyclic photophosphoryla- tion, the electrons cycle in a "closed" system within the photorecep- tor particle, whereas in noncyclic photophosphorylation the photo- receptor particle mediates an "uphill" electron transfer in an "open" system from an external electron donor to an external electron ac- ceptor. 1 Present address: Department of Biochemistry, Medical School, Kyoto Univer- sity, Kyoto, Japan. 2 Aided by grants from the National Institutes of Health and the Office of Naval Research. 175 176 METABOLISM AND PHYSIOLOGY If the basic photosynthetic mechanisms are the same in green plants and in photosynthetic bacteria, it would be expected that cyclic and noncyclic photophosphorylation of chloroplasts would have appropriate counterparts in photosynthetic bacteria. Such counterparts were indeed found, but only after other investigations revealed fundamental simi- larities that were at first obscured by differences between the chloro- plast and the bacterial systems. Thus, when Frenkel found (6) a light- induced phosphorylation by cell-free preparations of R. rubrum and pointed to its similarity to the photosynthetic phosphorylation dis- covered earlier in spinach chloroplasts (1) he observed that the bac- terial particles became substrate (cy-ketoglutarate) dependent after washing (6). This was at variance with the unique feature of photo- synthetic phosphorylation in chloroplasts, a feature which distinguishes it from oxidative phosphorylation in mitochondria and the anaerobic phosphorylations associated with fermentation: ATP formation in chloroplasts occurs without the contribution of energy by an oxi- dizable substrate and solely at the expense of the energy contributed by absorbed photons (1,2). However, in later experiments, Frenkel (7) and other investigators (8-10) found that the role of o-ketoglutarate and other organic acids in the bacterial system was catalytic and regulatory and not that of a substrate. Once this fundamental point was clarified, the basic similarity of what we now call cyclic photo- phosphorylation in chloroplasts and in bacterial chromatophores was no longer in doubt (4,11). Another basic difficulty surrounded attempts to find a noncyclic photophosphorylation in chromatophores. This difficulty seemed at first insurmountable because photosynthetic bacteria never evolve oxygen (12), whereas oxygen evolution was a part of noncyclic photo- phosphorylation in chloroplasts (Eq. 2), However, Losada, Whatley and Arnon (13) separated noncyclic photophosphorylation with chloro- plasts (Eq. 2) into two component photochemical reactions: (a) a photo- oxidation of water3 (0H~) to yield oxygen (Eq. 3), and (b) the noncyclic photophosphorylation reaction proper (Eq. 4) — a reaction not accom- panied by oxygen evolution, but one in which ATP formation is coupled with a light-driven "uphill" electron flow to TPN from an exogenous electron donor other than water (OH~). The two component reactions have been experimentally separated by using an indophenol dye in its reduced and oxidized forms (A~ and A) as an intermediary electron carrier, in accordance with Eqs. 3 and 4: 3 Water and OH will be used interchangeably. OH represents the hydroxyl ion at neutral pH as in the reaction: 40H" (lO"'^ n) --O^ + 2H2O + 4e"; E' = 0.815 V (pH 7) CYCLIC AND NONCYCLIC PHOTO PHOSPHORYLATION 177 40H + 4A O2 ^ ^^ ^ ^"2° 2TPN + 2ADP = Sum: 2TPN + 2ADP 4 2Pi + 4H* + 40H" 2TPNH, 2ATP + 4A (3) (4) 2TPNH, 2ADP O2 + 2H2O (2) This subdivision of noncyclic photophosphorylation of chloroplasts into two component reactions revealed that Reaction 4, the noncyclic photophosphorylation proper, was basically independent of oxygen evolution and might, therefore, conceivably occur in photosynthetic bacteria. According to this interpretation, noncyclic photophosphoryla- tion in bacteria would lack Reaction 3, the photooxidation of water, which in chloroplasts reduces an intermediate electron carrier, probably plastoquinone (14,15), corresponding to A in Equation 3, In photosynthetic bacteria, the intermediate donor (A inEq, 4) would not come from a photochemical reaction but would be supplied by the external medium, in accord with the well-known dependence of bac- terial photosynthesis on such external hydrogen (or electron) donors as thiosulfate or organic acids (12,16), The existence in photosynthetic bacteria of a noncyclic electron transport pathway was suggested by the evidence in Chromatium cells for a light- driven electron flow from thiosulfate as the electron donor to H and N2 as the electron acceptors, resulting in the evolution of H2 and reduction (fixation) of N2^ respectively (17,18). Likewise, the photoreduction of pyridine nucleotide by succinate or the ascorbate- DPIP couple in cell-free preparations of/?, rubrum (19-21) could also be interpreted as evidence for a light-driven noncyclic electron flow in photosynthetic bacteria. There was no evidence, however, that such noncyclic, light- dependent electron transport in photosynthetic bacteria could be coupled with a simultaneous ATP formation. Evidence for noncyclic photophosphorylation in chromatophores of R. rubrum has recently been presented by Nozaki, Tagawa and Arnon (22), They observed bacterial noncyclic photophosphorylation under conditions when cyclic photophosphorylation was experimentally sup- pressed, thereby making it possible to distinguish the ATP formed by a noncyclic electron flow mechanism from that formed by a cyclic mechanism. In chromatophores of R. rubrum, in which cyclic photo- phosphorylation was made inoperative, ATP formation was coupled with a light-dependent noncyclic electron flow from an ascorbate-DPIP couple as the external electron donor to DPN as the terminal electron acceptor. This article reports further work on the nature of bacterial cyclic and noncyclic photophosphorylation in R. rubrum chromatophores, and considers particularly the experimental conditions needed to demon- 178 METABOLISM AND PHYSIOLOGY strate a distinction between these two electron pathways in photo- phosphorylation. The results confirm and extend the evidence for non- cyclic photophosphorylation which we reported earlier (22) and do not support the recent contention of Bose and Gest (23) that the ATP formed concurrently with the photoreduction of DPN by the ascorbate-DPIP couple is the result of cyclic photophosphorylation. METHODS R. ruhrum was grown as previously described (22), at a tempera- ture of about 27°C in 4,5 1, Pyrex glass-stoppered bottles, completely filled to exclude air. A blower was used to prevent overheating of the culture bottles. An inoculum of approximately 1 per cent was used for each bottle. The inoculum was grown first for about 30 hours as a stab culture in illuminated 100 ml screw-capped tubes, using nutrient agar con- taining the components of the nutrient solution (22). At the end of this period some cells were transferred to another stab culture tube. The old stab culture tube was filled with a nutrient solution of the same composition, incubated again for 30 hours, and the liquid portion was used as the inoculum, (This technique was based on a suggestion by Dr, D, M. Geller.) At intervals of several weeks the purity of the cul- tures was checked by groAving the cells in Petri dishes on nutrient agar. The cells were harvested 40-45 hours after inoculation, the cultures yielding then about 4 g of wet packed cells per liter of nutrient solu- tion. The cells were sedimented by centrifugation for 5 min. at 4000 X ^ at 1°C, The supernatant solution was discarded and the sedimented cells were washed once by suspending them in 0,1 M tris/HCl buffer at pH 7,8. The washed cells were collected by centrifugation. The method of preparing chromatophores under anaerobic conditions was the same as that previously described (22), Washed chromatophore preparations were suspended in 0.1 M tris buffer, pH 7.8, and either used directly or stored at 4°C for several days under argon prior to use, Bacteriochlorophyll, ATP and DPNH2, and lactate were deter- mined as previously described (22). The experiments were carried out in Thunberg-type cuvettes or in Warburg vessels, which were illuminated by incandescent reflector spot lights. All experiments were carried out under argon unless otherwise specified. When aged chromatophores were used, hexokinase, d-glucose, and a catalytic amount of ADP were added to the reaction mixture as a trapping system for ATP, to prevent loss of ATP by ATPase activity, (The ATPase activity was negligible in fresh chromatophores but gradually increased upon aging,) The "aged" chromatophores were prepared by storage of fresh preparations at 2°C for 30 days under CYCLIC ANDNONCYCLIC PHOTO PHOSPHORYLATION 179 argon. The ATP or the glucose-6-phosphate formed was determined by the magnesium ammonium phosphate precipitation method (24) or by the siliconized celite column method (25). Crystalline lactate dehydro- genase (from muscle), crystalline alcohol dehydrogenase, antimycin A, and hexokinase (Type III) were purchased from Sigma Chemical Co. Hydrogenase was partially purified from Desulfo vibrio desulfuri- cans on a DEAE-cellulose column. RESULTS I. CYCLIC PHOTOPHOSPHORYLATION Inactivation by pretreatment with salt at pH 5.0 Acidification to pH 5, with acetate buffer, resulted in a coagulation of chromatophores without impairment of phosphorylating activity. However, when the acidification was accompanied by the addition of salt, an appreciable loss of phosphorylation occurred. As shown in Table 1, the loss of phosphorylation is due to the high ionic strength at pH 5, and not to the kind of salt. Ionic strength lower than 0,2 at pH 5 or high ionic strength at pH 7.8 had little effect. The inhibitory effect of high salt concentration at pH 5 was observed in the endogenous sys- tem as well as in the presence of phenazine methosulfate (Table 1). Attempts to reactivate phosphorylation in the salt-treated chromato- phores were unsuccessful. TABLE 1 Inactivation of cyclic photopliosphorylatioii by pretreatinoit ivith salt //moles ATP formed/mg Pretreatment . ■ . ,.u u Bchl/hr with salt lomc strength pH endogenous PMS None 0.05 7.8 140 400 NaAc 0.28 5.0 110 395 NaAc + NH4CI 0.63 7.8 - 386 NaAc + NH4CI 0.53 5.0 - 22 NaAc + KCl 0.78 5.0 25 36 NaAc + KNO3 0.53 5.0 - 18 Prior to thephotophosphorylation reaction the chromatophores were exposed for 5 minat 0°C to the indicated salt mixture, collected by centrifugation (5 min) and resuspended in 0.1 M tris buffer, pH 7.8. The reaction mixture for photo- phosphorylation included in a final volume of 3 ml, chromatophores (containing 100 yt/g bacteriochlorophyll), 1 mg hexokinase and the following in //moles: tris buffer, pH 7.8, 100; MgCl2, 5; ADP, 0.5; K2HP32O4, 10; D-glucose, 30; and where indicated, phenazine methosulfate (PMS), 0.3. The reaction was run at 20°C for 15 min in the light (20,000 lux). 180 METABOLISM AND PHYSIOLOGY Effect of aging and heat treatments As previously reported (see Table 1 in ref. 5), chromatophores from R. rubynm, when freshly prepared under anaerobic conditions, have a complete system for cyclic photophosphorylation. Cyclic photophosphorylation in chromatophores (unlike chloroplasts) proceeds without the addition of exogenous cof actors or water-soluble extracts. When chromatophores were prepared under argon (22), even washing them three times in succession had little effect on phosphorylating activity (cf. 26). The full phosphorylating activity of chromatophores stored anaero- bically (under argon) at ca. 4 °C was readily restored to original levels. As shown in Table 2, full phosphorylating activity was restored to chromatophores stored for 30 days by the addition of ascorbate, DPNH2 or phenazine methosulfate. In both fresh and aged chromato- phores, phosphorylation, in the presence of ascorbate or DPNH2 but in the absence of phenazine methosulfate, was completely inhibited by antimycin A. Table 3 shows that chromatophores heated in air for 10 min, at 50 °C lost about 60 per cent of their endogenous phosphorylating activ- ity. This loss of activity was restored by the addition of phenazine methosulfate or ascorbate. Heating at 50 °C under argon caused no impairment of phosphorylating activity. It appears that the loss of phosphorylating activity on aging or heating in air is probably the result of the oxidation of one or more electron carriers. Either this oxidation is reversed by the addition of a reductant (ascorbate or DPNH2), or the oxidized site may be by- passed by the addition of phenazine methosulfate. Effect of redox agents Horio and Kamen (26) have shown that a high phosphorylation activity of chromatophores depended on the presence of an appropriate amount of ascorbate to provide a redox potential of about volt. At lower concentrations of ascorbate the system was "overoxidized" and TABLE 2 Restoration of Cyclic Photopliosphorylatioii in Aged Clironiatophores /imoles ATP/formed/mg Bchl/hr Additions fresh aged None 5 X 10 "^ M ascorbate 135 41 200 135 3 X 10-3 M DPNH9 137 145 10-4 M phenazine methosulfate 352 304 Illumination and components of the phosphorylation reaction mixture were the same as given in Table 1. CYCLIC AND NONCYCLIC PHOTO PHOSPHORYLATION 181 TABLE 3 Effect of Heat Treatment on Cyclic Photophosphorylation ytimoles ATP formed/mg Bchl/hr Additions Heated in air Heated in argon Control none 77 201 192 ascorbate 176 204 203 phenazine 307 328 307 methosulfate Heat pretreatment was for lOmin at 50°C. The final concentration of ascor- bate was 5 X 10-3 M. The conditions and components of the phosphorylation reaction mixture were the same as given in Table 1. at higher concentrations the system became "overreduced" for opti- mum photophosphorylation. The effect of a particular ascorbate con- centration was influenced by the gaseous atmosphere. Thus, an over- reducing effect of a high concentration of ascorbate under anaerobic conditions was mitigated by the admission of air (cf. also 23), As shown in Table 4, we have confirmed the effect of ascorbate con- centration in argon and in air on cyclic photophosphorylation. ATP formation in air or argon was strongly inhibited by antimycin A. No appreciable antimycin A- resistant photophosphorylation occurred with the addition of ascorbate alone at any of the ascorbate concentrations tested (10"5 to 10*2 M). TABLE 4 Effect of Ascorbate aiirl Antimycin A on Cyclic Phosphorylation Knder Argon and Air Ascorbate concentration /imoles of ATP formed/mg Bchl/hr Argon Air antimycin A antimycin A present present none 182 0.4 37 0.7 10-5 M 248 0.4 57 0.6 10-4 M 332 1.3 86 0.4 10-3 jyj 373 2.0 160 1.0 10-- iM 178 2.2 224 1.2 50 ^g bacteriochlorophyll was used in all cases. Other components of the photophosphorylation reaction mixture were the sameas given in Table 1. Where indicated, 10 ^g of antimycin A was added. 182 METABOLISM AND PHYSIOLOGY Table 5 shows that the addition of DPNH2 fully restored the phos- phorylating activity of aged chromatophores. Although DPNH2 is a stronger reducing agent than ascorbate, it appears that the addition of DPNH2 alone, even at a concentration as high as 1,7 x 10-2 M, cannot result in a pronounced overreduction of the phosphorylating system. TABLE 5 Effect of DPNH9 oil Cyclic Photophosphorylation by Fresh and Aged Chromatophores _, . , TAT^-NTTT 4- + " //moles ATP formed/mg Chromatophores DPNH2 concentration ^ r hl/h Fresh none 135 " 3.3 X 10""^ M 137 Aged none 69 II 3.3 X 10-6 M 107 M 3.3 X 10-5 ^ 138 It 3.3 X 10-4 ]^ 167 II 3.3 X 10-3 jyj 162 It 1.7 X 10-2 M 140 The reaction mixture for fresh chromatophores lacked the hexokinase sys- tem but contained instead 10 ^ moles A DP. Other conditions and components of the phosphorylation reaction mixture were the same as given in Table 1. Overreduction with DPNH2, at the relatively low concentration of 3,3 X 10-3 y[^ ^a^s observed, however, in the presence of methyl viologen (Table 6), Similar results were obtained in the presence of benzyl viologen, but DPIP was not effective. Table 6 also shows that the inhibitory effect on photophosphorylation by overreduction with ascorbate was markedly enhanced by the presence of DFIP or phenazine methosulfate. Overreduction by H2 plus hydrogenase was greatly enhanced by the addition of methyl (or benzyl) viologen (Table 6); in this system, as with DPNH2, DPIP was much less effective. That these dyes have indeed enhanced overreduction is indicated by the re- versibility of their effects on admission of air to the system (Experi- ments B and C in Table 6), These results suggest that these dyes have a strong affinity for the phosphorylating system of chromatophores into which they promote an electron flow not only from a moderately reactive reductant such as DPNH2 but also from the more reactive ascorbate. The affinity of phenazine methosulfate and DPIP for the phosphory- lating system of chromatophores is so strong that, under certain con- ditions, they catalyze cyclic photophosphorylation in the presence of CYCLIC AND NONCYCLIC PHOTO PHOSPHORYLATION 183 TABLE 6 IitJiibitioii of Cyclic Pliotophosphorylation by Dyes under Reducing Conditions „ . , rr. . . « moles ATP formed /mg Expei'iment Treatment ^ r hi /h A 3.3 X 10-4 M DPNH2 137 10-4 M methyl viologen 154 DPNH2 + methyl viologen 7 B 6.7 X 10-3 jYI ascorbate 239 6.7 X 10-5 M 2,6-dichlorophenol indo- phenol (DPIP) + 6.7 x 10-3 ^ ascorbate 9 ascorbate, DPIP, air 240 C 10-4 jvi phenazine methosulfate (PMS) 314 10-4 M PMS + 6.7 X 10-3 m ascorbate 9 PMS, ascorbate, air 240 D H2 146 H2 + hydrogenase 55 H2 + hydrogenase + 10-4 m methyl viologen 2 A 16-day old chromatophore preparation was used in Experiments B and C. Other conditions and components of the phosphorylation mixture were the same as given in Table 1. Anaerobic conditions were employed except as otherwise indicated. antimycin A (10,23), It seems likely that DPIP, like phenazine metho- sulfate (10), acts as a b3rpass agent around an antimycin A- sensitive site in cyclic photophosphorylation (cf, 23), The ability of these dyes to act as bypass agents for the antimycin A- sensitive site of cyclic photophosphorylation is influenced by the redox status of the system. As shown in Table 7, phenazine metho- sulfate did not bypass antimycin A inhibition in the presence of 5 X 10-3 M ascorbate. Likewise, DPIP failed to catalyze an antimycin A-insensitive phosphorylation in the presence of 10-3 m ascorbate. The influence of ascorbate concentration on the bypass effect of DPIP is of special relevance to our subsequent discussion. As shown in Table 7, DPIP catalyzed an appreciable antimycin A-insensitive cyclic photophosphorylation in the presence of lO"'^ M ascorbate. However, this cyclic photophosphorylation was abolished at concentra- tions of ascorbate of 10-3 m or higher. Attention is called to these experimental conditions, since, as will be shown later, an antimycin A-insensitive noncyclic photophosphorylation was measured in the presence of DPIP but at concentrations of ascorbate at which cyclic photophosphorylation is excluded (Table 7), 184 METABOLISM AND PHYSIOLOGY TABLE 7 Reactivation of Autiuiyciii A -Inhibited Cyclic Photophosphorylatioii by Dyes ^ . ^ m 4. ^ Per cent of Experiment Treatment phosphorylating activity A Control Phenazine methosulfate (PMS) PMS + 5 X 10-3 jvi ascorbate 6 96 3 B Methyl viologen (MV) MV + DPIP + 6.7 X 10-3 M ascorbate 1 145 C DPIP DPIP + 10"^ M ascorbate DPIP + 10-4 jyi ascorbate DPIP + 10-3 M ascorbate 2 1 22 2 The phosphorylating activity of a similar system without antimycin A was designated as 100 percent. 10 jugoi antimycin A was included with all the treat- ments shown above. Where the dyes were added their respective concentrations were: phenazine methosulfate, 6.7 x 10-5 M; methyl viologen, 10-4 M; 2,6- dichlorophenol indophenol (DPIP), 6.7 x lO'^ M. Fresh chromatophores were used in Exper. B, 4-day old chromatophores in Exper. A and 30-day old chro- matophores in Exper. C. Other conditions and components of the phosphorylation mixture were the same as given in Table 1. II. PYRIDINE NUCLEOTIDE REDUCING SYSTEM IN CHROMATOPHORES Frenkel (19) and Vernon and Ash (20,21) have shown that chroma- tophores of R. rubrum can photoreduce DPN by succinate, reduced FMN or the ascorbate-DPIP couple. We have confirmed the photo- reduction of DPN, using succinate or the ascorbate-DPIP couple as the electron donor (22), The reduction of DPN by chromatophores is a photochemical reaction; our attempts to replace light with ATP were unsuccessful. DPNH2 is more reducing than the electron donors used, and photons absorbed by the photosynthetic pigments supply the addi- tional energy needed to drive the electron transfer against the thermo- dynamic gradient, by what appears to be a light-induced noncyclic electron flow. Washed chromatophores were found to be highly specific in their ability to photoreduce DPN; TPN was not reduced. The addition of ferredoxin from R. nihruyn, Chromatium, or spinach (27) did not change the rate or specificity of DPN photoreduction. However, we found that washed chromatophores were able to photoreduce TPN as rapidly as DPN when the reaction mixture included a water-soluble extract of R. rubnim cells (the supernatant solution from the chroma- tophore preparation). CYCLIC AND NONCYCLIC PHOTO PHOSPHORYLATION 185 The TPN- reducing factor in the water-soluble extract was found to be heat stable and was identified as DPN by paper chromatography after isolation withtheaidof aDuolite A-2 resin (28). Fig, 1 shows that chromatophores photoreduced TPN in the presence of either catalytic amounts of DPN or the boiled water-soluble extract of R. rubnim cells. The fact that TPN reduction was mediated by catalytic amounts of DPN indicated that the chromatophores contained a transhydrogenase. This conclusion is supported by the results shown in Fig. 2, Transhy- drogenase activity in washed chromatophores was determined by meas- uring the TPN reduced in the dark (as a change in optical density at 340 m^) in the presence of an added alcohol dehydrogenase system and catalytic amounts of DPN. TPN was not reduced without DPN. DPN was added either as the pure chemical or as a cell extract, i.e., the boiled and Duolite A 2-treated supernatant solution from the chro- matophore preparation. III. NONCYCLIC PHOTOPHOSPHORYLATION As already mentioned, we have interpreted the photo reduction of DPN as evidence for a noncyclic electron flow in R. rubrum chroma- tophores. To demonstrate noncyclic photophosphorylation in chroma- tophores, it was necessary to establish that the photoreduction of DPN was accompanied by ATP formation under conditions such that ATP formation by cyclic photophosphorylation was excluded. Effect of antimycin A Photoreduction of DPN by succinate is inhibited by antimycin A (22). However, antimycin A does not inhibit the photoreduction of DPN by the ascorbate-DPIP couple (22). It should be noted that the photore- duction of DPN by the ascorbate-DPIP couple in the presence of antimycin A, which we reported previously (see Fig. 2 in ref. 22), oc- curred at a concentration of ascorbate (6.7 x 10-3 m) at which cyclic photophosphorylation catalyzed by DPIP is suppressed (Table 7), Thus, the ATP formation (shown again in Table 8) which could not have occurred via cyclic photophosphorylation because of the high ascorbate concentration, could only have resulted from noncyclic photophosphorylation. Requirement for electron donor and acceptor Table 8 shows ATP formation by chromatophores in the presence of antimycin A, DPIP, and 6.7 x 10-3 m ascorbate. Appreciable photo- phosphorylation occurred only in a complete noncyclic electron trans- port system, i,e,, in the presence of both an electron donor (ascorbate + DPIP) and an electron acceptor (DPN), Little photophosphorylation occurred when either the electron donor or the electron acceptor, or both, were omitted. Table 8 also shows that little photophosphorylation 186 METABOLISM AND PHYSIOLOGY I I OJ Photoreduction of TPN via DPN ( R. rubrum chromatophores) + DPN (catalytic) boiled extract TPN only 9 m 10 20 30 40 50 minutes Fig. 1. Photoreduction of TPN via DPN by R. riibnou chromatophores. The reaction mixture included, in a final volume of 3 ml, chromatophores (containing 30 /yg bacteriochlorophyll) and the following (in //moles): tris buffer, pH 7.9, 100; magnesium chloride, 5; sodium ascorbate, 20; 2,6- dichlorophenolindophenol, 0.2; TPN, 2. Also added was 0.05 /ymole DPN or boiled extract (see text) as indicated. The reaction was carried out in Thunberg type cuvettes at 20°C. TPN reduction was measured at 15 min intervals as the increase in optical density at 340 m//. Illumination, 10,000 lux. CYCLIC AND NONCYCLIC PHOTO PHOSPHORYLATION 187 0.5 S 0.4 I ^ 0.3 ^ 0.2 I 0.0 TRANSHYDROGENASE ACTIVITY in R. rubrum chroma tophores +0.05 /imofes DPN Fig. 2. Transhydrogenase activity in R. rubrio)! chromatophores. The re- action mixture contained, in a final volume of 3 ml, alcohol dehydrogenase, chromatophores (containing 32 ^g bacteriochlorophyll) and the following (in //moles): tris buffer, pH 7.9, 100; magnesium chloride, 5; ethanol, 170; TPN, 2. Also present was 0.05 //moles DPN or extract (see text) as indi- cated. The reaction was carried out in Thunberg type cuvettes at room temperature in the dark. occurred in a complete system in which the photoreduction of DPN was inhibited by the addition of the inhibitor phenyl mercuric acetate (22). Stoichiometry The stoichiometry of the ATP formed and the DPN reduced in the course of noncyclic photophosphorylation is shown in Table 9, The theoretical ratio of ATP/DPNH2 = 1 was obtained when the DPNH2 formed was trapped by an added lactate dehydrogenase system. With- out the lactate dehydrogenase system, the ratio ATP/DPNH2 was 188 METABOLISM AND PHYSIOLOGY TABLE 8 NoncycUc PJwtopliosphorylatioii Additions //moles ATP formed/mg Bchl/hr Complete system Ascorbate, DPIP.DPN 44.8 Electron acceptor omitted Ascorbate, DPIP 5.4 Electron donor omitted DPN 7.6 Electron donor and acceptor omitted none 5.0 Complete system, PMA Ascorbate, DPIP,DPN, inhibited PMA 6.2 All vessels contained 10 /ug antimycin A. The final concentrations of the added components were: ascorbate, 6.7 x 10"*^ M; 2,6-dichlorophenol indophenol (DPIP), 6.7 X 10-5 M; DPN, 6.7 x 10-4 M; and phenyl mercuric acetate (PMA), 10-4 M. Other conditions and components of the phosphorylating mixture were the same as given in Table 1. greater than one, suggesting that the chromatophores might have con- tained a DPNH2 reoxidizing system. DPN dependence Since the concept of noncyclic photophosphorylation (22) envisages that the ATP formed by this pathway is obligatorily coupled with an electron flow from the ascorbate-DPIP couple to DPN, ATP formation TABLE 9 StoicliioDietry of ATP and DPNHp Formed in Noncyclic Photophosphorylation In the presence of DPNH trapping system In the absence of DPNH trapping system Time ATP formed DPNH2 formed ATP formed DPNH2 formed (min) (//moles) (//moles) (//moles) (//moles) 10 0.38 0.35 0.16 20 0.60 0.57 0.55 0.29 30 0.81 0.73 — 0.43 40 0.97 0.93 0.95 0.53 The reaction mixture included, in a final volume of 3 ml, chromatophores containing 40//gof bacteriochlorophyll, and the following in //moles: tris buffer, pH 7.8,100; MgCl2, 5; ADP, 5; K2HP32O4, 5; ascorbate, 20; 2-6 dichlorophenol indophenol, 0.2; DPN, 5. Where the DPN trapping system was used, DPN was reduced to 0.2 //moles, and 10 //moles of pyruvate and 25 //g of lactate dehydro- genase were added. All vessels contained 10 //g of antimycin A. The reaction was run in Thunberg type cuvettes at 20°C in the light (10,000 lux). CYCLIC ANDNONCYCLIC PHOTO PHOSPHORYLATION 189 should stop when DPN is completely reduced. But such a DPN- depen- dent formation would not be expected if the ATP formed in the experi- ments represented by Tables 8 and 9 is the result of a DPIP-catalyzed cyclic photophosphorylation. Fig, 3 shows a dependence of ATP formation on DPN availability that is consistent with the mechanism of noncyclic photophosphoryla- tion, A limited amount of DPN (0,3/imoles) was used as the electron 1 1 1 1 1 Noncyclic photophosphorylation with linnited onnounts of DPN 1.0 y^ - 'S y/^ 1 0.8 DPN added y or A> regenerated / 1^^ ^ 0.6 — / 1 t/^..^^^---^^*'^^ ^ 0.4 - ^ '03/imo/es DPN _ 0.2 A , , , " 10 20 30 40 minutes 50 Fig. 3. Dependence of noncyclic photophosphorylation on DPN. The re- action mixture contained, in a final volume of 3 ml, 1 mg hexokinase, 10 ^g antimycin A, chromatophores (containing 45 /ig bacteriochlorophyll) and the following (in/Umoles): tris buffer, pH 7.9, 100; magnesium chloride, 5; ADP, 0.5; K2HP3204, 10; sodium ascorbate, 20; 2,6-dichlorophenol indophenol,0.2; DPN, 0.3. Additional DPN (2 /^moles) or a DPN regenerat- ing system (lactate dehydrogenase plus 10 /imoles sodium pyruvate) was added at the time indicated by arrow. 190 METABOLISM AND PHYSIOLOGY acceptor for noncyclic photophosphorylation. After this quantity of DPN was completely reduced, the rate of ATP formation greatly de- creased. However, when substrate amounts of DPN or a DPN- regenerating system was then added to the reaction mixture, the rate of ATP formation became greatly accelerated. Effect of aging Another line of evidence which supports the existence of both cyclic and noncyclic photophosphorylation in chromatophores is the differ- ential stability of these two systems to aging. As was already shown in Table 5 and is again demonstrated in Table 10, chromatophores re- tained a capacity for cyclic photophosphorylation after at least 30 days of storage; such decrease in activity as occurred during this period was fully restored by the addition of ascorbate. By contrast, Table 10 shows that chromatophores stored for 30 days lost completely the capacity for noncyclic photophosphorylation and that this loss was irreversible. TABLE 10 Loss of Noncyclic Photophosphorylatioi in Aged Chronatophores /ymoles of ATP formed/mg Bchl/hr fresh aged system i^^^^^^.^^^.^ chromatophores chromatophores Photophosphorylating AHHf fresh aged Noncyclic ascorbate, DPIP, 43 2 DPN, antimycin A Cyclic none 133 41 Cyclic ascorbate (5 X 10-3 _ I35 M) Inhibited cyclic antimycin A (10 ^g) 9 5 Noncyclic photophosphorylation was carried out under the same conditions as given in Table 8. Experimental conditions and reaction mLxture for cyclic photo- phosphorylation are given in Table 1. We have suggested elsewhere (4,5) that cyclic and noncyclic photo- phosphorylation in chloroplasts and, by extension, in chromatophores, share a common site for ATP formation. If this hypothesis is correct, it would follow that the loss of noncyclic photophosphorylation on stor- age is the result of the inactivation of the DPN- reducing system without which ATP formation could not occur by the noncyclic pathway, but which would not affect ATP formation by the cyclic pathway. This interpretation is consistent with the experimental findings, Chromato- phores kept at ca 4°C for a week lost more than 90 percent of their CYCLIC AND NONCYCLIC PHOTO PHOSPHORYLATION 191 capacity for DPN reduction but, as mentioned previously, they retained their capacity for cyclic photophosphorylation even after storage for 30 days. DISCUSSION The results of this investigation confirm and extend the previous finding of noncyclic photophosphorylation in chromatophores of R. rubrum (22). The capacity of R. rubrum chromatophores to catalyze a noncyclic electron flow, i.e., a light-driven, "uphill," unidirectional electron transfer against the thermodynamic gradient, has already been seen in the photoreduction of DPN by the ascorbate-DPIP couple (19-21). Our experimental findings provide evidence that this non- cyclic electron flow in chromatophores is coupled with ATP formation under experimental conditions which exclude ATP formation by cyclic photophosphorylation . Noncyclic photophosphorylation is distinguished from cyclic photo- phosphorylation in chromatophores by its joint dependence on an ex- ternal electron donor system (ascorbate-DPIP couple) and an external electron acceptor (DPN). Losada et al. showed (13) that a similar "bacterial type" of noncyclic photophosphorylation, in which the ascorbate-DPIP couple is the electron donor and TPN is the electron acceptor, can be carried out by spinach chloroplasts once the use of the natural electron donor for chloroplasts, water (0H~), is experi- mentally suppressed. Bose and Gest have recently argued (23) that the noncyclic photo- phosphorylation which we have previously found in R. rubrum chroma- tophores is, in fact, a cyclic photophosphorylation catalyzed by the dye DPIP, which acts as a bypass for the antimycin A- sensitive site. They explain the joint requirement for an added reductant (ascorbate + DPIP) and oxidant (DPN) as resulting from "their action in estab- lishing a redox environment which permits efficient operation of cyclic LIP [photophosphorylation]" (23). The experiment reported by Bose and Gest (Exp. II, Table 6 in ref. 23) which comes closest to ours was carried out under hydrogen gas, in the presence of 1 /imole DPN, 0.2 //moles DPIP and 0.2 fimoles of ascorbate (in 3 ml), i.e., under conditions where the system was not overreduced. The relevance of this experiment and the accompanying arguments to our previous experiments and to those reported now is not apparent. Our experiments (for example. Table 2 in ref. 22) were carried out under argon gas, in the presence of 2 //moles DPN, 0,2 //moles DPIP, and 20/ 6- type cytochrome 1 ^ c-type cytochrome — *- Additional evidence for the participation of cytochrome c^in bacterial photophosphorylation was recently obtained by Horio and K amen (4), and of both cytochrome b and cytochrome c^ by Nishimura (5). Another similarity became apparent, when one of us (H. B.) obtained results indicating that a flavoprotein functions as electron carrier in the cyclic photophosphorylation system oiR. riibrum (6). In brief, the evidence was: 1) stimulation of photophosphorylation by rather high concentrations of FAD, 2) inhibition by rather low concentrations of atebrin and 3) reversal of the atebrin-induced inhibition by rather high concentrations of FAD. It was also shown that both the basal rate of photophosphorylation and the stimulation caused by added FAD were strongly inhibited by HOQNO and antimycin A (Table 1 in reference 6). This eliminated the possibility that a nonphysiological bypass had occurred from the added FAD to cytochrome c^ (Horio and Kamen (7) have recently emphasized that such bypass reactions may occur at high concentrations of FAD). Our data thus suggested resemblance between bacterial photophosphorylation and mitochondrial oxidative phosphorylation at the flavin level, giving experimental support for the occurrence of the following electron transport sequence in both systems: HOQNO — >-flavoprotein — > 6- type cytochrome 1 ^c-type cytochrome — ► The first direct evidence for a photochemical reaction sequence c-type cytochrome •►chlorophyll was obtained with the photosynthetic bacterium Chromatium by Chance and Nishimura (8), who demonstrated that this reaction is temperature- independent from 80° to 298°K. Recently, Clayton (9) reported that PHOTO PHOSPHORYLATION AND ELECTRON TRANSPORT 197 coenzyme Q can be photochemically reduced and thus may be a pri- mary electron acceptor in isolated chromatophores of photosynthetic bacteria. The presence of the electron carrier ferredoxin (the plant enzyme has long been known as photosynthetic pyridine nucleotide reductase or methemoglobin reducing factor) in photosynthetic bacteria has been reported by Tagawa and Arnon (10). To further compare bacterial photophosphorylation with mitochondrial oxidative phosphory- lation, one may endeavourtostate that the site of cytochrome c oxidase in oxidative phosphorylation is occupied by chlorophyll in bacterial photophosphorylation, where the pigment in the light may serve as a "cytochrome c^ photooxidase," While the general picture on the oxidizing side of the chain today appears to be quite similar to its mitochondrial counterpart, the present knowledge about the electron transport sequence at the reducing side is still rather limited. For example: is ferredoxin or coenzyme Q or another compound the primary acceptor of electrons from chlorophyll, or may more than one agent act in this capacity? The present uncertainty in this region is reflected in the multiple choices of our very tentative scheme for the electron transport re- actions in cyclic photophosphorylation of chromatophores from R. rubrum: Z? :--- Light ferredoxin- flavoprotein- ^ ^ , ■ -c-tvpe cytochrome-chlorophyll 6-type cytochrome -^ ^ pyridine nucleotide Recently one of us (H. B.) has tried to outline what similarities and differences one may find between the above and various other electron transport pathways (11), That outline and the above discus- sion may be taken as an expression of our conviction that the variabil- ity observed between different systems for biological electron trans- port is, in the final analysis, due to nothing but variations on a general and basically common theme. ABOUT THE PHOSPHORYLATION REACTIONS The fact that one may inhibit the "physiological" cyclic electron transport chain in chromatophores from R. rubrum with site-specific 198 METABOLISM AND PHYSIOLOGY agents such as HOQNO (2) and antimycin A (12) and obtain a bypass around the site of inhibition by adding phenazine methosulfate (12,13) (PMS, methyl-phenazonium sulfate) has made it possible to investigate the question of whether one or several phosphorylation sites are linked to the electron transport. Our evidence for the presence of two sites in the "physiological" chain and one in the "PMS" chain has been given (14-17), The discussion of the phosphorylation reactions will be limited to some possible implications of the results which we obtained with the uncoupling agent valinomycin, which is known to uncouple the phosphorylation at all three phosphorylation sites in the electron trans- port system in animal mitochondria (18), but which appears to uncouple only one of two existing phosphorylation sites in cyclic photophosphory- lation. Fig. 1 shows our general interpretation of the effects obtained with valinomycin on bacterial photophosphorylation (14-16), If this inter- pretation is correct, there is a valinomycin-sensitive phosphorylation site in a region of the cyclic electron transport chain containing only dark electron transport reactions and a valinomycin-insensitive phos- phorylation site in a region which contains the photochemical reactions and possibly one or several dark electron transport steps. It may well be more than a coincidence that, when sensitivity to valinomycin is used as an indicator, the bacterial photophosphorylation reactions are similar to those in oxidative phosphorylation in the re- gion where only dark reactions occur, but different in the region where the photochemical reactions aie localized. Obviously, a possible site (The arrows indicate the direction of the electron transport. A, B, C and D indicate physiological elec- tron carriers.) Chlorophyll Valinomycin-sensitive phosphorylation linked to the region: Valinomycin-insensitive phosphorylation linked to the region: D — ^Chlorophyll — ^A — ^B Fig. 1. Interpretation of the effects obtained with valinomycin on bacterial photophosphorylation. PHOTO PHOSPHORYLATION AND ELECTRON TRANSPORT 199 for this difference in a photosynthetic system is in the region of its unique component, the chlorophyll. The fact that a valinomycin- insensitive phosphorylation appears to be localized in the same part of the cyclic electron transport pathway as the chlorophyll brings into focus a recent suggestion by Calvin (19) that chlorophyll might be a site for an ATP-producing reaction pattern. He considered the pos- sibility that light energy would allow the 9-10 enol in chlorophyll to react with orthophosphate and that a subsequent dehydration reaction would give the necessary energy- rich configuration O I I II C=C-0-P-OH 6" to allow a phosphorylation of ADP to ATP, It is tempting to speculate that if such a reaction pattern does exist the rate of phosphate addition to the enol group and thus also the rate of ATP-formation at the chlorophyll level could be determined by the rate of electron transport at another part of the chlorophyll molecule (for example, at the 5 -bridge carbon atom (20)) over the conjugated double-bond system. Phosphorylation would here not be "coupled" to electron transport in the classical sense, but only "linked" to it to give a rate-dependency. The phosphorylation reactions at such a site may well have properties which are quite different from an ordinary elec- tron transport-coupled phosphorylation, as, for example, insensitivity to valinomycin, CONCLUDING REMARKS Our knowledge today about the electron transport and phosphoryla- tion reactions in bacterial photophosphorylation is far from complete. If indeed, asour evidence indicates, there are two phosphorylation sites in the cyclic electron transport chain oiR.rubrum, and if a difference in response to valinomycin means a basically different reaction pat- tern in an energy- transfer step, then it would seem to be important that the material under investigation should be as active and intact as possible in order to minimize erroneous results due to partial and perhaps selective inactivation. REFERENCES 1. Frenkel, A.W., Light induced phosphorylation by cell-free preparations of photosynthetic bacteria. J. Am. Chem. Soc, 76, 5568 (1954). 2. Smith, L., and Baltscheffsky, M., Respiration and phosphorylation in ex- tracts of Rhodospirillum rubrum. Fed. Proc, 15, 357 (1956). 200 METABOLISM AND PHYSIOLOGY 3. Smith, L., and Baltscheffsky, M., Respiration and light-induced phosphory- lation in extracts of RhodospiriUum nibrion. J. Biol. Chem., 234, 1575 (1959). 4. Horio, T., and Kamen, M. D., Optimal oxidation-reduction potentials and endogenous co-factors in bacterial photophosphorylation. Biochemistry , 1, 144 (1962). 5. Nishimura, M., Studies on the electron-transfer systems in photo synthetic bacteria. IL The effect of heptylhydroxyquinoline-N-oxide and antimycin A on the photosynthetic and respiratory electron-transfer systems. Biochim. Biophys.Acta, 66, 17 (1963). 6. Baltscheffsky, H., Flavin nucleotides and light-induced phosphorylation in cell-free extracts of RhodospiriUum riibnim. Biochim. Biophys. Acta, 40, 1 (1960). 7. Horio, T., and Karaen, M. D., Observations on the respiratory system of RhodospirilliDu riibrum. Biochemistry, 1, 1141 (1962). 8. Chance, B., and Nishimura, M., On the mechanism of chlorophyll-cytochrome interaction: The temperature insensitivity of light-induced cytochrome oxi- dation in Chromatium. P roc. Natl . Acad. Sci. U. S., 46, 19 (1960). 9. Clayton, R. K., Evidence for the photochemical reduction on coenzyme Q in chromatophores of photosynthetic bacteria. Biochem. Biophys. Res. Com- mun., 9, 49 (1962). 10. Tagawa, K., and Arnon, D. I., Ferredoxins as electron carriers in photo- synthesis and in the biological production and consumption of hydrogen gas. Nature, 195, 537 (1962). 11. Baltscheffsky, H., Elektrontransport-kopplad fosforylering i biologiska, speciellt fotosyntletiserande system. Svensk Kem. Tidskr., 75, 17 (1963). 12. Geller, D. M., Photophosphorylation by RhodospiriUum rubrum. p. 73 in Abstracts, Vllth. Int. Congress for Microbiology, Stockholm, 1958. 13. Baltscheffsky, H., and Baltscheffsky, M., On light-induced phosphorylation in RhodospiriUum rubrum. Acta Chem. Scand., 12, 1333 (1958). 14. Baltscheffsky, H., Baltscheffsky, M., and Arwidsson, B., On electron trans- port and phosphorylation in plant and bacterial light-induced phosphoryla- tion. Acta Chem. ScaiuL, 14, 1844 (1960). 15. Baltscheffsky, H., Electron transport and phosphorylation in light-induced phosphorylation, p. 431 in Biological structure and function (T. W. Good- win and O. Lindberg, eds.), Vol. 2. Academic Press Inc., New York and London, 1961. 16. Baltscheffsky, H., and Arwidsson, B., Evidence for two phosphorylation sites in bacterial cyclic photophosphorylation. Biochim. Biophys. Acta, 65, 425 (1962). 17. Baltscheffsky, H., Baltscheffsky, M., and Olson, J. M., The quantum effi- ciency of ATP production in bacterial light-induced phosphorylation. Bio- chim. Biophys. Acta, 50, 380 (1961). 18. McMurray, W. C, and Begg.R. W., Effect of valinomycin on oxidative phos- phorylation. Arch. Biochem. Biophys., S4, 546 (1959). 19. Calvin, M., Evolutionary possibilities for photosynthesis and quantum con- version, p. 23 in Horizons in Biochemistry (M. Kasha and B. Pullman, eds.). Academic Press Inc., New York and London, 1962. 20. Katz, J. J., Thomas, M. R., and Strain, H. H., Site of exchangeable hydrogen in chlorophyll a from proton magnetic resonance measurements on dcuterio- chlorophyll a. J. Am. Chem. Soc, 84, 3587 (1962). LIGHT-INDUCED AND DARK STEPS OF BACTERIAL PHOTO PHOSPHORYLATION MITSUO NISHIMURA Joluisoii Researcli Foundation , University of Pennsylvania. Philadelphia and Department of Biophysics and Biochei)iistry, Faeulty of Science, Unircrsitv of Tokyo, Tokyo INTRODUCTION The analysis of light- induced and oxygen- activated absorption changes in purple bacteria (1-9) and studies of heme proteins isolated from these organisms (10-17) have indicated that several electron transfer catalysts are functioning in light- and oxygen- activated oxidation- reduction systems. If photophosphorylation is coupled with the oxidation- reduction reactions between these electron carriers, there should exist at least two distinct phases in the process of photo- phosphorylation, namely, a light-induced primary step and light- independent dark processes (electron transfer and ATP synthesis). The flashing light technique has been effective in studying the kinetics of photophosphorylation and distinguishing between the primary light-induced step and the dark processes. By this technique, it was observed that photophosphorylation took place in the dark after short, flashing illumination. This "delayed" process of photophosphorylation was affected by certain reagents and temperature. Analysis of the delayed photophosphorylation, combined with spectroscopic studies of the cytochrome system, revealed the presence of two steps in the delayed photophosphorylation; viz., electron transfer and coupled phosphorylation. When the dark periods between flashes are sufficiently long, the amount of delayed photophosphorylation per flash is indicative of the amount of substance reacting during the flash. The maximal amount of the delayed process per flash is unaffected by temperature or chemical reagents. From the maximal amount of delayed photophos- phorylation observed, a value of two was suggested as the tentative number of ATP molecules formed per electron transfer in the oxidation- reduction chain. The comparison of relative quantum effi- ciencies in the presence of inhibitors and an activator (MPM) is also suggestive of two phosphorylating sites. Some of the data in this paper have been published (18-20), 201 202 METABOLISM AND PHYSIOLOGY MATERIAL AND METHODS Chromatophores of the purple bacterium Rhodospirillum ruhrum were isolated as reported previously (18). The rate of photophos- phorylation was measured under near-infrared illumination with a recording pH meter (21), or by phosphate analysis in the Reaction medium (18). The light source used was a direct-current incandescent lamp used in conjunction with an infrared filter (Wratten 88A) and a water layer (5 cm thick). This filter combination passed near-infrared illum- ination longer than 720 m/y in wavelengths. The light intensities used were all above the level of saturation under continuous illumination, unless otherwise stated. Flashing illumination was furnished by a rotating sector (18). For the experiments on the maximum amount of ATP formation by a single flash, a xenon flash tube (flash duration 0.5 msec) was trig- gered at intervals of 60 sec. The light from the flash tube was filtered through two Wratten 88A infrared filters. RESULTS Presence of delayed photophosphorylation after flash In the first type of experiments, the duration of light period was kept constant at 1,45 msec, and lengths of dark periods were changed. When the rates of phosphorylation were expressed in terms of phos- phate esterified per illuminated time, we observed a remarkable rise in the rate of phosphorylation with increasing dark period, indicating the presence of delayed photophosphorylation after the flash. From the curves (rate per illuminated time vs. dark period), we calculated by differentiation the rate of phosphorylation after the flash, as shown in Fig, 1, It was found that the decay of delayed photophosphorylation was dependent on light intensity. The half-life of decay was shorter with low intensity flashes and was longer with strong flashes. Light-induced phase of pJiotophosphorylation The rapid light-induced reaction which takes place in the short flash was studied in the second type of experiment, where sufficiently long dark periods were inserted between short flashes (181-1449 /isec), and the duration of a light-dark cycle was kept constant at 8.70 msec. The amount of phosphate esterified per minute per bacteriochlorophy 11 was plotted against the flash duration (Fig, 2), When the length of flashes was sufficiently short, the rate of phosphorylation was pro- portional to the flash length. The rate was greater under flashes of high light intensity, but under continuous illumination the rate was identical with these three light intensities (indicated in Fig. 2 by 8.70 msec light PHOTO PHOSPHORYLATION IN FLASHING LIGHT 203 APj/min/BCHL 8 6 n Rate Of Phosphorylation After Flash /^^-^ 1.45msec. Light Period ^ ^ \ '' \ \ • ■^ \ • \ \^ T \32,000lux + 88A \ \ 16,000 lux ♦88 A ^^.^^ \ \\ ^^^^ \ \ \ ^^ \4,000\ \ \B8A N^ ^--"^^^^ ill Dork*^ '\8.000lux*88A 20 25 30 35 msec. Fig. 1. Rate of phosphorylation after flash (Pi esterified in moles/min/mole bacteriochlorophyll). Type A illumination, 26°C. APi/min/BCHL 6 32,000 IUX + 88A 2 X ^/"Te.ooo ^^^ -^ / ^^^ lux* ^^ 88A^-^ ■ / ^'' X e.TOmsec. Cycles // // /^,000 IUX + 88A 27* /// 8 msec Light Period Fig. 2. Rate of photophosphorylation (Pi esterified in moles/min/mole bacterio- chlorophyll) and length of light period. Type B illumination, 27°C. 204 METABOLISM AND PHYSIOLOGY period). The initial tangents of the curves (ATP formation rate vs. flash duration) were proportional to light intensity. From these experi- ments it is indicated that the extent of the primary photochemical reaction is proportional to the amount of energy in the short flashes, but utilization of the first chemical product (s) is a dark process and requires a longer time than its production. Rate Of Phosphorylation After Flash A Pi/min/BCHL 8,000 lux + 88A 1.45msec il 28« 3.l6xlO"®M BCHL./H0QN0 = 2I8 15msec. Fig. 3. Effect of HOQNO on rate of phosphorylation after flash (Pi esteri- fied in moles/min/mole bacteriochlorophyll). # — #, no HOQNO; O — Q 3.16 X 10-8 M HOQNO. Type A illumination, 8000 lux f 88A filter, 28°C. Effect of HOQNO, ^ MPM and temperatures on the delayed photophos- phorylation The rates of delayed phosphorylation in the presence (3.16 x 10"^ m) and absence of HOQNO are compared in Fig. 3. It is apparent that the rate of decay of delayed photophosphorylation is markedly lowered by HOQNO. The lower level of phosphorylation under continuous illu- mination, as well as the slower decay of delayed photophosphorylation 1 Abbreviations: HOQNO, 2-^/-heptyl-4-hydroxyquinoline-iV-oxide; MPM, meth- ylphenazonium methosulfate ("phenazine methosulfate"). PHOTOPHOSPHORYLATION IN FLASHING LIGHT 205 in the presence of HOQNO, is probably due to inhibition of electron transfer between cytochromes b and c (8), The rate of photophosphorylation is markedly increased by MPM under sufficient illumination (22,23), The rates of phosphorylation after the flashes were calculated from a flashing light experiment. A remarkable acceleration of decay of delayed phosphorylation by the added MPM was noticed. The effect of MPM on the kinetics of delayed photophosphorylation was like a reversal of the effect caused by HOQNO. APj/min/BCHL 8 8.70msec. Cycles 16,000 IUX + 88A Control msec. Light Period Fig. 4. Effect of MPM on photophosphorylation in Type B illumination. Ordinate: Pi esterified in moles/min/naole bacteriochlorophyll. Abscissa: length of light period. no MPM: -,1.1 X 10--^ M MPM. 16,000 lux + 88A filter, 28°C. In the second set of experiments, the effect of MPM on the light- induced phase of photophosphorylation was investigated. The rate of phosphorylation was proportional to the length of the flash when the flash duration was sufficiently short. The tangents of these two curves were identical in the absence and presence of MPM (Fig. 4). 206 METABOLISM AND PHYSIOLOGY These data clearly indicate that MPM does not affect the primary light-induced phase of photophosphorylation. The kinetics of photophosphorylation under intermittent illumination at two different temperatures (15° and 26 °C) were also studied. The half- life of delayed photophosphorylation was lengthened at the lower temperature. It is concluded that a low rate of photophosphorylation at a low temperature under continuous illumination results from a slower dark process. Amount of ATP synthesis caused by a single flash The area covered by the curve of delayed photophosphorylation corresponds to the number of ATP molecules formed per chlorophyll molecule after a single flash. When the amounts of total delayed photophosphorylation at different temperatures were compared, it was found that the two values were approximately the same, though the rate of photophosphorylation under continuous illumination was lower at lower temperatures. The results are shown in Table 1. The presence of reagents such as HOQNO or MPM caused a marked change in the rate of photophosphorylation, but the amounts of total delayed photophosphorylation after a single flash in the presence and absence of these reagents were approximately the same (Table 1). These results indicate that the substances accumulating during the short flash are consumed through dark processes, and phosphorylation takes place accompanying the dark processes. Different temperatures, HOQNO, and MPM do not affect the rapid photochemical process which takes place during the flash. These factors influence the rate of dark reactions of photophosphorylation. When the dark period is sufficiently long, the total amount of delayed process is determined by the amount of first product formed by a photochemical process during flash, hence changes in rate of the dark process would have little effect on the amount of total delayed photophosphorylation. This concept agrees well with other data that indicated the presence of three steps for photo- phosphorylation; i,e,, a rapid photochemical process, a second slower process of electron transfer, and a third process of phosphorylation coupled with the second process. The maximum amount of ATP synthesis per flash was determined by combining the methods employing the sensitive recording pH meter and the xenon flash tube illumination (20), An example of pH record- ings is shown in Fig, 5, The single flash yield of photophosphorylation at the saturation level and the rate of phosphorylation under continuous illumination are tabulated in Table 2, Single flash yields are expressed as molar ratios by comparison with the bacteriochlorophyll and cyto- chrome concentrations. PHOTO PHOSPHORYLATION IN FLASHING LIGHT 207 -^ Ij fl ci n. -I CD ^ III- I 5-2 So "S o s= ? _ u ex w Tti ^ '* CO t~ CO c^ ■— I LO M LO CO o o o o o o XX X x: CO O] (N (N 00 00 00 00 00 00 00 00 00 00 00 00 § g § § II I o o o o o o o o o o o o 00 00 00 00 00 CQ UO g o ^ ffi CO S --I CO LO 00 00 00 00 lO UO (M ^ CVl (M oq (M eg eci Si OS ° !h ^ ni O 5-1 C c3 O a " o o 5 ^ * aw 0) 0) _ 'a; ^^ fi ^H ^(. d -S * >N 3 M *j m _o ^ ^ ° 13 ri T3 !=l T3 a> d rt 208 METABOLISM AND PHYSIOLOGY TABLE 2 Single flash yield of photophospJiorylatioii and rate of pliotophospJiorylation under continuous illumiiiatioii (both under saturating light intensities) Exp. no. Single flash yield AATP/BChl (mole/mole) AATP/cyto- chrome (mole/mole) Rate under contin- uous illumination AATP/min/BChl (mole/min/mole) 1 2 3 Mean 0.048 0.037 0.056 0.047 0.92 0.71 1.07 0.90 6.05 4.90 8.18 6.38 A: CONTINUOUS ILL. B: FLASHING ILL. Fig. 5. Recordings of pH change by illumination of R. ruhruni chromatophores. 25°C. A: continuous illumination, 8,000 lux + 88A filter, 0.1 pH unit full scale, rate of photophosphorylation -- 6.05 ATP/minA>acteriochlorophyll. B: flashing illumination, 3.58 x 10-2 jouie/cm2/flash + 88A filter, 0.01 pHunit full scale. Relative efficiency of light-energy utilization in different systems of phosphorylation The comparison of photophosphorylation rates for different phos- phorylating pathways under low intensity continuous illumination indi- cates the relative quantum efficiency of such systems. In this series of experiments the following four systems were studied under condi- tions of limiting light intensity: (a) untreated chromatophores, (b) PHOTOPHOSPHORYLATION IN FLASfflNG LIGHT 209 + MPM, (c) + MPM + antimycinA, (d) + MPM + HOQNO. The concen- trations of HOQNO and antimycin A were chosen to give both 100 per cent inhibition of photophosphorylation in the absence of MPM, and at the same time to give almost full recovery of phosphorylation when the inhibitor was added with MPM (under high light intensities). AATP/min/BCHL 15 10 + MPM \.ZOx\0-*M ^ — ' — ° — / ^-'" " ♦HOQNO 7.9lxlO-«M / ''^^ ^ + MPM ISOxlO-^M / ''' y^^^ + Ant. AII2xI0"''M / j^/ ^ " h> '/^ / . 4 6 Klux ♦ 88A Filter 10 Fig. 6. Rates of photophosphorylation in different systems. Ordinate: Pi esterified in moles /min /mole bacteriochlorophyll. Abscissa: light intensity. A: untreated chromatophores, B: + 1.30 x 10-4 M MPM, C: + 1.30 X 10-4 M MPM + 1.12 x 10-7 m ^timycin A. D: + 1.30 x 10-4 M MPM + 7.91 x 10-8 M HOQNO. Temperature 24°gs; The rates of photophosphorylation in these four systems in the low light intensity range are shown in Fig. 6. Curves A, B, C, and D cor- respond to systems (a), (b), (c), and (d) described above. The compari- son of tangents of these curves at zero light intensity shows the rela- tive quantum efficiencies for the different systems. It is seen from 210 METABOLISM AND PHYSIOLOGY the figure that the efficiency is slightly lowered in the presence of MPM as compared to untreated chromatophores. In the systems (c) and (d), where reactions of cytochromes of b and c types are inhibited REL. ACT. • ^ --- ^/^PM 130x10- ■^M A / / o / HOQNO 7.9lxlO~^M _ MPM I.30xl0-'»M / — ^ ^ ^ . — o- Ant A II2> lO-^M / MPM 1.30* IO"'*M ■ 4 6 8 Klux + 88A Filter 10 12 Fig. 7. Comparison of phosphorylative systems at different light in- tensities. Curves indicate ratios of phosphorylation rates. A: MPM- catalyzcd phosphorylation/normal phosphorylation. B: phosphorylation of MPM-HOQNO system/MPM-phosphorylation. C: phosphorylation of MPM-antimycin A -system/MPM-phosphorylation. Concentrations of reagents are the same as in Fig. 6. Temperature 24°C. and the flow of electrons is bypassed by MPM, the efficiencies are about half of the original value. This means that one of the phosphory- lating sites is located close to the oxidation- reduction site of cyto- chromes h and c, and bypassing this site with MPM reduces the PHOTO PHOSPHORYLATION IN FLASHING LIGHT 211 quantum efficiency to one half. Knowing the maximum amount of de- layed photophosphorylation from a single flash, we postulated two sites of phosphorylation on the electron transport chain in R. rubrum chromatophores as a tentative value (see Discussion), The absolute quantum efficiency measurement of photophosphorylation (24) and valiomycin experiments (25) by Baltscheffsky and others are generally in good agreement with our present studies. The operation of the different phosphorylating sites is a function of light intensity. For example, the MPM activation of photophosphoryla- tion is more marked under high light intensities. Likewise, the re- covery of antimycin A or HOQNO inhibition by MPM is greater under higher light intensities. These results are summarized in Fig. 7. Curve A is the ratio of MPM- catalyzed phosphorylation to normal phosphorylation. Curves B and C are ratios of HOQNO- (or antimycin A-) MPM phosphorylation/MFM phosphorylation, respectively. The concentrations of MPM, HOQNO, and antimycin A were identical to those in Fig. 6. These facts suggest that in low light intensity experi- ments the untreated chromatophores are more efficient in the energy utilization than other systems. However, MPM-bypassed electron transfer becomes greater under higher intensities, and the overall rate of phosphorylation can be high in the presence of MPM even if the normal electron transfer is blocked and one of the phosphorylating sites is lost. DISCUSSION It is concluded that the photochemical and dark processes in photo- phosphorylation can be separated by the technique of flashing illumina- tion. The first photochemical step is rapid and occurs only during il- lumination. The rate of steady state phosphorylation is limited by the rate of dark process, possibly by the rate of electron transfer. The amount of total delayed photophosphorylation is proportional to the amount of light absorbed when flashes are sufficiently short. The first step can be the formation and accumulation of oxidized cytochrome (which is rapid and temperature independent) (1-9) and some unknown reduced substance. The reduction of pyridine nucleo- tide is less rapid than cytochrome oxidation (26). The second process (light- independent phase of photophosphorylation) would include the transfer of electrons from the reduced low- redox potential system to the oxidized high- redox potential system (oxidized cytochrome) and the associated reactions which lead to phosphorylation of ADP. MPM and higher temperatures accelerate the decay of delayed photophos- phorylation, whereas HOQNO retards this decay. Under continuous illumination, higher temperatures and MPM raise the level of steady state photophosphorylation; HOQNO lowers the level. These data can 212 METABOLISM AND PHYSIOLOGY be explained in terms of an increase or decrease of electron flux at a bottleneck point. This point is most likely the HOQNO- and antimycin- sensitive site (between cytochromes b and c?). Evidence supporting this conclusion is as follows: (a) the reaction between cytochromes b and c is inhibited by HOQNO or antimycin A in light- and oxygen- activated oxidation- reduction reactions (8), (b) photophosphorylation is inhibited strongly by HOQNO or antimycin A (22,23), (c) inhibition of photophosphorylation by HOQNO or antimycin A is largely diminished in the presence of MFM (22,23), (d) the rate of photophosphorylation increases in the presence of MPM under sufficiently high light in- tensities, and (e) the decay of delayed photophosphorylation is acceler- ated by MPM and is decelerated by HOQNO. The mean value for the maximum single flash yield of ATP syn- thesis, 0.047 ATP/bacteriochlorophyll, is much higher than the amount of total delayed photophosphorylation by a single flash of light intens- ity of 32,000 lux (+ infrared filter) in the repeating flash experiments. For the maximal amount of delayed photophosphorylation, the light intensities used in the repeating flash work were apparently not suf- ficient. Since ATP synthesis during the flash is negligible as compared to the total delayed photophosphorylation when the flash is short and the light intensity is high, the maximum single flash yield obtained by the xenon flash can be regarded as equal to the amount of delayed photophosphorylation. The use of the infrared flashes for the activa- tion of chromatophores excluded the possibility of participation of carotenoids as the primary light- absorbing pigments. Therefore, the number of light quanta absorbed by chromatophores is limited by the number of bacteriochlorophyll molecules (except for nonspecific ab- sorption by chromatophore materials). If we assume that the maximal yield of ATP/bacteriochlorophyll is attained when all the bacterio- chlorophyll molecules are excited by the infrared flash, the minimum quantum yield for the delayed photophosphorylation will take the same value as the maximal value of single flash yield per bacteriochloro- phyll, i.e,, ATP/ht^ = ATP/bacteriochlorophyll = 0.047. There have been many discussions concerning the primary photo- chemical reaction in photosynthesis (6,26-31). Except for activated electronic states of assimilatory pigments, the first chemical process which takes place in bacterial photosynthesis is probably the light- induced oxidation of cytochrome. The rapidity and the temperature independence of the process suggest that the oxidation of cytochrome takes place during the short illumination and the rest of the photo- synthetic reactions proceeds in the dark. As the mechanism of photo- phosphorylation in photosynthetic bacteria, the following scheme seems most feasible. BChl + Cytochrome c (pe^ 7 ^BChf + Cytochrome c (Fe"^^) (l) PHOTOPHOSPHORYLATION IN FLASHING UGHT 213 BChf + Y + H^ ^BChl + HY (2) Cytochrome c (fc'^^) + HY ^Cytochrome c (fb^^) + Y + H^ (3) m^ADP + Pi + hyC')^ >.. m^ATP + H^o) (4) In this scheme, Eq, 1 indicates the photochemical oxidation of cyto- chrome c, and Eq, 2 shows the formation of the primary reduced sub- stance HY. Eqs. 3 and 4 indicate electron transfer between the oxidized cytochrome c and the reduced substance in the oxidation- reduction chain and the coupled phosphorylation. The maximal yield of delayed photophosphorylation found in this work is 0.90 ATP/cytochrome (Table 1). This figure is calculated on the basis of total cytochrome concentration in the chromatophores. If we assume the light-induced oxidation of c-type cytochrome only, the yield (ATP/photochemically oxidized cytochrome) becomes higher. It is calculated to be around 2 on the basis of relative concentrations of different heme protein species in R. rubruni cells (9), if the saturating yield is obtained when the cytochrome c is fully oxidized by the strong flash. This value (ATP/photochemically oxidized cytochrome = ~2) corresponds to the yield of ATP formation per electron transferred in the oxidation- reduction chain of chromatophores (factor w in Eq. 4). This suggests that the number of phosphorylating sites in the redox chain (Eq. 3) is probably two if the two sites are located in series on the chain. Yet this is a rather tentative value for the number of ATP molecules synthesized per electron transferred by the redox chain, and it remains to be scrutinized further. The increased rate of elec- tron transfer at the rate-determining site resulting from the addition of MPM would lead to an increase in the rate of overall electron trans- port, followed by an increased rate ofphotophosphorylation.lt must be noted, however, that in the presence of MPM (and HOQNO or anti- mycin A) the probable loss of one of the phosphorylating sites is expected. The lowering of the quantum efficiency of photophosphoryla- tion by the addition of these reagents (Baltscheffsky, Baltscheffsky and Olson, 24, and our present data) suggests the loss of a phosphory- lating site. The amount of total delayed photophosphorylation per flash was not appreciably affected by temperatures, MPM, or HOQNO when the dark periods were sufficiently long. Other data indicate that these factors affect only the dark steps of photophosphorylation. When the dark period is sufficiently long, the amount of total delayed process is de- termined by the amount of the primary product formed by the photo- chemical process during flash. Therefore, it is understood that the changes of the rate of dark processes (consumption of the primary 214 METABOUSM AND PHYSIOLOGY products and coupled phosphorylation) cause little effect on the amount of total delayed photophosphorylation, though the time required for the completion of the dark reactions changes markedly. REFERENCES 1. Duysens, L. N. M., Reversible photo-oxidation of a cytochrome pigment in photosynthesizing Rho do spirillum nibmm. Nature, 173, 692 (1954). 2. Chance, B., and Smith, L., Respiratory pigments of RhodospirilliDU rnbriou. Nature, 175, 803 (1955). 3. Chance, B., Oxygen-linked absorbancy changes in photosynthetic cells. Brookhaven Symp. Biol., 11, 74 (1959). 4. Smith, L., and Ramirez, J., Absorption spectrum changes in photosynthetic bacteria following illumination or oxygenation. ^ re//. Biochem. Biophys., 79. 233 (1959). 5. Olson, J. M., and Chance, B., Oxidation-reduction reactions in the photo- synthetic bacterium Chromatiuni. Arch. Biochem. Biophys., 88, 26 (1960), ibid., p. 40. 6. Chance, B., and Nishimura, M., On the mechanism of chlorophyll-cytochrome interaction: the temperature insensitivity of light-induced cj^tochrome oxi- dation in Chromatium. Proc. Natl. Acad. Set. U. S., 46, 19 (1960). 7. Nishimura, M.,and Chance, B., Studies on the electron transfer systems in photosynthetic bacteria. I. The light-induced absorption spectrum changes and the effect of phenylmercuric acetate. Biochim. Biophys. Acta, 66, 1 • (1963). 8. Nishimura, M., Studies on the electron-transfer systems in photosynthetic bacteria. II. The effect of heptylhydroxyquinoline-N-oxide and antimycin A on the photosynthetic and respiratory electron-transfer systems. Biochim. Biophys. Acta, 66, 17 (1963). 9. Nishimura, M., and Chance, B., Studies on the electron-transfer systems in photosynthetic bacteria. III. Spectroscopic studies of cytochrome sys- tems, p. 239 in Studies on Microalgae and Photosynthetic Bacteria (Japa- nese Soc. of Plant Physiol., ed.). University of Tokyo Press, 1963. 10. Vernon, L. P., and Kamen, M. D., Hematin compounds in photosynthetic bacteria. J. Biol. Chem., 211, 643 (1954). 11. Kamen, M. D., and Vernon, L. P., Comparative studies on bacterial cyto- chromes. Biochim. Biophys. Acta, 17, 10 (1955), 12. Bartsch, R. G., and Kamen, M. D., On the new heme protein of facultative photoheterotrophs. J. Biol. Chem., 230, 41 (1958). 13. Newton, J. W.,and Kamen, M. D., Chromatium cytochrome. Biochim. Bio- phys. Acta, 21,11 (1959). 14. Morita, S., Crystallization of /?//0(fo/7S'//?> r/n^?chromatophores by flashing light. Biochim. Biophys. Ada, 57, 88 (1962). PHOTOPHOSPHORYLATION IN FLASfflNG LIGHT 215 19. Nishimura, M., Studies on bacterial photophosphorylation. II. Effects of reagents and temperature on light-induced and dark phases of photophos- phorylation in Rliodospirillum rubruin chromatophores. Biochim. Biophys. Acta, 57, 96 (1962). 20. Nishimura, M., Studies on bacterial photophosphorylation. IV. On the max- imum amount of delayed photophosphorylation induced by a single flash, Biochim. Biophys. Acta, 59, 183 (1962). 21. Nishimura, M., Ito,T.,and Chance, B., Studies on bacterial photophosphor- ylation. III. A sensitive and rapid method of determination of photophos- phorylation. Biochim. Biophys. Acta, 59, 177 (1962). 22. Baltscheffsky, H., and Baltscheffsky, M., On light-induced phosphory- lation in Rliodospirillum rubnim. Acta Chem. Scand., 12, 1333 (1958). 23. Geller, D. M.,and Lipmann, F., Photophosphorylation in extracts of Rliodo- spirillum rubnim. J. Biol. Chem., 235, 2478 (1960). 24. Baltscheffsky, H., Baltscheffsky, M,, and Olson, J. M., The quantum effi- ciency of ATP production in bacterial light-induced phosphorylation. Bio- chim. Biophys. Acta, 50,380 (1961). 25. Baltscheffsky, H., and Arwidsson, B., Evidence for two phosphorylation sites in bacterial cyclic photophosphorylation. Biochim. Biophys. Acta, 65, 425 (1962). 26. Chance, B., and Olson, J. M., Primary metabolic events associated with photosynthesis. Arch. Biochem. Biophys., 88, 54 (1960). 27. Duysens, L. N. M., Transfer of excitation energy in photosynthesis. Druk- kerij enUitgevers-Maatschappif v/h Keminken Zoon N. V., Utrecht. (1952), 28. Calvin, M., Energy reception and transfer in photosynthesis. Rev. Mod. Phys., 31, 147 (1959). 29. Tollin, G., Solid-state phenomena and the primary quantum conversion process of photosynthesis. Brookhaven Symp. Biol., 11, 35 (1959). 30. Witt, H. T., and Moraw, R., Untersuchungen fiber die Primarvorgange bei der Photosynthese. Z. Phys. Chem. N. F., 20, 253 (1959), ibid., p. 283. 31. Arnold, W., and Clayton, R, K., The first step in photosynthesis: Evidence for its electronic nature. Proc. Natl. Acad. Sci. U. S., 46, 769 (1960). THE EFFECT OF UBIQUINONE2 ON PHOTOPHOSPHORY- LATION IN PARTICLES OBTAINED FROM RHODOSPIRILLUM RUBRUM GROWN IN MEDIA CONTAINING DIPHENYLAMINEl HARRY RUDNEY Deportment of Biochemistry, School of Medicine Western Reserve University, Clevclmid 6, Ohio It has previously been shown that cells of R. rubrum grown in the presence of diphenylamine contain greatly reduced amounts of ubi- quinoneio (coenzyme Qio)^ (1,2). Chromatophore particle preparations from these cells (DPA particles) carry on photophosphorylation at a lower rate than particles from normally grown cells. The DPA parti- cles require a catalytic amount of a reductant in order to maintain cyclic photophosphorylation. It was observed (2) that photophosphory- lation in the presence of 1 mM succinate was greatly stimulated by the addition of UQ2, however with lower concentrations of succinate (0.05 mM) addition of UQio completely inhibited phosphorylation. The stimu- lation of photophosphorylation by UQ2 was completely abolished by antimycin. Since reporting these results many experiments have been per- formed to elucidate the site of these various effects. The results were often puzzling in that it was soon found that depending on con- ditions one could obtain either stimulation or inhibition of light- induced photophosphorylation (LIP), and there was a great variation between preparations. It gradually became clear that what we were witnessing was the extreme sensitivity of the DPA chromatophore system to changes in redox balance brought about by the various agents added, i.e., there appeared to be very little poising capacity of the redox systems in these preparations. These experiments will be fully reported elsewhere but a few typical experiments shown in the follow- ing tables will suffice to show that UQ2 can overcome the inhibitory effect of antimycin merely by changes in the redox balance of the 1 This investigation was supported in part by a Public Health Service Research Career Program Award (GM-K-3-993-C3). 2 The following abbreviations are used: UQio = ubiquinoneio, UQ2 = ubiqui- none2, suffix indicating number of isoprenoid units in side chain; LIP = light- induced phosphorylation; DPA particles - chromatophore particles from cells grown in diphenylamine; PMS = phenazine methyl sulfate. 217 218 METABOLISM AND PHYSIOLOGY system. These results are in agreement with those reported by Kamen and collaborators (3,4), Vernon and Ash (5), and Geller and Lipmann (6) over the past few years. In particular, these results support the recent finding of Bose and Gest (7) showing that dyes such as DPIP may, under certain conditions, overcome an antimycin inhibition of light-induced, cyclic electron transfer. As shown in Table 1, when DP A particles are incubated with suc- cinate, one always observes a stimulation of LIP by the addition of catalytic amounts of UQ2. In this particular preparation PMS also greatly stimulated LIP, When succinate and PMS are both present one observes an inhibition which could be considered to result from the over- reduction of some carrier or of the dye. Similarly, in the case where UQ2 stimulated LIP, an oxidation of some component occurred. In agreement with this concept is the fact that UQ2 will relieve some of the inhibition due to over- reduction by succinate (compare vessels 3 and 6, Table 1), Further corroboration appears from a comparison of vessels 5 and 6, Table 1, Here again it is reasonable to consider that succinate is acting as an inhibitor by over- reducing a component of the system. In the presence of antimycin, LIP stimulation by succinate is com- pletely blocked, whether UQ2 is present or not, PMS- stimulated LIP TABLE 1 The Effect of Succinate, PMS, UQ2 and Antiniycin on LJP in DPA Particles Vessel No. Additions No Antimycin With Antimycin 1 Succinate 4.1 2 PMS 15.0 9.2 3 Succinate + PMS 4.7 2.3 4 Succinate <- UQ2 13.0 5 PMS + UQ2 23.2 16.5 6 Succinate + PMS + UQ2 12.8 10.6 Chromatophore particle preparations were prepared from RhodospiriUum nibruni grown in diphenylamine as previously described (2). Experiments were carried out as follows in Warburg manometer vessels. The main compartment contained 20 ^moles MgCl2,35 yWmoles KH2P04,pH 7.0,100 ^moles giycylgly- cine buffer, pH 7.4,40 //molesof glucose, 0.5 mgof hexokinase (Sigma type III), 0.8 mg of chromatophore particle protein and additions as indicated above. 10 //moles of ADP were in the side arm. Total volume, 2.5 ml. The cups were filled in the dark and gassed with pre-purified nitrogen for 5 minutes. Then the ADP was tipped and the light (1200 ft-candles on each vessel) was turned on. At the end of the incubation period (usually 45 minutes or 1 hour) the reaction was stopped and phosphate uptake determined as previously described (2). The fig- ures in the table represent micromoles of phosphate esterified in 1 hour. The additions consisted of 2 //moles succinate, 0.03 mg PMS, 3 fug antimycin, 0.08 //moles of UQ2 in 5 microliters alcohol. PHOTO PHOSPHORYLATION IN R. RUBRUM 219 is, however, inhibited about 38 per cent. This result is in agreement with the commonly observed bypass of antimycin inhibition by PMS (6). When succinate and PMS are added together one observes again an appreciable inhibition in the presence of antimycin which is by- passed in the presence of UQ2 (compare vessels 3 and 6, Table 1), Thus it would appear that UQ2 by affecting the redox balance of the system even in the presence of antimycin can stimulate or inhibit LIP, This concept is further borne out by a comparison of vessels 2 and 5 incubated in the presence of antimycin. Here again one could reason that UQ2 has affected the redox balance of the system being acted on by the light- reduced PMS, in the pathway bypassing antimycin, so that stimulation of LIP is obtained. If one now compares vessels 5 and 6 incubated in the presence of antimycin, one must conclude that suc- cinate has affected the redox balance towards over- reduction so that inhibition of LIP occurs. Similar results to the foregoing can be obtained from experiments using the ascorbate-DPIP couple which has been used in studies on photooxidation and photoreduction in chromatophore particles (3,5,8). Although the LIP of DPA particles is completely inhibited when high concentrations of ascorbate (4 mM) are used, it is maintained when the concentration is low (0,02 mM), as shown in Table 2. Addition of UQ2 or DPIP leads to complete inhibition of phosphorylation, again pre- sumably due to alteration in the redox balance of the system. Addition of UQ2 to the ascorbate-DPIP couple leads again to the establishment of a new redox balance which is favorable for increased LIP, Similarly, addition of DPIP to the ascorbate- UQ2 couple leads to the same effect. In the presence of antimycin the stimulating effect of ascorbate alone is completely abolished. In this sense the DPA particles behave dif- ferently from normal particles where in some cases ascorbate- induced LIP appears to bypass antimycin inhibition (7,8), Upon the TABLE 2 The Effect of Ascorbate, DPIP and UQ2 on LIP in DPA Particles Vessel No. Additions No Antimycin With Antimycin 1 Ascorbate 5.2 2 Ascorbate + UQ2 3 Ascorbate + DPIP 1.4 4 Ascorbate + DPIP + UQ2 9.2 3.8 Conditions same as for Table 1. Figures represent micromolesof phosphate taken up in 1 hour, using 0.6 mg of chromatophore particle protein. The addi- tions, where indicated, consisted of 0,5 //moles of ascorbate, 0.2 //moles of DPIP and 0.08 //moles of UQ2. 220 METABOLISM AND PHYSIOLOGY addition of DPIP it can be observed that the inhibitory effect of anti- mycin with ascorbate alone is partially relieved, and furthermore the inhibitory effect of UQ2 is bypassed, just as in the case without anti- mycin (compare vessels 2 and 4, Table 2). On the other hand, a com- parison of vessels 3 and 4 also shows that UQ2 acts to bypass an antimycin inhibition of LIP with the ascorbate-DPIP couple. The results outlined in Tables 1 and 2 indicate that, even in the presence of antimycin, agents affecting redox balance can produce stimulation or inhibition of LIP, depending on the particular redox balance established in the system. These findings extend the observa- tions of Bose and Gest (7) on the effect of DPIP in bypassing an in- hibition of electron transport by antimycin. The stimulatory effect of UQ2 on LIP in the presence of succinate appears to be due to an effect on the redox balance of an electron carrier in the particles. In this connection, it is of interest that succinate readily reduces UQ;i^o ^^ the dark in chromatophore particles (10). In the case of the stimulation by UQ2 in the presence of PMS the effect may also possibly be directly related to the new redox balance established as a result of interaction of UQ2 with the light- reduced dye. It seems reasonable to speculate that the effects observed with UQ2 and PMS in the presence of anti- mycin are due to this dye interaction and may not necessarily involve direct reaction with a second electron carrier site different from that where UQ2 and succinate interact. The quinone normally present in R. rubnim is UQiq- Since the Eq of UQ2 would not be expected to vary greatly from the +0.100+0.01 volts calculated for UQ^q (9), it could be assumed that the redox bal- ance at the UQjo site in the cyclic electron transport scheme would be directly affected by the addition of UQ2.^ On the basis of E^ values UQj^O would fit in somewhere between Rhodospirillum heme protein (+0,01 volts) and cytochrome c^ (+0.310 volts). Since the DPA particles have greatly reduced amounts of UQj^g ^he addition of UQ2 would cer- tainly influence the redox balance to a much greater extent than in the case of the normal particles, where little effect of UQ2 can be ob- served. UQio is insoluble in aqueous media, which may explain why the addition of this substance has a negligible effect on LIP. As in the case of PMS, it is difficult to distinguish whether the effects of UQ2 on the interaction of the ascorbate-DPIP couple in LIP are due to direct reaction with these reagents or to some action on an electron transport carrier in the particles. Nonetheless, a most important point resulting from the foregoing experiments, and one which requires emphasis, is that evidence for the existence of noncyclic photophosphorylation in bacterial systems (8) based on the inhibitory effect of antimycin must ^ Moret et al. (11) have measured the redox potential of a series of UQ analogues with varying isoprenoid side chains. They found that the length of the isoprenoid chain did not influence the final value of +0.098 volts at pH 7.4. PHOTO PHOSPHORYLATION IN R. RUBRUM 221 be carefully examined, to exclude the possibility that various reducing and oxidizing agents have not bypassed the antimycin sensitive site by changes in the redox balanceof the system. Such changes could lead to the re- establishment of cyclic photophosphorylation. REFERENCES 1. Sugimura, T., and Rudney, H., The effect of aerobiosis and diphenylamine on the content of ubiquinone in Rhodospirillum riibnim. Biochim. Biophys. Acta, 62, 167-170 (1962). 2. Rudney, H., The stimulation of photophosphorylation by coenzyme Q2 and Q3 in chromatophores oi Rhodospirillum ruhnim. J. Biol. Chem., 236, PC 39-40 (1961). 3. Newton, J. W., and Kamen, M. D., Photophosphorylation by subcellular particles from Chromatiion. Biochim. Biophys. Acta, 25, 462-474 (1957). 4. Horio, T., and Kamen, M. D., Optimal oxidation-reduction potentials and endogenous co-factors in bacterial photophosphorylation. Biochemistry, 1, 144-153 (1962). 5. Vernon, L. P., and Ash, O. K., Coupled photooxidation and photoreduction reactions and associated phosphorylation by chromatophores of Rhodo- spirillum rubrum. J. Biol. Chem. 235, 2721-2727 (1960). 6. Geller, P. M., and Lipmann, F., Photophosphorylation in extracts of Rhodo- spirillum rubrum. J. Biol. Chem., 235, 2478-2484 (1960). 7. Bose, S. K., and Gest, H., Bacterial photophosphorylation: regulation by redox balance. Proc. Natl. Acad. Sci. U. S., 49, 337-345 (1963). 8. Nozaki, M., Tagawa, K., and Arnon, D. I., Noncyclic photophosphorylation in photo synthetic bacteria. Proc. Natl . Acad. Sci. U.S., 47, 1334-1340 (1961). 9. (a) Redfearn, E. A., The possible role of ubiquinone in the respiratory chain, p. 353 in Ciba Foundation Symposium, Quinoiies in Electron Trans- port (G. E.W. Wolstenholme, and C. M. O'Connor, eds.). Little, Brown and Co., Boston, 1961. (b) E. C. Slater, General Discussion, Ibid., p. 415. 10. Sugimura, T., and Okabe, K., The reduction of ubiquinone (coenzyme Q) in chromatophores of Rhodospirillum rubrum by succinate. J. Biochem., 52, 235-236 (1962). 11. Moret, v., Pinamonti, S., and Fornassari, E., Polarographic study on the redox potential of ubiquinones. Biochim. Biophys. Acta, 54, 381-383 (1961). PHOTOS YNTHE TIC PHOSPHORYLATION WITH BACTERIAL CHROMA TOPHORES: CATALYSIS BY A NATURALLY OCCURRING FACTOR (PHOSPHODOXIN)l C. C. BLACK and A. SAN PIETRO Charles F. Kettering Research Laboratory Yellow Springs, Ohio A light- induced phosphorylation of ADP with cell-free preparations from the photosynthetic bacterium Rhodospirillum ruhnim was first observed by Frenkel (1-3). This light- induced process, which has been termed photosynthetic phosphorylation, has subsequently been observed with chromatophores from Chromatiiim (4-6) and Chloro- hium (4), Photosynthetic phosphorylation by bacterial chromatophores occurs in the absence of an exogenous electron acceptor at a rate which is compatible with the growth rate of whole cells. This is in sharp contrast to the low endogenous rate of photosynthetic phosphory- lation with spinach chloroplasts (7,8), which is only one to two per cent of the photosynthetic capacity of intact spinach leaves. Whereas the endogenous rates observed with chromatophores and chloroplasts are markedly different, both may be increased in the presence of suit- able cofactors. For example, photosynthetic phosphorylation by chromatophores and by chloroplasts is stimulated about 10-fold (5,9) and as much as 2,000-fold (10). respectively, in the presence of the dye methyl phenazoniummethosulfate. It seemed reasonable, therefore, that the intact organism might contain some component (s) which isolated chromatophores and chloroplasts either lacked or contained in low concentration. Thus, a study of the natural factors involved in photosynthetic phosphorylation was initiated. A water-soluble, heat-stable factor which stimulates the rate of photosynthetic phosphorylation by spinach chloroplasts as much as 200-fold (8) has been isolated from both spinach leaves and chloro- plasts. Further study revealed the presence of a similar factor^ in many photosynthetic organisms, including bacteria (8). In this report, 1 Contribution No. 116 of the Charles F. Kettering Research Laboratory. Sup- ported in part by a research grant (GM 10129-01) from The National Institutes of Health, U. S. Public Health Service. 2 It is not known whether the factor isolated from different organisms is the same or different. 223 224 METABOLISM AND PHYSIOLOGY additional data will be presented concerning this factor, which we propose to call phosphodoxin, e.g., spinach phosphodoxin or Rho do - spirillum nibnim phosphodoxin. METHODS Experiments on light-induced formation of ATp32 were conducted with a 1 ml reaction mixture containing the following components in /imoles: Tris-HCl buffer, pH 7.8, 48; MgCl2, 2; ADP, 1; Pi + Pi32 (containing from 0.5 to 1 //curie), l;bacteriochlorophyllor chlorophyll, less than 30 /igrams. ATp32 -^^s assayed as previously described (8). In some experiments ATP was determined spectrophotometrically with glucose, hexokinase, NADP, and Zwischenferment. The methods used for preparation of chromatophores and chloroplast fragments and for chlorophyll determination were those previously reported (8). A method of isolating the naturally occurring factor has been described elsewhere (8), The reaction mixtures were illuminated laterally in 1-cm cuvettes at 2,500 foot-candles. Light intensity was varied by varying the distance of the reaction mixtures from the light source (11). RESULTS Photosynthetic phosphorylation with R. r^ftrz^w chromatophores was markedly stimulated by the addition of the factor isolated from whole cells of the same organism (Fig. 1). The formation of ATP was more linear with time and fell off slower in the presence of the factor than in its absence. One characteristic of the factor (s) is its ability to stimulate photo- synthetic phosphorylation by chromatophores, regardless of the or- ganism from which it is isolated, i.e., an algae, a higher plant, a flagellate, or a bacterium (8). Further demonstration of these cross- reactions between organisms is given in Table 1 and Figs. 2 and 3. The rates of endogenous photosynthetic phosphorylation of both spinach chloroplasts and R. nibrum chromatophores were stimulated by the addition of the factor (s) isolated from spinach, Chroniatiiun, or R. rubnim (Table 1). Photosynthetic phosphorylation by spinach chloro- plasts was strikingly stimulated by the factor isolated from Chroma- tium (Fig. 2), and likewise Chromatium chromatophores were stimu- lated by the factor isolated from spinach (Fig, 3). In the presence of the factor, anaerobic conditions did not affect the rates of photosynthetic phosphorylation observed with 72. rubnim chromatophores during a two-minute illumination (Fig. 4, upper two curves). With longer illumination periods, a slight stimulation occurred PHOSPHORYLATION CATALYZED BY PHOSPHODOXIN 225 < .3 + FACTOR / / / ENDOGENOUS ■0- 2 4 6 8 10 12 TIME IN MINUTES Fig. 1. Time course of the effect of the factor isolated from R. ruhriou on photo phosphorylation with R. nibrnui chromatophores. 226 METABOLISM AND PHYSIOLOGY TABLE 1 Activity of chroinatophores and chloroplasts with factor from other organisms. Activity is expressed as fimoles of ATP per mg bacteriochlorophyll or chlorophyll per hour Source of factor Rhodospirillum rubnim chromatophores Spinach chloroplasts Endogenous PI us factor Endogenous Plus factor Spinach 121 194 1.5 196 Chromatium strain D 47 125 1.5 115 Rhodospirilluni rtibnim 68 206 1.0 54 120 ^100 o ^ n \ \- _ 80 < >« ^ ,»_ Q. o o 60 U) o a> x: o E " 40 :t. o» E ^ 20 - ^^•^ - /•^ X .05 .10 .15 ml of CHROMATIUM Factor .20 Fig. 2. F]ffect of concentration of the factor isolated from Chroindliiiiii on photophosphorylation with spinach chloroplasts. PHOSPHORYLATION CATALYZED BY PHOSPHODOXIN 227 .5 1.0 1.5 2.0 mg.of Spinach Factor /ml of Reaction Mixture Fig. 3. Effect of concentration of the factor isolated from spinach on photophosphorylation with Chromatimn chromatophores. 228 METABOLISM AND PHYSIOLOGY Argon + Factor Aerobic -I- Factor Argon, Endogenous Aerobic, Endogenous D— 4 6 8 10 TIME IN MINUTES 12 14 Fig. 4. Effect of anaerobic conditions and time on photophos- phorylation with R. rubrnni chromatophores in the absence and presence of the factor isolated from R. riibrioH. PHOSPHORYLATION CATALYZED BY PHOSPHODOXIN 229 .08 .06 a> o 04 £ .02 / + FACTOR ENDOGENOUS / / J I i_ 25 50 75 100 125 /i.grams Bacteriochlorophyll / ml Fig. 5. Effect of bacteriochlorophyll concentration on photophosphorylationby R. r?<6rz<»7 chromatophores in the absence and presence of the factor isolated from R. rubnim. under argon, probably resulting from a stimulation of the endogenous ATP production under argon (Fig, 4, lower two curves). Linearity of ATP production by R. nihrum chromatophores with increasing bacteriochlorophyll concentration (up to 30 //grams) in the presence of the factor is indicated in Fig, 5, The usual response of photosynthetic phosphorylation with bacteria to increasing light intensity was observed (Fig, 6) with R. ruhrum chromatophores in the presence of the factor isolated from R. rubrum. Saturation was reached at 500 foot- candles in both the absence (unre- ported data) and presence of the factor. This is about double the sat- uration intensity previously reported for R. rubrum (3) and for 230 METABOLISM AND PHYSIOLOGY 200 400 600 800 LIGHT INTENSITY , FOOT CANDLES 1000 Fig. 6. Effect of light intensity on photophosphorylation by R. rnhrum chromatophores in the presence of the factor isolated from R. nibrnm. Chromatium (5), It is pertinent to point out that photosynthetic phos- phorylation by spinach chloroplasts does not show proportionality with light intensity below about 50 foot- candles (11), A broad pH optimum between 7.4 and 8,4 has also been observed with chromatophores in the presence of the factor, Horio and Kamen (12) have reported that the low photosjmthetic phosphorylation capacity of "washed chromatophores" could be restored to a maximal rate by the addition of "chromatophore washings," One component of the "chromatophore washings" was identified as cyto- chrome 02- Heating, aerating, or freezing the "chromatophore wash- ings" resulted in a loss of photosynthetic phosphorylation-activation capacity. The factor reported in this paper does not appear to be the same as those studied by Horio and Kamen, since it is not destroyed by these treatments and does not contain cytochrome c^. PHOSPHORYLATION CATALYZED BY PHOSPHODOXIN 231 CONCLUSIONS Photosynthetic bacteria contain a water-soluble, heat- stable factor (phosphodoxin) which stimulates photosynthetic phosphorylation by both chromatophores and spinach chloroplasts. Spinach chloroplasts contain a similar factor which stimulates photosynthetic phosphorylation by bacterial chromatophores. The stimulated reaction by bacterial chro- matophores is linear with time, bacteriochlorophy 11, and light intensity. Anaerobic conditions do not affect the stimulated reaction during short illumination periods, REFERENCES 1. Frenkel, A. W., Light-induced phosphorylation by cell-free preparations of photosynthetic bacteria. J. Am. Chem. Soc, 76, 5568-5569 (1954). 2. Frenkel, A. W., Photophosphorylation of adenine nucleotides by cell-free preparations of purple bacteria. J. Biol. Chem., 222, 823-834 (1956). 3. Frenkel, A. W., Light-induced phosphorylation by cell-free preparations of RhodospiriUum mbriDu. pp. 303-310 in Research in Photosynthesis (H. Gaffron et al., eds.). Interscience Publishers, Inc., New York, 1957. 4. Williams, A.M., Light-induced uptake of inorganic phosphate in cell-free extracts of obligately anaerobic photosynthetic bacteria. Biochim. Biophys. Acta, 19, 570 (1956)". 5. Newton, J. W., and Kamen, M. D., Photophosphorylation by subcellular par- ticles from Chromatiiim. Biochim. Biophys. Acta, 25, 462-474 (1957). 6. Anderson, I. C, and Fuller, R. C, Photophosphorylation by isolated chro- matophores of the purple sulfur bacteria. Arch. Biochem. Biophys., 76, 168-179 (1958). 7. Arnon, D. I., Allen, M. B., and Whatley, F. R., Photosynthesis by isolated chloroplasts. Nature, 174, 394-396 (1954). 8. Black, C. C, San Pietro, A., Limbach, D., and Norris, G., Photosynthetic phosphorylation catalyzed by factors isolated from photosynthetic organ- isms. Proc. Natl. Acad. Set. U. S., 50, 37 (1963). 9. Geller, D. M., and Gregory, J. D., Light-induced oxidation- reduction changes in Rhodospirillum rubrum extracts. Fed. Proc, 15, 260 (1956). 10. Avron, M., Photophosphorylation by swiss-chard chloroplasts. Biochim. Biophys. Acta, 40, 257-272 (1960). 11. Turner, J. F., Black, C.C., and Gibbs,M., Studies on photosynthetic proc- esses. I. The effect of light intensity on triphosphopyridine nucleotide re- duction, adenosine triphosphate formation, and carbon dioxide assimilation in spinach chloroplasts../. Biol. Chem., 237, 577-579 (1962). 12. Horio, T., and Kamen, M. D., Optimal oxidation-reduction potentials and endogenous cofactors in bacterial photophosphorylation. Biochemistry, 1, 144-153 (1962), Top: B. Chance; Bottom: a section of the audience. ■ra^ III ELECTRON TRANSPORT PHOTOOXIDATION AND PHOTOREDUCTION REACTIONS CATALYZED BY CHROMATOPHORES OF PURPLE PHOTOSYNTHETIC BACTERIA^^^ LEO P. VERNON Charles F. Kettering Research Laboratory, Yellow Springs, Ohio Photooxidation Reactions: Oxidation and reduction reactions are the essence of both plant and bacterial photosynthesis. The photosynthesizing cell utilizes the energy obtained from light to effect oxidation and reduction reactions directed primarily toward control of the oxidation level of carbon compounds and an associated production of ATP. It was early recog- nized that under the influence of light, cells of the photosynthetic bacteria had the ability to oxidize molecules of either inorganic sul- phur or organic compounds and reduce carbon dioxide in a coupled reaction (1). French demonstrated a photooxidation of ascorbic acid by bacterial extracts of RJiodovibrio (2), and later Vernon and Kamen reported a photooxidation of cytochrome c and DPIPH2^ by chromato- phores of Rliodospirillum rubnim in the presence of air (3). Since these reactions were performed aerobically, they were subject to the criticism that the photooxidation could represent a nonbiological oxidation catalyzed by the chlorophyll itself. Subsequently, however, it was demonstrated that the reduced forms of DPIP, cytochrome c, methylene blue, and indigo carmine were photooxidized by chromato- phores of R. rubnim in the absence of oxygen if alternate electron acceptors were added, such as fumarate (4), Lindstrom has extended the studies with DPIPH2 and has shown a coupling of DPIPH2 photooxidation with the photo reduction of sulfate 1 Contribution No. 107 from the Charles F. Kettering Research Laboratory. 2 The term "chromatophore" is used in this presentation to mean thephotosyn- thetlcally active particle or fragment which is obtained upon rupture of the intact cell by sonic oscillation. There is some question concerning the origin of these particles, but whether they exist as separate entities in the cell or are part of the cell membrane does not significantly affect the conclusions drawn from these experiments. 3 In addition to the standard abbreviations, the following are used: TMPD, N,N,N',N'-tetramethyI-p-phenylenediamine; MB, methylene blue; MBH2, re- duced form of MB; PMA.phenylmercuric acetate; HQNO, 2-heptyl-4-hydroxy- quinoline-N-oxide; DTNB, 5,5'-dithiobis (2-nitrobenzoic acid). 235 236 ELECTRON TRANSPORT (5). He has also studied the stability of the photooxidase system in R. rubrum chromatophores (6). In a series of investigations, which were designed primarily for studying the photo reduction system of chromatophores, Frenkel dem- onstrated that a photoreduction of NAD could be coupled to a photo- oxidation of reduced FMN (7). Succinate also supported the photore- duction of NAD, and a photooxidation of this compound was implied, Vernon and Ash also studied the photoreduction of NAD, which was coupled to a photooxidation of succinate (8) . Ample evidence has accumulated during the past several years showing that whole cells of the photosynthetic bacteria are capable of catalyzing a photooxidation of intracellular cytochrome components. Duysens (9) and Chance (10) observed an oxidation of cytochrome upon illuminating cells of R. rubrum under anaerobic conditions. This ob- servation was later confirmed by the experiments of Chance and Smith (11). The light-induced oxidation of intracellular cytochrome has been extensively investigated for the bacterium Chromatium (12,13,14). Since this photooxidation proceeds at temperatures as low as 80 °K, this reaction is probably one of the primary photoreactions taking place after absorption of a light quantum by the chromatophore (12). Photoreduction Reactions: In the intact cell, the reducing phase of photosynthesis is evidenced in terms of carbon dioxide reduction. At the chromatophore level, the earlier experiments of French (2) and Vernon and Kamen (3) showed a photoreduction of oxygen. A photoreduction of NADP by chromato- phores of R. rubrum was reported by Vernon (15). The experiments of Frenkel (7) showed that NAD is photo reduced by R. rubrum chroma- tophores coupled with a photooxidation of added FMNH2. Subsequent experiments (8) showed that NAD photoreduction could be coupled with succinate, and that NAD was the nicotinamide nucleotide of choice in these reactions. Nozaki et al. (16) demonstrated that NAD photore- duction could also be coupled to the oxidation of DPIPH2 in the presence of ascorbate. Other photo reductions observed with /^. rub- rum chromatophores have involved methyl red and tetrazolium blue (17), the disulfide DTNB (18) and sulfate ion (5). A summary of these photoreactions and the rates which have been observed to date is shown below in Table 6. The photoreduction of intracellular NAD by R. rubrum cells was shown in the investigation of Duysens and Sweep (19). A similar photo- reduction was observed with Chromatium cells by Olson (20), and Amesz (21) has recently completed an extensive investigation on NAD photoreduction with R. rubrum cells. Evidence has been presented for a photochemical reduction of ubiquinone contained within the chro- matophores of Chromatium and Rhodopseudomonas spheroides (22), PHOTOOXIDATION AND PHO TOR EDUCTION REACTIONS 237 Also, Nishimura has reported a photo reduction of cytochrome b in cells of R. nibnim which have been poisoned with antimycin A or HQNO (23,24). In this case, the cyclic electron transport system is apparently blocked, allowing a direct demonstration of cytochrome b photoreduction. PHOTOOXIDATION REACTIONS CATALYZED BY R. RUBRUM CHROMATOPHORES It is apparent that a number of photochemical oxidation and re- duction reactions are now available for use in investigation of the electron transport system contained in chromatophores of the photo- synthetic bacteria. I would like to present some detailed information on one of these reactions, namely the photooxidation of DPIPH2. A preliminary report of some of these data has already appeared, and the experimental methods used to obtain the data reported here are essentially those which were described in this previous communication (25). The experiments were performed under anaerobic conditions with chromatophores prepared by sonic oscillation followed with two wash- ings by centrifugation of the particles sedimenting in the centrifugal range of 20,000 to 100,000 x g. The use of a modified Spectronic 505 recording spectrophotometer (25) permitted the photooxidation of DPIPH2 to be followed in detail. Fig. 1 presents the results obtained with and without an added oxidant present. With only chromatophores present in the reaction system, a fast initial reaction was observed which saturated after two to three seconds of reaction time. The presence of either NAD or fumarate in the reaction system allowed a secondary slower reaction to take place following the initial fast reaction. In all cases a dark back- reaction was observed when the light was turned off. In the reaction system containing NAD, the NADH formed in the reaction was immediately converted back to NAD by means of an enzyme system consisting of lactic dehydrogenase and pyruvate. In the absence of this trapping system the secondary slow reaction was not observed, since there is an active NADH-DPIP diaphorase present in the chromatophores. The secondary slow reactions which are coupled to NAD and fumarate are no different from the coupled photoreactions previously observed (4, 8,16), both in mechanism and rate, as shown below. The most interesting aspect of Fig. 1 is the initial fast photooxida- tion of DPIPH2. The rapid and definite saturation of this reaction in- dicates that the oxidation of the DPIPH2 is coupled to the reduction of components contained within the chromatophore. Fig. 2 shows the re- lationship of this initial fast reaction to the concentration of chromato- phores contained within the reaction system. Not only is the rate of the initial fast reaction proportional to the chromatophore concentra- 238 ELECTRON TRANSPORT MINUTES Fig. 1, Photooxidation of DPIPH2 in the absence and presence of added oxidants. The basic reaction system contained 33 mM Tris buffer, 66 flM DPIPH2 (reduced with ascorbic acid until a faint blue color remained) andR. nibrum chromatophores equal to the concentra- tion of 0.22 mg BChl in a final volume of 8.0 ml (except for the case where no DPIPH2 was present, in which experiment 0.48 mg BChl was present). When fumarate was present the pH was 8.0 and the final concentration of fumarate was 0.75 raM. For the experiment with NAD, the pH employed was 8.5 and the system also contained 0.5 mM DPN, 1.3 mgof lactic dehydrogenase, and 3.1 mM sodium pyruvate. Anaerobic conditions were obtained by three evacuations with alternate flushing with argon gas. Experiments were carried out using Thunberg tubes which were modified by joining the main tube through a T-joint with 2 one-cm Pyrex absorption cells. These cells were spaced to fit into the cell holder of a Spectronic 505 recording spectro- photometer which had been modified to allow one arm of the reaction vessel to be illuminated by a tungsten lamp moimtcd outside the housing of the spectrophoto- meter. The light intensity reaching the reaction vessel was 1.05 X 106 ergs per second per square cm. The experiments were run at 30° C for the fumarate sys- tem, and 20°C for the NAD system. For further details concerning the chromatophore preparation procedure and other techniques employed see reference 25. PHOTOOXIDATION AND PHO TOR EDUCTION REACTIONS 239 1 i,09 - OI8mg bacteriochlorophyll 1 °« cn •^,0 7 - r !< 06 5 05 >. 04 o 1 03 S 02 < 1 ■ i 009 0045 1 2 T me in minutes Fig. 2. Influenceof/i. ra6r?/;» chromatophore concen- tration upon the initial fast photooxidation of DPIPH2. Experimental conditions as for Fig. 1. tion, but the extent of the reaction is also directly proportional to the chromatophore concentration, supporting the idea that the photooxida- tion of the DPIPH2 observed in the fast reaction is coupled to the photoreduction of components contained within the chromatophore. The stoichiometry of the reaction, as shown below, is also consistent with this hypothesis. The data presented in Fig, 1 are traces made of the actual record- ing and show the noise inherent in the system. In subsequent figures a smooth curve has been drawn over the original tracing, so that the noise is not apparent. Another electron donor which reacts in a similar manner in this system is N,N,N',N'-tetramethyl-/)-phenylenediamine, abbreviated TMPD, whose response is shown in Fig. 3. It resembles DPIPH2 in all respects, including the initial fast reaction and the saturation in the absence of an added oxidant. It differs only in that it appears to react faster than does DPIPH2. This compound is an excellent donor of electrons, and has been shown to react with chloranil (26), and with cytochrome c contained in mammalian mitochondrial systems (27), The ability of/?, nibrum chromatophores tophotoxidize the reduced forms of methylene blue and cytochrome c was reported earlier (4), The kinetics of such oxidations obtained under tlie present conditions are shown in Fig, 4, For these experiments the methylene blue was reduced enzymatically by succinate in the presence of the chromato- phores prior to illumination, and fumarate was then added to poise 240 ELECTRON TRANSPORT .5 - LIGHT OFF NAD + FUMARATE LIGHT ON 2 3 4 MINUTES Fig. 3. Photooxidation of TMPD by R. rubnan chromatophores. The experimental conditions were as given for Fig. 1, except that 0.1 mM TMPD was substituted for DPIPH2. The BChl concentration was 0.196 mg in the 8 ml reaction system. the system. Illumination in this case resulted in a fast reaction fol- lowed by a secondary slower reaction which was coupled to the photo- reduction of the added fumarate. When the light was turned off, a bi- phasic reaction was observed. The initial fast back-reaction can be correlated with reduction of methylene blueby the reduced components in the chromatophore, while the secondary back- reaction is due to the enzymatic reduction of methylene blue by the succinate present in the system. In the case of ferrocytochrome c, a reaction was observed in the absence of added oxygen, but the initial reaction was slow and the extent of the reaction was less than that observed with DPIPH2. In this case also, a coupled photooxidation could be obtained when fumarate was added to the system. These reactions observed are in agreement with the data on photooxidations previously reported (4), but the present experiments show there are two phases for the photooxidation of both these compounds. The response of ferrocyanide in the R. riibrum chromatophore system was examined, and the results of this experiment are shown in Fig. 5. Although some absorbancy change was noted when the system was illuminated, this was largely due to absorbancy changes which take PHOTOOXIDATION AND PHOTOREDUCTION REACTIONS 241 CYTOCHROME c Fig. 4. Photooxidation of reduced methylene blue and ferrocytochrome c by R. ntbruDi chromatophores. The experimental conditions were those given for Fig. 1, except that DPIPH2 was replaced with the oxidants listed. Ferrocyto- chrome c was prepared by reduction with borohydride. When present, it was 5 mg per 8 ml and the BChl concentration was 0,190 mg. For the experiment in- volving MB, 3.1 raM succinate was present initially to reduce the MB via the succinic dehydrogenase contained in the chromatophore particles. Following the enzymatic reduction of MB, sufficient fumarate was added to make the so- lution 0.75 mM in fumarate. For this experiment the BChl concentration was 0.128 mg. place within the chromatophores themselves at this wave length. There was no coupled reaction with either fumarate or NAD present in the system. The inactivity of ferrocyanide in the present case is somewhat surprising, since ferricyanide has been shown to interreact with bac- terial chromatophores in two ways. Goedheer has shown that ferricy- anide will cause a bleaching of bacteriochlorophyll in/?, rubnim, which corresponds to an oxidation of the pigment (28). Calvin and Androes have also reported that mixtures of ferro- and ferricyanide of different redox potentials influence the photoinduced ESR signal observed with chromatophores of R. nthniui (29). One possibility for the inactivity of ferrocyanide in the present reaction is that it may be reduced as rapidly as it is oxidized by the bacterial chromatophores. If the ferro- cyanide must react through the cytochrome c^, whose standard redox potential is below that of the ferrocyanide system, the resulting slow 242 ELECTRON TRANSPORT 4. e <\j '^^ ^LIGHT H .Ir ON >- o ^ LIGHT CHROMATOPHORES ALONE y ^^^ + NAD . DPIPH; < 03 CO CD < + FUMARATE , DPIPH2 2 3 MINUTES Fig. 5. Inability of R. nibru >n chromatophores to photo- oxidize ferrocyanide. The experimental conditions were the same as those given for Fig. 1, except that 1 mM potassium ferrocyanide was substituted forDPIPH2. The BChl concentration was 0.212 mg per 8 ml reaction sys- tem. Fig. 6. Photooxidation of DPIPH2 by R. riibrio)! chromatophores under aerobic conditions. The experimental procedures given for Fig. 1 were employed with air present. No oxidant was added. PHOTOOXIDATION AND PHO TOR EDUCTION REACTIONS 243 rate of photooxidation could easily be balanced by reduction reactions from the photoreduced chromatophore components. One of the early reactions observed with /?, rubnim chromsitophores was the photooxidation of ascorbate in the presence of DPIP and molec- ular oxygen (3), The ability of chromatophores to photooxidize re- duced dye in the presence of oxygen is shown in Fig. 6. The oxidation of the dye in the presence of oxygen is a very stable reaction and is not appreciably influenced by heating the chromatophores to 60°C. The stability of this system has been studied by Lindstrom (6), By de- creasing the chromatophore content in the reaction system, it was possible to observe the usual biphasic reaction shown for the other photooxidations. It is interesting, however, that the initial photooxida- tion rate in the presence of oxygen was significantly lower than that observed under anaerobic conditions. The reason for this is not im- mediately apparent. PHOTOOXIDATION REACTIONS CATALYZED BY CHROMATIUM AND RHODOPSEUDOMONAS SPHEROIDES CHROMATOPHORES Other photosynthetic bacteria were investigated to see if their photosynthetically active particles could also photooxidize DPIPH2 in a manner similar to that observed with R. nibnim. Fig. 7 presents the LIGHT E .2 •" 1 NO OFF OXIDANT V + FUMARATE 1 . H < > z L^^^ "^-^ g.2 NO 1 OXIDANT T + NAD \ m 1 < LIGHT ON -\ 2 4*' 6 8 10 12 14 MINUTES Fig. 7. Photooxidation of DPIPH2 by Chromatium chromato- phores. The experimental conditions given for Fig, 1 were em- ployed with Chromatium chromatophores being used at a con- centration equal to 0.264 mg BChl. 244 ELECTRON TRANSPORT data obtained with Chromatmm chromatophores. In the absence of added oxidants, an initial fast reaction was observed. Although the rate of the reaction was less than that observed with R. nibnim, the back- reaction was faster than that observed with R. rubrum. The slower initial photooxidation rates may merely reflect the fact that a faster back-reaction obtains with these bacterial particles. The most significant difference observed between R. nibnim and Chromatium lies in the fact that the latter particles are unable to couple the photooxidation of the reduced dye with either fumarate or NAD reduction, as shown in Fig. 7, To further check on this problem, corollary experiments were done in which NAD reduction was attempted in the presence of either succinate or ascorbate-DPIP, which systems were designed for detection of NADH accumulation. All experiments of this nature were negative. Chromatium chromatophores were tested for their ability to photo- reduce methyl red, which is active in the photo reduction system of R. rubrum chromatophores (17). Fig, 8 shows that Chromatium chroma- tophores were able to photoreduce this dye, although the observed rate was less than that obtained with R. rubrum. All indications point to a relatively simple system being involved in the photo reduction of methyl red and tetrazolium blue in the presence of the ascorbate-DPIP couple. As shown below, this activity in the case of R. rubrum chromatophores is more stable than is the NAD or fumarate reducing systems. tV ""~-~ -^......^ CHROMATIUM ^"^^^^^Omg BCHL. E - LIGHT \ ON ^ . ^^^^ O t -2 \ 1- < \ 5 -.3 ■z < - \ ^ -.4 CD < R. rubrum \. .22mgBCHL. ^^^^^ I 2 3 4 5 MINUTES Fig. 8. Photoreduction of methyl red by R. rubrum and CliroiiiatiKiii chromatophores. The experimental conditions given by Ash et. al. (17) were employed. The BChl concentrations were 0.20 for R. ruhruiii and 0.20 for Cliro)iiot//tii, for the 8 ml reaction system. PHOTOOXIDATION AND PHOTOREDUCTION REACTIONS 245 The reason for the failure to show NAD reduction with Chromatium chromatophores is not immediately apparent. Arnon reported that he was able to obtain a photoreductionof NADin the presence of ascorbate and DPIP (30). In our laboratory, however, we have not been able to obtain a photo reduction of ISi AD with Chromatium chromatophores under any circumstance. This is surprising, since this bacterium would be expected to have an active system for NAD photo reduction, because it must reduce carbon dioxide via the photosynthetic route for all of its carbon compounds. Another photosynthetic bacterium, Rliodopseudomonas spheroides, was examined in the usual photooxidation systems with the results given in Fig. 9, Again, this organism showed the capacity to photo- oxidize DPIPH2 in a fast reaction which soon saturated in the usual manner. Like Chromatium, this bacterium did not have the ability to sustain the secondary coupled reactions with either NAD or fumarate, although there was a hint of a slow coupled reaction when fumarate was added to the system. In this case also, it was not possible to demonstrate directly a photoreductionof NAD when DPIP and an excess of ascorbate were present as the electron- donating system (31). The reason for inactivity in NAD reduction in the case of both Chromatium and Rliodopseudomonas spheroides is not known. One should probably look at the method of chromatophore preparation to see if inactivation of enzymes or other factors is involved in this situation. D I . NO OXIDANT ^ LIGHT ON -«• ^>i LIGHT OFF + FUMARATE y r NAD 4''6 8 10 MINUTES 12 14 Fig. 9. Photooxidation of DPIPH2by Rps. spheroides chromato- phores. The experimental conditions given for Fig. 1 were em- ployed, except that Rps. spheroides chromatophores equal tc 0.110 jumoles of BChl per 4 ml were employed. 246 ELECTRON TRANSPORT CHARACTERISTICS OF THE DPIPH2 PHOTOOXIDATION SYSTEM OF R. RUBRUM CHROMATOPHORES From the extent of the fast reaction observed with both R. nibnim and Chromatiiim, it is possible to calculate the amount of DPIPH2 converted in the photooxidation reactions. Table 1 presents the results of such calculations, showing that for R. nibnim there is one mole of DPIPH2 oxidized for every 6 moles of BChl contained in the chroma- tophore. The value for Chromatiiim is approximately one mole of DPIPH2 oxidized for each 7.5 moles of BChl. A logical compound to consider as the intrachromatophoral oxidant for DPIPH2 is ubiquinone, TABLE 1. Ratio of BChl molecules to DPIPH2 photooxidized by chromatophores. The experimental conditions for R. rubrum were those given for Fig. 1, with only slight variation in BChl content amojig the different experi- ments. The experimental conditions for Chromatiiim were the same as those given for Fig. 7. BChlA)PIPH2 R. rubrum Average of 57 samples 5.9 Chromatium Average of 5 samples 7,5 This compound is present in chromatophores of both Chromatium and R. rubrum in high concentrations. Fuller et al. have reported a ratio of about five for chlorophyll to ubiquinone in Chromatium (32). Lester and Crane (33) have reported a value of 4.3 /imoles ubiquinone per gram dry weight for R. rubrum, which compares with 2.9 for Chroma- tium. Nishimura (34) has recorded a value of 19 for the ratio of chloro- phyll to cytochrome in R. rubrum. When this ratio is coupled to the information presented byOeller (35) on the heme protein and ubiquinone content of R. rubrum chromatophores, a ratio of about three chloro- phylls to one ubiquinone can be calculated. Thus, there is sufficient ubiquinone present in the chromatophores to account for the observed photooxidation of DPIPH2 and TMPD in the absence of added external oxidants. The succeeding paper presents definite evidence that ubi- quinone is reduced as a function of added DPIPH2 with /?. nibnim chromatophores. Clayton (22) has previously presented evidence that a photoreduction of ubiquinone is a primary event following light ab- PHOTOOXIDATION AND PHOTOREDUCTION REACTIONS 247 sorption by Chromatium chromatophores. When all these data are con- sidered, it appears likely that the photooxidation of DPIPH2 and TMPD is linked to the photo reduction of ubiquinone contained within the chromatophore itself. The response of the R. nibnim chromatophores to varying concen- trations of NAD and fumarate have allowed the calculation of a Michaelis constant for these secondary slow reactions. In the case of NAD, a Km of 2.5 x 10"^ was calculated, while the corresponding Km for fumarate was 1 x 10"'* molar. These values are in the range ex- pected of ordinary enzymatic reactions, and agree fairly well with the 3 2 5 '2 I 5 o ^ 101- < Fast Reaction No Oxidant \ Reactionx20 6 5 9 Fig. 10. pH optima for the three reactions involving DPIPH2 photooxidation with R. nibnim chromatophores. The ex- perimental conditions given for Fig. 1 were employed utilizing 0,20 mg of BChl in the reaction system. Tris buffer was employed for the pH range from 7,0 to 9,0 and phosphate buffer was employed for the lower pH values. The rates for the slow reactions were multiplied by 20 to place them on the same scale as the fast reactions. 248 ELECTRON TRANSPORT response to NAD and NADH concentrations observed by Horio and Kamen in the caseof NADH- RHP reductase and the NAD photoreduction supported by succinate with R. rubnim chromatophores (36,37). The pH optima of these reactions are shown in Fig. 10. The effect of temperature on the photooxidation of TMPD was examined by fol- lowing the reaction at room temperature and at the temperature of liquid nitrogen. These experiments, shown in Fig. 11, were carried out in collaboration with and through the courtesy of Dr. Britton Chance, The photooxidation of TMPD did not proceed at the temperature of liquid nitrogen, indicating an ordinary chemical reaction was involved in this photooxidation. This distinguishes it from the physical process which results in cytochrome oxidation in Cliromatium (12). 440 480 520 560 600 WAVELENGTH, m/x 640 680 Fig. 11. Lack of TMPD photooxidation at 77°K. Experi- mental conditions given for Fig. 3 were used, except that the reactions were run aerobically and contained 0.20 mg of bacterial chlorophyll. These experiments were performed by Dr. Britton Chance using the spectro- photometer described previously (12). The curve cor- responding to 300° K was taken directly from the re- cording following 30 seconds illumination time with red light. The curve corresponding to 77° K was plotted by taking the difference between the tracing obtained before illumination and following 70-second illumination. The minimum observed at 600 ni^ is characteristic of light absorption changes which occur with chromato- phores alone. PHOTOOXIDATION AND PHO TOR EDUCTION REACTIONS 249 The effect of flavins upon the photooxidation of DPIPH2 was ex- amined. Fig, 12 shows the effect of adding FMN to the R. nibniDi sys- tem. At a level of 3 xlO-^M, FMN exerted an appreciable stimulation with the NAD slow reaction but had no effect upon the coupled fumarate reaction. FAD gave about one-third the stimulation in the NAD system and had no effect upon the fumarate system. Riboflavin was inactive in both systems. One striking feature of the FMN stimulation was the low concentration at which FMN was effective, with half maximal stimula- tion at a concentration of about 10"^ M FMN. The response in the present system to FMN should be compared with the stimulating effect this nucleotide has upon the DPNH- cytochrome c^ reductase system in the purified fractions obtained by Horio and Kamen (36). .6- .4 - LIGHT OFF \ /^^^^^-'"'^^^ ^^^^"^^[T ^^^^ ■ t LIGHT ON I 2 3 MINUTES Fig, 12. The effect of FMN on DPIPH2 photooxidation by R. nihyiim chromatophores. The experimental con- ditions for Fig. 1 were used with 0,21 mgof BChl present. This dark enzymatic reaction is stimulated half- maximally at about 10-'^ M FMN, FAD is less active than FMN, but riboflavin resembles FMN in its activity. It is possible that the same flavoprotein is in- volved in the present photooxidation reactions and in the dark enzymatic reactions studied by Horio and Kamen, Addition of quinacrine to the various oxidation systems produced the interesting results shown in Fig. 13. A clear separation of the coupled photooxidations with fumarate and NAD was observed, the NAD- coupled photooxidation being completely inhibited while the fumarate- coupled oxidation was markedly stimulated. This shows clearly that 250 ELECTRON TRANSPORT two systems are involved, one for NAD and one for fumarate. Quina- crine has been shown to have potent effects upon electron transfer systems in photo synthetic bacteria. Baltscheffsky has shown that this compound is an inhibitor of the photophosphorylation process (38). Furthermore, the inhibition observed in his experiments was partially relieved by the addition of FAD, while FMN was less active. It should be noted, however, that extremely high concentrations of FAD and FMN were required for these reactivations, and at such high concen- trations the flavin nucleotides are active in nonenzymatic reactions. E O If) 1- < .u .8 .6 LIGHT OFF > o < CD a: o en 03 < .4 .2 :/ ^^ \ NAD + QUINACRINE ' ^ LIGHT ON I 2 3 MINUTES Fig. 13. The effect of 2 x lO'^ M quinacrine on DPIPH2 photooxidation by R. ruhrum chromatophores. The ex- perimental conditions employed for Fig. 1 were used with 0.21 mg of BChI present. The experiments of Ash et al. (17) show that quinacrine has a strik- ing effect upon the photo reduction of methyl red and tetrazolium blue by R. riihnun chromatophores. The stimulating effect of quinacrine on a very similar system has been found by Bose and Gest (39). In their experiments, R. mbntin chromatophores catalyzed the photoreduction of fumarate by hydrogen gas. This photoreduction required the presence of DPIP, which was reducedenzymaticallyby the hydrogen gas. There- fore, the two systems are essentially the same, consisting of DPIPH2 and fumarate. In both cases DPIPH2 is photooxidized and fumarate is reduced, and in both cases quinacrine markedly stimulates the reac- tions. PHOTOOXIDATION AND PHO TOR EDUCTION REACTIONS 251 The effect of respiratory inhibitors upon both the fast and coupled slow reactions is shown in Table 2, Phenylmercuric acetate (PMA), a compound which inhibits by combination with sulfhydryl groups, in- hibited both the NAD and fumarate slow reactions. It was less effective on the fast reaction, and had little effect on the aerobic photooxidation. As shown below, the fast reaction and the aerobic photooxidation re- actions are stable activities which are resistant to heating. This indi- cates that the fast reactions and the aerobic reaction involves only a portion of the electron transport chain, and this portion is relatively inert to various treatments and outside agents. The inhibitions caused by antimycin A and HQNO also reveal that the coupled slow reactions are most sensitive. However, in general the TABLE 2. Effect of inhibitors upon the photoreactions ofR. rubrum chromatophores. The reaction conditions were as given for Fig. 1. The concentration of BChl was 0.20 mg. Per Cent Inhibition NAD Fumarate Aerobic Fast RX Slow RX Fast RX Slow RX PMA, 10-4 M 25 87 13 70 6 " 10-5 M 12 47 4 63 Antimycin A, 1.1 x lO"'* M 18 56 24 " 2.3 X 10-5 M 5 16 - HQNO, 1.4 X 10-4 M 9 46 10 19 " 2.9 X 10-5 M 5 21 — — - reactions reported here are less sensitive to both antimycin A and HQNO than either the photophosphorylation process or the photore- duction of NAD coupled to succinate. 10"^ M antimycin A inhibits over 90 per cent of the activity in theNAD-succinate system (16) and in the photophosphorylation process (16,40), HQNO is also over 90 per cent effective in these reactions at a concentration of 10"^ M (16,40), It would appear, therefore, that the antimycin A and HQNO inhibitions ob- served in the present case are of a different nature from the inhibitions observed in the photophosphorylation process. This has significance when considering the mechanism of the reduction of fumarate and NAD by DPIPH2. As discussed below, the inhibition pattern is not consistent with the idea that the reduction of these compounds is due to a re- versed electron transfer coupled to ATP utilization, since at the level 252 ELECTRON TRANSPORT of inhibitors used (lO-5 M) ATP formation is inhibited over 90 per cent and the reported photooxidation reactions are only slightly de- creased. The effect of ADP and Pi on the various reactions is shown in Table 3. Again a distinction is apparent between the NAD-coupled and the fumarate- coupled slow reactions. Whereas the NAD reaction was in- hibited by the addition of these components, the fumarate reaction was stimulated. If it is true that stimulation of a reaction by ADP and Pi indicates a phosphorylation in that reaction sequence, this would indi- cate that the photo reduction of NAD accompanying DPIPH2 oxidation does not involve a phosphorylation, whereas the photoreduction of fumarate is accompanied by ATP formation. These conclusions should be considered as only tentative, however, since ADP can have effects other than that of stimulating a reaction coupled to ATP formation, TABLE 3. Effect of ADP + Pi on the photo reactions of R. rubrum chromatophores. The reaction conditions were those given for Fig. 1 with a BChl concentration of 0.21 mg. Concentration Per Cent Stimulation NAD Fumarate Fast RX Slow RX Fast RX Slow RX 10-3 M^ p^Qp 10-3 M. Pi 22 -69 12 32 25 -40 22 38 However, it should be noted that the concentrations used in these experiments are below those reported by Horio to give inhibition of NAD reduction associated with succinate oxidation (37). The activity of heated chromatophores is shown in Fig. 14. For these experiments aliquots of chromatophores were heated at the in- dicated temperatures for five minutes, following which the reaction was run in the usual fashion. The rates of the slow reactions were multi- plied by ten in order to place them on the same scale as the fast re- actions. From this information it is apparent that heating to 40°C was sufficient to destroy the reaction coupled to NAD, and heating to 50°C inactivated the system involving fumarate. Heating to 60 °C had very little effect upon the fast reaction observed. Indeed, to inactivate the fast reaction with R. ruhnini chromatophores, a temperature of 80°C for five minutes must be employed. This emphasizes the stability of the fast reaction, and again implicates a very stable and perhaps frac- PHOTOOXIDATION AND PHO TOR EDUCTION REACTIONS 253 g 0.0 i -0.5 TEMPERATURE "C FUMARATE ADDED NAD ADDED Fig. 14. Heat stability of DPIPH2 photooxidation reactions with R. nthrion chromatophores. The experimental condi- tions outlined for Fig. 1 were employed with 0.20 mgof BChl present. The chromatophores were heated at the indicated temperature for 5 minutes before addition to the reaction system and commencement of the illumination period. The photooxidation reactions were run at the usual temperatures. The rates for the slow reactions were multiplied by 10 in order to place them on the same scale as the fast reactions. tional part of the entire electron transport system for the observed fast reaction. The stability of the fast reaction reported here agrees with the previously reported stability of the aerobic photooxidation of DPIPH2 as reported by Lindstrom (6). The rates of various related dark enzymatic reactions for chroma- tophores of R. nibnim and Chromatium are given in Table 4. These are the various dark enzymatic reactions which might be expected to influence the photoreactions under investigation. The R. rubrum chromatophores have a very potent NADH-DPIP diaphorase, which is most likely the enzyme which has been studied in some detail by Horio and Kamen (36). R. nibnim has very weak DPIPH2 oxidase activity and a moderate NADH oxidase activity. Chromatium also has a potent NADH-DPIP diaphorase activity, and has a very active DPIPH2 oxidase activity. This is surprising because this organism lives under anaerobic conditions and does not practice a respiration involving oxygen, where a terminal oxidase would be expected to function. 254 ELECTRON TRANSPORT TABLE 4. Rates of dark enzymatic reactions of chromatophores . The reaction mixtures (4 ml total volume) contained chromatophores equivalent to 0.1 and 0.165 mgof BChl for R. rubrum and Chromatium respectively, ivith the buffer concentrations listed in Figs. 1 and 7. The ex- perimental system for NADH-DPIP diaphorase also contained 1.5 fimoles of NADH and 0.27 /umoles of DPIP, that for DPIPH2 oxidase contained 0.27 /umoles of DPIPH2 and that for NADH oxidase contained 0.6 ju moles of NADH. The diaphorase assay ivas performed under anaerobic conditions at 25°C. //molesAr/mg Chlorophyll R. rubrum Chromatium NADH-DPIP Diaphorase 18 17.5 DPIPH2 Oxidase 1 15 NADH Oxidase 7.7 2.4 CORRELATION OF KNOWN PHOTOREACTIONS IN R. RUBRUM It would be well to compare the rates of DPIPH2 photooxidation with other photo re actions which have been reported in the literature. Table 5 reports the rates we have observed for both R. rubrum and Chroma- tium. The rates reported are average values, and the fastest photooxi- dation rate observed to date was the fast reaction with NAD present, which amounted to 553 jumoles DPIPH2 oxidized per hour per mg of BChl, This reaction, then, is one of the fastest reactions reported for the photochemical system of the bacteria. In general, Chromatium particles were less active on a chlorophyll basis. The data again show the lack of a coupled slow reaction with either NAD or fumarate. The rates for photophosphorylation reported in the literature usually fall between 100-300 /umoles per hour per mgof chlorophyll (16,41,42, 43). However, M. Baltscheffsky has reported a photophosphorylation rate of 620 /imoles per hour per mg of chlorophyll (44). The rates of the two reactions, the photooxidation of DPIPH2 and ATP formation under the influence of light, are certainly very similar. Again this lends support to the idea that the photooxidation of DPIPH2 observed is mediated by the central closed electron transfer system contained in R. rubnim chromatophores, and that the same light- activating system is active in both cases. The rates of other photooxidation and photo reduction reactions re- ported in the literature fori?. r?^/n7^;;/ chromatophores have been com- PHOTOOXIDATION AND PHOTO REDUCTION REACTIONS 255 TABLE 5. Rates of DPIPH9 photooxidatiou. The values given for R. rubrum are average values obtained from 50 experiments, with the exception of the aerobic reaction. The values given for Chromatium are average values taken from 5 experiments . ^molesAr/mg Chlorophyll R. rubrum Chromatium Fast Reaction with NAD Fast Reaction with Fumarate Slow Reaction with NAD Slow Reaction with Fumarate Aerobic Reaction 404 224 220 287 12 7.3 47 16.5 piled in Table 6, Perusal of this table shows that the fastest rate re- ported is that by Amesz for the photo reduction of NAD by whole cells (21). This rate of 360 ^moles per hour per mg of BChl places this photo reduction within the same range asphotophosphorylation, and just below the rate of the fast DPIPH2photooxidation reported in this paper. The rate reported by Amesz was the initial rate obtained when the cells went from a nonilluminated to an illuminated condition, and represents the maximal rate at which electrons could be transferred from cyto- chrome through chlorophyll to NAD. The rate of 145 reported by Ash et al. for the photo reduction of methyl red in the presence of quinacrin is higher than the usual re- action reported. It appears from the evidence at hand that methyl red photo reduction (in the presence of ascorbate and DPIP as electron donor) is one of the more stable reactions catalyzed by chromato- phores and involves only a portion of the chromatophore system. None of the ordinary enzymatic components involved in NAD photo reduction and fumarate photo reduction are required in the case of methyl red (17). The remainder of the photoreactions have rates between the range of 4 to 45 /^moles per hour per mg of BChl. The slow photooxidation reactions coupled to NAD and fumarate reported in this investigation fall in this range also. These apparently, then, are the photooxidation reactions which are coupled to the enzymatic components within the electron transfer system. The information available on rates of the various photoreactions is consistent with the idea that an initial fast photooxidation of DPIPH2 is representative of the fast photochemical reactions induced following the absorption of light quanta by the chromatophore. It is generally thought that the bacterial chromatophore contains a closed electron 256 ELECTRON TRANSPORT TABLE 6. Rates of other photoreactions of R. rubrum chromatophores. Photoreaction Rate and Reference //molesAir/mg Chlorophyll Oxidations H2 Oxidation (with DPIP) FMNH2 Oxidation (with NAD) Reductions NAD (with FMNH2) NAD (with Succinate + CN ) NAD (with Succinate) NAD (whole cells) Reductions Coupled with Ascorbate and DPIP NAD Disulfide Methyl Red Methyl Red + Quinacrine Sulfate 4 (Bose and Gest, 39) 9.5 (Frenkel, 7) 9.1 (Frenkel, 7) 45 (Vernon and Ash, 8) 24 (Nozaki et al., 16) 360 (Amesz, 21) 37 (Nozaki et al., 16) 11 (Newton, 18) 24 (Ash et al., 17) 145 (Ash et al., 17) 43 (Ibanez and Lindstrom, 5) transfer system in the sense that electrons travel along this electron transport chain in the usual fashion (from the compound with the lowest oxidation potential to that with the highest potential) and are then re- cycled through the light- activated chlorophyll back to the low potential compound once again. Isolated chromatophores have the capacity to carry out the process of photophosphorylation with no added cofactors, and this phosphorylation is very sensitive to the extent of reduction of the various components within the chain, since either overreduction or overoxidation inhibits the cyclic phosphorylation process (42,55, 57), Thus, we can imagine the electron transport system of the chroma- tophore as consisting of a closed electron transport system which yields ATP when electrons are cycled around the system. There must be several points of entry onto the system to allow the photochemical reactions to be coupled with the chemical oxidation and reduction re- actions which must take place in the whole cell (NAD reduction and substrate oxidation). The fastest rates observed with isolated chromatophores involve the photophosphorylation process and the photooxidation of DPIPH2 in the initial fast reaction. The fast reaction of DPIPH2 must be coupled to the photo reduction of components contained within the cyclic electron transfer system. The most likely candidate for such a photo reduction is ubiquinone. The secondary slow reactions coupled to DPIPH2 oxida- tion appear to involve other enzymatic components on the chromato- PHOTOOXIDATION AND PHOTOREDUCTION REACTIONS 257 phore which are not ordinarily associated with the cyclic electron transfer process. This accounts for the slower rates observed in gen- eral for the photooxidation and photo reduction reactions with/?, ruhruni chromatophores. The same general pattern applies to C/zrowa^mw, with the exception that in this case the slow enzymatic reactions are not observed at all. Some essential cofactor must be lost or destroyed during preparation of the chromatophore. There is an alternative explanation for the data presented above. There is ample evidence for "reverse electron flow" in mitochondrial systems, in which case the electrons are transferred from either suc- cinate or cytochrome to NAD (45). This transfer being against the thermodynamic gradient, energy is required and the energy is supplied by ATP, Thus the requirements for electron transfer in the reverse direction are the necessary enzymatic components plus energy sup- plied in the form of ATP. In the present case, these conditions are present. Thus one could explain the photooxidation of DPIPH2 as being due to reversed electron transport via the reagents contained in the chromatophore with the energy for this reverse electron transport being supplied by the ATP formed in the process of photophosphoryla- tion. Although on the surface this explanation appears to be tenable, there are a number of reasons which lead me to believe that this can- not be the case. These reasons are as follows: 1. The photophosphorylation process is very sensitive to the re- spiratory inhibitors, antimycin A and HQNO, At 10~^ M inhibitor con- centration, the photophosphorylation process is almost completely in- hibited, while at this concentration of inhibitor the photooxidation reactions proceed with very little depression in rate. A more striking example is given by the compound quinacrine which inhibits photophos- phorylation by 80 per cent at a concentration of 5 x 10-4 m (40), yet has no appreciable effect on the DPIPH2 fast reaction at this concen- tration and actually stimulates the coupled slow reaction with fumarate. 2. When R. riibniDi chromatophores are heated to 60°C, their abil- ity to carry out photophosphorylation is completely inhibited. Never- theless, such chromatophores still have the ability to catalyze the fast oxidation of DPIPH2 as shown in Fig. 14. 3. The photo reduction of NADP by DPIPH2 in the plant system (utilizing the long wavelength system in chloroplasts) does not appear to proceed by reverse electron flow. Thus, the photo reduction of NADP by DPIPH2 results in the formation of ATP, and removal of plasto- quinone from the chloroplasts prevents the ATP formation but does not prevent NADPH formation (46). 4. Addition of ATP in the dark to the system containing chromato- phores, DPIPH2 and NAD does not result in the formation of NADH and DPIP. The results of the experiments are shown in Fig. 15, Whereas 258 ELECTRON TRANSPORT ATP was inactive, the subsequent illumination caused the immediate photooxidation of DPIPH2 in the typical fast reaction. The possible ob- jection to this reasoning is that ATP may not enter into the chromato- phore and be able to affect the reactions as does ATP formed directly in the photophosphorylation process. This argument would seem in- valid since the chromatophores used in this experiment are capable of coupling with external ADP and Pi to form ATP which is contained in the medium. Thus the phosphorylation site is available to added reagents and a reversed electron flow, if operative in the above ex- periments, should have been capable of demonstration. .4 - 54. E 10 - ^^-^ 1- < ^.-^ >-2 z < OD to + ATP f CD < i \ T -.^ \ JLIGHT ON 3 MINUTES Fig. 15. Effect of ATP addition upon DPIPHg oxidation with NAD in the presence of /?. r«6n<>» chromatophores. The experimental conditions given for Fig. 1 were em- ployed with 0.21 mg BChl present. At the arrow, 8 ^moles of ATP were added to the reaction system. The reasons given above for supporting a direct electron transfer from DPIPH2 to NAD are based on the fact that ATP formation and DPIPH2 photooxidation can be separated by means of inhibitors and various treatments. The possibility remains, however, that the reac- tion does proceed via reverse electron transfer, with the energy being supplied not by ATP itself , but by some high energy intermediate which under normal circumstances would lead to ATP formation. K the in- hibitors act at a late stage in ATP formation, and allow the formation of intermediate high energy compounds, then the conclusions based on PHOTOOXIDATION AND PHOTOREDUCTION REACTIONS 259 these inhibitor studies are not valid. However, one cannot resolve this problem at present, since it is not possible to detect and experi- mentally manipulate such high energy intermediates. ELECTRON TRANSFER SEQUENCE Since there is considerable information becoming available on the various photooxidation and photo reduction reactions catalyzed hy R. rubrum, it would be well to attempt to correlate all this information. Figs. 16, 17, and 18 are an attempt to do this and are presented here not with the intent of being an authoritative statement on this matter, but rather with the idea of using this means to bring the information to the attention of workers in the field with the hope that it may stimulate thinking and experimentation to solve some of the problems which now face us. Fig. 16 is a representation of the components of the chroma- tophore electron transfer system arranged in a probable sequence for electron transfer. Fig, 17 has imposed upon this basic electron trans- fer system the sites of action of the various compounds known to be effective in photooxidation and photo reduction reactions. Fig, 18 lists the possible sites of action of the inhibitors known to affect the photo- chemical reactions in R. rubnini chromatophores. It immediately becomes apparent that the various photochemical reactions demonstrated by R. rubnim chromatophores are couched in terms of only one electron transfer system. In discussing Fig. 17 it will be pointed out that a portion of this electron transfer scheme could be operative in the respiratory reactions which take place in the dark with this organism. ELECTRON TRANSFER COMPONENTS in Ki rubrum e^- Cy^to. C2 I Cyto. b Light ▼ t ^vN~v-*.B. Chi. RHP ►02 Ubiquinone "* » A, /t Ferredoxin-*-Flavo-"*z:z^ nao Protein (FMN) Fig. 16. Electron transfer components and sequence of reaction in R. riibnoji chromatophores. 260 ELECTRON TRANSPORT PROBABLE REACTION SITES in R. rubrum Light CytO. C2 ^ — DPIPHa. TMPD, A Cyfo. c, MBHj CytO. b t RHP ► 02 t Ubiquinone / Production e- * *^^ Ferredoxin-*'Flavo- - ^i Protein , (FMN), 02, Methyl Red, Tetrazolium Blue, S04 = Disulfide {With Viologen Dyes) Succinate "* Fumorate Fig. 17. Reaction sites for photochemical reactions observed with R. rubrum chromatophores. INHIBITION SITES in R. rubrum Succinate "*■ Fumorate (FMN) ' ^■^ -Qu(nacri Fig. 18. Sites of inhibitor action with R. rubrum chromatophores. The components contained within the R. nibnim chromatophores are presented in Fig. 16 in a logical sequence. There is ample evidence that cytochrome C9 contained in this organism is photooxidized in a primary photochemical reaction by the BChl (9-13,28). This organism also contains a compound whichappears to be similar to the ferredoxin isolated by Tagawa and Arnon from Chro)uatium (47), Preliminary evidence indicates that the ferredoxin is tightly bound to the chroma- tophore in the case oiR. riibrn nt andis not readily extracted. The first compound shown to be reducedby the action of light in the case of plant PHOTOOXIDATION AND PHOTOREDUCTION REACTIONS 261 chloroplasts is the corresponding compound photosynthetic pyridine nucleotide reductase (48). For these reasons ferredoxin is presented as the compound being reduced by light- activated chlorophyll. It should be mentioned, however, that in the experiments designed to demon- strate a ferredoxin in R. nibnim fractions, the assay employed was the photoreduction of NADP by spinach chloroplasts. No stimulation has been observed in the R. rubrum photo reactions by addition of R. nihnim ferredoxin fractions. The position of the other compounds in this scheme is less certain. The inclusion of a flavoprotein between ferredoxin and NAD is made on the basis that the NAD- coupled photooxidation of DPIPH2 is stimu- lated by FMN at levels which are consistent with its functioning as a coenzyme to a flavoprotein. In a similar system involving the ferre- doxin from Clostridium , a flavoprotein was shown to be required for the reduction of NADP (47), Fig, 16 represents two pathways leading to ubiquinone, one coming from ferredoxin and the other from the flavoprotein. The evidence presented above on the slow photooxidation of DPIPH2 coupled to fu- marate and to NAD indicates separate pathways for these two reactions. Thus, FMN stimulates the NAD- supported reaction and has no effect on the fumarate- supported reaction, Quinacrine inhibits the NAD re- action and stimulates the fumarate- supported reaction. There is considerable evidence that ubiquinone occupies a central position in accepting electrons on the reducing side of the scale. Thus Clayton has shown that in the case of Chromatiiim one of the primary reactions which occurs following illumination is most likely the photo- reduction of ubiquinone (22). The high level of ubiquinone contained in R. nibntni chromatophores (33), and especially the large increase observed in light-grown cells over dark-grown cells indicates a prom- inent role for this compound (35). The scheme presents ubiquinone as transferring electrons to RHP. RHP is unique to photosynthetic bac- teria, is oxidized by molecular oxygen (49), and has been shown to be essential for the photophosphorylation process (36). The assignment of sequence of electron transfer from ubiquinone to RHP is not in agreement with the reported oxidation potentials of -0.008 volts for RHP (49) and 0.098 volts for ubiquinone (50). The potentials observed in the isolated compound may not accurately represent the potential of the same compound in situ in the chromatophore structure, however. The main reason for choosing this sequence is that it allows ubiquinone to act very near the chlorophyll system, which is in agreement with present information. Cytochrome b is included in the pathway between RHP and cyto- chrome cp. This is the usual sequence encountered in mammalian tissues, and the spectroscopic evidence obtained by Nishimura, al- though not conclusive, indicates that cytochrome b contained in/?, nibnim chromatophores can be reduced under the influence of light 262 ELECTRON TRANSPORT when inhibitors such as antimycin A are present to block electron transfer between cytochrome b and cytochrome c^ (23,24), Fig. 17 presents the basic electron transfer system proposed in Fig. 16, and has included the possible sites of action of the various compounds which enter into the photochemical reactions demonstrated by these particles. For discussion purposes, let us begin at the lower end of the potential scale. The experiments of Tagawa and Arnon im- plicate ferredoxin in the hydrogen metabolism of selected bacteria, both photo synthetic and nonpho to synthetic (47), To date, no evidence has been presented to show that hydrogen evolution can be produced with isolated R. nibnim chromatophores, although it is made in copious amounts by the intact cell under various conditions (51). Be- cause of the implication of ferredoxin in other hydrogen-producing systems, and because of the suitability of the potential of the two sys- tems, it is indicated as the precursor of hydrogen in Fig. 17. A number of compounds are shown as reacting with the chromato- phore system at the level of ferredoxin and/or flavin. These compounds include oxygen, methyl red, tetrazolium blue, sulfate ion, and the di- sulfide bond contained in DTNB. The primary reason for including these compounds at this position is that the photo reductions involving these compounds are all heat stable and also relatively insensitive to the inhibitors which affect the other photo reactions. Lindstrom has shown that the photooxidation of DPIPH2 in the presence of air is very stable to heat (6). Information presented above shows that aerobic photooxidase is also insensitive to the inhibitors tried, such as anti- mycin A, HQNO, and quinacrine. The methyl red and tetrazolium blue photoreductions have also been shown in our laboratory to be stable to heat. Heating chromatophores to 60 °C for five minutes destroys the ability of these particles to carry out photophosphorylation and the secondary slow photooxidation of DPIPH2 coupled with NAD and fumar- ate, but does not appreciably affect the methyl red photoreduction. Likewise, the photoreduction of methyl red and tetrazolium blue is also relatively insensitive to inhibitors known to affect the other photo- reactions (17). Photoreduction of sulfate and DTNB are also fairly stable to heating, although heating does have more effect upon these reactions than upon the others in this group (5,18). RHP is shown in Fig. 17 as reacting with molecular oxygen (49), Succinate and fumarate are shown as reacting with the electron trans- fer system through ubiquinone. This assignment is logical in view of the known reaction of succinate with ubiquinone in mitochondrial sys- tems (52) and in view of the recent report that ubiquinone can be re- duced enzymatically in the dark by succinate with chromatophores from R. rubnim (53), The portion of the electron transfer chain up to RHP could very well be functioning as an NADH oxidase system in the dark respiratory activities of this organism. The phosphorylation coupled to this span PHOTOOXIDATION AND PHOTOREDUCTION REACTIONS 263 would allow for the phosphorylation accompanying aerobic oxidation of substrate molecules via NAD. Since this organism is lacking a cyto- chrome oxidase in the traditional sense, one looks for another com- pound to complete the electron transfer span to molecular oxygen, and the logical candidate is RHP (36). For reasons discussed above, the flavin listed in this scheme may well be common to both the photo- chemical reduction of NAD and the various enzymatic reductase ac- tivities observed in extracts of R. nihnun in the dark (36). It has been reported by Cohen- Bazire and Kunisawa that the particles called chromatophores, which can be isolated after the cell is broken, are actually segments of the cell membrane which has been ruptured dur- ing the process, and that in the intact cell the photosynthetic apparatus is laid down upon this limiting membrane (54). The implication is obvious, therefore, that this same membrane is involved in both the oxidative reactions and the photosynthetic reactions. One would expect, therefore, to find both of these functions combined in the one structural component, and theutilizationof a portion of the photosynthetic electron transfer chain in the oxidative metabolism would meet this goal. The compounds which are photooxidized in the presence of NAD and fumarate are shown to interact with the electron transfer scheme at the cytochrome c^ level. One of the prime reasons for proposing this site of entry into the chain is that the photooxidation of these com- pounds is not affected appreciably by the respiratory inhibitors anti- mycin A and HQNO at concentrations which almost completely inhibit the photo phosphorylation process. The data of Nishimura point toward the site of action of these inhibitors somewhere between cytochrome b and cytochrome C2 (23,24). The entry of DPIPH2 and the other photo- oxidizable compounds at the cytochrome c^ locus is also consistent with the fact that this cytochrome is not firmly bound in the chromato- phore system and can be removed by relatively easy treatments such as extraction with citrate buffer, etc. This indicates that the cyto- chrome is exposed to the aqueous medium in the chromatophore and would logically be a site of action for these compounds. Furthermore, Jacobs has shown that in a rat-liver mitochondrial system, TMPD reacts by reducing the cytochrome c on the particle (27). Fig. 17 shows two sites for ATP formation along the electron transfer chain. This is in accord with the data of Baltscheffsky and Arwidsson (56), who have studied the effect of the inhibitor valinomycin upon R. nibnim chromatophores in the photophosphorylation reaction. Also, the data of Nishimura (34) on the amount of ATP formed per flash of light with R. nibnim chromatophores indicate that two sites for ATP formation are to be found. If PMS overcomes antimycin A inhibition by means of serving as a bypass for the inhibited site be- tween cytochrome b and cytochrome c^, then there must be a phos- phorylation site before the site where PMS is reduced. The scheme in Fig. 17 accommodates this. 264 ELECTRON TRANSPORT Fig. 18 shows the probable sites of action of various inhibitors which are effective with R. nibrum chromatophores. PMAis shown as inhibiting NAD reduction, since it was shown to be an effective in- hibitor in the photoreduction of NAD by both succinate and ascorbate- DPIP by Nozaki et al. (16). Also, as shown above, this was an effective inhibitor of the NAD-supportedphotooxidationofDPIPH2. Amytal would also be expected to inhibit at this position. Geller showed amytal to be an inhibitor of the dark aerobic oxidation of ^ ADU hy R. nibnim chromatophores, while it has no effect upon the photophosphorylation reactions (35). Since quinacrine has an inhibitory effect upon the photophosphory- lation process, it probably acts at another site in addition to the flavo- protein designated in Fig. 18, Quinacrine is shown as an inhibitor at the flavoprotein level, since it inhibits the photooxidation of DPIPH2 coupled to NAD, yet stimulates the reaction coupled to fumarate. The stimulating effect of FMN on the NAD-supported photooxidation of DPIPH2 also indicates a flavoprotein acting at this site. The spectroscopic evidence of Nishimura and Chance (23,24) points to the site of action of antimycin A and HQNO as being between cyto- chrome b and cytochrome c^. In addition, PMA has been shown to re- sult in a photooxidation of cytochrome c^anda photoreduction of cyto- chrome b in R. nibrum cells, and probably acts at this locus also. Since PMS has been shown to overcome the inhibition of photophos- phorylation by antimycin A, it bypasses the site which is inhibited by antimycin A. The most logical mechanism for this requires that PMS be reduced by ubiquinone and reoxidized by cytochrome c^. Recent evidence obtained by Bose andGest (57) indicates that DPIP, when present in the oxidized form, can also serve as a bypass and overcome the inhibition of antimycin A upon photophosphorylation. This again is consistent with its photooxidation at the cytochrome C9 level and its photoreduction at a prior point, most likely at the ubiquinone point in the electron transfer chain. It is anticipated that someof the reaction sequences and components listed in Figs. 16-18 will be changed as additional information becomes available. Not all of the evidence available could be reconciled with the scheme as presented. Thus, Nishimura says that RHP is probably not located between ferredoxin and cytochrome b, since carbon monoxide (which does combine with RHP) does not affect the reduction of cyto- chrome b (23,24). Nishimura also states that quinacrine does not have any effect upon the absorption changes caused by illumination of /?. nibrum chromatophores, as would be expected if quinacrine does in- hibit at some site other than the flavoprotein designated in Fig. 16, One main question to be resolved is whether there is one electron transport system serving both the photochemical and oxidative path- ways in R. nibrum, or whether separate pathways are involved. As stated above, the schemes outlined in Figs. 16-18 accommodate the PHOTOOXIDATION AND PHOTOREDUCTION REACTIONS 265 idea that one electron transport chain functions in both areas, with the oxidative chain involving those components from NAD through RHP. Although Geller favors separate electron pathways for the photochem- ical and oxidative metabolism (35), his data on the effect of inhibitors correlate very nicely with the scheme proposed above. Thus, oxida- tive phosphorylation was not affected by antimycin A or HQNO, whereas photophosphorylation was strongly inhibited. Furthermore, amytal strongly inhibited NADH oxidation aerobically, but did not affect the photophosphorylation process. ACKNOWLEDGMENTS The author wishes to acknowledge the very capable technical assistance of Mrs. Ann Tirpack and Miss Dorothy Limbach in the course of the investigations reported above. REFERENCES 1. van Niel, C. B., The bacterial photosyn theses and their importance for the general problem of photosjmthesis. Advaii. Eiizyuiol., 1, 263-328 (1941). 2. French, C. S., The pigment-protein compound in photosynthetic bacteria. I, Extraction and properties of photosynthin. J. Gen. Physiol., j'J, 469-481 (1940). 3. Vernon, L. P., and Kamen, M. D., Studies on the metabolism of photosyn- thetic bacteria. XV. Photoautoxidation of ferrocytochromc c in extracts of Rhodospir/Uidii nibntni.Arch. Biochew. Diophys., 44, 298-311 (1953). 4. Vernon, L. P., Photooxidations catalyzed by chromatophores of Rhodo- spirilliou rtibnim under anaerobic conditions. J. Biol. Clicni., 234, 1883- 1888 (1959). 5. Ibanez, M. L,, and Lindstrom, E. S., Photochemical sulfate reduction by Rhodospi rill /oil rubrum. Biochein. Biophvs. Res. CoiiiDiiin., 1, 224-227 (1950). 6. Lindstrom, E. S., Photooxidase activity of heated chromatophores of Rhodo- spirillum rubrum. Plant Physiol., 37, 127-129 (1962). 7. Frenkel, A. W., Simultaneous reduction of diphosphopyridine nucleotide and oxidation of reduced flavin mononucleotide by illuminated bacterial chroma- tophores. J. Am. Chem. Soc, 80, 3479 (1958). 8. Vernon, L. P., and Ash, O. K., The photoreduction of pyridine nucleotides by illuminated chromatophores oi Rhodospirillum rubrum in the presence of succinate. J. Biol. Chem., 234, 1878-1882 (1959). 9. Duysens, L. N. M., Reversible photooxidation of a cytochrome pigment in photosynthesizing /?/wr/os/j/r/////;;? rubrum. Nature, 173, 692-694 (1954). 10. Chance, B., Spectrophotometry of intracellular respiratory pigments. Science, 120, 767 (1954). 11. Chance, B., and Smith, L., Respiratory pigments of Rhodospirillum rubrum. Nature, 175, 803-810 (1955). 12. Chance, B,, and Nishimura, M., On the mechanism of chlorophyll-cytochrome interaction: The temperature insensitivity of light-induced cytochrome oxidation in Chromatium. Proc. Natl. Acad. Sci. U. S., 46, 19-25 (1960). 266 ELECTRON TRANSPORT 13. Olson, J. M., and Chance, B., Oxidation-reduction reactions in the photo- synthetic bacterium Chromatiuni. I. Absorption spectrum changes in whole cells. Arc/?. Biochem. Biophys., 88, 26-39 (1960). 14. Chance, B., and Olson, J. M., Primary metabolic events associated with photosynthesis. Arch. Biochem. Biophys., SS, 54-58 (1960). 15. Vernon, L. P., Photoreduction of triphosphopyridine nucleotide by chroma- tophores oi Rhodospirillum rubrum. J.Am. Chem. Soc, 80, 246 (1958). 16. Nozaki, M., Tagawa, K., and Arnon, D. I., Noncyclic photophosphorylation in photosynthetic bacteria. Proc. Natl. Acad. Sci. U. S., 47, 1334 (1961). 17. Ash, O. K., Zaugg, W. S., and Vernon, L. P., Photoreduction of methyl red and tetrazolium blue by spinach chloroplasts andchromatophores of Rhodo- spirillum rubrum. Acta Che?)}. Scand., 15, 1629 (1961). 18. Newton, J. W., A disulfide photoreduction system in chromatophores of Rhodospirillum rubrum. J. Biol. Chem., 257,3282 (1962). 19. Duysens, L. N. M., and Sweep, G., Fluorescence spectrophotometry of pyridine nucleotide in photosynthesizing cells. Biochi)n. Biophys. Acta, 25, 13 (1957). 20. Olson, J. M., Fluorometric identification of pyridine nucleotide changes in photosynthetic bacteria and algae. Brookhaven Symp. Biol., 11, 416 (1958). 21. Amesz, J., Kinetics, quantum requirement and action spectrum of light- induced phosphopyridine nucleotide reduction in Rhodospirillum rubrum and Rhodopseudomonas spheroides . Biochim. Biophys. Acta, 66, 22-36 (1963). 22. Clayton, R. K., Evidence for the photochemical reduction of coenzyme Q in chromatophores of photosynthetic bacteria. Biochem. Biophys. Res. Com- muu., 9, 49 (1962). 23. Nishimura, M., and Chance, B., Studies on the electron-transfer systems in photosynthetic bacteria. I. The light-induced absorption-spectrum changes and the effect of phenylmercuric acetate Biochim. Biophys. Acta, 66, 1-16 (1963). 24. Nishimura, M., Studies on the electron-transfer systems in photosynthetic bacteria. II. The effect of heptylhydroxyquinoline-N-oxide and antimycin A on the photosynthetic and respiratory electron-transfer systems. Biochim. Biophys. Acta, 66, 17-21 (1963). 25. Vernon, L. P., Photooxidation of reduced 2,6-dichlorophenolindophenol by chromatophores of Rhodospirillum rubrum under anaerobic conditions. ^.20^ in Studies on Microalgae ami Photosynthetic Bacteria (Japanese Soc. of Plant Physiol., ed.). Univ. of Tokyo Press, 1963. 26. Eastman, J. W., Engelsma, G., and Calvin, M., Free radicals, ions and donor-acceptor complexes in the reaction: chloranil ^ N,N-dimethylaniline ^ crystal violet. J. Am. Chem. Soc, 84, 1339 (1962). 27. Jacobs, E. F., Phosphorylation coupled to electron transport initiated by substituted phenylenediamincs. Biochem. Biophys. Res. Co))imun., 3, 536 (1960). 28. Goedheer, J. C, Spectral and redox properties of bacteriochlorophyll in its natural state. Biochim. Biophys. Acta, 38, 389 (1960). 29. Calvin, M., and Androes, G, M., Primary quantum conversion in photo- synthesis. Sc?e»ce, 138, 867 (1962). 30. Arnon, D. I., Cell-free photosynthesis and the energy conversion process, p. 489 in A Symposium on Light and Life (W. D. McElroy and B. Glass, eds.). The Johns Hopkins Press, Baltimore, 1961, 31. Hood, S. L., Unpublished experiments. 32. Fuller, R. C, Smillie, R. M., Rigopoulos, N., and Yount, V., Comparative studies of some quinones in photosynthetic systems. A re/?. Biochem. Bio- phys., 95, 197 (1961). PHOTOOXIDATION AND PHO TOR EDUCTION REACTIONS 267 33. Lester, R. L., and Crane, F. L., The natural occurrence of co-enzyme Q and related compounds. J. Biol. Chem., 234, 2169 (1959). 34. Nishlmura, M., Studies on bacterial photophosphorylation. IV. On the max- imum amount of delayed photophosphorylation induced by a single flash. Biochim. Biophys. Acta, 59, 183-188 (1962). 35. Geller, D. M., Oxidative phosphorylation in the extracts of Rlio do spirillum rubnim. J. Biol. Chem., 237, 2947-2954 (1962). 36. Horio, T., and Kamen, M. D., Observations on the respiratory system of Rhodospirillum nibrum. Bioche)n., 1, 1141 (1962). 37. Horio, T., Yamashita, J., and Nishikawa, K., Photosynthetic adenosine tri- phosphate formation and photoreduction of diphosphopyridine nucleotide with chromatophores of Rhodospirillum rubrum. Biochim. Biophys. Acta, 66, 37-49 (1963). 38. Baltscheffsky, H., Flavin nucleotides and light-induced phosphorylation in cell-free extracts of Rhodospirillum rubrum. Biochim. Biophys. Acta, 40, 1-8 (1960). 39. Bose, S. K., and Gest, H., Electron transport systems in purple bacteria. Hydrogenase and light-stimulated electron transfer reactions in photo- synthetic bacteria. Nature, 195, 1168-1171 (1962). 40. Baltscheffsky, H., and Baltscheffsky, M., Inhibitor studies on light-induced phosphorylation in extracts oi Rhodospirillum rubrum. Acta Chem. Scand., 14, 257-263 (1960). 41. Nishimura, M., Studies on