~ . ‘V ‘ V 311,. mfl’m UNIVERSITY .3‘ v « . .'.n* 3 MAR 0 4 2002 £13 1 9 0 3 W: 25¢ per do per item RETURNING LIBRARY MATERIALS: —__‘__ Place in book return to any charge from circulation ncol INVESTIGATIONS ON THE MECHANISM OF RIBULOSE-1,S-BISPHOSPHATE CARBOXYLASE/OXYGENASE BY Stephen D. McCurry A DISSERTATION Submitted to Michigan State Univeristy in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1979 ABSTRACT INVESTIGATIONS ON THE MECHANISM OF RIBULOSE-1,S-BISPHOSPHATE CARBOXYLASE/OXYGENASE BY Stephen D. McCurry Several investigations concerning the active site and properites of the purified ribulose-1,5-biSphosphate (ribulose-Pz) carboxylase/ oxygenase (E.C. 4.1.1.39), a chloroplast protein, are discussed in this thesis. This enzyme catalyzes the addition of C02 or 02 to ribulose- P2 to yield either two molecules of phosphoglycerate or one molecule of phosphoglycerate and one of phosphoglycolate. When a 50% saturated ammonium sulfate slurry of this enzyme was stored by freezing into tiny beads in liquid nitrogen, the activity could be restored to 90% of the freshly prepared enzyme. Enzyme frozen in this manner was stored in a -80° freezer for up to six months without appreciable loss in activity. Similar results have been obtained for enzyme frozen in the absence of ammonium sulfate. When the substrate for this enzyme, ribulose-PZ, was treated with base, pH 9, or heated, it was found to degrade to inhibitory dicarbonyl compounds, and to epimerize to xylulose-PZ, a potent inhibitor of this enzyme. This instability of the substrate may be responsible for the substrate inhibition which has been claimed for this enzyme. Both xylulose-P2 and xylitol-P2 are potent inhibitors of ribulose- P2 carboxylase/oxygenase. The enzyme was 50% inhibited at a xylulose-P2 concentration of .56 HM at an active site concentration of .112 UM, but 100% inhibition could not be obtained with this compound. The inhibition due to xylitol-PZ and xylulose-P2 is competitive with respect to ribulose-Pz when these compounds are added to the enzyme with the substrate. If the enzyme is preincubated with either of these compounds, the inhibition appears noncompetitive because of extremely tight binding. Incubation of this enzyme with both C02 and Mg++ prior to initiation of the assay is required for the maximum rate of either activity. Attempts to stabilize activator C02 on the enzyme with diazomethane were not successful. While other compounds, such as 2,3-diphosphoglycerate, could not replace the requirement for C02, there was some stimulation observed at suboptimal C02 concentrations. O-Methylisourea did not activate the enzyme. Evidence for a free radical associated with the enzyme has been found through examination of the enzyme during catalysis with EPR spectroscopy at 13° K. The radical was present in all samples of the enzyme even in the absence of substrate. The origin of the radical has not been determined. All samples examined showed the presence of 2 mol of iron/mol of enzyme. Removal of 60% of the iron by dithiothreitol treatment did not alter either activity. A copper signal was observed in only one sample of enzyme and it had been prepared from spinach obtained from Florida. Rapid reaction kinetics performed in conjunction with this study demonstrated a 1 s lag in the carboxylase activity as measured by 14C02 fixation. Modification by tyrosyl residues with tetranitromethane resulted in complete inhibition of both activities. This inhibition could be partially prevented through preincubation of the enzyme with ribulose-Pz before addition of the tetranitromethane. Efforts to demonstrate a histidyl residue in the active site through modification with dibromoacetophenone were unsuccessful in that no inhibition was observed. Claims that the carboxylase and oxygenase activities could be separated through chromatography on Sepharose 63 were not supported. Similarly, no support was found for the claims that differential regulation of the two activities could be achieved through modification with iodoacetamide in the presence of ribulose-P2 or through treatment with hydroxylamine. No support was found for the notion that 02 is required for the activation of both activities. The addition of superoxide dismutase to enzyme assays was found to have no effect on either the carboxylase or the oxygenase activities. ACKNOWLEDGEMENTS I must express my gratitude to Dr. N.E. Tblbert who, in addition to supporting me financially, has given a great deal of support, advice, and encouragement. Many of the experiments in this thesis were performed with the cooperation of others in the laboratory, Dr. Christian Paech, Dr. Frederick Ryan, and John Pierce, deserve Special thanks in this regard. Thanks are due to Dr. W.A. Wood, Dr. C.H. Suelter, and Dr. R. Barker for many useful discussions and suggestions. I must thank my friend Bridget for her patience and assistance, especially during the preparation of the thesis. Finally, I must thank my parents whose encouragement and support have enabled me to endure the past few years. ii The flowers fall, for all our yearning; Grasses grow, regardless of our dislike Sengtsan from the Hsinhsinming 6th Century AODO iii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . LIST OF FIGURES. . . . . . . . . . LIST OF ABBREVIATIONS. . . . . . . INTRODUCTION 0 O O O O O O O O O 0 LITERATURE REVIEW. . . . . . . . . Structure and Properties of Ribulose-Pz Carboxylase/ Oxygenase. . . . . . . . . Biosynthesis of Ribulose-Pz Carboxylase/Oxygenase. . . Mechanism of Action. . . . . ACtivation o o o o o o o o 0 Active Site Characterization CHAPTER I. BASIC TECHNIQUES . . . . Introduction . . . . . . Purification of Ribulose-PZ Carboxylase/Oxygenase. Synthesis of Ribulose-P2 . . . . . . . . . . . . . Assay of Ribulose-P2 Carboxylase . . . . . . . . . Assay of Ribulose-Pz Oxygenase . . . . . . . . . . II. STABILITY AND STORAGE OF OXYGENASE O O O O O O O 0 Introduction . . . . . . Materials. . . . . . . . Methods. . . . . . . . . Results and Discussion . III.. STABILITY AND STORAGE OF IntrOdUCtion o o o o o 0 Results and Discussion . RIBULOSE-Pz iv CARBOXYLASE/ Page vii viii dU'lle‘w [9—5 24 24 24 31 33 34 39 39 39 39 40 43 43 43 CHAPTER Page IV. INHIBITOR STUDIES. . . . . . . . . . . . . . . . . . . . 46 Introduction . . . . . . . . . . . . . . . . . . . . . . 46 Materials. . . . . . . . . . . . . . . . . . . . . . . . 47 Methods. . . . . . . . . . . . . . . . . . . . . . . . . 47 Xylulose-P2 and Xylitol-PZ: Discussion of Results . . . 54 Discussion of Other Inhibitors . . . . .'. . . . . . . . 67 V. ACTIVATION STUDIES . . . . . . . . . . . . . . . . . . . 69 Introduction . . . . . . . . . . . . . . . . . . . . . . 69 Materials. . . . . . . . . . . . . . . . . . . . . . . . 69 Methods. . . . . . . . . . . . . . . . . . . . . . . . . 69 Discussion of Results with Diazomethane. . . . . . . . . 71 Results with O-Methylisourea . . . . . . . . . . . . . . 73 Results with Neodymium . . . . . . . . . . . . . . . . . 75 Results with Phorphorylated Compounds. . . . . . . . . . 77 VI. EPR SPECTROSCOPY, IRON, AND RAPID REACTION KINETICS. . . 81 Introduction . . . . . . . . . . . . . . . . . . . . . . 81 Materials. . . . . . . . . . . . . . . . . . . . . . . . 81 Methods. . . . . . . . . . . . . . . . . . . . . . . . . 81 Discussion of the EPR Spectroscopy . . . . . . . . . . . 82 Iron Content of the Enzyme . . . . . . . . . . . . . . . 88 Copper Content of the Enzyme . . . . . . . . . . . . . . 93 Rapid Reaction Initial Kinetics. . . . . . . . . . . . . 97 VII. ACTIVE SITE MODIFICATION . . . . . . . . . . . . . . . . 100 Introduction . . . . . . . . . . . . . . . . . . . . . . 100 Materials. . . . . . . . . . . . . . . . . . . . . . . . 100 Methods. . . . . . . . . . . . . . . . . . . . . . . . . 100 Results and Discussion . . . . . . . . . . . . . . . . . 101 VIII. PROJECTS TO VERIFY THE WORK OF OTHERS. . . . . . . . . . 103 Introduction . . . . . . . . . . . . . . . . . . . . . . 103 Ribulose-Pz Carboxylase/Oxygenase, One Enzyme or Two . . 103 Oxygen Activation. . . . . . . . . . . . . . . . . . . . 105 Differential Regulation. . . . . . . . . . . . . . . . . 105 Effect of Superoxide Dismutase . . . . . . . . . . . . . 109 BIBLIOGRAPHY I O O O O O O O O O O O O O O O O O O O O O O O O O O 11 1 APPENDICES (Reprints) I. Active site of Ribulose-1,S-Bisphosphate Carboxylase/ Oxygenase. In Photosynthetic Carbon Assimilation, Siegelman and Hind eds. 1978. Christian Paech, Stephen D. McCurry, John Pierce and N.E. Tblbert. APPENDICES (Reprints) II. III. IV. Inhibition of Ribulose-1,5-Bisphosphate Carboxylase/ Oxygenase by Ribulose-1,S-Bisphosphate Epimerization and Degredation Products. Biochem. Biophys. Res. Comm. 83: 1084-1092, 1978. Christian Paech, JOhn Pierce, Stephen D. McCurry, and N.E. Tolbert. Inhibition of Ribulose-1,S-Bisphosphate Carboxylase/ Oxygenase by Xylulose-1,S-Bisphosphate. J. Biol. Chem. 252: 8344-8346, 1977. Stephen D. McCurry and N.E. Tolbert. Ribulose-1,S-Bisphosphate Carboxylase/Oxygenase From Parsley Biochem. Biophys. Res. Comm. 84: 895-900, 1978. Stephen D. McCurry, Nigel P. Hall, John Pierce, Christian Paech, N.E. Tolbert. vi LIST OF TABLES Table 1. 2. 3. 6. 10. Properties of ribulose-P2 carboxylase/oxygenase from Spinach Reaction sequence of the coupled oxygenase assay . . . . . . Comparison of various storage methods for purified ribUIOSe-PZ CarbOXYIase I o I o o o o o o o o o o o o o o o o Modification of ribulose-P2 carboxylase with O-methylisourea at different PH. 80 O O O O O O O O O O O O O O O O O O O O O Modification of ribulose-PZ carboxylase with Odmethylisourea in the presence Of ribUIOse‘PZ o o o o o o o o o o o o o o o The effect of neodymium chloride on activation and assay of ribUIOSe-PZ CarbOXYIaBeo o o o o o o o o o o o o o o o o o o The effect of 2,3-diphosphoglycerate on activation of ribu1086-P2 carbOXYIaseo o o o o o o o o o o o o o o o o o 0 Iron content of reagents for assay of ribulose-P2 carbOXYIase/Oxygenase O I O O O O O O O O O O O I O O O O O O Modification of ribulose-P2 carboxylase/oxygenase with 10doacetamide O O I O O O O O O 0 O O O O O O O O O O O O O 0 Treatment of ribulose-P2 carboxylase/oxygenase with hYdtOXYlamineo o o o o o o o o o o o o o o o o o o o o o o 0 vii Page 36 41 74 76 78 79 92 107 108 LIST OF FIGURES Figure 1. 2. 3. 5. 10. 11. 12. 13. 14. Postulated reaction mechanisms of ribulose-PZ carboxylase and oxygenase O O O O O I O O I O O O O O O O O O O O O O O O O Postulated reaction mechanism of ribulose-PZ carboxylase inVOlVing a Covalent intermediate. 0 o o o o o o o o o o o o 0 Model for the catalytic and activator sites of ribulose-P2 carbOXYlaseocoo.0.0000000000000000... Elution profile of ribulose-PZ carboxylase from Sepharose 4B COl‘m‘n. O O O O O O O O O O I O O O O I I O O O O O O O O O Elution profile of ribulose-P2 carboxylase from DEAE cellulose column . . . . . . . . . . . . . . . . . . . . . . . Reaction scheme for the synthesis of 14C-xylulose-P2 from 14C-U-G-D-glucose. o o o o o o o o o o o o o o o o o o o o o o Elution profile of 14C-xylulose-P2 from a Dowex-1 C1 column. . a)The proton decoupled 15.08 MHz 13C-NMR spectrum of glycoladehyde-P revealing contamination by formaldelyde and an unknown compound. . . . . . . . . . . . . . . . . . . b)13C-NMR spectrum of glycolaldehyde-P after further purification by chromatography on DEAE-acetate . . . . . . . Inhibition pattern of ribulose-Pz carboxylase when xylitol-PZ and ribulose-P2 were added to the enzyme simultaneously. . . . Inhibition pattern of ribulose-P2 carboxylase when the enzyme and xylitol-P2 were incubated together for 20 min prior to initiation of the reaction with ribulose-PZ . . . . . Cblumn profile from Sephadex G-25 (fine) showing complete separation of bound and unbound 3H-xylitol-P2. Column profile from Sephadex G-25 (fine) showing separation of bound and unbound 3H-xylitol-P2 with and without SDS. . . . Column profile from Sephadex G-25 (fine) showing separation of bound and unbound 14C-xylulose-P2 to the carboxylase in the presence and absence of SDS. . . . . . . . . . . . . . . . The EPR spectrum 0f ribulose-P2 carboxylase at 20 mg/ml. . . . viii Page 13 18 28 30 48 50 53 53 56 58 61 64 66 85 Figure Page 15. 16. 17. 18. EPR spectra of ribulose-P2 carboxylase under three conditions . 87 EPR spectra of ribulose-Pz carboxylase after chemical mOdification O O O O O O O O O O O O O O O O O O O O O O O O O 90 The EPR spectrum of ribulose-P2 carboxylase purified from SpinaCh Obtained from Florida. 0 o o o o o o o o o o o c o o o 96 Reaction progress curve of ribulose-P2 carboxylase for very short time assays o o o o o o o o o o o o o o o o o o o o 99 ix Arabinitol-P2 Bicine Carboxyarabinitol-PZ Carboxyribitol-Pz P-CMB DEAE O-dianisidine DTT EDTA Fructose-P2 P-gluconate 2,3-glycerate P-glycerate P-glycolate NADPH Pydridoxal-P Ribulose-P2 SDS SOD Sugar-P Xylitol-P2 Xylulose-P2 LIST OF ABBREVIATIONS arabinitol-1,S-bisphosphate N,N-bis(2-hydroxyethyl) glycine 2-carboxyarabinitol-1,5-bisphosphate 2-carboxyribitol-1,S-bisphosphate P-choromercuribenzoate diethylaminoethyl 3,5-dimethyl benzidine dithiothreitol ethylenediamine tetracetate fructose-1,6-bisphosphate 6-phosphogluconate 2,3-diphosphoglycerate 3-phosphoglycerate 2-phosphoglycolate nicotinamide adenine dinucleotide phosphate pyridoxal-5'-phosphate ribulose-1,5-bisphosphate sodium dodecyl sulfate superoxide dismutase sugar phosphate xylitol-1,5-bisphosphate xylulose-1,5-bisphosphate INTRODUCTION Ribulose-1,S-bisphosphate (ribulose-PZ) carboxylase/oxygenase (E.C. 4.1.1.39) catalyzes the addition of C02 at C2 of ribulose-PZ to form two molecules of 3-phosphoglycerate (P-glycerate), as well as the addition of 02 at C2 of ribulose-P2 to form one molecule of phosphoglycolate (P-glycolate) and one of P-glycerate. This enzyme accounts for up to 50% of the total leaf protein in some plants. It has been more than thirty years since Wildman and Bonner (1) described this protein and termed it "Fraction-I" protein. At the time of its discovery, its function was unknown, and ribulose-Pz was unknown. It was not until the path of carbon in photosynthesis was elucidated through work in the laboratories of Calvin, Horecker, Ochoa, Racker, and others, that the carboxylase function of this enzyme was discovered (2,3,4,5). Otto Warburg observed in 1920 (6) that oxygen inhibited C02 fixation in Chlorella, and further work on Chlorella led Tamiya and Huzisige (7) to conclude that ..."02 competes, in some way or other, with C02 (or its derivative RHCOZ) for the enzyme.” It was more than twenty years before the in yitrg oxygenase activity of ribulose-PZ carboxylase was demonstrated in the laboratories of Ogren and Tolbert (8,9,10,11). This oxygenase activity was immediately linked with the "Warburg Effect" and with photorespiration, the light-dependent release of C02 and concomitant uptake of 02. Because of the fact that some plants have substantially avoided the loss of C02 from photorespiration, the so called C4 plants, and because this process apparently wastes energy gained in the photosynthetic light reactions, many authors have suggested that photorespiration serves no useful purpose and should be eliminated. These suggestions have stimulated a great deal of work on photorespiration in general, and on the regulation of ribulose-PZ carboxylase/oxygenase in particular. The approach of enzymologists has been to attempt to elucidate the reaction mechanism of each activity, and then determine whether they can be regulated. The research reported in this thesis has been mostly directed towards further understanding of the reaction mechanisms of this enzyme. In addition, efforts to repeat claims of differential regulation of the enzyme are reported in Chapter VIII. Chapter III contains research concerning the instability of the substrate ribulose-PZ. LITERATURE RIEVIEW Structure and Properties of Ribulose-P, Carboxylase/Oxygenase The general properities of this enzyme are summarized in Table 1. In higher plants and most algae, ribulose-P2 carboxylase is a very large enzyme with a molecular weight of 560,000 (12,13,14,15). In such plants, it is composed of 16 subunits (A838), 8 large (A) subunits of 56,000 molecular weight, and 8 small (B) subunits of 12,000 molecular weight (16,17,18). From calculations based on the size and amount of this enzyme, the concentration in the chloroplast is estimated at 0.4-0.5 mM or 3-4 mM in ribulose-P2 binding sites (19). This in 3232 concentration of up to 250 mg protein/ml has not been obtained in vito. Ribulose-P2 carboxylase is present in all known photosynthetic organisms. In most cases it is an B enzyme, but in Rhodospirillum 8 rubrum the structure is A2 (14,20), in Chlorobium thiosulfatophilum it is A6 (21), and in Thiobacillus intermedius (22) and Anabaena gylindrica (23) the structure is A8. It has been suggested that small subunits could be lost during purification because of their size (24). Large subunits from a variety of sources have a great deal in common. Antibodies to large subunits will cross-react to large subunits from quite different plants (25, 26) and their amino acid compositions are quite similar (12,24,27). Differences in the small subunit, however, are quite pronounced (25). Different members of the genus Nicotiana have been studied extensively in this regard (27,28,29,30). Table 1 Properties of Ribulose-P2 Carboxylase/Oxygenase from Spinach Molecular weight: 560,000 Daltons Subunit Composition: 8 large (A) 56,000 Daltons 8 small (B) 12,000-14,000 Daltons Substrates: D-ribulose-1,S-bisphosphate (ribulose-Pz), C02, 02 Essential cofactor: Mg2+ Reaction catalyzed and assay procedure: Mg2+ (a) carboxylase: Ribulose-P2 + C02-----iD2,3-phosphoglycerate (1) determination of 14C-3-phosphoglycerate with NaH14co3 as substrate (2) coupled enzyme assay "92+ (b) oxygenase: Ribulose-Pz + 02 % 3-ph03phoglycerate + phosphoglycolate (1) oxygen uptake by oxygen electrode or respirometer (2) phosphoglycolate determination Activation: C02 and "92+, required for both enzyme activites and selected sugar phosphates under suboptimal C02 conditions pH-Optimum: 8.2 - 8.6 Km-Values (a) carbOXYlase= Km(c02) 1o - 15 pM; Km(Ribulose-P2) 20 - 25 “M (b) oxygenase: Km(02) ca. 0.4 mM; Km(Ribulose-P2) 25 - 35 UM Inhibitors: xylulose-PZ, xylitol-Pz, ribose-S-phosphate, 2-carboxyarabinitol-P2, 2-carboxyribitol-P2, and 6-phosphog1uconate Fuctional groups: sulfhydryl, lysyl, and arginyl residues It has been claimed by Akazawa's group (3,32,33,34) that ribulose-P2 carboxylase from spinach could be separated into subunits, and that the isolated large subunit maintained its catalytic properties, albeit at a low rate. This separation was accomplished with p-chloromercuribenzoate (p-CMB). Several other groups (Tolbert, Lorimer, Chollet, personal communication) have tried to repeat these experiments without success. Recently Chollet (35) has reported that p-CMB separation of subunits could be accomplished, but upon addition of Mg++ (essential for activation and catalysis) the purified large subunits precipitated. Since the enzyme from R. rubrum is an A2 enzyme but still retains all of the control mechanisms of the A888 enzyme, the function of the small subunit remains unclear (36). From the evidence accumulated in Wildman's group (37) the primary structure of the small subunit determines whether the enzyme from tobacco can be easily crystallized. In the systems which have been considered, studies concerning the three dimensional structure support the ABAB model. The electron microscopic evidence of McFadden and coworkers (38) suggest a cubic or pseudorhombic dodecahedron with a 4:2:2 symmetry. This has been supported by the x-ray crystallographic evidence of Baker and coworkers (39). This topic has been recently reviewed by Eisenberg (40). Serological evidence suggest that the small subunits are prevented from reacting with antibodies by the large subunits in the native enzyme (41). Chemical crosslinking agents such as tetranitromethane, dimethyl suberimidate, dimethyl adipimidate, and methyl-4-mercapto butryimidate have been used with this enzyme (42,43). Mild crosslinking treatment results in a substantial number of AB pairs, and pairs and triples of B (42). Pretreatment of the enzyme with HCO3 produced no effect on the crosslinking pattern, but pretreatment with ribulose-P2 resulted in a substantial decrease in the amount of crosslinking. Preincubation with both Mg++ and HCOj- resulted in a substantial increase in the amount of crosslinking (42). Grebanier and coworkers (42) also found a shift in the CD spectrum upon preincubation with Mg++ and HC03: which are the conditions for activation of the enzyme (discussed later). Recently it has been suggested that the carboxylase and oxygenase activities of this enzyme could be separated through chromatography on Sepharose 68 (44). Vigorous efforts by us to confirm this failed (45). Several other groups also could not confirm this report (Randall, Bahr, Chollet, personal communication). While the idea of these two activities being separable is not new, Branden (44) published the first data in support of this notion. Many lines of evidence, from many laboratories provide a proponderance of data that these two activities are inseparable, and that one active site catalyzes both activities (46). Biosynthesis g: Ribulosefgz Carboxylase/Oxygenase Considering the complexity of the structure of this enzyme, the complexity of its biosynthesis is to be expected. The large subunit is coded for by chloroplast DNA and is synthesized on chloroplast ribosomes (28,47,48,49,50,51). The small subunit is coded for by nuclear DNA and is synthesized on cytoplasmic ribosomes (48,50,52). The small subunit is made as a longer precursor protein and is cleaved upon insertion into the chloroplast (53,54). Formation of the large subunit is not required for synthesis of the small subunit (55). The portion of chloroplast DNA which codes for the large subunit has been identified and cloned (49,56). Since the large subunit has thus far proven too difficult to sequence, it is hoped that the sequence of the DNA will soon be determined. Mechanism 2: Action The most widely accepted of the proposed carboxylase mechanisms is shown in Figure 1. This is essentially the mechanism suggested by Calvin (57). A divalent metal, preferably Mg++ is required for assay (3,4). The standard free energies for the forward reaction have been estimated at -8.4 to -12.4 KCal/Mol (58,59) and there are no reports of successful efforts to reverse the reaction. The pH optimum for both activities (of the fully activated enzyme) is approximately 8 (60,61). Considering the proton release (Figure 1), the pH undoubtedly aids in the irreversible nature of the reaction. C02 and not HCOS is the substrate for the reaction (62), it is also the species required for activation (60). While most carboxylating enzymes use HCOi, others are known which use If the mechanism shown in Figure 1 is correct, ribulose-P2 probably binds to the enzyme before C02. The postulated oxygenase mechanism is also shown in Figure 1 (11). A comparison of these two mechanisms (Figure 1) has led to the suggestion that the oxygenase reaction is simply an ”inevitable consequence" of the carboxylase mechanism (64). Experiments designed to determine the order of substrate binding are very difficult to perform for several reasons: a) C02 has a dual role as a substrate and as an activator (60,65), so it must always be added first for maximal activity. b) In the carboxylase reaction, both of the products are identical (P-glycerate), yet only one arises from the region of the newly fixed C02. c) Since C02 is required for activation of the .Ahm.pwv omccwmhxo can ommamxonumo mmtomoadnau mo mEchmzooE ceauocmu omumHsumom .F mufimam $1.80.sz Iko- I 008 oooo :-w-oz mxnoaofo j of. . mxdaofo - . . 01.80.10 01.9.0.6 . 61.0.5.8 I r 3: 10-0-: cum . 190-com aauo IOAWI ouw 10-00 W o uohxd 2H: .s mxnoaofw :99: one. :ooboom m: noaofo - am: oznoaouzw Io-w-: :o-w mos aznoaouxw A: mznoaofm bun. 1 9:- IO-H_V-I IouwsI Ono mznoaofo 10 oxygenase, it is not possible to run the assay with no C02. d) The binding constants of P-glycerate and P-glycolate are high. When an enzyme mechanism involves a step such as proton abstraction (first step Figure 1), it is often possible to isolate this part of the reaction and demonstrate proton exchange. This has been attempted for ribulose-P2 carboxylase by Rose's group (66). Ribulose-PZ, labeled at C-3 with 3H, was incubated with enzyme in the absence of C02. The results showed no 38 exchange with the medium. Three possible explanations were offered: a) the mechanism does not proceed as suggested; b) the abstracted proton is never free in solution, i.e. it stays attached to the protein; c) C02 must bind first. This research was performed before the C02 requirement for activation was established. It probably will not be possible to decide among these three possibilities until some method to activate the enzyme in the absence of C02 is discovered. An analogue of the 3-keto branched chain intermediate (III in Figure 1) would be expected to inhibit the enzyme if the postulated reaction mechanism is correct. Lane's group discovered that the carboxylation reaction was inhibited by HCN (67). This was due to reaction of HCN with ribulose-Pz. Lane's group then went on to make 2-carboxyribitol bisphosphate from the HCN and ribulose-P2 adduct (68). This compound was a potent inhibitor of both ribulose-P2 carboxylase (68,69) and oxygenase (70). Their product was actually an isomeric mixture of 2-carboxyribitol-bisphosphate and 2-carboxyarabinitol-bisphosphate, but it has consistently been referred to as 2-carboxyribitol-bisphosphate. The KD for the mixture is 10"8 M (69), and divalent metal is required for 11 maximal inhibition (69). Ribulose-P2 (69,71) and carboxyribitol-Pz (71) have similar effects on UV absorption spectrum of the carboxylase. Mildvan's group demonstrated that the addition of carboxyribitol-PZ to enzyme which had been incubated with H13C0; and Mn++ abolished the effect of divalent metal on the relaxation rate of water protons, or of H13CO§ (72). This would suggest that either the H13C05'was forced from the region of the Mn++ or its rapid dissociation was blocked. It has been demonstrated that carboxylase treated with H14CO§ and carboxyribitol-Pz binds 14C02 stoichiometrically, and the enzyme-14C02- carboxyribitol-PZ complex is stable to passage over Sephadex G-75 (73). Since this experiment was done under conditions required for activation of the enzyme, and because of the similarity of carboxyribitol-PZ to the postulated intermediate III (Figure 1), this work has been used as evidence that the activator and catalytic C02 sites are distinct. Recently, the mixture has been resolved into its two isomers, carboxyribitol-P2 and carboxyarabinitol-P2 (74,75). Of the two compounds, it is the carboxyarabinitol-P2 which is the better inhibitor. The Koff of carboxyribitol-P2 is less than 10"11 M. If one assumes that the better inhibitor most closely resembles the reaction intermediate III (Figure 1), then the intermediate probably has the arabino- rather than the ribg-configuration (75). This raises a potential problem with the stereochemistry of the P-glycerate which arises from the top half of the ribulose-Pz. It is known that the configuration is D, but if the intermediate is in the arabino- form, the P-glycerate formed with C02 would appear to be L. This problem is under investigation by John Pierce in Tolbert's laboratory. 12 The postulated intermediate III has been synthesized (76). The authors stated that the compound, 2-carboxy-3-ketoribitol-1,5- bisphosphate, rapidly decomposed in the absence of carboxylase either through loss of C02, or to a mixture of D and L P-glycerate. The intermediate was also a substrate for the enzyme, but the stereochemistry of the products was not determined (76). It is not clear whether the synthesized intermediate was really the ribg- form or whether it could have been the arabino- form. Rabin and Trown (77,78,79) have suggested a somewhat different mechanism (Figure 2) for the carboxylase reaction. In their model there is a covalent attachment of the substrate, ribulose-Pz, at C-2 to a sulfhydryl group on the protein. This mechanism would involve loss of the oxygen at C-2. Two separate laboratory groups have recently demonstrated that the oxygen atoms at both C-2 and C-3 are retained during the reaction (80,81). These reports would suggest that either no covalent attachment exists at these carbons, or that the enzyme returns to the product the same oxygen that it removes from the substrate. As far as the oxygenase mechanism is concerned (Figure 1) it has been shown conclusively that P-glycolate arises from the top two carbons of ribulose-P2 (82). Studies with 180 have shown that the oxygen fixed in the oxygenase reaction is incorporated into the carbonyl carbon of P-glycolate (11). The first step in the postulated mechansim of Figure 1 is the abstraction of a proton from C-3 of ribulose-P2 by a base on the protein. Sugar bisphosphates which do not have the ribo- configuration at C-2, C-3, have been found to be potent inhibitors of the enzyme. These are 13 CHZOPO3H- CH20PO3H- CH20PO3H- -S C=O -S -C-OH -S-C H-C-OH H-C-OH HO-C -H +H20 _ F H-C-OH H-C-OH H- -OH CH20P03H' CH20P03H‘ CHZOP O=C=O CH20P03H“ CH20P03H' O -S H-C-OH -S-C- l ‘0' C00” ‘2 i=0 H -OH €00- CH20PO3H- H-C-OH CHZOPOBH- Figure 2. Postulated reaction mechanism of ribulose-P2 carboxylase involving a covalent intermediate (Rabin and Trown 77,78,79). 14 xylitol-1,5-bisphosphate (xylitol-Pz) (83), xylulose-1,5-bisphosphate xylulose-PZ) (84), and arabinitol-1,S—bisphosphate (arabinitol-PZ) (John Pierce, personal communication). Besides their obvious similarities to ribulose-P2 in size and bisphosphate nature, these compounds all posess a hydroxyl at C-3 which could allow hydrogen bonding in the region where proton abstraction occurs. The importance of the stereochemistry at C-4 is not known. The oxygenase mechanism shown in Figure 1 was proposed (11) by analogy to the carboxylase mechanism of Calvin (57). There is a major problem with the mechanism as written; 02 is in the triplet state, i.e. it has two unpaired electrons, while the intermediate II (Figure 1) is singlet. Such a reaction is termed "spin-forbidden” (85). The energy level of triplet 02 is 22 kcal/mol below the lowest energy singlet state (85). To perform this transition, the oxygen must be reduced to the level of H20. This could be done one electron at a time generating 02‘ as the first product, two electrons at a time proceeding through H202, or four electrons at a time breaking the O=O bond. Two known ways of solving this problem are: complexing the 02 with a transition metal to unpair the electrons, or forming a resonance-stabilized free radical. Such a free radical would need to be stabilized by a cofactor such as a flavin, or some similar prosthetic group on the protein (85,86). Hamilton states this very strongly (85); "Thus, every oxygenase has one or both of the following characteristics: (a) the enzyme requires a transition metal ion (usually iron or copper) for activity; (b) the enzyme has a cofactor or substrate, the reduced state of which gives a highly resonance stabilized free radical by the loss of an electron or the equivalent of a hydrogen atom.” (emphasis in original). 15 Wildner (87) suggested that the form of oxygen involved in the oxygenase reaction was 02', however he offered no data to support this. The only data published to support this claim is from Bhagwat and Sane (88). Their data were that superoxide dismutase (SOD), purified by them from bovine erythrocytes, inhibited ribulose-Pz oxygenase, they offer no data concerning the carboxylase function. We tried to repeat their conditions, using SOD from Sigma and found no effect on either activity (11, Paech and McCurry unpublished observations, Chapter VIII, this thesis). Lorimer (personal communication) also has attempted this experiment using copper penicillamine, a small molecular weight SOD, and found no effect. Wildner (personal communication) claims to have confirmed Bhagwat and Sane (88). If 02' is the real substrate, then it must be made by the enzyme from 02, because its concentration in free solution is much too low to account for the observed rates. If it were being made by the enzyme then it could simply be unavailable for reaction with SOD. Data will be presented later in this thesis (Chapter VI) concerning EPR spectroscopy performed during catalysis. These spectra fail to show any 02‘ signal. Mg++ is not capable of performing the role of a transition metal in regards to the required spin-inversion. The potential involvement of 02‘ remains to be resolved. Activation There are two ways in which C02 is involved in the mechansim of this enzyme. It is a substrate, and it is essential for maximal activation of either carboxylase or oxygenase activity. The requirement of C02 for activation was discovered by Pon and coworkers (65,89) in 1963. This early work described 002 as a homotropic effector for 16 ribulose-P2 carboxylase, and stated that prior to initiation of the reaction, the enzyme required incubation with C02 amd Mg++ for maximal activity. It was not suggested that C02 was required at more than one site. This report was relatively ignored by others in the field. Akazawa's group confirmed the effect (90), but it still did not become routine to activate the enzyme before each assay. It was through the work of Bahr and Jensen (91,92) on enzyme prepared from freshly ruptured chloroplasts that the problem of activation became fully recognized. They observed that carboxylase prepared from freshly ruptured chloro- plasts had a lower Km for C02 than purified enzyme did. This increase in Km was traced to loss of C02 from the enzyme. The requirements for full activation of both activities was then worked out by Lorimer and coworkers (60,61). One of the most fundamental features of this discovery was that the pH Optimum for fully activated ribulose-P2 oxygenase was found to be 8.0-8.2 and 9.2-9.4 (61). Fully activated enzyme did not show any differential regulation with respect to sugar phosphates (93), as had been claimed for the unactivated enzyme (70). Several compounds have been shown to stimulate the enzyme to varying degrees, these include 6-P-gluconate, fructose-P2, NADPH, P-glycerate (94,95,96,97) and pyridoxal-P (98). All of these compounds activate only under certain conditions: a) they must be present only in small amounts, b) they must be incubated with the enzyme prior to the addition of ribulose-PZ, and c) they must be used at suboptimal concentrations of C02 or the effect will not be observed. In no case is the activation from these compounds as great as the activation from saturating amounts of C02. When these compounds are present in the assay in high concentrations, they are inhibitory (96,97). It has been suggested that 17 a separate sugar-P site exists for the activator role of these compounds, with their inhibitory role being at the ribulose-PZ catalytic site (95,97,98). Similarly, it has been argued that ribulose-Pz, under conditions of low C02 will bind to this sugar-P site and inhibit the enzyme (94,98,99). A model for control of this enzyme, involving five distinct sites for various compounds has been proposed (97). This model involves the catalytic site for the binding of ribulose-PZ, C02, and 02, an allosteric site for the binding of ribulose-PZ, a second allosteric site for the binding of Mg++ and HCOj, a third allosteric site for the binding of 6-P-gluconate and fructose-P2 and a fourth allosteric site for NADPH, with some lack of specificity of the sites. This model seems unnecessarily complicated. A simpler model has been proposed by Tolbert's group (100) and is shown in Figure 3. In this model, there is one active site for ribulose-P2 and C02, and one activator site for Mg++ and C02. The sugar-P effectors exert their effect at the catalytic site, and the positioning of 02 is uncertain. It has been known for some time that incubation of the carboxylase with ribulose-PZ in the absence of C02 results in inhibition of the enzyme activity (97) and thus the proposed allosteric site for ribulose-Pz. Recently we have discovered that the substrate ribulose-P2 is quite labile with respect to base (101). under mild base treatment, including storage at pH 8.0 and -20°, ribulose-P2 can epimerize to xylulose-PZ, as well lose the phosphate group from C-1 (101). That these products may account for ribulose-P2 inhibition is discussed in more detail in Chapter III. The necessity of a divalent metal (preferably Mg++) for the activity Large Subunit Activator site Catalytic site Lys ( T NIH Arg Lys ’C\ 6 ® 0’; O 3 6 .29 NH3 Ma ' \ é) 9H2 9 7 0:0 c e i II S H-C-OH O H-(ll-OH E J 9H2 7 ®9 ._@__ Arg Figure 3. Model for the catalytic and activator sites of ribulose-P2 carboxylase (100). 19 of this enzyme has been known since the identification of the function of this protein (3,4). Pon observed that Mg++ was required for the homotropic effect of C02 (65). Two groups have considered the role of metal through the use of Mn++ and EPR spectroscopy (72,102). Chu and coworkers (102) found a stoichiometry of 3 Mn++/mol of enzyme, a rather surprising ratio for an enzyme with 8 active sites. The explanation may lie in the fact that these workers failed to supply any C02 to their enzyme-Mn++ incubation mix. If it had not been for the presence of endogenous C02 in their samples the ratio would have been smaller. Miziorko and Mildvan (72) were aware of the effect of C02 and found that when the carboxylase was incubated with sufficient 002 for complete activation, a stoichiometry of 8 mol of Mn++/mol of enzyme was found. They were able to determine the distance from Mn++ to H13C05'to be 5.4 A. This was determined by measuring the effect of Mn++ on the longitudinal relaxation rate of the H13C05 With 13C-NMR. It is clear from their data that the 13C signal is not a carbamate. Addition of carboxyribitol-P2 (probably the isomeric mixture) abolished the effect of Mn++. My intrepretation of this data is that the C02 in the enzyme-COZ-Mn complex exchanged so slowly that no signal for this C02 was observed (discussions with John Pierce helped clarify this point). This data also implies that carboxyribitiol-P2 binds very close to the Mn. It is clear that carboxyribitol-P2 requires M++ for maximal binding (69). Miziorko (73) has recently demonstrated that enzyme activated with Mg++ and H14CO§ and then mixed with carboxyribitol-P2 (the isomeric mixture) incorporates 14C in stoichiometric amounts. If carboxyribitol-Pz is really an analogue of the reaction intermediate (III Figure 1) then it should exclude C02 from the catalytic site. Using this 20 reasoning then the bound C02 represents activator C02. One likely way for C02 to be bound for activation would be through a carbamate to the E-amino group of a lysyl residue (Figure 3). Such a carbamate could be stabilized by M++, perhaps through some additional anion at the active site. Using the enzyme from R. rubrum, O'Leary and coworkers (103) have demonstrated carbamate formation with 13C02 using 13C-NMR. While they did not rule out the possibility of a Tris-C02 carbamate in their paper, they have since run this control and ruled it out as a possible source of the signal (F. Hartman, personal communication). Since catalytic C02 would not be expected to be bound as a carbamate, this data can be used as further evidence of two distinct C02 sites. If there are two sites for C02 and the activator C02 does not participate in catalysis, then the activator species might be identifiable by means of an isotope trapping experiment. Lorimer (104) has performed such an experiment. The enzyme was fully activated using Mg++ and high Specific activity H14C05. The activated enzyme was injected into a large excess of H12CO§ and ribulose-PZ and allowed to catalyze for a short period. Subsequently, the enzyme was rapidly separated from the smaller molecular weight species by gel filtration in the presence of a large excess of Mg++. The resulting enzyme fraction contained a 40 fold higher specific activity (of the 14C) than would have been expected if the activator C02 were used in catalysis. This is evidence for a distinct activator site for C02, and argues against the suggestion that the bound activator C02 is subsequently fixed (65,89). In an effort to irreversibly activate the carboxylase, two groups 21 have used the isoelectronic analogue of C02, cyanate (105, 106). In each case it was found to be a potent inhibitor, although the mode of inhibition was not agreed upon (105,106). Active Site Characterization Armed with some knowledge of the mechanism of an enzyme, it is often possible to make some assumptions about the nature of the catalytic site. With much less data available to them than is available today, Rabin and Trown (77,79) made some speculations about the active site of ribulose-Pz carboxylase. They proposed a sulfhydryl group adjacent to C-2 of ribulose-P2 for covalent attachment during catalysis. There may be such a sulfhydryl, but it doesn't function as suggested (80,81). They also suggested that some sort of base was involved in the binding of C02. This has been shown to be the case for C02 activation and is likely to be true for the catalytic C02 as well. Lastly, they suggested the presence of a base to function in the proton abstraction at C-3 of ribulose-Pz. They offered some data on the presence of SH groups. Using 5,5'-dithiobis-(2-nitrobenzoic acid) Ellman's reagent (107), they found 4 fast-reacting SH groups/mol enzyme. Two of these could be protected by preincubation with ribulose-PZ. No explanation for this stoichiomentry was offered. Pyridoxal-5'-P has been found to be a potent inhibitor of ribulose-P2 carboxylase/oxygenase (108,109). This compound is known to react with primary amino groups, such as the e-amino groups of a lysyl residue, through formation of a Schiff base. Paech and Tolbert (110) demonstrated that at pyridoxal-P concentrations sufficient to cause 100% inhibition of both activities of the enzyme purified from spinach, only 22 the E-animo group of lysyl groups were modified by reduction of Schiff base with NaBH4. The inhibition by pyridoxal-P is competitive with respect to ribulose-P2 and C02 (108,110). There is clearly one class of rapidly reacting lysyl groups amounting to 8 mol per mol of enzyme (110). The enzyme from spinach requires the binding of 16 mol of pyridoxal-P per mol of enzyme for complete inhibition (110), however proteolysis of carboxylase modified by pyridoxal-P followed by reduction with NaB3H4, shows only one significant peptide and several minor ones (111,112). Data obtained from 5; rubrum, an A2 enzyme, suggest that modification of only one lysyl group per enzyme molecule results in complete inhibition (113,114). Pyridoxal-P is an especially good probe for the active site since it can be used either as a covalent modifier, or as a non—covalent competitive inhibitor for kinetic studies. There has been research to design chemical compounds which resemble the substrate ribulose-Pz, but form stable complexes with the enzyme. These affinity labels have been made and used with success by Hartman's group (115). The best results have been with compounds such as 3-bromo-1,4-dihydroxy-2-butanone-1,4-bisphosphate (116,117,118) and N-bromoacetyl ethanolamine phosphate (119). These reagents modify both cysteine and lysine residues, and activation causes a shift towards modification of lysyl residues. The modification of a sulfhydryl lends credence to earlier reports of SH groups in the active site (77, 120). Incubation of the enzyme with 3-bromo-1,4-dihydroxy-2-butanone-1,4- bisphosphate resulted in the modification of two different lysyl residues (121). Tryptic peptides containing these modified residues have been sequenced (121). Activation of the enzyme before modification results in 23 an increase in the amount of one of these peptides, termed "peptide I" by the authors. Recently, a tryptic peptide has been obtained and sequenced from enzyme modified with pyridoxal-P and NaBH4 (111,112). Both the pyridoxal-P and the butanone derivative modify the same lysyl residue on "peptide I”. Lorimer has had some success in attaching 14C02 to the enzyme with diazomethane (personal communication). It is attached to a lysyl residue, and the sequence is being determined. Essentially all of the compounds which compete with ribulose-P2 for the catalytic site are phosphate esters. Several groups have designed experiments to determine the nature of the residues responsible for binding the phosphates. In other proteins arginyl residues perform this function (122). The reagents of choice for arginyl residues are diketo compounds, such as cyclohexanedione and glyoxal (123). Results with butanedione (124,125) indicate the presence of 2 to 3 reactive arginyl residues per large subunit, which can be protected by preincubation with ribulose-Pz. This work was done with R; rubrum (124) and Pseudomonas oxalaticus (125). Paech and Spellman (Tolbert's group, unpublished) have found similar results with spinach, and Chollet (126) has repeated the experiments on tobacco. To date, no one has reported any data concerning arginyl residues at a distinct activating site. CHAPTER I BASIC TECHNIQUES Introduction The work discussed in this thesis concerns aspects of research on ribulose-P2 carboxylase/oxygenase. While the specialized methods are discussed with each set of experiments, the basic techniques common to all sections are described in this chapter. Purification of ribulose-PZ carboxylase/oxygenase Glycylglycine, HEPES (N-2-hydroxyethylpiperazine-N'-2-ethanesul- phonic acid), Bicine (NN-bis-(Z-hydroxyethyl)-glycine), B-mercaptoethanol, and polyvinylpolypyrrolidone were from Sigma Chemical Company. Ultra-pure ammonium sulfate was from Schwarz-Mann. The spinach used came from a variety of sources: grown in our growth chambers, green house, or field plots; purchased locally; flown in from Oklahoma, California, Florida, or Georgia. No differences were found in the activity of different preparations of enzyme which could be unequivocally attributed to the source of the spinach. Ribulose-PZ carboxylase/oxygenase was pruified from spinach leaves in the following manner: 500-1000 g of washed, deveined leaves were ground with a two fold excess (v/w) of grinding buffer. The buffer at pH 7.8 to 8.2 was 25mM Bicine (HEPES or glycylglycine were used occasionally) with 1 mM EDTA, 10-50 mM B-mercaptoethanol, and 0-2% 24 25 polyvinylpolypyrrolidone (w/v). The grinding and subsequent steps were performed at 4°. The leaves were ground in a 4 l waring blender at slow speed for 40-60 sec. The resulting homogenate was allowed to stand for 10-15 min to facilitate decanting by allowing the cellular debris to rise to the top of the homogenate. The homogenate was filtered through 2-8 layers of cheesecloth and 1-4 layers of miracloth. The excluded debris was discarded. The homogenate was centrifuged at 10,000-11,000 x g for 30-45 min. The resulting supernatant was poured through either miracloth or glass wool for removal of dark green lipid material. The supernatant was made 37% saturated with a solution of saturated ultrapure (NH4)ZSO4. The saturated solution had been previously prepared, stored at 4°, and the pH adjusted to 7-8 with NH4OH. The enzyme solution was then stirred for 5-20 min and allowed to stand for an additional 10-40 min before centrifugation as before. The supernatant from the second centrifugation was made 50% saturated in (NH4)ZSO4 to precipitate the carboxylase and allowed to stand and then centrifuged as before. The precipitate from the 50% (NH4)ZSO4 centrifugation was resuspended in 25 mM buffer (Bicine, HEPES, etc.), 1 mM EDTA, 10 mM B-mercaptoethanol, pH 7.8-8.2. The resuspension volume was kept as low as possible; from 500 g of leaves, the volume was 25-30 ml. Great care was taken in resuspending the precipiate employing gentle stirring with a glass rod. The resuspended precipitate was centrifuged at 27,000 x g for 10-15 min for clarification. The supernatant from this spin was either loaded onto a Sepharose 4B column (2.5 x90 cm) and chromatographed overnight, or stored overnight for zonal centrifugation. In the latter 26 case a 10-30% (w/w) linear sucrose gradient was used in a 750 ml zonal rotor and spun at 30,000 rpm for 4 hr. The buffer in either case was the resuspension buffer. In either case fractions were colected and monitored by their absorbance at 280 nm. A sample Sepharose 4B profile is shown in Figure 4. The fractions containing carboxylase (the first peak after the void volume) were pooled. The pool was loaded onto DEAE cellulose (Whatman DE 52) in a 6 x 40 cm column. The column had been previously equilibrated with the resuspension buffer. After application of the protein sample, the column was washed with 500-1000 ml of resuspension buffer. The sample was eluted with a 0-0.4 M NaHCO3 2 1 gradient. The fractions were monitored as before. A sample profile is shown in Figure 5. The peak fractions were pooled and usually stored as 50% ammonium sulfate slurry at 4°. under these conditions the specific activity remained relatively constant for 3-4 weeks. The yield of purified carboxylase from 500 g of spinach leaves was 0.7-1.0 9. To prepare enzyme for assay, the desired amount of slurry was centrifuged at 15,000-20,000 x g for 10-20 min. The precipitate was usually resuspended in a buffer at pH 8.0 to 8.2 of 100 mM Bicine, 1 mM DTT, and 0.4 EDTA. The resuspension volume was adjusted so that the resulting protein concentration was 20-40 mg/ml. The sample was placed in dialysis tubing and dialyzed overnight against the resuspension buffer. Improved storage methods are discussed in Chapter II. 27 .H5 moww md3 omen coauomum .casaoo mcma on» now ooHoom mama vaNN mace» locum .5: omm um coccouomnm Haws» mo coHumcwsuoumo Mow Homwsn cadaoo on» spas oaom om oousaflo mums mcoHuomuw are .casaoo mv omoumnmmm Baum ommamxonumo NmIomOHanu mo oHHmoum coausHm .v assent 28 9v mmmZDz 20_._.O<¢m DIN O. 1n. 29 .Ha om no: omen coHuomuu och .omwuoum new ovumasm Edficoaam ca omuMHsumm wom moms can omaoom mums mm 1mm oceauomum .5: cmN um cocooHOmnm mean» no cofluMCHEuouoo you “moosez ozs gotten escaoo can sue; tact op emusHHt mums meoauoeum och .uxou 0:» ca ownwuomoo mum mGOSUSocOU cadaoo on? .casaoo mmoasHaoo Home Scum onwamxonumo mmlomoasnau mo oHHmoum coausHm .m 9.3on m «mass 2 zo_bo P-Glycerate + P-Glycolate Mg++ P-Glycolate P-Glycolate phosphatase > Pi + Glycolate Glycolate + 02 Glycolate oxidase :>. Glyoxylate + H202 Horseradish H202 + o-Dianisidine (reduced) peroxidase 3} H20 + o-Dianisidine (oxidized) 37 coupled system will work without running assays. Because of the requirement that ribulose-P2 oxygenase be activated with C02, and because this C02 rapidly dissociates from the enzyme under the low C02 conditions necessary for the oxygenase assay, the oxygenase only fixes 02 at a linear rate for 1-2 min. This requires that the lag period be kept as short as possible. Considering the amount of time required for mixing, making the lag shorter than 5 S would be unnecessary. The final enzyme in the series, horseradish peroxidase has a rate constant for H202 of 108 (M'1 x 8‘1) (133). As a result its Km is essentially zero. According to the catalogue from Boehringer (Biochemica Information) the peroxidase couples to o-dianisidine with a rate 8-10 times higher than other electron donors. These two facts mean that a few units of the peroxidase in an assay makes the lag time of this step essentially zero. Therefore a lag of 2.5 s could be allowed for P-glycolate phosphatase and glycolate oxidase. The Km's of these two enzymes, in this system have been determined to be 160 HM and 400 uM respectively (Christian Paech, unpublished). The required amount of each of these two enzymes is given by 0=Km/T, where U is the number of units of activity, and T is the acceptable lag in min. For P-glycolate phosphatase, 2.5 sec = .042 min, and Km = 160 uM so U = 3.8 Mol/min.ml. For glycolate oxidase, the required number of units is 9.6. We have been able to prepare P-glycolate phosphatase of high specific activity. George Santora in Tolbert's lab has prepared this enzyme from tobacco with a specific activity of 70 U/ml; this is clearly enough for the coupling assay purposes described here. Glycolate oxidase has been more difficult. Initially it seemed that the method of Harris and Stern (134) would be the way to purify the enzyme. They reported a specific 38 activity of 3092.5 UMol/min/mg protein. We were unable to reproduce this work. Harris (personal communication) belatedly informed us that they had made a 1000 fold error in reporting their data and that the correct activity was 3.0925 UMol/min/mg protein. As of this writing, we have not been successful in obtaining glycolate oxidase of adequately high specific activity to make this coupled assay a reality. Nicola Selph in Tolbert's lab is working on this problem and hopefully the enzyme will be available in the near future. CHAPTER II STABILITY AND STORAGE OF RIBULOSE-P2 CARBOXYLASE/OXYGENASE Introduction This enzyme is generally stored as a 50% (NH4)ZSO4 slurry at 4°. Under these conditions it is relatively stable for 3-4 weeks. After this period of time it rapidly loses activity down to 5-10% of the activity of the freshly purified enzyme. This loss does not seem to be due to the cold lability of this enzyme. The nitrogenase protein is subject to similar losses in activity but it has been stored with success by slowly dripping it into liquid nitrogen and freezing it as small "peas" followed by storage at -80° (W.H. Orme-Johnson, personal communication). this chapter contains the results of storage under a variety of conditions including the liquid nitrogen method, and a discussion of a method for restoration of lost activity through treatment of the enzyme with high concentrations of DTT. Early results of this work have been reported (135). Materials See Chapter I. Methods The enzyme was purified and assayed as described in Chapter I. The studies with glycerol, sucrose, and (NH4)ZSO4 were done by mixing the reagents being tested with the enzyme and storing it in the manner 39 40 indicated. For the liquid nitrogen studies, the enzyme was frozen by dripping it slowly into an erlenmeyer flask containing liquid nitrogen. The flask was positioned in an ice bucket at the base of a ring stand under a separatory funnel containing the enzyme. Care was taken to freeze the enzyme into individual beads ("peas") rather than into chains. This required frequent additions of liquid nitrogen to the flask to replace that which had boiled off. When high DTT concentrations were used, the enzyme was made 50-100 mM DTT for several hours and then dialyzed against 100 mM Bicine pH 8.0. Because of the absorbance of oxidized DTT at 280 mM, it must be removed before determination of the protein concentration by 280 nM absorbtion. For routine determinations of the protein concentration, the method of Paulsen and Lane (136) was used: AZBO/ml x 0.61 = the protein concentration in mg/ml. Results and Discussion Storage of the carboxylase for six weeks under a variety of conditions is shown in Table 3. The sample frozen in 50% (NH4)ZSO4 retained its activity the best, but it was still lower than the activity of the freshly prepared enzyme by 40%. In this experiment, the samples were dialyzed overnight, activated with 2 mM DTT, 20 mM MgC12, and 10 mM NaHCO3, and held at 30° for 1 hour before assay. Recently, Nigel Hall in Tolbert's lab has discovered that incubation of aged enzyme of reduced specific activity could be restored to a specific activity approximating that of the freshly prepared enzyme, by incubation with 50-100 mM DTT for a period of hours (135,137). Neither 41 Table 3 Comparison of Various Storage Methods for Purified Ribulose-PZ Carboxylase Additions for storage Storage 50% saturated temperatures No additions 30% sucrose 20% glycerol ammonium sulfate 4° 3.3 i 6% 4.1 i 2% 3.3 :11% 4.0 i 2% 22° 3.3 + 6% 1.9 3% 3.0 + 7% 3.5 + 4% n|+ Enzyme from spinach leaves was stored under the conditions indicated for 6 weeks, dialyzed overnight, and activated before assay. Conditions are described in the text. The assay volume was 0.25 ml and the specific activity of the NaH14CO3 was 220 cpm/nMol. values Shown are cpm x 10'”3 and are the average of 3 assays. The original activity before storage was about 8,000 cpm. 42 B-mercaptoethanol nor Na-ascorbate worked as well as DTT. This means that enzyme can be conviently stored as a 50% (NH4)ZSO4 slurry at 4° for 2-3 months without substantial loss in restorable activity. The possibility of storing the enzyme by freezing it in small beads in liquid nitrogen was investigated. Enzyme frozen in this manner was subsequently stored in a -80° freezer. Two storage conditons were investigated. In one case, a 50% (NH4)2804 slurry was frozen, in the other case, a dialyzed protein solution of 30 mg/ml was frozen. In the first case, enzyme stored for 6 months retained 70% of its original activity, addition of 50 mM DTT restored the activity to 85 to 95% of its original activity. In the second case, enzyme stored for seven weeks retained 35% of its original activity and addition of 50 mM DTT restored the enzyme to 105% of its original activity. It is expected that when the optimum conditions for the freezing are determined, it will be possible to store the enzyme in this manner for months to years with no adverse effects on the enzyme activity. CHAPTER III STABILITY AND STORAGE OF RIBULOSE-P2 Introduction In the more than 20 years since its discovery, many workers have discussed the substrate inhibition of the carboxylase/oxygenase by ribulose-P2 (3,65,97,136). This phenomenon has been attributed to an allosteric binding site (97), or competitive binding for ribulose-P2 and C02 (108), and such binding can produce a conformation change (71,138). Since the elucidation of the required conditions for activation (60), it seems clear that in some cases (97) the ribulose-Pz was inhibiting by simply preventing activation. An additional reason for this inhibition is the presence of certain inhibitors in various preparations of ribulose-PZ. Because of the discovery that with different ribulose-Pz solutions of apparently identical concentrations, different initial rates of carboxylation were obtained, a study was undertaken to determine the stability of ribulose-PZ. These results have been reported at two meetings (Appendix 1 (100),139) and have been published (101). The publication is included in this thesis as Appendix 2, and therefore only the results are summarized in this chapter. Results and Discussion Since the ribulose-P2 had been prepared in our laboratory (Chapter I) chance contamination seemed unlikely and substrate degradation was suspected. One possible source of contamination was tested first 43 44 however. Because of a report that at least one batch of ATP from the Sigma Chemical Company had been contaminated with NaVO3 (140) this compound was tested and found to have no effect on carboxylase activity at concentrations up to 10 mM. When one sample of ribulose-PZ was incubated at pH 11 and 30° for varying amounts of time (Appendix 2, Figure 2) and then tested for inhibition of the carboxylase, it became clear that the degree of inhibition was dependent on the duration of the high pH treatment. The inhibitory features of these treatments were tested in an experiment designed by Christian Paech. The enzyme was mixed with the treated sample of ribulose-PZ and catalysis was allowed to proceed until the substrate concentration was zero. Then untreated ribulose-P2 was added and the enzyme was assayed. This was possible since the product from the first incubation (P-glycerate) is known to have little effect at these concentrations. After realizing that the inhibitory compounds could be generated through base and or heat treatment, a careful analysis of the ribulose-Pz was carried out. The analysis revealed that xylulose-Pz (a known inhibitor of this enzyme (84)) and a dicarbonyl compound, were present. Other dicarbonyl compounds have been used to demonstrate the presence of arginyl residues in this enzyme, and they are good inhibitors (124,125,126). It was determined that both of these compounds could be formed by exposure of ribulose-P2 to strong base, even for a brief period. The conversion of ribulose-P2 to xylulose-Pz is a reversible reaction simply involving epimerization at C-3. The dicarbonyl compound arises from an irreversible B-elimination of the phosphate at C-1. This 45 explained why frequently the amount of inorganic phosphate increased as a ribulose-P2 solution aged. Uhder the conditons of the assay the phosphate at C-1 is lost at the rate of 1.5% per hour. During the course of this study the ribulose-Pz solution that had the most inhibitory effect was prepared from ribulose-P2 which had been purchased from Sigma. At that time we attributed this to the fact that it was an old solution. At present Sigma claims that their ribulose-P2 is 95% pure by TLC, phosphate and pentose phosphate assays, but then they go on to say that with their assay using the carboxylase, the substrate is only 85-90% pure 0 One conclusion from this research was a new hypothesis concerning the condition of the enzyme in the chloroplast. At the pH and temperature in a chloroplast, the instability of ribulose-P2 could pose a problem. In the light, the internal pH can exceed 8 and on a hot day the temperature can be above 40°. However, partial enzyme inactivation does not seem to occur in the chloroplast. Even when the ribulose-PZ is made to accumulate to concentrations higher than the concentration of active sites of the carboxylase, activity is fully restored within 10 min after the additon of bicarbonate (141). It has been suggested that only a portion of the carboxylase in the chloroplasts is functional at any one time (19), and while this has been disputed (D. walker, personal communication), it leaves open the possibility that under normal conditions, when the amount of enzyme active sites probably exceeds the substrate, ribulose-P2 is not free in solution long enough to be broken down or epimerized. It may be bound immediately to the enzyme. CHAPTER IV INHIBITOR STUDIES Introduction Frequently, when a new enzyme activity is discovered, several compounds are assayed for their efficacy as substrates to determine the Specificity of the enzyme. Ribulose-PZ is the only known substrate for the carboxylase/oxygenase (130). A knowledge of substrate specificity often allows one to make certain assumptions about what parts of the substrate are most important for its binding to the active site of the enzyme, but when only one substrate is known, these assumptions cannot be made. One route for determining the requirements for substrate binding is through the use of competitive inhibitors which closely resemble the substrate. The analogue of the proposed reaction intermediate (III in Figure 1), 2-carboxyribitol-P2, has been known for many years to be a potent inhibitor (68,69). Recently it has been shown that the epimer of carboxyribitol-Pz' 2-carboxyarabinitol-P2, is an even better inhibitor (74,75). This probably means that carboxyarabinitol-P2 is more like the intermediate. Xylitol-P2 (83) has been reported to be a good inhibitor and arabinitol-P2 has also been found to be inhibitory (F. Ryan, personal communication, 1975; J. Pierce, personal communication, 1979). Ribitol-P2 has been reported to be not inhibitory (83), but this has not been confirmed with a different preparation of ribitol-Pz. The report on xylitol-P2 claimed that the pattern of inhibition was noncompetitive with respect to ribulose-Pz (83). This chapter reports the results of 46 47 experiments with xylitol-Pz and xylulose-P2 and presents data in support of the hypothesis that both of these compounds are competitive, tight-binding, non-covalent inhibitors of ribulose-PZ carboxylase/ oxygenase. Materials The 3H-xylitol-P2 was synthesized for us in the laboratory of R. Barker according to the method of Hartman and Barker (142). The xylulose-P2 was synthesized according to the method of Byrne and Lardy (143), as described previously (84). The enzyme and substrate were prepared in the manner described in Chapter I. The assays were performed in the manner described in Chapter I. Amersham was the source of (U-14C) G-D-glucose. Glycolaldehyde phosphate and dihydroxyacetone phosphate were from Calbiochem. The compounds, 2-pyridinecarboxaldehyde, 3- pyridinecarboxaldehyde, 4-pyridinecarboxaldehyde, and carboxymethoxyla- mine hemihydrochloride were all obtained from Aldrich. Iminodiacetic acid was from Sigma. The 2-phosphate and the 2-sulfate derivatives of ascorbic acid were gifts from B. Tolbert (U. of Colorado). Methods 12C-Xylulose-P2 was prepared from the condensation of glycolalde- hyde-P and dihydroxyacetone phosphate with aldolase (84). Xylulose-P2 was prepared with a 14C label (at C-1, C-2, and C-3) from U-14C- a-D-glucose, according to the scheme in Figure 6. The radioactive products were purified by ion exchange chromatography on a Dowex-1 Cl‘ column (2.1 x 50 cm), and eluted with a series of HCl concentrations. The elution profile is shown in Figure 7. In order to get a satisfactory 48 *CHO *cno *iHZOH H*-C-OH H*-C-OH *c=o | 1 | 2 HO-*C-H Mg++ HO*-C-H HO*-C-H x x H*-C-OH “g I H*-C-OH '7 H*-C-OH ATP ADP | | H*-i-OH H*-C-OH H*-C-OH *CH20H H2*-C-OPO3H- H2*—C-OPO3H- ATP ‘ 3 Mg++ ADP .mmuaouaasx 0: Am “amtaouaasx z: m E “~183st z: 3 8 “managed? 2: em 3 “SASS? 21 av Ad .0>Huaummfioo on ou CBOnm ma coauananca mo cumuumm one .wamsomcouHSEwm mahuco on» on omooo mum3 mmtamoasnau ocm mmtaouaaax cons onwamxonuoo NmtmmoHanH mo chopumm coHpHnancH .m ensues 56 as Hut -3385 2 on 3 o. o o... p _ p . '01 x (All) 57 .SMmmmchs\emxau «ounceP doze masses > .mmnflouaasx on Am ANA IHOuaamx z: o. .4 .NmtomoHsnau sues coauoomu 0:» mo coaumauaca on HOSHQ GAE om How Honuomou Capabdoca ouo3 NmIHouaahx one Deanna and c053 omoathAMMD mmtomoHdnHH mo cuouuom coauabw£CH .o. magmas 58 :5 nae -3332... 2 .0. |°|"(I\/l) 59 competitive, then preincubation of the enzyme with ribulose-P2 should prevent xylitol-Pz binding, at least partially. To test this hypothesis, ribulose-P2 was incubated with carboxylase, in the absence of Mg++v for 20 min and then incubated with 3H-xylitol-P2 for an additional 20 min. This sample was chromatographed through Sephadex G-25 (Figure 11) to separate bound from unbound xylitol-Pz. The difference between the sample with and without ribulose-Pz was quite small, suggesting that the binding constant for xylitol-P2 is somewhat lower than the 1 HM suggested for ribulose-P2 (148). Incubation of the enzyme with iodoacetamide (IAA) completely prevented the binding of xylitol-P2 (Figure 11). Treatment of the enzyme with IAA also prevents the binding of pyridoxal-P (110). Both xylulose-P2 and xylitol-Pz are potent inhibitors of the carboxylase. When the xylulose-PZ concentration was 0.56 DH and the enzyme was 0.15 pM (1.12 MM in sites) the enzyme was 50% inhibited (84). The carboxylase Specific activity was 0.5 UMol/min/mg protein. While most investigators are working with a pure protein, there is a wide variety in the specific activities. The highest specific activity ever seen in this laboratory was 2.8 uMol/min/mg protein for the enzyme purified from spinach leaves, this was an isolated case, and the average is 1.0-2.0 uMol/min/mg protein. Lorimer (personal communication) has found a specific activity of 2.9 in one isolated case, also from spinach. The specific activity values for the enzyme prepared from other organisms are even more variable. The relationship between the specific activity of the enzyme and its ability to bind either xylitol-Pz or xylulose-P2 remains to be elucidated. Since both xylitol-P2 and xylulose-P2 bind so tightly to the enzyme, 60 .oeuueUm me3 Houoeaaoo noHuoenm on» euoueb oeuoeaaoo me3 ennao> oao> no as NP one .HE m.o me3 euwm n0wuoeuu on» one nEnHoo en» no omoeoH we: as N emeo noem nH ..XIx. mmlaouaamx mo noHuHooe an» op Hoaum nan on now moaseueoeoooa spas oouennona me3 mahnne one unesanemxe omen» e nH ..meaonao oemoHo. metaouaamx 0:» mo noauono touuna any euowen nae om Mom automoHnowu nus: oeuennona me3 oaauno on» unonfluwmxo onooem e nH ..meHoHHo name. oenmeumouesouno nonu one nae on now Henuemou oeuebnona eum3 mmtaouaawx one meanne on? .En\amo mum me3 mmlaou teasxtem on» «0 mua>auot oaueommm was .:As on you as: sees NmIHouwHAXImm nuaa oeuennona me3 HE\mE m be eshune one .AmmusouAmetmm wagons some emanu men can: eoeaoeaoo .semme an no an owm ue eonenuomne an Honuae oenasueueo me .naeu Ioum 0:9 .NmtaouaH>XI=m onnonnn one onnon mo nofiueuemem eueamnoo mnaaonm .onam. m~10 xwoenmom Scum eawmonm nsnaoo .FF magmas 61 mumZDZ 202.041.... 00 0.. 00 00 0e. ION W60 le 62 the possibility that they were attached covalently was considered. To test this hypothesis, an experiment was run to determine whether sodium dodecyl sulfate (SDS) would release the inhibitor from the enzyme. In one case the carboxylase was incubated with 3H-xylitol-P2 and the bound and unbound portions were separated with the Sephadex G-25 column as before. The fractions with the protein-xylitol-P2 complex were pooled and made 1% SDS, dipped into boiling water for 2-3 min and rerun on the Sephadex G-25 column (with 1% SDS added to the buffer). The results are shown in Figure 12. Since the SDS treatment resulted in the complete removal of 3H from the protein fraction it was concluded that xylitol-P2 binding was not covalent. A similar experiment was conducted with xylulose-Pz. In this case 14C-xylulose-P2 was incubated with three different samples of activated enzyme. The control sample was chromatographed as before. Both of the remaining samples were made 1% SDS, one sample was heated at 65° for 2 hr, and the remaining sample was placed in a boiling water bath for 2 min. Since the label did not remain bound to the protein that had been boiled (Figure 13), it was concluded that the interaction of xylulose-P2 with the enzyme was not covalent. The fact that the sample which had been heated to 65° in 1% SDS for 2 hr still bound significant xylulose-P2 attests to the remarkable heat stability of this enzyme. It is not known whether this bound portion represents A888 enzyme or individual large subunits. It should be possible to isolate large subunits and determine whether they will bind either of these bisphosphates. While the isolated large subunit is insoluble in the presence of Mg++ (35), Mg++ is not required for xylitol-P2 binding (McCurry, unpublished), and therefore is probably not required for the binding of xylulose-PZ. 63 .nnu mes be enamen mom w. woes me3 omen onooom end nw Homunn nsnaoo ooh ..meaouao oomoHo. naeme nnu one nEnHoo en» on ooooe mes H000 oeaaon on» no «He: «no .eae m “on emaaon tee .msm w. wees .emnoon mums mate. mn0auoeum .FP enemas na me eEem ecu euo3 mnoHuaonoo nEnHoo one muw>auoe oamaoomm mmlaouaahxtm one ..meHoHHo name. cannot e on emcee use use o~ sow mmuaouaasxmm :5 m spat oouennonw me3 Hn\os op ue esannm .mom unonua3 one nua3 NQIHOHHHAXIEM onnonnn one onnon mo noaueuemom mnH3onm .enam. mmtw xeoenmem noun eaamoum nEnHou .~_ masons 64 m moZDz 20:05.“. 0.. 00 on 0e 0» 0M 10.. ..0.N X r01 was r0.» 10.0 65 .FP onnmnm nn me eno3 mnonnnonoo nEnHoo on» emnsnonuo .oenamme we: onSem no as one .zn\Emo comm me3 NmtomoHnamxlovp on» no mnn>nuoe onmnoomm one .NmtemoHnme z: om one HE\nneuonm on m ene3 unannnonoo n0nuennonw 0:3 .mnmenmouenonno enamob .meaonao nomo. nnfi N now ooHHOQ no .xtx. nn N non omo ue oouennonn nonnne one mom mp ooen mez ennuxan many no eamnem nenu tone ..moHonno oomoao. mNtw xeoenmom no oonmenmonenonno one eSMNne oene>nnoe nun3 oonebnonn me3 NmtemoHnahxtovp .mom «a monomne one oonemenm ecu nn omeahxonneo won on leomoanaaxtoep onnonnn one onnon no nannenemom unnaonm .enaw. mNIO xooenmem Eonu ennwonm nanaou .m. masons 66 Ob 7../ . ./ 00 cumin... 20.._.onuot onnnowau e not moosezéeF as o. one «Home as on «net oeonunocoo Stone one .mhemme n no enema one nsonm menHeP one nun» one o.m mm ue ennonm SE cop no momneno oennu unnaeme mnmhaeao he oeaoaaom .oom we neon .oeueoaonn me nan on now oeue>auoe p now an oeueononn on» we eennomnahnuesro z v.0 unannna no nuaa oonennonn enoa moamnem eahnno 0:9 oe_H m.o en.“ o.o. ap-H ..o om H.e.o as.“ ..o an H.m.° 1 o.o mo_H o.m em.H m.mp an.“ ~.o om.H o.o wm_H ~.o so H.... + m.o an.“ m.~m a. .H m.o.F so.“ m.o an.“ o.~ sm-H s.o on.“ m.m 1 m.a oe.H o.mm on .H o.om an.“ m.o on.“ o.~ s..H o.o no.“ ~.m + m.m on.“ ~.~e om.H m.R~F as H o.o am_H m.~ we H.m.o mm.“ m.s 1 m.m am.H ~.~e on.“ o.so. on.“ m.o an.“ e.~ as.“ 5.0 mo_H o.n + m.m .mudp x Emu moonuz as P nonmez as F 1°, x emu m m 111111h. m m new: as on Home as on onoez as o. oomoz 2s a. onenoonamlm 25 P Nave: Naomz an oN a moonnnesoo connotnuoa an nonuennonH m.nn unonmnnnn no 0 OHQeB monsoonnnouoz1o sun: outnnxonnoo mm1moonsonm no eonueonnnooz 75 resulted in substantial inhibition, a long term modification was tried at pH 9.5. Enzyme was incubated in the presence and absence of O-methylisourea, and with and without substrate, ribulose-Pz, at pH 9.5 and 30° overnight (Table 5). From these data, it appears that incubation with ribulose-P2 protected the enzyme from inactivation at the high pH, but ribulose-P2 may have slightly increased the inhibition from the O-methylisourea. This may be due to exposure of more lysyl residues through a conformational change caused by the binding of the ribulose-P2 at the catalytic site. Experiments with cross-linking reagents suggest that the binding of ribulose-PZ causes the protein to "loosen” resulting in less cross-linking (42). Since the exploratory experiments indicated that the potential for o-methylisourea as an irreversible activator was small, this project was not pursued. Results with Neodymium While the requirement of the carboxylase for Mg++ has been known for some time, a recent report demonstrated that Mn++ could replace Mg++ in the oxygenase assay, but not in the carboxylase assay (158). This result has been confirmed by N. Hall in Tolbert's lab (159). Hall found that the effect of the metal appeared to be in activation. It seems likely that Mn++ would stabilize a carbamate as well or better than Mg++. This data could be used to support the hypothesis that Mg++ is required in the carboxylation reaction as well as for activation, but not in the oxygenase reaction, and that Mn++ fails in this catalytic role. Because of the variation in the effects of various metals it was suggested (J.F. Morrison, personal communication) that an element of the lanthanide series might have an interesting effect on activation. Neodymium, in the 76 Table 5 Modification of Ribulose-P2 Carboxylase with O-Methylisourea in the Presence of Ribulose-P2 o-Methylisourea Treated Untreated + Ribulose-Pz cpm x 10-3 + Ribulose-P2 A 1.9 :_9% 5.6 i_20% 12.3 :,8% 9.7 $.13% B 0.5 + 9% 1.5 + 7% 3.9 + 9% 3.2 + 3% The treated samples were made 50 mM in O-methylisourea in 25 mM Bicine buffer and incubated overnight at pH 9.5 and 30°. The ribulose-Pz concentration for incubation was 1 mM and without added Mg++° The samples were dialyed overnight to remove excess o-methylisourea. The numbers shown are cpm x 10'3, and are the means of 3 assays. A was activated and assayed with 20 mM MgC12 at 10 mM NaHCO3. B was activated and assayed with 20 mM MgC12 at 1 mM NaHCO3. The specific activity of the 14C-NaHCO3 was 230 cpm/nmol. 77 form of NdCl3, was tried (Table 6). Nd+3 did not replace Mg+2 for catalysis or activation. Nd+3 was inhibitory, probably by interfering with Mg+2 binding. Similar results were obtained for the oxygenase reaction, i.e. Nd+3 behaved differently than either Mg++ or Mn++. Results with Phosphorylated Compounds Hemoglobin is another protein which binds C02 as a carbamate (160). It also binds 02 and sugar phosphates, and the sugar phOSphate binding competitively regulates the C02 binding (160,161). The C02 and the phosphates bind at the N-terminus of the subunits (160). The best regulator for C02 binding to hemoglobin is 2,3-diphosphoglycerate. Because of carbamate formation in activation of ribulose-P2 carboxylase (103), the known effect of sugar phosphates on activation of the carboxylase (97), and the high concentration of 2,3-diphosphoglycerate in soybean leaves (162), this compound was investigated for its effect on the activation of the carboxylase. A representative sample of the data from the 2,3-diphosphoglycerate experiments is shown in Table 7. There is some small amount of stimulation at sub-optimal C02 concentrations and even the hint of stimulation at 10 mM NaHCO3. While these results were reproducible, the standard deviations of replicate experiments were large enough to account for the difference. This compound is not an activator in the absence of added C02. Based on these data it is unlikely that 2,3-diphosphoglycerate plays a role in the in zizg activation of this enzyme 0 Because of the fact that a variety of phosphorylated compounds, 78 Table 6 The Effect of Neodymium Chloride on Activation and Assay of Ribulose-P2 Carboxylase Activation Assay NdCl3 MgC12 * NdC13 + * MgClZ + MgClZ NdCl3 cpm x 103:; MgClz 0.8 :.2% 24.1 :_1% 14.4 :_3% 22.6 :_1% The concentrations of MgClz and NdCl3 were both 20 mM. Each assay contained 360 Ug/ml protein. The specific activity of the NaHCO3 was 230 cpm/nmol. values shown are the means of 3 assays. Addition of NdC13 to the oxygenase assay gave similar results, and there was no activity in the absence of Mg++. *The compound listed first was incubated with the enzyme for 10 min in the presence of 10 mM NaHCO3, followed by 10 min with the compound listed second. 79 Table 7 The Effect of 2,3-Diphosphoglycerate on Activation of Ribulose-P2 Carboxylase Activation conditions Results cpm x 10"3 1. Enzyme alone 0.3 i_4% *2. Enzyme + MgC12 + NaHCO3 6.8 :_9% *3. Same as 2 + 1 uM 2,3-glycerate-P2 7.3 1.9% *4. Same as 2 + 10 UM 2,3-glycerate-P2 7.4 i 10% *5. Same as 2 +100 HM 2,3-g1ycerate-P2 8.8 :.3% *6. Same as 2 + 1 mM 2,3-glycerate-P2 8.4 i'7% 7. Enzyme + MgClZ + NaHCO3 9.7 i'4% 8. Same as 7 + 1 HM 2,3-glycerate-P2 10.1 1.6% 9. Same as 7 + 10 MM 2,3-glycerate-P2 9.6 i_1% 10. Same as 7 +100 HM 2,3-glycerate-P2 10.2 :_3% 11. Same as 7 + 1 mM 2,3-glycerate-P2 10.0 i'1% 12. Enzyme + 10 NM 2,3-glycerate-P2 1'2 i.1% 13. Enzyme + 1 mM 2,3-glycerate—P2 1.7 i_6% The samples were activated for 30 min before initiation of the reaction. The assay time was 30 s, and the results are the means of 3 assays. In all cases the MgC12 concentration was 10 mM for activation (if used) and 10 mM for assay. All samples were assayed at 10 mM NaHCO3. The specific activity of the 14C-NaHCO3 was 230 cpm/nmol. *In these cases the NaHCO3 concentration for activation was 1 mM, in all other cases it was 10 mM (if used). 80 including phosphate and pyrophosphate (M. Spellman, personal communication) cause some degree of activation of the carboxylase under sub-optimal C02 concentrations, a reorientation of our working hypothesis is probably in order concerning the activation of this enzyme. It seems certain from the work of Lorimer and others that one aspect of C02 activation is the formation of a carbamate which is stabilized by M++. The effect of sugar-P's clearly cannot replace this phenomenon. Perhaps the effect of sugar phosphates is a non-specific anion activation at the ribulose-P2 site, and this manifests itself only under sub-optimal C02 conditions, because this activation is normally performed by HCOS. The rationale behind this hypothesis is that the binding energy of effector causes a conformational change in the protein which allows it to bind substrate readily. This idea for activation of an enzyme has been discussed by Koshland (163). If this explanation is representative of the real condition then the inhibition due to sugar-P is due to their relative life time in the active site. One might expect then that ribulose-P2 would activate the carboxylase, but it might be too fast to observe under normal assay conditions. Data will be presented in Chapter VI with rapid reaction kinetics showing an initial lag, as if such a ribulose-P2 activation did occur. CHAPTER VI EPR SPECTROSCOPY, IRON, AND RAPID RECTION KINETICS Introduction The proposed oxygenase mechanism (Figure 1) contains an apparently spin-forbidden step (see Literature Review). Therefore, the possibility that the reaction could proceed through a free radical was explored in a joint research project between Dr. Tolbert's group and Dr. W.H. Orme-Johnson at the University of Wisconsin. Dr. Peach and I made two trips to Dr. Orme-Johnson's laboratory to make use of his expertise and equipment for EPR spectroscopy and rapid-quenched enzyme reactions. This Chapter contains the results of these two research trips. Materials The enzyme was prepared in the manner described in Chapter I, the spinach was obtained locally, flown in from California, or flown in from Florida. Ultrapure: Mg(NO3)2, KDH, HCl, and NaHCO3 were from the Alfa Division of the Ventron Corporation. BathOphenanthroline, tetranitromethane, and hydroxylamine hydrochloride were from Sigma. Methods The EPR spectroscopy was performed on a Varian V-4500 spectrometer at 13°K. The individual settings for the spectrometer are described in the appropriate Figure legends. The rapid-quench apparatus was from 81 82 Update Instruments. The instrument consisted of two syringes: one with substrate, and one with activated enzyme. At time = 0 the enzyme and substrate were forced simultaneously into a mixing chamber and subsequently into either 2N HCl (for the fast reaction enzyme assays) or into isopentane at 77°K (for the EPR studies). The duration of the assay was a function of the length of the tube connecting the mixing chamber and the quenching solution. The samples that were quenched in isopentane were stored in liquid nitrogen until the EPR spectroscopy was performed. Conducting enzyme assays for very short reaction times (e.g. 10 ms) poses serious mixing problems, so the ratio of volumes of enzyme and substrate had to be kept close to 1 to 1. In an effort to get the final sample for the EPR as concentrated as possible, the starting concentration of enzyme was 40-60 mg/ml. The sample volume was about 400 pl so at least 10 mg of enzyme was used in each reaction. The assay times varied from 3 ms to 10 3. Discussion of the EPR Spectroscopy The initial approach was to determine the EPR spectrum of the enzyme during catalysis. The rationale behind this was to determine whether a radical signal was present, and if it was, whether the signal changed during the course of the reaction. The rapid-quench apparatus was used to prepare frozen samples from the initial portion of a reaction progress curve, and then to study them in the EPR. Since the activity under study in this case was the oxygenase reaction, care was taken to saturate both the enzyme and the substrate with 02 before beginning the reaction. Even with this precaution it was not possible to optimize the oxygenase reaction. Normally in an oxygenase assay, activated enzyme is diluted 25 83 fold, so that the carryover of inhibitory HCOS from activation is only 0.4 mM. However in this case, the ratio of volumes required for adequate mixing resulted in a carryover of around 5 mM HCOS. As a result, it was impossible to eliminate the carboxylase reaction and the oxygenase was substantially inhibited. The immediate results were the discovery of a radical signal in the g=2 region of the spectrum and an iron signal in the g=4 region (Figure 14). A radical signal was present in every enzyme sample that was run. control samples were prepared by mixing the activated enzyme with water instead of substrate and waiting a defined time before quenching the "reaction”. Figure 15 shows a 3 s assay, a 3 8 control, and a difference spectrum created by storing the first two assays in a computer and displaying the difference. The radical has a certain amount of hyperfine splitting, but the experiments to date do not reveal much more. According to Dr. Orme-Johnson, the signal probably does not represent superoxide since it is missing certain hyperfine features present in the spectrum of superoxide. Because the radical was always present in the enzyme samples, it seemed reasonable to consider possible sources of radicals on the enzyme surface. Two potential sources of radical signals were cysteine and tyrosine. In an effort to determine the source of the radical signal four samples of enzyme with a concentration of 20 mg/ml were prepared. a) A control sample was prepared with no additions, but was otherwise treated exactly like the test samples. b) An enzyme sample was treated with C(NO3)4 which reacts with tyrosine residues, in a competitive manner with respect to ribulose-P2 (164, and Chapter VII this thesis) and 84 .cofimwu mum onu ca Havana“ a ma mums? .AHHHV mm «o w>wuooaosfi ma scammu vnm on» cw Hosbwm one .Ammsmm mum oNIoP x mNth.ov coumcmma Mnom onu ma m .Aoom mum hmuop x Pm~®.mv acoumcoo m.xocmHm ma 2 .hocmsvmuw m>m3 aouowe on» ma >_.ox F.VIF.o Scum omwum> Anumswuum oaoam usuv m ouo£3 Amy mmuAanv scum ooumasoamo mum3 modam> m was .somp um .Fncas coco? «0 mums cmom m can .umx so. no mocwsvum cOHDMHSGOE .w or no mosuflamam coaumasooa .Hm3om o>m30HOHE 35 or .ucmuwcoo mad» m m.o .wa memo.m hocmskuw o>m30uoas "macauflocoo mewsoaaom 0:» moves umuosouuoomm oom¢1> cowum> m ca ooEHOMme mos mane .HE\mE om um ommamxonumo NmImmoHanu mo Eduuoomm mmm mnfi .v. magmas 85 man; omn.~ mhod o_n.¢ - r F mm:4<>-o oo_.~. 86 .owsaahmuwo mm3 asuuoomm wocmuwwmao man can Housmsoo o as omuoum wum3 m can < Bonn onuoomm on» no snow AU .< cw mm mcwnocosv snowmn oo no u m How omoooum on om3oHHm mos :OHuome man can ouon mafia um mfihuco on» nuHB omxfla mm3 NmImmoHanm Am .ep madman mu mach may mum3 mcofluaocoo Monuo Ham .wx mvw.MImvm.N no «menu zoom m can .m m «o ucmumsoo 02H» .35 P um stom w>m3onoae .wa cmmo.m no? wocosvmum m>mBOROHE och .Mohb um osmucwmomw aw mcflnocwsv whomwn oo um m m Mom conundocfi can Chou was» as Dawnsm may spas omxfla was umumz Ad .mcofipwocoo moss» Moons womahxonuoo «mlomoasnwu mo muuowmm mmm .mp shaman 87 vamp; Gum .. 0mm; nnOMN NhO.N r _ mu=4<>-o 4 88 totally inactivated the enzyme. c) An enzyme sample was treated with sufficient iodoacetic acid to totally inhibit the enzyme by modification of sulfhydryl groups at the active site. d) Since, hydroxylamine has been reported to inhibit this enzyme (165, see also Chapter VIII this thesis), it was mixed with the enzyme, incubated, and then NaBH4 was added in an attempt to reduce any oxime which might have been formed during the incubation. The above portions of this experiment were carried out at Michigan State, the samples were frozen in EPR tubes in liquid nitrogen and sent to Orme-Johnson at the University of Wisconsin in a liquid nitrogen shipping dewar for the EPR spectroscopy. The EPR spectra of these samples are shown in Figure 16. The results were totally unexpected. The only treatment which altered the EPR radical signal (the NHZOH, NaBH4 treatment) was also the one which had no effect on the enzyme activity. Why the hydroxylamine produced this effect is unclear. It is possible that if this sample had been activated before the determination of the EPR spectrum the results could have been different. Iron Content of the Enzyme The discovery of a signal in the EPR spectrum corresponding to iron raised such questions as: was the iron simply a contaminant, could this iron be the required transition metal for the oxygenase, and was iron present in all sources of the enzyme. Experiments to answer these questions were carried out at Michigan State. The method used for quantitation of the iron was the analytical technique of Van de Bogart and Beinert (166). This method is basically a wet ash procedure with strong acids followed by reduction of any Fe+++ to Fe++, neutralization, 89 .0 ca mm wuw3 macauaocoo Had .woanohnouon ESHoOm flaws ucoeumwuu an om3oHHom mowEdHaxonoan :DHB neweuooua an .m CH mm ouo3 msowuflocoo Honuo Ham .mmvo.m mos mucoswmum m>m3ou0fla one .ucmeumouu o: .mcon mahucm AU .< ca mm mum3 mcoaufiocoo HH¢ .0cmcumfionuacmuuou nuw3 ucmEummHB Am .vF whamwm ca mm mum: m:OHUHocoo Monuo Has .38 m mm3 Hmzom m>m30u0HE on» .Nmo mhvo.m mm3 hocmsvmum o>m30uoHE one .oofifioumomooofl spa: ucoeuwoua Ad .coHHMOflMHoOE HmowEmso umumo ommaxxonumo NmIomoHdnHH mo muuowmm mam .o. magmas 90 anfi 050.» _ mu=4<>-o moms? «BK 91 and extraction into isoamyl alcohol through the complexing of the iron with bathophenanthroline. The resulting pink color was read in the spectrophotomer at 535 nm against an isoamyl a1cohol-bathophenanthroline blank. The standards were linear up to 25 nmol of iron, and 5 nmol gave an absorbance of approximately 0.100. The stoichiometry of the iron proved to be 2 mol of iron/mol of enzyme, or one iron atom for every 4 active sites. This number was found in all samples examined and was confirmed independently in Dr. Orme-Johnson's lab where this particular technique is frequently used. This stoichiometry represents total iron, whereas the EPR data is only for Fe+++ (Fe++ is not paramagnetic). Usually Fe+++ is readily reduced if it is free in solution, but up to 4 mM DTT had no effect on the EPR signal suggesting that this iron was bound to the protein, perhaps in a manner analogous to transferrin (167). To evaluate whether the iron was present in the enzyme samples as a contamination, iron free water was obtained by passing glass distilled water over a Chelex 100 column and buffer through Dowex Chelating Resin. After the iron content of bicine was determined to be low this latter column was eliminated due to materials leaching out of the Dowex (see Chapter IV). Table 8 shows the amount of iron present in some of the reagents required for the ribulose-P2 carboxylase assays. Iron free MgClz was prepared by ion exchange chromatography of ultrapure Mg(NO3)2 on a Dowex-1-Cl‘ column prepared with ultrapure HCl. Buffers for assay were adjusted with ultrapure KOH. Dialysis of the enzyme at pH 8, against 10 changes of iron free buffer (1 to 100 volume ratio), resulted in no change in the iron content of the protein. The dialysis 92 Table 8 Iron Content of Reagents for Assay of Ribulose-P2 Carboxylase/Oxygenase Reagent Iron (nmol Fe/ nmol reagent) Bicine Sigma >0010 Boehringer >0.10 Ribulose-P2 >0.10 MgClz >o.o1 NaHCO3 >0.02 93 tubing had been prepared by boiling for 5-10 min in 5 mM EDTA. Similar results were obtained when the dialysis was conducted at pH 5.9 but this pH caused the protein to precipitate after several hours. With many iron binding proteins and transferrin (167) the bound iron can be removed if it can first be reduced. Dr. Hall in Tolbert's lab had noticed that treatment of the enzyme with high concentrations of DTT produced a pink color which could be dialyzed away (135). Such a pink color has been attributed to iron-DTT complexes (Henry Lardy, personal communication). Because of this, an enzyme sample of 10 mg/ml was made 100 mM DTT for 4 hr and then dialyzed agasinst several changes of buffer containing 100 mM DTT. This procedure reduced the iron concentration from 2 mol/mol of enzyme to 0.7 mol/mol of enzyme. Assay of this enzyme in iron-free media revealed no change in either activity. In a related experiment, stirring the enzyme in 50 mM ascorbate over Chelex produced no change in the iron EPR signal. From these data, conclusions concerning the role of iron in the mechanism of this enzyme, if any, are impossible. Since removal of 60% of the iron had no effect on either of the two catalytic activities of the enzyme, and since the amount of iron present is quite small relative to the number of active sites, one is tempted to rule out any role for iron. On the other hand it is difficult to understand why different sources of enzyme all had the same amount of tightly bound iron if it doesn't have a role. Copper Content 2: the Enzyme The role of copper, if any, in ribulose-P2 carboxylase has been extensively investigated and hotly debated. Lane's group reported that 94 copper was present, but removal of it had no effect on the carboxylase activity (174,168). Chollet 32 El (169) reported finding essentially no copper, iron, or flavin associated with crystalline tobacco enzyme. Lorimer 25 El (11) found similar results from spinach. On the other hand there is the report of Branden (44) which claims that the oxygenase is a copper protein and that it can be separated from the carboxylase by Sepharose 68 column chromatography (see Chapter VIII). During our first trip to Orme-Johnson's facility, we were able to use Michigan grown spinach as the source of the enzyme and saw no indication of copper in the EPR. Our second trip was made in the winter, and since the experiments were to require large amounts of enzyme, spinach was shipped in from California and Florida. The spinach from the two different sources were clearly different varieties, but their specific activities were quite similar, 1.8 pMol/min/mg for the enzyme from California Spinach, and 1.6 pMol/min/mg for the enzyme from Florida spinach. The enzyme prepared from California spinach contained no copper signal in the EPR, but the enzyme prepared from Florida spinach did (Figure 17). No analytical determinations of copper were made and since only Cu++ will show up on EPR, it is possible that the California enzyme contained Cu+. The copper was easily reduced, addition of 2 mM DTT completely abolished the signal (Figure 17). Whether copper has a physiological role in the mechanism of this enzyme remains a moot question, but it is a potentially useful probe for the conformational state of the protein. The addition of Mg"'+ and C02 for activation resulted in the splitting of the copper signal (Figure 17). This suggests that the copper is interacting with the protein in some region which changes with the activation process. 95 .m ca mm 05mm 0:» ohms msowuaosoo Ham .ousumuwmamp Soon um ass on how BBQ 25 N aua3 scaumndocw “mums wamfimm mahnao 05mm on? A0 .< cw mm memo was whoa mcoauwocoo Monuo Ham .wa vmmo.m mm3 hocosvoum o>m30u0HE one .++mz mo snowman ca mawucm Am .vw musmwm ca on menu on“ mums mooHuHocoo nonuo Add .wa vmvo.m mos mocmswoum o>m3ouows 0:9 .35 m mm3 Hmzom m>m3ou0fle 0:9 .oouflsvmu mm3 ++m2 .Hmsmam so on» ca umem wnu mudooum no: oao mcoam N00 .++mz ocm Nov an ooum>fluom mm3 mamsmm ofihnco mane Am .AHHvso mo m>aumofloca mum sodomy mum on» ca mocamuomhn ooumHoommm on» was Hannah mmumH on? .mowuoam Baum owcflmuno nomcwmm Scum ooawwusm onwaxxonuoo NmIomoHsbwu mo Esuuoomm mmm 0:8 .np mucosa 96 00a; amnfi 000.» p . b mm:4<>-o o (i. _.n.¢ P no.‘ 97 Rapid-Reaction Initial Kinetics Access to the rapid-quench apparatus provided us with the opportunity of determining a rection progress curve for the carboxylase at very short times (Figure 18). The curve indicates a lag of 0.5 to 1 s in the reaction. While this seems negligible in terms of normal enzyme assay times, it raises some questions about the carboxylase mechanism. Unless these results can be attributed to some flaw in the experimental design (a repeat experiment gave the same results) they may indicate substrate activation (see Chapter V) or perhaps some form of positive cooperativity for the carboxylase reaction. These assays were performed at 10° with a very high enzyme concentration. As a result, the leveling off after a few seconds may be due to the lack of available C02 for the carboxylation reaction. This experiment should be extended with the addition of carbonic anhydrase to the reaction mix to speed up the COz-Hcoj'equilibrium. This topic will be pursued next year when I have regular access to the rapid-quench apparatus. 98 .HE\mE om mos soaumuucmocoo mahucm was .uxou on» as emnfluommv ma endeavoum was .Hom z~ no coauomncw onu an owaaax wum3 meoHuoomu 0:9 .mhdmmm wed» uuonm mum> now mmmamxonumo mmlwmoHsnflu mo 0>Hdo mmmumonm sowuommm .mF mucosa 99 “my MEI. CHAPTER VII ACTIVE SITE MODIFICATION Introduction Several laboratories have been successful in modifying amino acid residues in the active site region of ribulose-P2 carboxylase. Cysteine (107), arginine (124,125), lysine (108,119), tyrosine (164) have all been shown to be in the active site region of ribulose-P2 carboxylase. My efforts in the direction of active site modification were limited to trying to explain two other phenomena uncovered in the work of this thesis: a) what group on the protein is responsible for the tight binding of xylitol-PZ and xylulose-PZ: and b) what group is reSponsible for the radical signal observable in the EPR? Materials Tetranitromethane and dibromoacetophenone (p-bromophenacylbromide) were purchased from Sigma. NaB3H4 was obtained from New England Nuclear. Methods The enzyme assays were performed in the usual manner. Assays with dibromoacetophenone were in 0.25 ml reaction volume and all other were 0.5 ml volume. The dibromoacetrophenone was dissolved in acetone before use. The reduction with NaB3H4 was carried out in a hood for the venting of the 3H2 given off during the reaction. 100 101 Results and Discussion One hypothesis offered for the strong inhibition by xylitol—P2 and xylulose-P2 is that the hydroxyl at C-3 reacts with an adjacent base in the active site to form a hydrogen bond. Histidine could be such a base, so modification of histidyl residues was attempted with dibromoacetophenone. The protocol was to incubate activated enzyme with dibromoacetophenone for varying periods of time, up to 6 hr, and then assay the carboxylase activity. The assumption was made that if such a modification occurred the enzyme would be catalytically inhibited as well as having a reduced affinity for the two inhibitors. Since the reagent is insoluble in water and had to be dissolved in acetone, acetone treated enzyme controls were also run. The dibromoacetophenone inhibitied the carboxylase reaction less than 10% and was indistinguishable from the acetone controls. The active site of phospholipase A2 contains an essential histidine and while the pH optimum for catalysis is around 8, the pK for inactivation with dibromoacetophenone is 6.1 (170). I attempted modification of ribulose-P2 carboxylase both at pH 7.0 and at pH 8.1, but there was no effect different from the acetone control. Either there are no reactive histidyl residues in the active site region, or this was not the reagent of choice. This later possibility is supported by the fact that C. Paech (personal communication) found a reaction between diethylpyrocarbonate and the carboxylase by following the reaction spectrophometrically, but he did not correlate this effect with enzyme activity. Tetranitromethane is a powerful nitrosylating reagent which is rather specific for tyrosyl residues. Because of the radical signal in 102 the EPR spectrum of this enzyme, this compound was tested for its effect on enzyme activity. C(NO3)4 had no effect on the radical signal (Chapter VI), but it was a potent inhibitor of both activities. In one preliminary experiment, ribulose-PZ was found to partially prevent the inactivation due to C(NO3)4. This was run at only one concentration of C(NO3)4. Before I had the opportunity to continue these experiments a paper appeared from Tabita's group (164) discussing the inhibition of the carboxylase activity in enzyme prepared from R; rubrum. They found C(NO3)4 to be a competitive inhibitor with respect to ribulose-Pz. Because of this and the negative results in the EPR experiments, futher work with tetranitromethane has been postponed. After solid data was obained for a lysyl residue in the active site region, Dr. Paech and I made some abortive efforts to look for Schiff base formation between this residue and ribulose-P2 and xylulose-PZ. The procedure was to incubate the enzyme in the presence of each of the keto diphosphates for 20 min and then add NaB3H4 in an effort to reduce the Schiff base if one was present. While there was incorporation of label into the protein, acid hydrolysis and chromatography of the residues revealed no incorporation of the sugar phosphates. CHAPTER VIII PROJECTS TO VERIFY THE WORK OF OTHERS Introduction One of the basic requirements of science is that the experiments be reproducible and that the phenomena discovered must be features of the material under study rather than a feature of the way in which the experiment was performed. This has not been the case for many of the papers on ribulose-P2 carboxylase. It is for this reason that experiments performed in one laboratory must be repeated in different laboratories and from different directions. In the few years that I have been associated with ribulose-PZ carboxylase/oxygenase several papers have been published which have yet to be successfully confirmed in other laboratories. Some of these papers presented data of such potential significance that those of us in Tolbert's group felt it necessary to attempt to repeat the experiments. This chapter contains the results of five such projects. Ribulose-P; Carboxylase/Oxygenase, One Enzyme 2E Two Since the discovery of the oxygenase activity of ribulose-Pz carboxylase (8,9,10,11) many efforts have been made to determine whether both activities are really present in one protein. Many different laboratories have found the two activities to copurify through ammonium sulfate fractionation, Sepharose chromatography, ultracentrafugation, cocrystallization, ion exchange chromatography, electrophoresis, 103 104 isoelectric focusing, and hydroxyl apatite chromatography. In addition there is competitive inhibition between C02 and 02. Numerous compounds have been tested, all of which (with the exception of Mn++ (158)) inhibit both activities equally. Both activities have a break point in the Arrhenius plot at the same temperature (171). The compound 2-carboxyribitol-P2, an analogue of the proposed reaction intermediate inhibits both activities equally. Because of this accumulation of data which suggested that both the carboxylation and oxygenation reactions occurred at the same active site, a report of the separation of these activities (44) was received with a great deal of interest. The decision was made to repeat this work, initial skepticism notwithstanding. Our results have been published (45) and are included in this thesis as Appendix 4. We repeated Branden's (44) work as closely as possible, and talked to him to clear up specific points about the methods. We totally failed to reproduce his results. He was plagued with a high endogenous rate of oxygen consumpton and as a result, incubated the enzyme in the assay mix and initiated the reaction with ribulose-Pz. Our explanation for his data is that this failure to use activate enzyme caused the oxygenase activity to be undetectable in the column fractions which contained carboxylase. When we tried running the oxygenase assays in this manner our activity all but disappeared. Why he found a ribulose-P2 dependent uptake of oxygen in a different region of the Sepharose profile remains unclear. This activity was associated with a large amount of copper (44) and iron (personal communication). We postulated that due to the instability of the substrate ribulose-PZ, a non-specific oxygen uptake 105 catalyzed by a cu++-protein complex would not be surprising. The role of copper in this enzyme, if any, remians to be determined (Chapter VI). Oxygen Activation Recently Wildner reported that oxygen was required in a non- catalytic capacity for both activities of this enzyme (172). His data were that enzyme prepared from spinach, in Tris-sulfate buffer, suffered a substantial loss of both activities after being gassed for several hours with Ar gas. Also, there was a substantial change in the intrinsic fluorescence of the protein. Both of these results were completely reversible following the reintroduction of 02 to the system. To repeat this work. R. Gee prepared the enzyme, gassed it with Ar, and determined the intrinsic fluoresence, N. Hall ran oxygenase assays, and I ran the carboxylase assays. In our hands, the sample treated with argon was exactly the same as the untreated control in all respects. We were unsuccessful in two attempts to confirm Wildner (172), but I can offer no explanation for the differences between his results and ours. Differential Regulation If one accepts the argument that photorespiration is purely a wasteful process and therefore should be eliminated, differential regulation of the carboxylase and oxygenase would be one way to achieve this goal. This section deals with claims from two different laboratories that oxygenase activity could be reduced by certain chemical treatments. Wildner (87) reported that preincubation of the enzyme with 106 ribulose-PZ and subsequent treatment with iodoacetamide resulted in complete retention of the carboxylase activity and a total loss of oxygenase activity. Omission of the ribulose-PZ resulted in complete inactivation of both activities. I attempted to repeat this experiment on three occasions, and a sample of the results is shown in Table 9. Both activities were protected when the enzyme was preincubated with ribulose-Pz before the IAA treatment, and complete inactivation of both activities occurred when this preincubation was not done. C. Paech also performed this experiment and obtained results identical to mine. N. Hall aided in the performance of the oxygenase assays. Bhagwat and coworkers (165) have reported that 5 mM hydroxylamine completely inhibited the oxygenase without affecting the carboxylase activity. They did not use activated enzyme in their experiments. It is not clear from their paper whether the enzyme or the substrate was exposed to the NHZOH before the initiation of the reaction. In our effort to repeat this work, C. Paech and I used activated enzyme and studied the effect of NHZOH when it was preincubated with the enzyme, and when it was preincubated with the substrate (Table 10). Our data shows equivalent inhibition of both activities under each of these two conditions. The oxygenase may have been slightly more inhibitied than the carboxylase when the enzyme was preincubated with the hydroxylamine, but the carboxylase was certainly inhibited also. Since the appearance of this paper, there has been a second report of differential regulation with hydroxylamine (173). This later work was done with partially purified enzyme from Anabaena, and both activities were assayed in the same reaction vessel. While this seems like a good approach it means that it is impossible to Optimize each of the activities. Treatment of 107 Table 9 Modification of Ribulose-P2 Carboxylase/Oxygenase with No Treatment 5 mM iodoacetamide 5 mM iodoacetamide after 20 min with 25 mM ribulose-P2 Iodoacetamide Carboxylase 100% 0% 85-95% Oxygenase 100% 0% 60-75% 108 Table 10 Treatment of Ribulose-P2 Carboxylase/Oxygenase with Hydroxylamine NH20H Concentration Carboxylase Oxygenase mM A 0.00 100% 100% 0.63 99% ___ 1.25 97% ____ 2.50 89% ____ 5.00 85% ____ 6.25 80% 58% 10.00 69% ‘___ 12.50 46% 39% 25.00 11% 23% B 0.00 100% 100% 2.50 90% 87% 5.00 84% 84% 10.00 ___ 65% 25.00 41% ____ In A, hydroxylamine was incubated with ribulose-P2 for 6-10 min before initiation of the reaction by the addition of the enzyme. The reaction was run at the same concentration of hydroxylamine as the incubation. In B, hydroxylamine was incubated with the enzyme for 20 min before initiation of the reaction with ribulose-Pz. The reactions were run at the same concentration of hydroxylamine as the incubation. All assays were run as described in Chapter I. 109 the enzyme with NHZOH and NaBH4 did produce a change in the radical signal present in the enzyme (Chapter VI), but this could not be correlated to any change in activity. Hopefully these differences in data will eventually be resolved. Effect pf Superoxide Dismutase Since the discovery of the oxygenase reaction the possibility of the involvement of Oi’has been considered. Ryan and Tolbert (129) demonstrated that on polyacrylamide gels the enzyme acted as a superoxide dismutase. More recently, Wildner and Henkel (87) postulated that Oi'was the true substrate for the oxygenase reaction. Bhagwat and Sane (88) have recently offered data in support of this hypothesis. Their data suggested that the oxygenase was inhibited by the addition of superoxide dismutase or nitroblue tetrazolium to the assay mix. This is contrary to published work from Tolbert's lab (11). They purified their own SOD from bovine erythrocytes. No evidence was offered concerning the effect of SOD on the carboxylase reaction. Since 05 is such a reactive species, its concentration in free solution is quite small. To account for observed rates of the ribulose-P2 oxygenase reaction, the oxygenase would have to generate the 05 from 02, and for the oxygenase reaction to be inhibited by SOD this 05 would have to be released from the enzyme before it was used. C. Paech and I attempted to repeat this experiment using SOD from Sigma. The SOD was very active as checked by us in the xanthine oxidase-cytochrome C assay of Fridovitch (174). No inhibition of either ribulose-P2 carboxylase or oxygenase activity was noted upon addition of 100 U of SOD to each assay mix. Using copper penicillimine, a small molecular weight SOD Lorimer (personal communication) also found no effect on either activity. In a similar vein, we found no signal in the EPR indicative of 05 (Chapter VI). until the mechanism of the oxygenase is elucidated, the role of 0;, if any, will remain unclear. B IBLIOGRAPHY 4. 5. 6. 9. 10. 11. 12. 13. 14. 15. 16. 17. BIBLIOGRAPHY Wildman, S.G. and J. Bonner, (1947) Arch. Biochem. 14: 381. Quayle, J.R., Fuller, R.C., Benson, A.A. and M. Calvin, (1954) J. Amer. Chem. SOC. 7_63 36100 Weissbach, A., Horecker, B.L., and J. Hnwitz, (1956) J. Biol. Chem. 218: 795. Jakoby, W.B., Brummond, 0.0., and S. Ochoa, (1956) J. Biol. Chem. Srere, P.A., Cooper, J.R., Klybas, V., and E. Racker, (1955) Arch. Biochem. Biophys. 29; 535. Warburg, 0., (1920) Biochem. Z. 103; 188. Tamiya, H. and H. Huzisige, (1949) Acta. Phytochimica 15: 83. Ogren, W.L. and G. Bowes, (1971) Nature 230: 159. Bowes, G., Ogren, W.L., and R.H. Hageman, (1971) Biochem. Biophys. Res. Commun. 25; 716. Andrews, T.J., Lorimer, G.H., and N.E. Tblbert, (1973) Biochem. 12; 11. \ Lorimer, G.H., Andrews, T.J., and N.E. Tblbert, (1973) Biochem. 12: 18. Kawashima, N. and S.G. Wildman, (1970) Ann. Rev. Plant Physiol. 21; 325. Siegel, M.I., Wishnick, M., and M.D. Lane, (1972) in "The Enzymes” (editor P.D. Boyer) Acad. Press, N.Y. Anderson, L.E., Price, G.B., and R.C. Fuller, (1968) Science 161: 482. Givan, A.L. and R.S. Criddle, (1972) Arch. Biochem. BiOphys. 149: 153. Kawashima, N., (1969) Plant Cell Physiol. 12’ 31. Rutner, A.C. and M.D. Lane, (1967) Biochem. Biophys. Res. Commun. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 112 Moon, K. E. and E. O. P. Thompson, (1969) Aust. J. Biol. Sci. 22; 463. Jensen, R. G. and J. T. Bahr, (1977) Ann. Rev. Plant Physiol. 28; 379. Tabita, F. R. and B. A. MCFadden' (1974) Jo BiOlo Chem. 249’ 3459. Tabita, F. R., McFadden, B. A., and N. Pfennig, (1974) Biochim. Biophys. Acta. 341; 187. Purohit, K., McFadden, B. A., and A. L. Cohen, (1976) J. Bacteriol. 127; 505. Tabita, F. R., Stevens, S. E. Jro, and J. Lo Gibson, (1976) J. Bacteriol. 125; 531 McFadden, B. A. and F. R. Tabita, (1974) Biosystems 6: 93. Takabe, T. and T. Akazawa, (1975) Plant cell Physiol. 16, 1049. Gray, J. C. and R. G. O. Kekwick, (1973) Biochem. Soc. Trans. 1; 455. Kung, S. D., (1976) Science 191; 429. Kung, S. D., Gray, J. C., Wildman, S. G., and P. S. Carlson, (1975) Science 187: 353. Gray, J. C., Kung, S. D., Wildman, S. G., and S. J. Sheen, (1974) Nature 252: 226. Kawashima, N., Tanabe, Y., and S. Iwai, (1974) Biochim. Biophys. ACtao 371: 4170 Nishimura, M., Takabe, T., Sugiyama, T., and T. Akazawa, (1973) J. Biochem. 14: 945. Nishimura, M. and T. Akazawa, (1974) Biochem. Biophys. Res. Commun. .52: 584. Nishimura, M. and T. Akazawa, (1974) J. Biochem. 16: 169. Nishimura, M. and T. Akazawa, (1973) Biochem. Biophys. Res. Cbmmun. £1; 842. Brown, H. M. and R. Chollet, (1979) Plant Physiol. supplement pg; No. 361. McFadden, B. A. and K. Purohit, (1978) in "Photosynthetic Carbon Assimilation“ (editors H. W. Siegelman and G. Hind) Plenum Press, N. Y. and London. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 113 Sakano, K0, Kung, S. D., and S. G. Wildman, (1974) Plant Cell Physiol. 15: 611. McFadden, B. A., Iord, J. M., Rowe, A., and S. Dilks, (1975) European J. Biochem. 54; 195. Baker, T. S., Eisenberg, D., and F. Eiserling, (1977) Science 196: 293. Eisenberg, D., Baker, T. S., and S. W. Sub, (1978) in "Photosyn- thetic Carbon Assimilation" (editors H. W. Siegelman and G. Hind) Plenum Press, N. Y. and London. Gray, J. C. and R. G. O. Kekwick, (1974) European J. Biochem. 44; 481. Grebanier, A. E., Champagne, D., and H. Roy, (1978) Biochem. 11; 5150. Roy, H., Valeri, A., Pope, D. H., Rneckert, L., and K. A. Costa, (1978) Biochem. 11; 665. Branden, R., (1978) Biochem. Biophys. Res. Commun. g; 539. McCurry, S. D., Hall, N. P., Pierce, J., Paech, C., and N. E. Tolbert, (1978) Biochem. Biophys. Res. Cbmmun. 84; 895. Siegelman, H. W. and G. Hind, (1978) editors "Photosynthetic Carbon Assimilation" Plenum Press, N. Y. and London. Chan, P. H. and S. G. Wildman, (1972) Biochim. Biophys. Acta. 277: 677. Kawashima, N. and S. G. Wildman, (1972) Biochim. Biophys. Acta. 262; 42. Coen, D. M., BedbrOOk' J. R., Bogorad, Lo, and A. RiCh, (1977) Proc. Nat. Acad. SCio 7_4} S4870 Iwanij’ V0, Chua, N. H., and Po SiekeVitz, (1975) J. Cell 3101. _6_4; 572. Sakano, K., Kung, S. D., and S. G. Wildman, (1974) Mol. Gen. Genet. Roy, H., Patterson, R., and A. T. Jagendorf, (1976) Arch. Biochem. Biophys. 172; 64. Dobberstein, B., Blobel, G., and N. H. Chua, (1977) Proc. Nat. Acad. Sci. 14; 1082. Chua, N. H. and G. We SChIBidt’ (1978) Proc. Nat. Acad. 5010 LS; 6110. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 114 Feierabend, J. and G. Wildner, (1978) Arch. Biochem. Biophys. 186; 283. Gelvin, S., Heizmann, P., and S. H. Howell, (1977) Proc. Nat. Acad. Sci. 15’ 3193. Calvin, M., (1954) Fed. Proc. 13; 697. Bassham, J. A. and G. H. Krause, (1969) Biochim. Biophys. Acta. 189; 207. "‘ Bassham, J. A., (1963) Advan in Enzymol. 25; 39. Lorimer, G. H., Badger, M. R., and T. J. Andrews, (1976) Biochem. 15; 529. Badger, M. R. and G. H. Lorimer, (1976) Arch. Biochem. Biophys. 175; 723. Cooper, T. G., Filmer, D., Wishnick, M., and M. D. Lane, (1969) J. Biol. Chem. 244: 1081. Cooper, T. G., Tchen, T. T., Wood, H. G., and C. R. Benedict, (1968) J. 81010 Chem. 243} 38570 Lorimer, G. H. and T. J. Andrews, (1973) Nature 243: 359. POD, No Go, Rhin, B. R. and Mo cal-Vin, (1963) BiOChem Z. 338; 70 Fiedler, F., Mullhofer, G., Trebst, A., and I. A. Rose, (1967) European J. Biochem. 1: 395. WighniCk, Mo and M. D. lane, (1969) J. 8101. Chem. 244, 550 Wishnick, M., Lane, M. D., and M. C. Scrutton, (1970) J. Biol. Chem. 245: 4939. Siegel, M. I. and M. D. Lane, (1972) Biochem. Biophys. Res. Cbmmun. Ryan, F. J. and N. E. Tolbert, (1975) J. Biol. Chem. 250: 4234. Kwok, S. Y. and S. G. Wildman, (1974) Arch. Biochem. Biophys. 161: 354. MiZiorkO’ H. M. and A. S. Mildvan, (1974) J. 81010 Cl'lemo 249; 27430 Miziorko, H. M., (1979) J. Biol. Chem. 254; 270. Pierce, J., Barker, R., and N. E. Tolbert, (1979) 178th Am. Chem. Soc. Meeting, Abstract No. 2. Pierce, J., Barker, R., and N. E. Tolbert, (1979) In Manuscript. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 115 Siegel, M. I. and M. D. Lane, (1973) J. Biol. Chem. 248; 5486. Rabin, B. R. and P. W. Trown, (1964) Nature 292; 1290. Rabin, B. R. and P. W. Trown, (1964) Proc. Nat. Acad. Sci. 51; 497. Trown, P. W. and B. R. Rabin, (1964) Proc. Nat. Acad. Sci. 52; 88. Sue, J. M. and J. P. Knowles, (1978) Biochem. 19; 4041. Lorimer, G. H., (1978) European J. Biochem. 82; 43. Pierce, J., Barker, R., and N. E. Tolbert, (1979) J. Biol. Chem. In press. Ryan, F. J., Barker, R., and N. E. Tolbert, (1975) Biochem. Bio- PhYSo Res. Commun. 65; 390 McCurry, S. D. and N. E. Tolbert, (1977) J. Biol. Chem. 252: 8344. Hamilton, G. A., (1974) in "Molecular Mechanisms of Oxygen Activa- tion” (editor 0. Hayaishi) Academic Press, N. Y. Hamilton, G. A., (1969) Advan. in Enzymol. 23: 55. Wildner, Go F0, (1976) Ber. DeutSCh BOto GeSO Ede-22; 3490 Bhagwat, A. S. and P. V. Sane, (1978) Biochem. Biophys. Res. Commun. 84; 865. Akoyunoglou, G. and M. Calvin, (1963) Biochem z. 338; 20 Sugiyama, T., Nakayama, N., and T. Akazawa, (1968) Arch. Biochem. Biophys. 126; 737. Bahr, J. T. and R. G. Jensen, (1974) Plant Physiol. 53: 39. Bahr, J. T. and R. G. Jensen, (1974) Arch. Biochem. Biophys. 164: 408. Chollet, R. and L. L. Anderson, (1976) Arch. Biochem. Biophys. 17 ; 344. Buchanan, B. B. and P. Schurmann, (1973) J. Biol. Chem. 248; 4956. Buchanan, B. B. and P. Schurmann, (1973) Curr. pr. Cell Regul. 73 1o Chu, D. K. and J. A. Bassham (1974) Plant Physiol. 54: 556. Chu, D. K. and J. A. Bassham (1975) Plant Physiol. 25; 720. Whitman, W. B., Oolletti, C., and F. R. Tabita, (1979) FEBS Lett. 101, 249. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 116 Vater, J. and J. Salnikow, (1979) Arch. Biochem. Biophys. 194: 190. Paech, C., McCurry, S. D., Pierce, J., and N. E. Tolbert, (1978) in “Photosynthetic Carbon Assimilation” (editors H. W. Siegelman and G. Hind) Plenum Press, N. Y. and London. Paech, C., Pierce, J., McCurry, S. D., and N. E. Tblbert, (1978) Biochem. Biophys. Res. Cbmmun.{§3; 1084. Chu, D. K., Chang, R., and L. E. Vickery, (1974) Biochim. Biophys. Acta. 334; 438. O'Leary, M. H., Jaworski, R. J., and F. C. Hartman, (1978) Proc. Nat. Acad. SCio‘ZE; 6730 Lorimer, G. H., (1979) J. 3101. Chem. 254} 55990 Bahr, J. T., (1978) Plant Physiol. supplement 61: 538. Chollet, R., and L. L. Anderson, (1978) Biochim. Biophys. Acta. 525: 455. Ellman, G. L., (1958) Arch. Biochem. Biophys. 113 443. Paech, C., Ryan, F. J., and N. E. Tolbert, (1977) Arch. Biochem. Biophys. 179; 279. Whitman, W. and F. R. Tabita, (1976) Biochem. Biophys. Res. Commun. .11; 1034. Paech, C. and N. E. Tolbert, (1978) J. Biol. Chem. 253; 7864. Spellman, M., Hartman, F. C., and N. E. Tolbert, (1979) 178th Am. Chem. Soc. Meeting, Abstract No. 3. Spellman, M., Hartman, F. C., and N. E. Tolbert, (1979) In manu- script. Whitman, W. and F. R. Tabita, (1978) BiOChem‘.lZ' 1282. Whitman, W. and F. R. Tabita, (1978) Biochem. 17; 1288. Hartman, Fe Co, Norton, I. L., Stringer, C. Do, and J. V. SCthSS, (1978) in ”Photosynthetic Carbon Assimilation" (editors H. W. Siegelman and G. Hind) Plenum Press, N. Y. and London. Hartman, F. C., welch, M. H., and I. L. Norton, (1973) Proc. Nat. Acad. SCiO 12} 3721. Norton, I. L., Wélch, M. H., and F. C. Hartman, (1975) J. Biol. Chem. 250; 8062. Schloss, J. V. and F. C. Hartman, (1977) Biochem. Biophys. Res. Commun. 15; 320. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 117 Schloss, J. V. and F. C. Hartman, (1977) Biochem. Biophys. Res. Commun. 11; 230. Sugiyama, T., Akazawa, T., and N. Nakayama, (1967) Arch. Biochem. Biophys. 121: 522. Stringer, C. D. and F. C. Hartman, (1978) Biochem. Biophys. Res. Commun. 80; 1043. Riordan, J. P., McElvany, K. 0., and C. L. borders, Jr., (1977) Science 195; 884. Barker, R., (1971) ”Organic Chemistry of Biological Compounds” Prentice-Hall, New Jersey. Schloss, J. V., Norton, I. L., Stringer, C. D., and F. C. Hartman, Lawlis, V. B. and B. A. McFadden, (1978) Biochem. Biophys. Res. Commun. 80: 580. Chollet, R., (1979) Plant Physiol. supplement 6;; No. 852. Horecker, B. L., Hurwitz, J., and A. weissbach, (1958) Biochem. Preparations 6: 83. Horecker, B. L., (1957) Methods Enzymol. 3; 105. Ryan, P. J. and N. E. Tolbert, (1975) J. Biol. Chem. 250; 4229. Racker, E., (1962) Methods Enzymol. 5; 266. Easterby, J. S., (1973) Biochim. Biophys. Acta. 223; 552. McClure, W. R., (1969) Biochem. 83 2782. Paul, K. G., (1963) The Enzymes 8; 227. Harris, G. C. and A. I. Stern, (1978) J. Exp. Bot. 22; 561. Hall, N. P., McCurry, S. D., and N. E. Tolbert, (1979) Plant Physiol. supplement 6;: No. 357. Paulsen, J. M. and M. D. Lane, (1966) Biochem. 5; 2350. Hall, N. P., McCurry, S. D., and N. E. Tolbert, (1979) In manuscript. vater, J., Salnikow, J., and H. Kleikauf, (1977) Biochem. Biophys. Res. comuno 12} 1618. McCurry, S. D., Paech, C., Pierce, J., and N. E. Tolbert, (1978) Plant Physiol. supplement 61: No. 543. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 118 Cantley, L. C., Jr., Josephson, L., Warner, R., Yanagisawa, M., Lechene, C., and G. Guidotti, (1977) J. Biol. Chem. 252; 7421. Jensen, R. G., Bahr, J. T., and R. C. Sicher, (1977) Proc. 4th Int. Cbng. Photosynthesis, Abstracts (editor J. Cbombs) UKISES, London, p. 179. Hartman, F. C. and R. Barker, (1965) Biochem. 4; 1068. Byrne, W. L. and H. A. Lardy, (1954) Biochim. Biophys. Acta. 14: 495. Serianni, A. 5., Pierce, J., and R. Barker, (1979) Biochem. 18; 1192. Segel, I. H., (1975) "Enzyme Kinetics" John Wiley & Sons, N. Y. Paulus, H., (1969) Anal. Biochem. 32; 91. Wishnick, M., Lane, M. D., and M. C. Scrutton, (1970), J. Biol. Chem. 245: 4939. Pritchard, G. G., Griffin, W. J., and C. P. Whittingham, (1962) J. Exp. BOto E} 1760 Zelitch, I. and G. A. Barber, (1960) Plant Physiol. 35} 205. Pawlizki, K. H., Kelly, G. J., and E. Latzko, (1976) Anal. Biochem. 73: 434. BIO-RAD Technical Bulletin No. 114, May 1970. Ornstein, L., (1964) Ann. N. Y. Acad. Sci. 121; 321. Davis, B. J., (1964) Ann. N. Y. Acad. Sci. 131; 404. Weber, K. and M. Osborn, (1969) J. Biol. Chem. 244: 4406. Anker, H. s., (1970) FEBS Lett. 1; 293. Akoyunoglou, G., Argyroudi-Akoyunoglou, J. H., and H. Methenitou, (1967) Biochim. Biophys. Acta. 132; 481. Sheen, S. J., (1978) Plant Physiol. supplement 61: No. 540. Wildner, G. F. and J. Henkel, (1978) FEBS Lett. 21; 99. Hall, N. P. and N. E. Tolbert, (1979) In manuscript. Benesch, R. E., Benesch, R., Renthal, R. D., and N. Maeda, (1972) Biochem. 11; 3576. Perella, M., Kilmartin, J. V., Foss, J., and L. Rossi-Bernardi, (1975) Nature 256: 759. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 119 Aronoff, S., (1951) Arch. Biochem. Biophys. 32; 237. Koshland, D. E., Jr., (1970) The Enzymes l: 341. Robison, P. D. and F. R. Tabita, (1979) Biochem. Biophys. Res. Commun. 88; 85. Bhagwat, A. S., Ramakrishna, J., and P. V. Sane, (1978) Biochem. Biophys. Res. Commun.‘§3; 954. van de Bogart, M. and H. Beinert, (1967) Anal. Biochem. 20: 325. Aisen, P. and E. B. Brown, (1977) Seminars Hematol. 14: 31. Wishnick, M., Lane, M. D., Scrutton, M. C., and A. S. Mildvan, (1969) J. Biol. Chem. 244: 5761. Chollet, R., Anderson, L. L., and L. C. HOvsepian, (1975) Biochem. Biophys. Res. Commun. 64; 97. Vblwerk, J. J., Pieterson, W. A., and G. H. de Hass, (1974) BiOChem.‘l§; 14460 Badger, M. R. and G. J. Collatz, (1977) Carnegie Inst. Yearbook 76: 355. Wildner, G. F. and Jo Henkel, (1979) FEBS Letto 103; 2460 Okabe, K. I., Oodd, G. A., and W. D. P. Stewart, (1979) Nature 279; 525. McCord, J. M. and I. Fridovich, (1969) J. Biol. Chem. 244: 6049. APPENDICES J O From: PHOTOSYNTI TITK‘ CATT‘VAETI .‘-..C733.‘:HI,.~’\TIOF~I {T.Htcd by ii I; ;iK§C Hicvsiv1zn 33.1(Jocffrcy lihld (ihctkn1}¥fim3lfing LlfpgidLun,1978) ACTIVE SITE OF RIBULOSE 1,5-BISPHOSPHATE CARBOXYLASE/OXYGENASE* Christian Paech, Stephen D. McCurry, John Pierce, and N. E. Tolbert Department of Biochemistry, Michigan State University East Lansing, Michigan 48824 The properties, distribution, biogenesis, function, and regu- lation of ribulose bisphOSphate carboxylase/oxygenase, as described elsewhere in this Symposium, are mainly integrated around the mech- anism of action of the carboxylase and oxygenase reactions. Re- cently, interest has been stimulated by the discovery of the oxygen- ase reaction (1, 2) and its role in the glycolate pathway of photo- respiration (3), and by the desirability of differential regulation of the carboxylase and oxygenase activities, if possible, in favor of the former in order to increase photosynthetic productivity. A prerequisite for such regulation is a thorough knowledge of the relationships between structure and function in this protein. RuBP carboxylase, in various molecular forms (4), is present in all photosynthetic organisms. To date, most of the work has been done either with the crystalline enzyme from tobacco leaves or with an enzyme from spinach whose ease of purification to homogene- ity in large amounts has often been reported. Purified preparations can be obtained in one day by precipitation from the homogenate with polyethylene glycol followed by DEAE cellulose chromatography (N.P. Hall, unpublished). However, the complexity of this large protein and the inability to dissociate it into active subunits have been intimidating, and unfortunately little is known about RuBP carbox- ylase/oxygenase relative to other key enzymes. *Supported in part by NSF grant PCM 7815891. Abbreviations: RuBP, ribulose 1,5-bisphosphate; XuBP, xylulose 1,5-bisphosphate; xylitol BP, xylitol 1,5-bisphosphate; FBP. fructose 1,6-bisphosphate; CRBP, 2-carboxy-D-ribitol 1,5- bisphosphate; bromodihydroxybutanone BP, 3-bromo-l,4-dihydroxy-2- butanone 1,4-bisphosphate. 227 228 C. PAECH ET AL 6 pnzoposu e HO-C-H e CH20P03H .cooe ooc-g-OH +H20 Scheme l C|Z=0 o 1’ H-C-OH ”H cooe CH20P03H6 . H-C-OH 9 (n1) CH20P03H 3.” ch=0 9 CH20P03H9 gH22503H9 epH20P03H gzo 'He (5’0 9‘0” H—C-OH C—OH .__. . 9:0 H—c—OH *“ H-C-OH H-C-OH cnzoposne CHzoposHe CH20P03H9 (I) (11) 0:0 CH opo He 6 CHZOPOSHG 60:39 3 Scheme 2 O-O-C-OH *H O c=o . ———1—o + I _H0 e H-C-OH 9 900 CHZOPOBH H_$_OH 6 (IV) CHZOPOSH . The chemical mechanism of the carboxylation reaction predicted by Calvin (5) (Scheme 1) even before discovery of the enzyme is soundly based on organic chemistry, but progress has been slow in showing how the functional groups at the active site enable the enzyme to catalyze this reaction. The initial step is a base- catalyzed enolization (II) at C3 of RuBP (I), which could be the rate-limiting step of the overall reaction (6). After addition of 002, which is the actual substrate for the carboxylase (7), to the intermediary carbanion, at CZ of RuBP (II), C-C bond cleavage of the intermediate B-keto acid (III) occurs between C2 and C3 (8). On the basis of the enzyme's sensitivity to sulfhydryl-modifying reagents, and in order to explain unsuccessful attempts to isolate the hypothetical B-keto acid (III), a catalytic mechanism involving hemimercaptal formation of a sulfhydryl group with RuBP at carbon 2 was proposed (9). Later, support for the existence of the B-keto intermediate (III) was sought by attempts at its synthesis (10, 11). The finding that the stable transition-state analog of (III), 2-carboxy-D-ribitol 1,5-bisphosphate (CRBP) inhibits by competition with RuBP (12, 13) has been considered proof for intermediate (III) . in the reaction. Calvin's original hypothesis for the mechanism of the carbox- ylase reaction was also used to explain the 18O-labeling pattern in ACTIVE SITE 229 phosphoglycolate produced by the oxygenase reaction (2) (Scheme 2). Although this concept is appealing, the enzyme does not contain any of the cofactors normally associated with oxygenase essential for activating the 02 (2,13), such as metal ions or flavins. 02 is normally in the triplet state, which is 22 kcal/mol lower in energy than singlet 02. A direct reaction of triplet 02 with a singlet molecule (II) is a Spin-forbidden process. Therefore, the enzymic mechanism for the oxygenation of RuBP must circumvent the mechanism described by Scheme 2, or use some other one, and it remains unknown. 2+ Pon et al. (14) showed in 1963 that preincubation with C02 and Mg was essential to the enzyme assay, and this requirement has been further ascribed to a homotrOpic effect of 002 (15, 16). Pon et al. also noted that the reaction had to be initiated with RuBP in order to get a maximal rate. But it was not clearly established until 1976 that CO and Mg 2+ activation is a necessity for obtaining a realistic Km(C02§ of 10 to 20 HM and a Km(02) of about 200 EM, and that without C02 and Mg 2+ pretreatment the enzyme is active (17,18). The activity of RuBP carboxylase lg vivo is thought to be controlled also b several other mechanisms such as shifts in stromal pH and Mg + concentration due to the H+ gradient created by electron transport in the light. In addition, some of the intermediates of the reductive and oxidative pentose phosphate pathways are effectors for isolated RuBP carboxylase/oxygenase. Among these are 6-phosphogluconate (see Bassham et al., this volume), fructose 1,6-bisphosphate (FBP), ribulose S-phosphate, ribose 5- phosphate, and NADPH. The physiological function of these effectors is uncertain, and their mechanism and site of action on the carbox- ylase are unknown. A working model for the active site (Figure l) of RuBP carbox- ylase/oxygenase must take into account a binding site for RuBP and the mechanism of the base-catalyzed enolization, a C02 substrate site near carbon 2 of RuBP, and the possiblity of an 02 site, as well as amino acid residues essential for catalysis. The C02 acti- vator site must bind both C02 and Mg2+ in a manner to activate or affect the catalytic site. In addition, there is a need for ef- fector sites. ESSENTIAL LYSYL RESIDUES AND SULFHYDRYL GROUPS Pyridoxal 5'-ph08phate has been used as a probe to demonstrate that e-amino groups of lysyl residues are important for the mecha- nism of RuBP carboxylase/oxygenase from spinach (19-21) and Rhodo- spirillum rubrum (22-24). Pyridoxal phosphate has some advantage over other affinity labels employed so far (ZS-29) (see Hartman et al., this volume) in that it not only reacts reversibly with RuBP carboxylase forming a Schiff base, but also can be used as an irre- versible inhibitor through subsequent NaBH4 reduction. In addition to the characteristic spectral properties of both the Schiff base and its reduced form, a radioactive label can be easily introduced 230 C. PAECH ET AL. Activator site Catalytic site Lys Arg or Lys Lys 5 l @9 I ': 9H2 .. C=O C G I II S H—C-OH O I . H- -OH 2 ‘3 ——J CH2 ? I @e Arg or Lys Figure 1. Working model for active site of RuBP carboxylase/oxygenase. IOO 3: so- ' IE ° 8 o 60 > E °}. >‘ I g \o 40» 3 ° \ 8 o 20 r 8 \--~3: C O O A .. $.\0~a_i 2 4 6 8 IO l2 I4 16 [pyridoxal phosphate bound] [R u 8P carboxylase/oxye nose] Figure 2. Activity of RuBP carboxylase/oxygenase with increasing amounts of bound pyridoxal phosphate after NaBHh reduction. Open circles represent carboxylase activity and squares, oxygenase activity. The enzyme was incubated with increasing concentrations of pyridoxal phosphate and reduced with NaBH4, and the amount of bound pyridoxal phosphate was determined spectrophotometrically (21). iii ACTIVE SITE 231 by NaB3H4 reduction, but the stoichiometry is subject to an isotOpe effect (C. Paech, unpublished). Pyridoxal phosphate appears to have an exclusive specificity for primary amino groups, whereas the affinity labels 3-bromo-l,4-dihydroxy-2-butanone 1,4—bisphos- phate and N-bromoacetylethanolamine phOSphate exhibit greatest activity with sulfhydryl groups (26, 28). RuBP carboxylase/oxygenase from spinach, which has an A838 structure, was completely inactivated when 16 lysyl residues were blocked by pyridoxal phosphate (21). The stoichiometry of pyridoxal phosphate incorporation was determinid spectrophotometrically with an extinction coefficient of 4800 M" cm"1 for the reduced Schiff base, as derived for this system (21). This is different from the coefficient used for this purpose as found in the literature, but it is in the same range as that derived for aspartate transcarbamyl- ase (30). This value for the reduced Schiff base should not be con- fused with the value 5800 M'1 cm’1 derived for the Schiff base be- tween RuBP carboxylase and pyridoxal phosphate. Loss of enzyme activity with increasing amounts of irreversibly bound pyridoxal phosphate was linear but biphasic (Figure 2). The two linear por- tions extrapolated to 8 and 16 mols of bound pyridoxal phosphate per mol of enzyme. Complete inactivation was not reached until all 16 mols of pyridoxal phosphate were bound. The carboxylase and the oxygenase activities declined in parallel. SDS gel electrophoresis of enzyme-pyridoxal phosphate complexes reduced with NaB3H4 revealed no tritium incorporation into the small subunit under conditions producing complete inactivation. Introduc- tion of covalently bound pyridoxal phosphate resulted in a net charge change on the subunits, so that specificity of pyridoxal phosphate in- corporation could also be monitored directly by gel electrophoresis in 8 M urea. Even under conditions such that 16 pyridoxal phosphate molecules were bound, only the mobility of the large subunit was changed. For both activities, modification of the enzyme decreased Vmax values but did not change Km values, which indicates only a decrease in the amount of catalytically active enzyme and not a change in the catalytic properties of the enzyme. Bicarbonate and Mg2+ competitively reduced the pyridoxal phos- phate inhibition of the spinach carboxylase (Figure 2 of ref. 20). This is in contrast to the enhancement of inhibition reported for the enzyme from R; rubrum (23). In the absence of Mg2+, RuBP, alone or with bicarbonate, provided full protection against pyridoxal phosphate inactivation. Although the number of pyridoxal phosphate molecules incorporated was reduced by 16, a small amount of non- specific binding of pyridoxal phosphate still occurred (21). RuBP was at first reported to protect about half the reactive lysyl residues against pyridoxal phosphate (20). This was an error caused by the experiments being done in the presence of Mg + and incubated for 20 min to establish the Schiff base equilibrium. Under these conditions the RuBP was slowly lost through catalytic reaction with 02 and C0 , which had not been excluded from the in- cubation mixture, and tgerefore the protective effect was lost. 232 C. PAECH ET AL. On the basis of these data, the simplest explanation is that the 16 essential amino groups are located in the active-site re- gion. A more complicated explanation would be that RuBP exerts its protective effect both at the catalytic site directly and at an allosteric site through a proposed RuBP-induced conformational change of the enzyme. To gain further information about the function of the two lysyl groups, the reactivation pattern of the pyridoxal phosphate-inhibited enzyme during dissociation of the Schiff base complex was studied (21). From low initial inhibition, reactivation proceeded in a first- order reaction, independent of whether enzyme activity was measured directly during reactivation or after samples taken at different times had been fixed by NaBH4 reduction. From high levels of pyri- doxal phosphate inhibition, the reactivation followed first-order kinetics with the same rate constant only in the experiment that included NaBHa reduction. Direct assay showed a clear delay in the reactivation process, as if a rate-limiting step were imposed on the overall process. This suggested that the lysyl groups reacting only at high pyridoxal phosphate concentrations are at the activator sites, and that the delay in regaining activity is due to slow acti- vation of the enzyme with C02 and Mg2+ after liberation of the pyri- doxal phosphate. This has to take place before any particular sub- unit is catalytically active. The 8 amino groups reacting at low pyridoxal phosphate concentration appear to be at the catalytic sites. This is consistent with the reactivation kinetics and with the unaltered Km values of the pyridoxal ph08phate-modified enzyme. The structural similarity between RuBP and pyridoxal phosphate, i.e., that the carbonyl group and one phosphate are 3 carbons apart, may account for a higher affinity for pyridoxal phosphate at the substrate site. Differential labeling of the enzyme first with p-chloromercuri- benzoic acid and then with pyridoxal phosphate indicated that one group of 8 of the lysyl residues was protected from the pyridoxal phosphate. Removal of the mercuribenzoate group by dithiothreitol left the enzyme inactive with only 8 pyridoxal phosphate molecules bound to it. Although both p-chloromercuribenzoic acid and iodo- acetamide inactivate the enzyme, and the binding of each can be pre- vented by RuBP, only the large p-chloromercuribenzoate group blocks pyridoxal phosphate binding. Since modification of the sulfhydryl group(s) prevents binding of RuBP, a sulfhydryl.group(s) must be very close to the binding site for RuBP. Therefore, the 8 lysyl groups excluded from reaction with pyridoxal phosphate in the presence of p-chloromercuribenzoic acid are thought to be located at the cata— lytic site, and the other group of 8 that are still available for Schiff base formation are possibly located in the vicinity of the activator site. The function of the e-amino group of lysine at the catalytic site is most likely noncovalent binding of C02, since pyridoxal phosphate appears as a competitive inhibitor for C02 (20). However, lysyl groups, there, may also be involved in proton transfer and ACTIVE SITE 233 binding of phosphate groups. An e-amino group of lysine at the activator sites could provide the requirements for allosteric binding of C02 and Mg2+ that would be exothermic, reversible, and pH dependent according to the model in Scheme 3, which is based on the model for C02 activation (17): +Mg NH Enz-NH2 + co2 ———*.___ Enz-NH-COO- .——————-‘ Enz / \coo'. ..MgZ+ '”"M92+n’ Supporting the model in Figure l for two types of primary amino groups is the establishment of two different lysyl residues, also, with the use of bromodihydroxybutanone bisphosphate (29). Since this affinity label is one carbon atom shorter than RuBP, it may have greater mobility at the active site and therefore react with either lysyl group; but, once bound to one of the two, the first molecule may preclude binding of another to the second amino group because of a high charge density through the phosphate groups. This does not explain the low stoichiometry of 4 to 5 moles of reagent per mole of enzyme for inhibition, when the stoichiometry of inhibition with CRBP clearly indicates the re- quirement of 8 equivalents per mole of enzyme (11). The determin- ‘ ation of bound bromodihydroxybutanone BP is based either on 32F measurement of the incorporated affinity label or on 3H measure- ments after NaB3H4 reduction of the enzyme- -inhibitor complex. Both methods are subject to errors due to the known lability of the affinity label (loss of one or two phosphate groups) (28) and to possible isotopic effects during reduction with NaB H4. The A2 enzyme from R; rubrum binds two pyridoxal phosphate molecules per molecule or one per subunit, but is inactivated after blocking of the first catalytic site (24). This agrees with data obtained with bromodihydroxybutanone BP for the same enzyme (25), but is in contrast to the finding that two pyridoxal phOSphate molecules must be bound to each subunit of the spinach enzyme. One suggested reason for this difference is that RuBP carboxylase from R; rubrum may behave like a half-site enzyme. Although the amino acid compositions of the large subunits of the carboxylaSe from all its various sources are apparently similar, there should be some differences between the A2 and the A838 enzyme, expressed in catalytic and regulatory properties. ARGININE AT BINDING SITE FOR RuBP Arginine residues in enzymes serve a general function as bind- ing sites for negatively charged groups such as phosphate.or car- boxylate ions (31). Similarly, RuBP carboxylase/oxygenase is 234 C. PAECH ET AL. inactivated by reagents specific for arginyl residues, i.e., 2,3- butanedione and phenylglyoxal. McFadden's group (32), using 2,3- butanedione with borate to bind a third of the total arginines, noted enzyme inhibition that was partially protected against by the product 3-phosphoglycerate, as if the modified arginines were at the active site. Phenylglyoxal, which does not require borate, appears to be a better reagent for arginyl residues in this enzyme. Inactivation follows saturation kinetics, and ~.25 to 30 arginyl residues are blocked, as judged by amino acid analysis. RuBP has given about 70% protection against phenylglyoxal inhibition (C. Paech and M. Spellman, unpublished), although Hartman's group (33) reports finding no such protection. These preliminary results indicate that one or more arginyl residues at the active site probably function in binding the phos- phate of RuBP. Further work is necessary to decide whether both phosphate groups of RuBP are bound by arginine or whether some other cationic group such as a lysyl e-amino group may also be involved. STUDIES OF THE RuBP BINDING SITE WITH SUBSTRATE ANALOGS RuBP carboxylase/oxygenase is specific for RuBP as a substrate for carboxylation and oxygenation. A very low dissociation constant for the enzyme-RuBP complex has been estimated: 0.5 “M (34) or sl‘ “M (12). In fact, the RuBP-enzyme complex can be separated from excess RuBP by gel filtration, and the bound RuBP subsequently de- tected by the carboxylase reaction. Whereas most sugar mono- and biSphOSphates are poor inhibitors, xylitol BP (S. D. McCurry, un- published, but see ref. 35) and xylulose bisphosphate (XuBP) (36) are powerful, competitive inhibitors with respect to RuBP. The maximum inhibitory effect of either one is manifested after'”20 min, and the binding constant for xylitol BP is <1 ”M (McCurry, unpub- lished). Like CRBP, 3H-xylitol BP could not be dialyzed away from the enzyme. XuBP and xylitol BP differ from RuBP stereochemically at C3, and their severe inhibitory effect may reside in hydrogen— bonding between that hydroxyl group and the base at the catalytic site that participates in enediol formation for RuBP. The energy required to break this hydrogen band could contribute substan- tially to stability of the enzyme-inhibitor complex, as evidenced by ribitol BP being a poor inhibitor (F. J. Ryan, unpublished). XuBP has not been reported to occur in nature, but it could be readily formed 33 situ, and it is prepared by aldolase lg vitro from dihydroxyacetone phosphate and glycolaldehyde phosphate. Apparently the formation of free glycolaldehyde phosphate would be lethal, and this could explain why all C2 transformations in carbohydrate metab- olism involve a glycolaldehyde-thiamin pyrophosphate complex with transketolase rather than free glycolaldehyde phosphate. Conceiv- ably XuBP could be formed by epimerization from RuBP in the active site of the carboxylase from the enediol intermediate if it were allowed to revert, but the absence of such reversion is evidenced ACTIVE SITE 235 by the lack of 3H exchange in the absence of C02 with RuBP labeled with 3H at C3 (6). The question of whether or'not such an epimeri- zation would be permitted by the enzyme remains moot. The most effective inhibitor known for carboxylase/oxygenase is the transition-state analog CRBP (ll, 12), a competitive, time- dependent, irreversible inhibitor. Each enzyme molecule has 8 bind- ing sites for this compound, and a divalent cation is required for binding. Synthesis of CRBP from cyanide and RuBP produces two epimers, CRBP and 2-carboxy-D-arabinitol 1,5-bisphosphate, and the proportion of inhibition due to each is unknown. These two com- pounds have been synthesized and purified in our laboratory (J. Pierce, unpublished), and their effects on RuBP carboxylase/ oxygenase are under investigation. In addition, the compounds 2-carboxy-D-xylitol 1,5-bisphosphate and 2-carboxy-D-1yxitol 1,5- bisphosphate have been synthesized and purified. Determination of the relative effects of these compounds on the enzyme, coupled with knowledge of the stereochemistry of known inhibitors of the enzyme, may allow estimation of the geometry of the active site. It has been suggested that carboxylase inhibition by RuBP is an allosteric effect (37) or is due to an interaction with lysine at the catalytic site (19). Attempts to detect a Schiff base be- tween either RuBP or XuBP at a lysine by reduction with NaB3H4 have not been successful (Paech and McCurry, unpublished). Summa- rized below is new evidence that the RuBP inhibition is caused by other inhibitory compounds, leaving unanswered the question whether RuBP itself has any regulatory control of enzyme'hctivity. RuBP DEGRADATION PRODUCTS AS INHIBITORS Inhibition of the initial rate of carboxylase activity.by high concentrations of RuBP has been known for many years (14). Pre-incubation of the enzyme with RuBP in the absence of 002 re- sults in inhibition. Therefore the enzyme assay is initiated with RuBP after first activating with 002 and Mg2+. Although RuBP in- hibition has been ascribed to an allosteric effect (37) or to an interaction at the active site (19), it does not occur in assays with whole chloroplasts (38). We have obtained data indicating that the "substrate inhibition" phenomenon produced with solutions of RuBP is due to the presence of inhibitors formed from the RuBP by epimerization at C3 or by B-elimination of the C1 phosphate (39). The idea that an inhibitory compound was present in RuBP arose from the observation that different batches of RuBP had different initial reaction rates although all contained nearly the same amount of RuBP as judged by enzyme, phosphate, and carbohydrate analyses. The initial catalysis rate, when a reaction is started with RuBP, is almost unaffected by the suspected impurities; but, with time or in the absence of RuBP after the substrate is used up (Figure 3), time-dependent inhibition of these impurities becomes apparent. Addition of a second aliquot of substrate to the same enzyme results 236 C. PAECH ET AL. C .5 I- cl 8 0. 04L , (35w, E A II A \ v n - v "D Q) I- «II .5 c..— ONZP ‘ O I a . . E 1 I2 3"8 9 IO ll l2 Ihne (nun) Figure 3. Time course of RuBP carboxylase reaction with a limit- ing amount of RuBP. After the reaction had gone to completion, a second aliquot of substrate was added, and the change in the ini- tial rate was recorded as percent inhibition (39). in a much slower, or inhibited, rate of carboxylation. Product inhibition was ruled out (39). In RuBP solutions, the inhibitors build up with increasing storage time, particularly at pH above 8 or elevated temperatures. Incubation of RuBP solutions at pH 11 and 30° results in total loss of phosphate at carbon 1. Under the condition of the carbox- ylase assay, RuBP decomposes at the rate of 1.25% per hr. Mild base treatment causes substantial inhibition without substantial loss of substrate. Fresh solutions of enzymatically synthesized RuBP have essentially no inhibitory components, but RuBP from commercial sources is inhibitory because of degradation products formed during the isolation steps. Two inhibitory products have been found in RuBP solutions. One is XuBP, which arises from nonenzymic epimerization, as shown in reaction Scheme 4: ACTIVE SITE 237 .1 _ F - CH2 oPo,2 c'Hfd‘PoE CH2 0P0,2 c=o - (c-o'.) - c=o . OH H-C-OH :- 66” OH HO-C-H H-C-OH H-C—OH 2_ H-C-OH CHzopof‘ _ CHZOPO3 J CHZOPOE' PO43 9H2 913 I 900' coo‘ C-OH c=o - - ,_ __ ._ +0H’ H,c COH + H0 90H, 9-0 ~— 9-0 ——> H-C-OH 2_ H-C-OH 2_ ”‘94)” zH'Q‘OH 2_ CHZOPO3 CHZOPO3 CH20P03 CHZOPOZU (Y) (121) (III) l-deoxy-D-glycero- 2-C-methyl-D - 2-C-methyI-D - 2,3-pentodiulose- erythrotetronic acid threoietronic acid 5 - phosphate 4 -phosphate 4 - phosphate Proof of its presence and structure has been obtained by thin-layer and gas-liquid chromatography after reduction and phosphatase treat- ments. XuBP preparations used as substrate with isolated enzyme cause some initial C02 incorporation into 3-phosphoglycerate (39), but this catalysis is due to the presence of a little RuBP arising from epimerization rather than to XuBP acting as a substrate. The other inhibitory compound arising from RuBP appears to be a diketo degradation product formed from one or the other or both of the bisphosphates by B-elimination. Its structure has not been unequivocally established, but the UV absorption spectra, NMR spec- tra, and phosphate elimination support the idea that it is l-deoxy- D-glycero-Z,3-pentodiulose S-phosphate. Additional support comes from treatment of RuBP solutions with o-phenylenediamine resulting in partial removal of the inhibitory effect: compound V is rela- tively unstable and probably undergoes rearrangement to a stable end product having properties consistent with those of a branched- chain compound (Scheme 4), clearly recognizable by NMR spectroscopy after extensive base treatment of RuBP. Since the diketo compound, which appears to be the inhibitory component, is rather unstable, it accumulates to only a small extent and does not persist in the reaction mixture. The physiological implication of these data is that RuBP is an unstable substrate with toxic degradation products, and therefore 238 C. PAECH ET AL. free RuBP must not occur in chloroplasts. Under normal condi- tions in chloroplasts the estimated concentration of RuBP is 0.2 to 0.4 mM but that of RuBP binding sites in the carboxylase is 8 times as high: 3 mM (38, 40). Therefore, because of the very low dissociation constant (0.5 nM) for the enzyme-RuBP complex (11), it is unlikely that this labile substrate ever exists free in solution in the chloroplast. In chloroplasts, only 40 to 60% of the carboxylase is estimated to be active at any one time (41). Our hypothesis is that one function for the large amounts of the carboxylase protein in chloroplasts is storage of RuBP to prevent its breakdown to inhibitory products. CARBOXYLASE/OXYGENASE RATIO AND 02 SITE Since RuBP carboxylase activity initiates the photosynthetic carbon cycle and RuBP oxygenase activity diverts carbon flow into phosphoglycolate formation and photorespiration, the regulation and control of these two overlapping and competing carbon pathways appears to be a major consideration for net photosynthesis. Com- parison with other systems suggests that regulation may depend on a combination of substrate availability and alteration in enzyme activity for reactions that initiate metabolic pathways or are at branch points. Indeed, compelling evidence indicates that RuBP carboxylase and oxygenase activities are competitively dependent on the C02 and O substrate concentrations (42), but some evidence suggests that the amount of carbon flow through the glycolate path- way in the leaf is in part independent of the COz/Oz ratio. A change in enzyme activity at branch points is the general way of dealing with alterations in limiting amounts of protein, through feedback regulants or effectors. The carboxylase/oxygenase, however, is present in large excess. Effectors for this enzyme have been reported, but they alter the carboxylase and oxygenase activities similarly (43). All the inhibitors or substrate analogs so far examined also affect the carboxylase and oxygenase activities similarly. So far there is no confirmed evidence for any alteration in the carboxylase/oxygenase ratio by any mechanism other than C02 and 02 competition. We cannot confirm the claim that RuBP carbox- ylase and RuBP oxygenase are two different proteins (44) (manuscript in preparation). There have been reports of variations in the activity ratio of preparations from different stages of leaf de- velopment, from leaves versus fruit, from different plants or mu- tants, after treatment with different inhibitors or effectors, and at different degrees of purification (literature not reviewed). However, many of these reports cannot be confirmed; until such re- ports are confirmed by several groups, a change in the carboxylase/ oxygenase ratio must be considered unproven. One of the difficul- ties is that the oxygenase assay is relatively insensitive compared with the carboxylase assay. For instance, the enzyme must be C02 activated, but C02 is an inhibitor of the oxygenase. The C02 and ACTIVE SITE 239 MgZI-activated enzyme, upon dilution into a CO -free oxygenase assay medium, loses activity with a half-life of l min (N. P. Hall, unpublished). The nature of the RuBP oxygenase reaction, the possible exis- tence of an 02 site or an undiscovered cofactor for the oxygenase, and whether the reaction can be regulated are critical unanswered questions. It is very unlikely that the lysyl COZ-binding sites could bind or activate 02. Despite intensive search, no cofactor has been found that is required for the oxygenase reaction. An early report that Cu2+ is present in the enzyme but has no effect on the carboxylase activity (45) has twice been reexamined without finding Cu2+ or any effect of Cu2+ on oxygenase activity (2, 13). The speculation that inherent amino acid functions of the enzyme, such as sulfhydryl groups, could serve as activator and/or binding site for the oxygen (46) was based on a claim that differential regulation of carboxylase and oxygenase could be achieved by mod- ifying sulfhydryl groups with glycidate or iodoacetamide; but this could not be confirmed by several laboratories including ours. The basis for C02 and 02 being competitive inhibitors of each other may be competition for a reaction with the other substrate, RuBP, rather than for a reactive-site component. In fact the most impor- tant property of this enzyme may be that it catalyzes the carboxyla- tion reaction in the face of an unavoidable oxygenase activity. The unknown nature of the oxygenase is of great interest mechan- istically as well as physiologically. RuBP oxidation is an enzymic property of the RuBP carboxylase/oxygenase protein. Without the activated enzyme RuBP oxidation is relatively slow. Not only does RuBP rearrange and decompose to inhibitory products as described above, but it may also be attacked by strong oxidants equivalent to those found in the chloroplasts (F. J. Ryan, unpublished; J. Pierce, unpublished; N. P. Hall, unpublished). Just as we are pro- posing that RuBP is stabilized against rearrangement in the chloro- plast by being stored at the RuBP binding site on the enzyme, it may also need protection against oxidation in the chloroplasts. The enzyme-RuBP storage complex is inactive as either a carboxylase or an oxygenase, but C02 and Mg2+ activate both activities (18). This activation by C02 or by effectors, which is necessary for ca- talysis, is to be avoided at other times in the presence of limit- ing amounts of CO because of the oxygenase reaction. Upon activa- tion, the 02 attack may be an essential part of the overall photo- synthetic process, or it may be an unavoidable reaction of an intermediary enolate form of the RuBP in the active site of the enzyme. But exactly how the oxygenase reaction occurs is not known. SUMMARY The catalytic region of each large subunit of RuBP carboxylase/ oxygenase from spinach must contain two anionic binding sites for RuBP, of which one or both appear to be arginine. This site may 240 C. PAECH ET AL. serve for storage of labile RuBP and as a catalytic site upon ac- tivation. The 00 substrate site includes a lysyl residue which might take part in orienting the CO adjacent to carbon 2 of the RuBP. The lack of evidence for an 2 binding site suggests that the oxygenase activity may proceed by direct oxidation of the ac- tivated RuBP-enzyme complex. The CO /Mgz+ activator site is also in the large subunit and inyglves another lysyl residue for the formation of a carbonate-Mg complex. The conformational change induced during the formation of this complex is conceptually an orientation of the base opposite C3 of RuBP for catalyzing the rate-limiting enolization step. 11. 12. 13. 14. 15. l6. l7. 18. 19. REFERENCES Bowes, G., Ogren, W. L., and Hageman, R. H., Biochem. Biophys. Res. Commun. 42, 716-22 (1971). Lorimer, G. H., Andres, T. J. and Talbert, N. E., Biochemistry 12, 18-23 (1973). Talbert, N. E. and Ryan, F. J., in 999 Metabolism and Plant Productivity, pp. 141-59, R. H. Burri§ and C. C. Black, Editors, University Park Press, 1976. McFadden, B. A., Bacterial. Rev. 31, 289-319 (1973). Calvin, M., Fed. Proc. 13, 697-711 (1954). Fiedler, F., Mfillhofer, G., Trebst, A., and Rose, I. A., Eur. J. Biochem. 1, 395-9 (1967). Cooper, T. G., Filmer, D., Wishnick, M., and Lane, M. D., J. Biol. Chem. 244, 1081-3 (1969). Mfillhofer, G. and Rose, I. A., J. Biol. Chem. 249, 1341-6 (1965). Rabin, B. R. and Trown, P. W., Nature London 292, 1290-3 (1964). Sdein, B. and Vestermark, A., Biochim. Biophys. Acta 221, 165-73 (1973). Siegel, M. I. and Lane, M. D., J. Biol. Chem. 248, 5486-98 (1973). Wishnick, M., Lane, M. D., and Scrutton, M. C., J. Biol. Chem. 245, 4939-47 (1970). Chollet, R., Anderson, L. L., and Hovsepian, L. C., Biochem. Biophys. Res. Commun. 93, 97-107 (1975). Pan, N. G., Rabin, B. R., and Calvin, M., Biochem. Z. 338, .7-19 (1963). Sugiyama, T., Nakayama, N., and Akazawa, T., Arch. Biochem. Biophys. lZQ. 737-45 (1968). Murai, T., and Akazawa, T., Biochem. Biophys. Res. Commun. 49, 2121-6 (1972). Lorimer, G. H., Badger, M. R., and Andrews, T. J., Biochemistry 15, 529-36 (1976). Badger, M. R. and Lorimer, G. H., Arch. Biochem. Biophys. 115, 723-9 (1976). Paech, C., Ryan, F. J., McCurry, S. D., and Talbert, N. E., Plant Physiol. Suppl. 21, 54 (1976). ACTIVE SITE 241 20. Paech, C., Ryan, F. J., and Talbert, N. E., Arch Biochem. Biophys. 111, 279-88 (1977). 21. Paech, C. and Talbert, N. E., J. Biol. Chem., in press (1978). 22. Whitman, W. and Tabita, F. R., Biochem. Biophys. Res. Commun. 11, 1034-9 (1976). 23. Whitman, W. and Tabita, F. R., Biochemistry 11, 1282-7 (1978). 24. Whitman, W. and Tabita, F. R., Biochemistry 11, 1288-93 (1978). 25. Norton, I. L., Welch, M. H., and Hartman, F. C., J. Biol. Chem. 129, 8062-8 (1975). 26. Schloss, J. V. and Hartman, F. C., Biochem. Biophys. Res. Commun. 12, 320-8 (1977). 27. Schloss, J. V. and Hartman, F. C., Biochem. Biophys. Res. Commun. 11, 230-6 (1977). 28. Hartman, F. C., Welch, M. H., and Norton, I. L., Proc. Natl. Acad. Sci. USA 19, 3721-4 (1973). 29. Stringer, C. D. and Hartman, F. C., Biochem. Biophys. Res. Commun. 19, 1043-8 (1978). 30. Blackburn, M. N. and Schachman, H. K., Biochemistry 12, 1316- 22 (1976). 31. Riordan, J. F., McElvany, K. D., and Borders, C. L. Jr., Science 111, 884-6 (1977). 32. Lawlis, V. B. and McFadden, B. A., Biochem. Biophys. Res. Commun. 19, 580-5 (1978). 33. Schloss, J. V., Norton, 1. L., Stringer, C. D., and Hartman, F. C., Fed. Proc. 11, 1310 (1978). ‘ 34. Vater, J., Salnikow,,J., and Kleinkauf, H., Biochem. BiOphys. Res. Commun. 15, 1618-25 (1977). 35. Ryan, F. J., Barker, J. R., and Talbert, N. E., Biochem. Biophys. Res. Commun. 9;, 39-46 (1975). 36. McCurry, S. and Talbert, N. E., J. Biol. Chem. 121, 8344-6 (1977). 37. Chu, D. K. and Bassham. J. A;, Plant Physiol. 11, 720-6 (1975). 38. Jensen, R. G. and Bahr, J. T., Annu. Rev. Plant Physiol. gg, 379-400 (1977). 39. Paech, C., Pierce, J., McCurry, S. D., and Talbert, N. E., Biochem. BiOphys. Res. Commun., submitted (1978). 40. Hitz, W. D. and Stewart, C. R., Plant.Physiol. Suppl. Q1, 100 (1978). 41. Bahr, J. T. and Jensen, R. G., Arch Biochem. BiOphys. 111, 39-48 (1978). 42. Bowes, G. and Ogren, W. L., J. Biol. Chem. 131, 2171-6 (1972). 43. Chollet, R. and Anderson, L. L., Arch. Biochem. Biophys. 119, 344-51 (1976). 44. Branden, R., Biochem. Biophys. Res. Commun. 81, 539-46 (1978). 45. Wishnick, M., Lane, M. D., Scrutton, M. C., and Mildvan, A. S., J. 3101. Chem. 1%, 5761-3 (1969). 46. Wildner, G. F., Ber. Dtsch. Bot. Ges. 11, 349-60 (1976). 242 C. PAECH ET AL. DISCUSSION McFADDEN: There are about 80 to 100 sulfhydryl groups per molecule of the higher plant enzyme. How extensively modified by mercurial reagent (PCMB) was your RuBP carboxylase when tested for RuBP binding? If extensive modification had occurred, it would be important to probe directly for RuBP binding (i.e., by gel filtra- tion or equilibrium dialysis). Was this done? PAECH: The enzyme was incubated with a 50-fold molar excess of p-chloromercuribenzoic acid for 20 min. This resulted in modi- fication of'~30 SH groups, according to T. Sugiyama et al. (Arch. Biochem. Biophys. 112, 98-106), and complete loss of enzyme activity. RuBP binding was determined directly by gel filtration (see ref. 21). MIZIORKO: You mentioned that a very tight complex is formed with both XuBP and xylitol BP. Is this type of complex stable to treatment by passage over a Sephadex column? McCURRY: Yes, very stable. MIZIORKO: Is there a divalent ion requirement for formation of the complex? McCURRY: No. MIZIORKO: Is a covalent adduct formed in either or both cases? McCURRY: We cannot envision one. We tried to reduce the complex, as it had been suggested that RuBP might form a Schiff base in the active site with carbonyl, and we thought of applying the same hypothesis to XuBP. We could not show any reduction, but that is no guarantee it will not occur. MIZIORKO: Do you have to age the incubation mixture in order to get a tight complex that is stable to Sephadex chromatography or can you just do an initial mixing? McCURRY: Do you mean running over the column within 1 or 2 min? MIZIORKO: Yes. You mentioned that it was a slow process. McCURRY: Yes, «90% inhibition occurs within 2 min but complete inhibition takes 20 min. MIZIORKO: According to a model which requires two C02 sites- per subunit of enzyme, modification of either the substrate C02 site or the activator C02 site should result in loss of catalytic activity. The modification data clearly show two types of site, with substantial enzyme activity remaining in the region of pyri- doxal phosphate titration where virtually complete modification of one type of site is expected. Is the residual activity consistent with the two-site model? TOLBERT: According to our proposal, binding of the first group of pyridoxal phosphate molecules is facilitated because pyridoxal phosphate is structurally similar to RuBP. At higher pyridoxal phOSphate concentrations, significant amounts also begin to form Schiff base at the second site. Because of the increasing competition between the two groups of pyridoxal phosphate binding sites, which becomes significant after 6 pyridoxal phOSphate molecules ACTIVE SITE 243 are bound to the enzyme, the reaction with the first set of amino groups would not go to completion when 8 pyridoxal phosphate molecules were bound, nor would full inactivation occur until saturation of all 16 amino groups with pyridoxal phosphate. HARTMAN: The lack of protection by bicarbonate against modi- fication by pyridoxal phosphate appears inconsistent with your hypothesis that the site labeled is involved in binding of C02. PAECH: The data from kinetic experiments in Figure 2 of ref. 20 clearly demonstrate protection by bicarbonate against pyridoxal phos- phate inhibition. In order to measure spectrophotometrically Schiff base formation between the enzyme and pyridoxal phosphate, the en- zyme was equilibrated at 1 mg/ml with 0.5 mM pyridoxal phosphate and 0 to 300 mM bicarbonate. Although dissociation constants are not available for the enzyme-C02 complex and the enzyme-pyridoxal phos- phate complex, it is clear from the kinetic experiments in ref. 20 that the two competing ligands were present in an unfavorable ratio, i.e., too little bicarbonate. Therefore, this second research ap- proach indicated incorrectly a lack of protection by CO . WILDNER: What is the chemical evidence that the activation of RuBP carboxylase is based on carbamate formation at the lysine of the activation site? TOLBERT: Carbamate formation with a basic amino group, in this case e-amino group of lysine, has been the mechanism generally cited for C02 modification of enzyme activity. As pointed out in the text, carbamate formation is exothermic and pH sensitive, as required for RuBP carboxylase activation. Further chemical and physical proof for this carbamate formation will be forthcoming. CHOLLET: Are the products of RuBP degradation irreversible inhibitors? If you assay the enzyme after passage over a Sephadex 625 column, or after dilution, is it still inhibited? TOLBERT: The xylitol BP reaction is essentially irreversible. CHOLLET: In doing chemical modification studies, we, and perhaps the Oak Ridge people and Bob Divita, have used substrate protection, as you have. If we use control enzymes with no modifier, plus or minus 5 mM RuBP, preincubated for l to 1% hr, then dilute out or pass over a Sephadex G25 column, the two enzymes (plus or minus RuBP) have identical activity. Since the time is sufficient for degradation products to form, either they are readily reversible or we do not see them; the data of Fred Hartman and Bob Divita in- dicate that they do not see them either. PIERCE: It has not been shown whether or not the diketo com- pounds are reversible. Enzyme incubated with what we have proposed to be inhibitors arising from RuBP have not been passed over columns. HARTMAN: I am not aware of any evidence suggesting that ionized sulfhydryl is the essential base obstructing the C3 proton. TOLBERT: We do not know what base is there. The only thing we do know is the sulfhydryl protection. It has been proposed that there are two sulfhydryl groups. Vol. 83, No. 3, 1978 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS August 14, 1978 Pages 1084—1092 INHIBITION OF RIBULOSE-1,S-BISPHOSPHATE CARBOXYLASE/OXYCENASE BY RIBULOSE-1,S-BISPHOSPHATE EPIMERIZATION AND DEGRADATION PRODUCTS Christian Paech, John Pierce, Stephen D. McCurry, and N. E. Talbert Department of Biochemistry, Michigan State University, East Lansing, MI 48824 Received June 26,1978 Summary: Xylulose-l,S-bisphosphate: in_ preparations. of ribulose-1,5-bisphos- phate (ribulose-P ) arises from non-enzymic epimerization and inhibits the enzyme. Another Anhibitor, a diketo degradation product from ribulose-P ,is also present. Both compounds simulate the substrate inhibition of ribulose- P carboxylase/oxygenase previously reported for ribulose-P . Freshly prepared ribulose-P had little inhibitory activity. The instability of ribulose-P may be one reason for a high level of ribulose-P carboxylase in chloroplasts where the molarity of active sites exceeds that of ribulose-P . Because the K of the enzyme/substrate complex is :1 11M, all ribulose-P generated. in situ may' be stored as this complex to prevent decompositIon. The hysteretic substrate inhibition (H? ribulose-P2 carboxylase/oxygenase (EC 4.1.1.39) by ribulose-P2 (1-4) has been attributed to an allosteric binding site (3) or competitive binding for ribulose-P2 and C02 (5), either of which could produce a conformational change (6,7). However purified carboxylase cannot be fully reactivated upon addition of Mg2+ and bicarbonate once it has been preincubated with ribulose-P2 (3). Partial inactivation does not occur for enzyme activity in whole chloroplasts where ribulose-P2 may accumulate in the absence of CO2 to a 16 fold molar excess over the carboxylase binding sites, but the carboxylase activity is fully restored within 10 min after addition of bicarbonate (8). Much of the substrate inhibitory effect and the difference between the isolated enzyme and the chloroplasts can be explained by 2 impurities which we find in ribulose-P2 preparations used for in vitro assays. One is xylulose-P2, an inhibitor of the enzyme (9), and the other apparently is :1 diketo compound, most likely l-deoxy—D-glycero-Z,3-pentodiulose—S-phosphate. Other o-dicarbonyl compounds are also inhibitors of this enzyme (10,11). The carboxylase inhibition by' newly prepared ribulose-P2 was small, as if there had been insufficient time for inhibitor formation. Abbreviations: ribulose-P2,ribulose—l,S—bisphosphate; deoxypentodiulose-S- P, l-deoxy-D—gZyoero—Z,3-pentodiulose—S-phosphate; xylulose-P2, xylulose-l, S-bisphosphate. 0006-29lX/78/0833-1084$Ol.OO/O Copyright © 1978 by Academic Press, Inc. All rights of reproduction in any form reserved. 1084 Vol. 83, No. 3, 1978 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS MATERIALS AND METHODS Ribulose—P’ Carboxylase/oxygenase from spinach leaves was. prepared as described elsewhere (12). Ribulose-P2 was prepared enzymatically from ribose-S-P (13). Xylulose-P was synthesized with rabbit muscle aldolase from glycolaldehyde-P and dihydroxyacetone-P (14). Proof of structure and purification have been described (9). (1-13C) xylulose-P was synthesized from (1-13C) fructose-1,6-P , which was a gift from A. S. Serianni of Mich. State Univ., an glycolaldehyde-P by aldolase. NaH14CO and NaB3H were from Amersham-Searle; other chemicals were from Sigma Chemical Co. and were used without further purification. Ribulose—P carboxylase activity was determined using a radiometric assay (15). The enzyme was incubated at 2mg/ml for at least 1 h in the assay buffer composed of 0.1 M N,N-bis(2-hydroxyethyl)glycine (Sigma, "Bicine") at pH 8.0 and 30° , 10 mM MgCl , 0.2 mM Na -EDTA, and then made to 10 mM NaHCO and kept for 30 min at 30 . Aliquots from this stock solution were equilibrated in the assay .buffer at 20 mM NaH14CO (0.16 Ci/mol) and a final protein concentration of 80 ug/ml (0.14uM). The reaction was initiated by addition of ribulose-P to a final concentration of 0.5 mM and was run for 20 to 60 sec, stopped by acid and the fixed 14C counted as“ described elsewhere (5). Under these conditions the specific activity of the enzyme preparations was 1.0 to 1.4 nmol CO2 fixed per min per mg protein. Protein determinations were performed according to a modified Lowry procedure (16). Carbon-13 n.m.r. spectra were obtained with a Bruker WP-60, 15.08 MHz Fourier-transform spectrometer equipped with quadrature detection. The spectrometer was locked to the resonance of D O in a capillary. Chemical shifts are given relative to external tetramethyfsilane (0 ppm) and are accurate to within 0.1 ppm. RESULTS AND DISSCUSSION Evidence for Inhibitors in Ribulose—PZ Preparations- Early reports (1,2) of ribulose-P2 inhibition of the carboxylase indicated that saturating (n: higher concentrations had to be used. In our experience the extent of inhibition was not reproducible. Different batches of ribulose-1"2 solutions at apparently identical concentrations, as judged by phosphate (17), carbohydrate (18), and enzymatic analyses with an excess of ribulose-P2 carboxylase and NaH14C03 of a known specific activity gave different initial rates of carboxylation (data not shown). These differences became insignificant when ribulose-P2 concentrations below 5 times the Km were used in the assay. At these low concentrations, short reaction times of 10-40 sec are required, and the inhibition, which is time-dependent, is not significant. Nevertheless the inhibitory effect from low concentrations of ribulose-P2 can be demonstrated (Fig. 1). After complete utilization of a known amount of substrate, severe inhibition occurred during a second reaction period initiated by adding ribulose-P2 to the same concentration as in the first stage. Several sources of inhibition could be ruled out. Product inhibition under these conditions was negligible; 1 mM 3-P-glycerate had no effect and 0.5 mM P-gycolate caused a 42 inhibition. of the inital rate. Also mono— and divalent cations (e.g. 1085 Vol. 83, No. 3, I978 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS 8* M I 08%) .E 6 I- -4 o '8 5. B E (35%) ~\ 4.. 1 - '0 h ‘ - C g: 3 . v ‘: (54%i N O i.) 2 " ‘- c c c ‘ I '6 E J a. I E 3 8 9 I0 II I2 tune (nun) Fi ure 1. Progress curve of ribulose-P2 carboxylase reactions with (A) 0.5, (B? 0.25, (C) 0.125 mm ribulose-P2. 0.3 m1 of a ribulose-P solution was added to 7.2 m1 of the assay buffer containing the enzyme at 80 ug/ml and NaH18003 at 20 mM final concentration. Ten aliquots of 0.25 ml were withdrawn at indicated times and added to 0.2 m1 2 N HCl. The arrows indicate initiation of the second progress curve by addition of a similar amount of ribulose-P2(O.2 m1 into 4.8 ml) as in the first reaction. A corresponding volume was removed prior to addition of ribulose-P2 in the second run. Thus, the zero point is 42 lower and a maximum of 962 of the first initial rate in the second progress curve would represent no inhibition. The values in parentheses represent the observed percent inhibition of the first initial velocity. Ba2+, Ca2+, Sr2+) as well as NaVO3 (19) were not inhibitory up to 10 mM. Stability of Ribulose—Pz- Treatment of ribulose--P2 solutions at 30° and pH 11 increased the inhibition to a maximum of about 70%, presumably from reaching an inhibitor/substrate equilibrium. Prolonged base treatment or more elevated temperature caused more rapid substrate loss and appearance of orthophosphate, but no further inhibition (Fig. 2). Loss of ribulose-P2 during alkaline treatment correlated with the formation of inorganic phosphate (20) (Fig. 3). When ribulose-P2 was no longer detectable-enzymatically, approximately 502 of its total phosphate had been liberated. No ribulose-P2 was regenerated by incubating samples, taken during the time course of alkaline treatment, with a mixture of phosphoriboisomerase, phosphoribulokinase, and ATP; therefore neither ribulose-S-P nor ribose-S-P were formed during the alkaline treatment. 1086 Vol. 83, No. 3, I978 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS I rw‘r‘ t I I l I r 5~ J 0min I min 5min 4 - < c IO - 3 20mm 2 a. 40min 05 E 3 ’ -1 \ _‘O a) .5 ”'N I60 min C) r o 2 ' I '3 E :L I '- -1 L l #1 1 l 1 l l l I 2 6 7 8 9 IO II time (min) Figure 2. Time course of the ribulose-P2 carboxylase reaction with 0.2 mM ribulose-P2. Ribulose-Pz, 5 mM, was incubated at pH 11 and 30° for 1, 5, 10, and 20 min as indicated. The pH was then adjusted to 8.0 and an aliquot was used to run two consecutive progress curves (0—0) as described in the legend to Fig. 1. Another sample was incubated for 40 and 160 min at pH 11 and 40° and used for the same experiment (A—A). When ribulose-P2 was generated from ribose-S-P enzymatically, and used directly the initial rate in the second progress curve reached 882 of the initial rate, which because of the dilution was over 90% of the maximum expected (I---I). Even at pH 8.3 and 30° (the enzyme assay conditions) loss of ribulose-P2 amounted to 1.251 per hour (data not shown). ' Ribulose-P2 coupled assay in which ribulose-P2 was generated from ribose-S-P or Solutions essentially free of inhibitor were obtained in a ribulose-S-P with phosphoriboisomerase, phosphoribulokinase, ATP and used immediatelyi I----I in Fig. 2). In this case there was little effect from inhibitor accumulation during the second cycle of the repeat assay. . Mechanism of Degradation and Identification of Inhibitor-s - Ribulose-P2 is usually prepared enzymatically from ribose-S-P and stored as the barium salt which is unstable even at 4°(13). The usual way to‘ prepare the potassium or sodium salt is by mixing the barium salt with Dowex 50/H+, filtering, and neutralizing with 6 N K011. During neutralization, epimerization at C-3 induced by hydroxyl ions would form xylulose-P2, which is a potent inhibitor 1087 Vol. 83, No. 3, I978 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS of ribulose-P2 carboxylase (9). Alkaline treatment may also lead to phosphate elimination at C1 from the enediol intermediate formed from either ribulose-P2 or xylulose-P2 , to yield deoxypentodiulose-S-P . This unstable compound (I) should undergo a rearrangement to form (II) and (III). cnzopoz ' CHZOPOE I cnzopog $80 fi—O‘00. H $.0 H—C-OH —-.___°H‘ (It-o / ___—0H {IO-('I-H I H-C-OH H-C-OH H-c-ou cuzoroi L cnzopog cnzopog ribulo se-P2 xylulo se-P 2 P03- 8‘ 913 I c-on c=o $00 coo c-o ===e $80 ;* OH , H3C-fi-0H + Ho-c-cua H-o-on H-c-ou H-fi-OH H—C-OH cuzopoi cuzopog CHzoPoa CH20P03 (I) (II) (III) ( I ) l-deoxy-D- glycerO-Z , 3-pentodiulose-S-phosphate (II) 2-C-methy1-D-37'yth7'0 -tetronic acid 4-phosphate (III) 2-C-methy1—D-thr'eo -tetronic ac id 4-phosphate Evidence for a phosphate elimination mechanism was the appearance of ortho phOSphate (Fig. 3) and a time course of degradation of (1-13C) xylulose-P2 at pH 11 and 40° (Fig. 4). The starting compound was 91% 13C enriched at C-1. As the reaction proceeded, the area under the doublet, centered about 69.4 ppm, decreased as methyl resonances at 24.0 and 23.4 ppm increased, reaching their maximum values when no further xylulose-P2 remained. The methyl resonances are assumed to arise from the two major degradation products of the proposed diketo compound via the described mechanism. Similar mechanisms for non-phosphorylated sugars have been proposed (21). . To identify the inhibitors, samples’ of ribulose--P2 ware reduced with NaBH4, pH 7, and treated with acid phosphatase. The products were converted to.the pentaacetyl pentitols and subjected to gas liquid chromatography (22). The major peaks were identified as ribitol and arabinitol, present in approximately the same quantities. These peaks were followed by a small xylitol peak. After treatment with 11383114 and phosphate ester cleavage, the tritiated reduction products were separated by thin layer chromatography on 1088 Vol. 83, No. 3, I978 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS I I 1 1'; f 1‘; I a 6~ ‘6 I A A . 1 2 I 2 E E 4» *4 o E N O. .c o ‘ 3 3% .2 é; :1 _Q 2" -2 2 .: ~ 5 O I o O O l l l '1': I 1; A I 2 3 5 l8 time (hours) Figure 3. Loss of ribulose-P2 at pH 11 and 40°, and formation of inorganic p osphate as a function of time. The ribulose-P2 content was measured enzy- matically with a pure preparation of ribulose-P carboxylase and NaH1“C03 of a known specific activity. borate-impregnated silica gel plates. The two major labeled spots corresponded to arabinitol and ribitol, and a minor spot corresponded to xylitol. These compounds were recrystallized with the corresponding unlabeled alcohols to constant specific activity. The presence of approximately 12 xylulose-P2 in Preparations 0f ribulose-92 Could account for the inhibitory phenomenon of ribulose-P2 on, carboxylase activity. That xylulose-P2 solutions may also epimerize to ribulose-P2 was demonstrated by assaying them with the carboxylase. 3-P-G1ycerate was formed, but the reaction reached completion with most of the xylulose-P2 remaining. Thus it is unlikely that xylulose-P2 Was acting as a substrate for ribulose-P2 carboxylase. The B-elimination. product from ribulose-PZ’ has not been characterized unequivocally but is proposed to be l-deoxy—D- glycero- 2,3-pentodiulose-5- phosphate for the following reasons: (a) B-elimination is kinetically favored over hydrolysis under alkaline conditions. The deoxypentodiulose-S-P would not accumulate and its further degradation to the end product shown in the scheme would explain why the inhibitory effect from alkaline treatment did not further increase. (b) Incubation of ribulose-P2 preparations with o-penylenediamine to complex the dicarbonyl compound reduced inhibition of the enzyme by the ribulose-P2 preparation. (c) Dicarbonyl derivatives of carbohydrates are sensitive to oxygen under alkaline conditions and undergo rapid degradation and rearrangement (23). (d) Deoxypentodiulose-S-P belongs to a class of reagents specific for' arginyl residues in Tenzymes (24), and it has been shown that 1089 Vol. 83, No. 3, 1978 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS MWHW W MI l I 7? 7o 63 6'0 2'5 2'0 7'5 70 63 (3‘0 25 20 Figure 4. Time course of degradation of (1-13C) xylulose-P2 at pH 11 and 40°. (A) 4-16 min, (B) 47- 58 min, (C) 125-160 min, (D) 245-295 min. The doublet centered about 69.4 ppm is due to couplingP of the 1P nucleus with the 13C nucleus at C-1 of (l- 3C) xylulose-P2 (JC— = 4. 4 Hz). ribulose-P2 carboxylase is inhibited by such compounds, e.g. phenylglyoxal and 2,3-butanedione (10,11). (e) Dicarbonyl compounds have characteristic absorption spectra which are pH dependent and include peaks in the region 400-460 nm (25). Such spectral properties have been observed with ribulose- P2 solutions'(data not shown). Physiological Implications - The rapid formation of inhibitors in solutions of ribulose-P2 is important with respect to kinetic studies and the carboxylase assayu Recognition of these inhibitors may also be the key to some yet unsolved questions concerning the enzyme in vivo. The average concentration of ribulose-P2 in chloroplasts is less than the total concentration of tunding sites of ribulose-P2 carboxylase (26). Therefore, ribulose-P2 probably never occurs in free solution in the chloroplasts (R.C. Jensen, personal 1090 Vol. 83, No. 3, I978 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS communication). Conditions fbr‘ chemical epimerization or B-elimination of ribulose-P2 could occur in the chloroplasts, for example during the day when temperatures may reach 40° and the pH over 8. The presence of a large excess of protein (ribulose-P2 carboxylase) would account for the binding of almost all the ribulose-P2 as soon as it is formed by phosphoribulokinase, as the dissociation constant for the enzyme/ribulose-P2 complex is :1 uM (5,27). If xylulose-P2 and deoxypentodiulose-S-P were formed in situ there would be sufficient ribulose-P carboxylase to bind these inhibitors without any 2 loss in rate of the whole photosynthetic carbon cycle.' Bahr and Jensen (28) have estimated that in the chloroplast only 40 to 602 of the carboxylase is functioning at any given time; the rest may be a storage reserve or trap for ribulose-P or these inhibitory products, respectively. The deoxypento— diulose-S-P seems too unstable, to accumulate in chloroplasts, so that the removal of xylulose-P2 would be of more serious concern. A phosphatase catalyzing the conversion of xylulose-P2 to xylulose-S-P has not been reported, but such an enzyme would effectively recycle the inhibitor through the photosynthetic carbon cylce. Acknowledgement: This work was supported by NSF Grant PCM 78—15891 and is pub- lished as article number 8573 of the Michigan Agricultural Experiment Station. REFERENCES I-‘ Weissbach, A., Horecker, B.L., and Hurwitz, J. (1956) J. Biol. Chem. 118, 795-810. Paulsen, J.M., and Lane, M.D. (1966) Biochemistry 5, 2350—2357. Chu, D.K., and Bassham, J.A. (1975) Plant Physiol. 25, 720-726. Pon, N.G., Rabin, B.R.,and Calvin, M. (1963) Biochem. Z. 128, 7-19. Paech, C., Ryan, F.J., and Tolbert, N.E. (1977) Arch. Biochem. Biophys. 112, 279-288. Kwok, S.Y., and Wildman, S.G. (1974) Arch. Biochem. Biophys. 161, 354-359. 7. Vater, J., Salnikow, J., and Kleinkauf, H. (1977) Biochem. Biophys. Res. Commun. 14, 1618-1625. 8. Jensen, R.C., Bahr, J.T., and Sicher, R.C. (1977) Proc. 4th Int. Congr. Photosynthesis Abstracts, Coombs, J., ed., UKISES, 21 Albemarle St., London, p. 179. 9. McCurry, S.D., and Tolbert, N.E. (1977) J. Biol. Chem., 252, 8344-8346. 10. Paech, C., McCurry, S.D., and Tolbert, N.E. (1977) Proc. 4th Int. Congr. Photosynthesis Abstracts, Coombs, J., ed., UKISES, 21 Albemarle St., London, p. 289. 11. Lawlis, V.B., and McFadden, B.A. (1978) Biochem. Biophys. Res. Commun. 89, 580—585. ‘ 12. Ryan, F.J., and Tolbert, N.E. (1975) J. Biol. Chem., 159, 4229-4233. 13. Horecker, B.L., Hurwitz, J., and Weissbach, A., (1958) Biochem. Prep. 6, 83-90. , l4. Byrne, W.L.,and Lardy, H.At (1954) Biochim. Biophys. Acta 14, 495-501. 15. Lorimer, G.H., Badger, M.R., and Andrews, T.J. (1976) Biochenistry 1;, £11wa 00 so 0‘ o 1091 Vol. 83, No. 3, I978 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS l6. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. Bensadoun, A., And Whinstein, D. (1976) Anal. Biochem. 19, 241-250. Leloir, L.F., and Cardini, 0.8. (1957) Methods Enzymol. 1, 840-850. Horecker, B.L. (1957) Methods Enzymol. 1, 105-107. Cantley, L.C., Jr., Josephson, L., warner, R., Yanagisawa, M., Lechene, C., and Guidotti, G., (1977) J. Biol. Chan. 25;, 7421-7423. Horecker, B.L., Hurwitz, J., and Weissbach, A. (1956) J. Biol. Chem. 218, 785-794. Nef, J.U. (1910) Ann. 116, 1-119. Sloneker, J.H. (1972) Methods Carbohydr. Chem. 6, 20-24. Isbell, R.S. (1973) Advances in Chemistry Series 111, pp. 70—87, Am. Chem. Soc., pp. 70-87. Riordan, J.F., McElvany, K.D., and Borders, C.L., Jr. (1977) Science 125, 884-886. Scott, A.I. (1964) Interpretation of the Ultraviolet Spectra of Natural Products. Pergamon Press, Oxford, pp. 36-37. Jensen, R.C., and Bahr, J.T., (1977) Annu. Rev. Plant Physiol. 18, 329-400. Wishnick, M., Lane, M.D., and Scrutton, M.C. (1970) J. Biol. Chem. 345, 4939-4947. Bahr, J.T., and Jensen, R.C. (1978) Arch. Biochem. Biophys. 185, 39-48. 1092 Communication Inhibition of Ribulose-l,5- bisphosphate Carboxylase/Oxygenase by Xylulose l,5-Bisphosphate* (Received for publication, July 25, 1977, and in revised form, October 18, 1977) Susan: D. McCunnv AND N. E. Torssn'r From the Department ofBiochemistry, Michigan State University, East Lansing, Michigan 48824 SUMMARY Xylulose 1.5-bisphosphate is a potent inhibitor of ribulose- 1,5-bisphosphate carboxylase/oxygenase. The enzyme was inhibited 50% at a xylulose 1.5-bisphosphate concentration of 0.56 pm and an enzyme concentration of 0.14 pm. When both ribulose 1.5-bisphosphate and xylulose 1,5-bisphos- phate were added simultaneously to the enzyme, this inhi- bition appeared to be competitive. A preincubation of 20 to 30 min was needed for maximum inhibition. The inhibitory effect of this compound is probably exerted through its binding at the ribulose 1.5-bisphosphate active site, since both the carboxylase and oxygenase activities were inhibited similarly. Ribulose-l,5-bisphosphate carboxylase/oxygenase (EC 4.1.1.39) catalyzes the addition of CO, to ribulose-P2l to give 2 molecules of 3-phosphoglycerate, as well as the addition of 0., to ribulose-P2 to give 1 molecule of 3-phosphoglycerate and l of phosphoglycolate. These two reactions compete to divide carbon flow between photosynthesis and photorespira- tion, and therefore become a logical point for control. Studies on the mechanism of action of ribulose-P2 carboxylase/oxygen- ase contribute to the determination of whether differential regulation of the two activities is possible. Xylulose-P, was tested because of its similarity to ribulose-P2, for the two differ only in the stereochemistry at carbon 3. Both the carboxylase and oxygenase functions are known to be in- hibited by ribulose-P2 (1, 2) and xylitol-P2 (3). Likewise, we have now found that xylulose-P2 is a competitive inhibitor with respect to ribulose-P2, and thus both the carboxylase and oxygenase activities were inhibited. MATERIALS AND METHODS Ribulose-P2 carboxylase/oxygenase was prepared from spinach plants grown in the greenhouse on a 10-h day as described previously (3) and stored at 4° in 50% saturated (NI-10280,. The purified and desalted enzyme was diluted to a 2 mg/ml stock solution for assay. The enzymatic preparation of ribulose-P, has been described previously (4). Xylulose-P, was prepared by an aldolase-catalyzed ‘ This work was supported in part by National Science Founda- tion Grant PCM 77-09089 and is published as journal article number 8189 of the Michigan Agricultural Experiment Station. The research is part of a Ph.D. thesis by S. D. M. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement” in accord- ance with 18 U.S.C. Section 1734 solely to indicate this fact. ’ The abbreviations used are: ribulose-P2, ribulose 1,5-bisphos- phate; xylulose-P2, xylulose 1,5-biphosphate; xylitol-E, xylitol 1,5. biphosphate; arabitol-P2, arabitol 1,5-bisphosphate. Tm Jounmu. or BIOLOGICAL Carmel-av Vol. 252, No. 23, Issue of December 10, pp. 8344-8346, 1977 Printed in USA. condensation of glycolaldehyde phosphate (Calbiochem) and dihy- droxyacetone phosphate (Calbiochem) (5). The proof of structure of xylulose-P, was established by a NaBH, reduction of the aldolase condensation product to form a mixture of epimeric pentitol 1,5- bisphosphates, presumably xylitol-P, and arabitol-P2. After acid phosphatase (Sigma) treatment, the free pentitols were chromato- graphed against known standards on silica gel plates impregnated with 0.1 M boric acid. The spots were visualized by alkaline perman- ganate (6) and coincided with xylitol (R, 0.08) and arabitol (R, 0.15) standards (gifts from R. Barker). The concentration of xylulose-P, was determined by total phosphate analysis (7). Carboxylase assays were run in stoppered scintillation vials at 30° with 0.25 ml final volume composed of 100 mM N,N-bis(2-hydroxy- ethyl)glycine (Sigma, Bicine) at pH 8.0, 10 mm MgCl,, 10 mm NaH“C03 (252 cpm/nmol), 0.25 mm EDTA. 1 mM dithiothreitol, 20 pg of enzyme (0.14 am), and 0.5 mm ribulose-P, unless otherwise indicated. The enzyme was activated in a stock solution composed of 2 mg of protein/ml of 50 mm N ,N -bis(2-hydroxyethyl)glycine at pH 8.0, 0.40 mM EDTA, 10 mM MgCl,, 1 mm dithiothreitol, and 10 mM NaHCO, for at least 10 min prior to its use. The reaction was initiated by ribulose-P, and after 20 s to 1 min was stopped by the addition of 0.2 ml of 2 N HCI. Activity was expressed as total counts per min or nanomoles of ”CO, fixed per min into acid-stable reaction product (carbon 1 of 3-phosphoglycerate). The oxygenase assays were run in a 0.5-ml volume in the reaction chamber of a Rank oxygen electrode at 30° with a continuous recording of 0, uptake. The assay solution was 100 mm N ,N -bis(2- hydroxyethyl)glycine at pH 8.3, 20 mM MgCl,, 0.5 mM ribulose-1),, 0.25 mM EDTA. 1 mm NaHCO,, 1 mm dithiothreitol, enzyme, and atmospheric oxygen. The enzyme was preincubated as above with 10 mM NaHCO;. for at least 10 min, and then assayed at a concentra- tion of 400 ug/ml by initiation with the ribulose-P1. RESULTS AND DISCUSSION A Dixon plot for xylulose-P, inhibition of ribulose-P2 carbox- ylase, when the enzyme was preincubated with the inhibitor for 20 min, is presented in Fig. 1. When the molar ratio of inhibitor to enzyme was equivalent to 4 inhibitor molecules per 8 active sites, the carboxylase was about 50% inhibited. The Dixon plot is biphasic, in that it consists of two different slopes. This implies two types of inhibition. For given amounts of xylulose-P2, the inhibition of the carboxylase reached a maximum after 20 to 30 min of preincubation (Fig. 2A). We do not know if this hysteretic response to the inhibitor is an opposite effect from the hysteretic CO2 activation. A plot of log residual activity versus time (Fig. 28) was linear and indicated that the inactivation was an apparent first order process. This line did not pass through the point for 100% activity at zero time, as if there were a very fast initial inhibition phase or an initially inadequate mixing of the inhibitor and the enzyme. When xylulose-P2 and ribulose-P, were added simultane- ously, the inhibition was apparently competitive with respect to ribulose—P2 as shown by double reciprocal plots (Fig. 3). When the data are replotted as the slope versus [xylulose-P2] (Fig. 3, inset), a straight line is obtained, indicative of pure competitive inhibition. Similarly, a plot of reaction velocity versus [xylulose-P2] at fixed [ribulose-P2] also indicates pure competitive inhibition (figure not shown). Experiments run after preincubation of enzyme for 20 min with similar concen- trations of xylulose-P2, before addition of ribulose-P2, produced 90% inhibition. Since this long preincubation was required for equilibrium of enzyme and xylulose-P2, one cannot expect 8344 Inhibition of Ribulose-P, Carboxylase/Oxygenase I I F coz mod/ mln) l/ ( nmol 1 1 J. 1 l20 ISO [Xylulosr P2] 994 FIG. 1. Xylulose-P, inhibition of ribulose-P2 carboxylase after a 20-min preincubation with xylulose-P2. 6h- 8 8 minutes preincubation FIG. 2. A, time course of inhibition. Enzyme was preincubated in 8 an xylulose-P, for indicated times and then assayed for 1 min. V is measured as counts per min of “C fixed per assay. B, first order plot of the inhibition, log ((V, — V.,,.,,)/(V, — V.,...» X 100 versus time of preincubation. V, = velocity at any time; V.,,“ = the minimum velocity at this xylulose-P, concentration; V, = the initial velocity. ribulose-P2 and xylulose-P2 to come to equilibrium with the enzyme during a 1-min assay. Because of the hysteretic response of . the carboxylase to the inhibitor, it is expected that the kinetics would appear noncompetitive after a 20-min preincubation, and this has been observed with xylitol-P2 (data not shown and Fig. 3 of Ref. 3). Since CO, is needed for activation of the enzyme, it is not possible to preincubate the enzyme with ribulose-P2 and xylulosech and then use CO, to initiate the reaction. Xylulose-P, inhibited the ribulose-P2 oxygenase activity with approximately the same stoichiometry as it did the carboxylase. The inhibition was also dependent upon the duration of the preincubation. Inhibition of both the carboxyl- ase and oxygenase activity is to be expected if xylulose-P, and ribulose-P2 are competing at the same site on the protein. Xylulose-P, has been considered as a substrate for ribulose- P, carboxylase (9), but being inactive, it was apparently not further investigated. Its structure makes it almost indistin- guishable from ribulose-P2 by the paper chromatographic techniques used originally to identify the components of the 8345 8):“ o " ISuM 4.6uM . I I O 12pM s Q i ‘ 9 . - t 2 It 5 no 20 so > 0 x I I - u E 4’ I ' u o" '2] H v t ’ no Xylulou-Pz 3 f’ o //" I -30 ~81 -IO 0 K) 20 30 I/ [RIbqusc-le mM FIG. 3. Double reciprocal plot of activity versus ribulose-P, con- centration when xylulose-P, and ribulose-P, were added simultane- ously. The lines were calculated by a computer program (8). V is measured as counts per min of “C fixed per assay. photosynthetic carbon cycle. While xylulose-P2 can be easily synthesized by aldolase, we know of no reports concerning the natural occurrence of this compound. This apparent ab- sence of xylulose-P, is consistent with the lack of confirmed reports of the occurrence of glycolaldehyde phosphate in na- ture. Xylulose-P, and xylitol-P, are potent inhibitors of ribulose- P, carboxylase/oxygenase, because of their structural similar- ities to the substrate. cnzopof 01,090,: 0120903: I20 (:0 HCOH HIGH HOI2H min HCOH HCOH HIGH CH20P03: IZI‘I20I’03: 0420903: Ribulose 1,5- Xylulose 1,5- Xylitol 1.5- bisphosphate bisphosphate bisphosphate Xylulose-P, has a potential for three types of interactions at the ribulose-P, active site. (a) The phosphate groups can be bound in a manner similar to the binding of ribulose-P2. (b) The hydroxyl at carbon 3 may interact with a base in the active site which is essential for enediol formation of ribulose- P, during catalysis. This binding is probably the unique characteristic of both xylulose-P2 and xylitol-P, inhibition. (c) The carbonyl group at carbon 2 of xylulose-P, may interact with or form a Schiff base with an essential 6 amino group of a lysyl residue which is involved in CO, binding (10). This possibility is supported by the fact that xyluloseoP, partially blocks the reaction of pyridoxal 5’-phosphate with this lysyl residue (11). Acknowledgments - W. A. Wood advised us on the synthesis and chromatographic identification of xylulose-P2. We thank Christian Paech and Frederick J. Ryan for discussions of the experiments. 8346 REFERENCES . Weissbach, A., Horecker, B. L., and Hurwitz, J. (1956) J. Biol. Chem. 218, 795-810 . Paulsen, J. M., and Lane, M. D. (1966) Biochemistry 5, 2350— 2357 . Ryan, F. J., Barker, R., and Tolbert, N. E. (1975) Biochem. Biophys. Res. Commun. 65, 39-46 . Ryan, F. J., and Talbert, N. E. (1975) J. Biol. Chem. 250, 4229— 4233 . Byrne, W. L., and Lardy, H. A. (1954) Biochim. Biophys. Acta 14, 495-501 6. Inhibition of Ribulose-Pz Carboxylase/Oxygenase Lemieux, R. U., and Bauer, H. F. (1954) Anal. Chem. 26, 920— 921 . Leloir, L. F., and Cardini, C. E. (1957) Methods Enzymol. 3, 840~850 . Wilkinson, G. N. (1961)Biochem. J. 80, 324-332 . Racker, E. (1962) Methods Enzymol. 5, 266—270 . Paech, C., Ryan, F. J., and Tolbert, N. E. (1977)Arch. Biochem. Biophys. 179, 279—288 . McCurry, S. D., Paech, C., and Tolbert, N. E. (1977) Plant Physiol. 59 (suppl.), 90 Vol. 84, No. 4, I978 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS October 30, I978 Pages 895-900 RIBULOSE-1,S-BISPHOSPHATE CARBOXYLASE/OXYGENASE FROM PARSLEY Stephen D. McCurry, Nigel P. Hall, John Pierce, christian Paech, and N. E. Talbert I Department of Biochemistry, Michigan State University, East Lansing, Michigan 48824 Received August 28, 1978 SUMMARY I Ribulose-l,S-bisphosphate carboxylase/oxygenase from parsley leaves was purified by Sepharose 68 gel filtration at pH 8.3 as a single, colorless peak containing both activities. Approximately 0.2 9 atom copper per mole enzwne was detected by atomic absorption spectroscopy, but this copper was not detectable by EPR spectrometry. INTRODUCTION Since the discovery of the oxygenase activity of ribulose-P2 carboxylase/ oxygenase (E.C. 4.1.1.39), much evidence has been accunulated which indi- cates that both the carboxylase and oxygenase activities are due to one enzyme. Therefore, the recent report (1) that the two activities could be separated by gel filtration on Sepharose 6B at pH 8.3, and that the oxygen- ase fraction was blue and contained copper required fbrther investigation. Small and varying anounts of copper have been reported to be asociated with the A838 Enzyme from spinach leaves: 1 g atom/mole (2), 0.11 to 0.14 g atom/mole (3), 0.2 g atom/mole (4), and in this report 0.18 9 atom in the enzyme-from parsley. An EPR spectrun of Cu(II) bound to the spinach enzyme has been published (2). This paper presents unsuccessful attempts to i separate the oxygenase from the carboxylase activities by the procedure of Branden (l). The Specific activity of ribulose-P2 carboxylase/oxygenase prepared from parsley was similar to that obtained from spinach or tobacco leaves. MATERIALS AND METHODS Ribulose-P2 Carboxylase/oxygenase was prepared from the leaves of par- sley (from local grocery store) according to Br'a‘nd'en (1). One hundred g of Abbreviations:. Ribulose-Pz for ribulose-1,5-bisphosphate. 0006-291X/78/0844-0895$Ol . 00/0 Copyright © 1978 by Academic Press, Inc. 895 All rights of reproduction in any form reserved. Vol. 84, No. 4, I978 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS tissue was ground in a Haring blender at high speed for 60 s with 200 ml of 50 mM Tris-HCl at pH 8.3 and 4°. This buffer was prepared in 2 ways for different experiments; in one the pH was adjusted at 4°, the other at room temperature (22-24°) according to Branden (personal communication). The homogenate was filtered through several layers of cheese cloth and centri- fuged at 11,000xg for 30 min. The supernatant was filtered through Mira- cloth and solid (NH4) $04 was added slowly, with constant stirring, to 30% saturation (16.4 g/ 00 ml). The suspension was stirred for 15 min, and stood for 45 min at 4° before centrifuging at 11,000xg for 40 min. This precipitate was discarded, and the supernatant was made 50% saturated with solid (NH4)ZSD4 (11.8 g/l00 ml) to precipitate the enzyme. The stir- ring, standing, and centrifugation were repeated. The 30-50% (NH4)ZSO4 precipitate was resuspended in 25 ml or less of buffer. In one case the buffer was 50 mM Tris, pH 8.3 (adjusted at 4°), and in the second case the buffer was 5 mM Tris, adjusted at room temperature to pH 8.3. Usually the resuspended and redissolved precipitate was clarified by centrifugation at 25,000xg for 15 min. The supernatant was filtered through a Sephadex G-25 (mediun), 6 x 25 cm col mm, which had been previously equilibrated with the resuspension buffer. The peak fractions, as determined by absorbance at 280 nm, were pooled and concentrated to 5-7 ml with a PM 30 membrane in an Amicon ultrafiltration chamber, pressurized with argon at 60 psi. The concentrate was chromatographed on a Sepharose 6B col mm (2.6 x 90 cm) equilibrated with the same resuspension buffer. All steps were carried out at 4°. The entire procedure was repeated with 2 mM B-mercaptoethanol in all buffers. The spinach enzyme was prepared by our usual procedure (5). Ribulose-Pz was prepared enzymatically (6). Sephadex G-25 and Sepharose 68 were from Pharmacia Fine Chemicals, NaH 4(:03 from Amersham and other reagents from Sigma. The enzyme was activated by adjusting preparations to 20 mM MgCl2 and 10 mM NaHC03 and incubating them at 30° for at least 10 min prior to assay- ing. The carboxylase assays were run at 30° for 1 min and initiated by the addition of 10 pl aliquots of activated enzyme to 240 pl of 100 mM N,N-bis— (2-hydroxyethyl)glycine (Sigma, Bicine) at pH 8.2, 20 mM MgCl2, 10 mM NaH1 C03, 1 mM dithiothreitol, and 0.5 mM ribulose-P2. The 14C ac1d stable product was measured by scintillation counting. The oxygenase assay was based on the ribulose-P2 dependent oxygen uptake in a Rank Brothers oxygen electrode. The reactions were normally initiated by the addition of 20 pl of activated enzyme to 480 pl of 100 mM N,N-bis(2- hydroxyethyl)glycine at pH 8.2, 20 mM MgCl2, 1 mM dithiothreitol, and 0.5 mM ribulose-P2. The reactions were monitored for 50-605 at 30°. Assays were also run according to Br‘a‘ndén (1) by initiating with ribulose-P . after a 5 to 10 min incubation of the enzyme in the assay buffer in t e oxygen electrode chamber. A Varian Node] 175 Spectrophotometer was used for the atomic absorption spectroscopy using copper sulfate as a standard. EPR spectra were recorded at 77K with a Varian E-4 spectrometer Operating at 9.22 GHz; microwave power, 20 mwatts; modulation fre uency, 100 KHz; modulation amplitude, 5 gauss; field set, 2900 gauss; fie d sweep, 2000 gauss; time constant, 0.1 sec; instrunental gain, 2 x 103. RESULTS AND DISCUSSION Ribulose-PZ carboxylase and ribulose-P2 Oxygenase activity from parsley leaves chromatographed as a single peak on Sepharose 68 at 4° in 50 mM Tris 896 Vol. 84, No. 4, 1978 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS I I2 [I r. Isolation with x Isolation with mercoptoethanol {I no mercaptoethmol to» ‘ I. 5 f ‘ c E I . o C C : e e g I e 8 ~0- . I— N 8' I I- ?! E B '4 : ~ I o ,’ I - ,E i . i‘ ,5 E 8 6» . . .' E .96 E C I Q I \ \ O I it I CS‘ .0 5 I I". I (5' o I B 4 : I'. I 8 .442 3 < l I t : B (D 9 ' I I a .Q.’ I I I I E O I 9 4 3 q E ‘ 2,. I: I : .del a l b" I I I ~\ I ‘v-.. f \‘----O I, 3 L --.A |60 ieo Fraction Number Figure 1. Separation of ribulose-P2 carboxylase/oxygenase from parsley leaves on a sepharose 6B col mm. The buffer was 50 mM Tris adjusted to pH 8.3 at 4°. buffer at pH 8.3 (Fig. 1). Similar results were obtained with or without 2 mM B-mercaptoethanol. These procedures and results are similar to those previously observed for enzyme for spinach or tobacco leaves (5), as well as parsley (Br'andén, personal conmunication). To separate the two activities Br'a'ndén used 5 mM Tris, adjusted to pH 8.3 at room temperature, for the Sepharose 6B filtration step. This modification of the procedure with or without B-mercaptoethanol , did not result in the separation of the carboxy- lase/oxygenase activities in our experiments (Fig. 2). When the clarifica- tion step preceding the 6-25 col mm was anitted (Methods), to bring the present procedure in line with that in reference 1, a shoulder absorbing at 280 nm, presumably nucleic acids, preceded the enzyme activity off the Sepharose 6B colunn (Fig. 2). This omission had no influence on the subsequent distribution of the carboxylase and oxygenase activities. Br'a'ndén's oxygenase assays were plagued by a substantial endogenous rate of oxygen consumption which was not dependent on the presence of 897 Vol. 84, No. 4, l978 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS IZI' c ‘3 .9 a 2 g g 5 o :I.O<~5.é CD _ N E s ‘6 8- .‘s 8 4 g 5 8“ g 6- 3'61-3: \- W a 3 ‘ 2 2 2 4~ I E 4 2 E 9 :I 1 i 2 \EA . rI I ....... ‘ r I20 Fraction Number Fi ure 2. Separation of ribulose-P carboxylase/oxygenase from parsley leaves on a sepharose 6B colunn.The bu fer was 5 mM Tris adjusted to pH 8.3 at 22°. The oxygenase assays were initiated by the addition of activated enzyme to the assay mix containing ribulose-P2 (o—o) or alternatively by the addition of ribulose-P2 to oxygenase preincubated in the assay mix for 10 min (0— —o). ribulose-P2. This forced him to use a 5-10 min preincubation of the enzyme in the assay buffer before initiation of the reaction with ribulose-P2. Nith partially purified enzyme preparations from spinach, tobacco, and now parsley, we did not observe an endogenous rate of 02 uptake. Bra'ndén (personal comnunication) has attributed his endogenous rate to oxidation of the dithiothreitol in the activating and assay buffers. Copper has been reported to accelerate the oxidation of dithiothreitol (7), but addition of small amounts of Cu(II) to our system did not produce a significantly faster endogenous rate, so we are at loss to explain his endogenous rate in the assay. In our experiments ribulose-P2 oxygenase assays were initiated by addition of an aliquot of the concentrated, activated enzyme into an assay buffer without NaHC03. An initially linear, substrate and enzyme dependent rate of 02 Uptake was measured over the first 30 5. Without substrate or before adding enzyme the endogenous rate was zero. If a 898 Vol. 84, No. 4, 1978 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS purified enzyme from spinach leaves was first activated by preincubation with 10 mM NaHC03 and MgClz. and then assayed for oxygenase activity at various times after its addition to the assay buffer, rapid inactivation occurred with a half time of approximately 1 min (data not shown). This rapid decline in activity has been analyzed by Laing and Christeller (8) and Lorimer e_t i]: (9) and attributed to a dissociation of the active enzyme— 2+ complex in the final assay mixture which contained only 0.4 COZ-Mg mM NaHC03 from carryover in the aliquot of activated enzyme in 10 mM NaHCO3. Performance of the oxygenase assay in the manner described by Br'a'ndén (1) resulted in much lower activites due to dissociation of the activated complex. Thus 10 min after dilution of the activated enzyme from parsley' into the assay buffer, the oxygenase activity was severely reduced (o--o in Fig. 2). Nevertheless, this small amount of remaining ribulose- P2 oxygenase coincided with the ribulose-P2 carboxylase peak from the Sepharose 6B col unn. Additional observations by BrEndén (1) were a blue color in the frac- tions with oxygenase activity and a strong EPR signal characteristic of "blue" oxidases. Our atomic absorption spectrosc0py revealed trace amounts of copper, 0.18 g atom/mole enzyme, in the peak fractions (assuming a mole- cular weight of 550,000) and a similar amount of c0pper per mg protein in the shoulder fractions. No EPR signal characteristic of copper was found in any of the Sepharose 6B col unn fractions with carboxylase/oxygenase activity that contained approximately 10 pM copper by atomic absorption, even though the limits of detection in the EPR experiments were estimated to be between 1-2 pM for Cu(II). Addition of 1 mM ferricyanide to convert Cu(I) to Cu(II) was without effect on the EPR spectrun. Thus it seems that any Cu(1) pre- sent in the protein samples was inaccessible to ferricyanide or that the copper was EPR inactive. Cu(II) (10 pH) was added to fractions from Fig. 2, but it had no effect on the rate or on the location of the oxygenase activity. 899 Vol. 84, No. 4, I978 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS In conclusion, ribulose-P2 carboxylase/oxygenase from parsley leaves seems to be a single protein similar to that from spinach and tobacco leaves. Brandén‘s (1) reported separation of these two activities has not been confirmed. He also emphasize that in the ribulose-P2 oxygenase assay, the rate must be determined quickly before the enzyme is inactivated in the absence of C02. The difference in the copper content of the pre- parations may be in the source of parsley. In view of the instability of ribulose-P2 (10), a nonspecific copper-protein complex might also oxidize this substrate. Until the mechanisn of action and the cofactor requirement for ribulose-P2 Oxygenase are known, a continuing concern about copper in the enzyme can be expected. ACKNOWLEDGEMENTS Supported in part by NSF grant PCM 78 15891 and published as journal artical number 8715 of the Michigan Agricultural Experiment Station. We thank B. A. Averill for determination of the EPR Spectra and R. H. Leucke far the copper absorption spectra. REFERENCES 1. Branden, R., (1978). Biochem. Biophys. Res. Commun. 81, 539-546. 2. Wishnick, M., Lane, M. D., Scrutton, M. C., and Mildvan, A. 5., (1969). J. Biol. Chem. gig, 5761-5763. 3. Lorimer, G. H., Andrews, T. J., and Tolbert, -N. E., (1973). Biochemistry 12, 18-23. 4. Chollet, R., Anderson, L. L., and Hovsepian, L. C., (1975). Biochem. Biophys. Res. Commun. 64, 97-107. 5. Ryan, F. J., and Tolbert, N. E. (1975). J. Biol. Chem. 259, 4229-4233. 6. Horecker, B. L., Hurwitz, .J., and Weissbach, A., (1958). Biochem. Prep., 6, 83-90. 7. Takabe, T., Ishikawa, H., Miyakawa, M., Nikai, S., (1978). Agric. Biol. Chem. g, 593-598. 8. Laing, N. A., and Christeller, J. T. (1976). Biochem. J. l§2, 563-570. 9. Lorimer, G. H., Badger, M. R., and Andrews, T. J., (1977). Anal. Biochem. 18, 66-75. 10. Paech, C., Pierce., J., McCurry, S. D., and Talbert, N. E., (1978). Biochem. Biophys. Res. Commun., in press. 900