MSU LIBRARIES M \— RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES wi11 be charged if book is returned after the date stamped be1ow. EVIDENCE FOR A SPIN-COUPLED BINUCLEAR IRON UNIT AT THE ACTIVE SITE OF THE PURPLE ACID PHOSPHATASE FROM BOVINE SPLEBN (PART I) and ISOLATION FROM HUMAN AND BOVINE SPLEEN OF A GREEN HEME PROTEIN WITH PROPERTIES OF MYELOPEROXIDASE (PART II) BY James C. Davis A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1982 ABSTRACT EVIDENCE FOR A SPIN-COUPLED BINUCLEAR IRON UNIT AT THE ACTIVE SITE OF THE PURPLE ACID PHOSPHATASE FROM BOVINE SPLEEN (PART I) and ISOLATION FROM HUMAN AND BOVINE SPLEEN OF A GREEN HEME PROTEIN WITH PROPERTIES OF MYELOPEROXIDASE (PART II) BY James C. Davis A new purification scheme has been developed for the pur- ple acid phosphatase from beef spleen. Metal analyses show the presence of two iron atoms per 40,000 molecular weight enzyme molecule. Upon mild reduction the enzyme becomes more active and undergoes a purple (Ama 550 nm) to pink (Amax x 505 nm) conversion. Resonance Raman spectra of the pink, purple, and phosphate-inhibited forms of the enzyme show tyrosyl vibration modes, implicating a ligand-to-metal charge transfer band as the chromophore. The purple form is EPR silent while the pink form exhibits a complex 9 1.77 EPR 305%7 ’0‘ (1 ( James C. Davis spectrum equivalent to 0.9:0.1 spins per two iron atoms. Upon addition of saturating phosphate concentrations, the purple color returns and the enzyme becomes EPR silent. Magnetic measurements show the purple form to be diamagnetic (4 to 300 K); however, the pink form behaves as a Curie law S=% system (per two iron atoms). Anaerobic reductive titrations show that the shift from purple to pink is a one electron process as is conversion to the EPR active form. Phosphate has been shown to bind to the purple but not to the pink form, suggest- ing that covalently bound phosphate is formed upon phosphate ester hydrolysis. These spectral and magnetic data, when con- sidered with kinetics data, represent unequivocal evidence that the iron center in beef spleen purple phosphatase is a spin-coupled binuclear iron (III) unit. Corroborating evidence for the binuclear nature of the iron center is provided by the formation of the Fe-Zn, Fe-Cu, and Fe-Cd forms of the enzyme. The Fe-Zn form exhibits EPR and magnetic behavior consistent with high spin Fe (III), while exhibiting 100% of the p-nitrophenylphosphate hydrolysis activity of the native enzyme. Also reported here is the isolation of green heme per- oxidases from human and bovine spleen. Both enzymes are similar in terms of spectral characteristics to myeloper- oxidase (EC 1.11.1.7) but differ in several respects. Both Spleen enzymes have a much lower molecular weight than myelo— peroxidase and exhibit different substrate specificity. To Dianne, Scott, and Stephanie for their love and support ii ACKNOWLEDGMENTS I would like to express my appreciation to my preceptor Professor Bruce A. Averill for providing his unique blend of humor, incentive, and guidance that made three years seem like only one; thanks are also due to Professor Averill for providing financial support. I am grateful to the members of my guidance committee, Professors T. Pinnavaia, G. T. Babcock, and C. K. Chang for providing expert professional advice, stimulating discussion, and creative criticism. Special thanks are due to Professor Tanis for acting as substitute in the absence of Professor Chang. I gratefully acknowledge the assistance of Les Szabo of the Biochemistry Department in the gel electrophoresis work and the use of equipment and supplies provided by Professors Willis Wood and Karel R. Schubert. Special thanks are due to S. Tonsager and M. B. Martin for the plasma emission metal assays, to R. Ingle for assistance in obtaining EPR and resonance Raman spectroscopy, to Professor J. A. Fee of the University of Michigan, Ann Arbor, for confirmation of EPR spectral data and to Professor J. Dye of Michigan State for helpful discussions and as- sistance with magnetic measurements. Appreciation is also expressed to Professor P. Debrunner of the University of Illinois, Champaign, for obtaining Mossbauer spectra and iii to Professor P. Aisen of Colombia University, New York, for providing a sample of porcine uteroferrin. I would also like to thank my fellow graduate students in Professor Averill's group for their friendship, support, and stimulating discussions. Dr. H. C. Silvis, Dr. R. H. Tieckelman, M. R. Antonio, W. E. Cleland, S. M. Kauzlarich, and P. Lamberty provided companionship and professional assistance whenever needed. A very special acknowledgment is due to Ms. M. Johnson who processed hundreds of spleens with no complaint and treated my research project as if it were her own. Much of the data collected herein would not have been possible without the careful purification of enzymes provided by Ms. Johnson. I would also like to acknowledge the financial support of the Chemistry Department, the National Institutes of Health (Grant GM 28636), andtme Standard Oil Co. (Ohio). Thanks are also expressed to Mrs. Warstler for her assist- ance in preparing this dissertation. iv TABLE OF CONTENTS Chapter Page LIST OF TABLES. . . . . . . . . . . . . . . . . . . x LIST OF FIGURES . . . . . . . . . . . . . . . . . . xiii LIST OF ABBREVIATIONS . . . . . . . . . . . . . . . xxii PART I. EVIDENCE FOR A SPIN-COUPLED BINUCLEAR IRON UNIT AT THE ACTIVE SITE OF THE PURPLE ACID PHOSPHATASE FROM BOVINE SPLEEN INTRODUCTION. . . . . . . . . u . . . . . . . . . . 1 LITERATURE REVIEW . . . . . . . . . . . . . . . . . 4 Phosphatases. . . . . . . . . . . . . . . . . . 4 Purification and Properties of Bovine Spleen Purple Acid Phosphatase. . . . . . . . . 8 Kinetics, pH Activity, and Specificity of Bovine Spleen Purple Acid Phosphatase. . . . 11 Spectral Characteristics of the Bovine Spleen Purple Acid Phosphatase. . . . . . . . . 12 Magnetic Properties of Purple Acid Phosphatases. . . . . . . . . . . . . . . . . . 15 Metal Substitution Studies of Purple Acid Phosphatases . . . . . . . . . . . . . . . 17 CHAPTER 1 - PURIFICATION AND CHARACTERIZATION OF BOVINE SPLEEN PURPLE ACID PHOSPHATASE . . . . . . . . . . . . . . 19 Materials and Methods . . . . . . . . . . . . . 20 Chemicals . . . . . . . . . . . . . . . . . 20 Analytical Methods and Instrumenta- tion. . . . . . . . . . . . . . . . . . . . 21 Chapter Page Extraction and Purification of the Enzyme. . . . . . . . . . . . . . . . . 23 Molecular Weight Determinations . . . . . . 24 Results . . . . . . . . . . . . . . . . . . . . 25 Purification. . . . . . . . . . . . . . . . 25 Protein Assay . . . . . . . . . . . . . . . 31 Molecular Weight Data and Electro- phoretic Purity . . . . . . . . . . . . . . 31 Discussion. . . . . . . . . . . . . . . . . . . 37 CHAPTER 2 - MOLECULAR PROPERTIES OF BOVINE SPLEEN PURPLE ACID PHOSPHATASE. . . . . 40 Materials and Methods . . . . . . . . . . . . . 41 Results . . . . . . . . . . . . . . . . . . . . 42 Metal Content . . . . . . . . . . . . . . . 42 Kinetics Data . . . . . . . . . . . . . . . 43 Reactivity of Purple Acid Phosphatase with Neuraminidase. . . . . . . . . . . . . 48 Covalently Bound Phosphate in Purple Acid Phosphatase. . . . . . . . . . . . . . 54 Discussion. . . . . . . . . . . . . . . . . . . 54 CHAPTER 3 - SPECTROSCOPIC AND MAGNETIC PROPERTIES OF BOVINE SPLEEN PURPLE ACID PHOSPHATASE . . . . . . . . 62 Materials and Methods . . . . . . . . . . . . . 63 Electronic Spectra. . . . . . . . . . . . . 63 EPR Spectra . . . . . . . . . . . . . . . . 64 Resonance Raman Spectra . . . . . . . . . . 64 Mdssbauer Spectroscopy. . . . . . . . . . . 65 vi Chapter Page Magnetic Susceptibility Measure- ments. . . . . . . . . . . . . . . . . . . 66 Anaerobic Titrations . . . . . . . . . . . 66 Results . . . . . . . . . . . . . . . . . . . 67 Visible Spectra of Native and Reduced Enzyme in the Presence of Inhibitors . . . 67 Resonance Raman Spectroscopy of the Oxidized (Purple), Reduced (Pink), and Phosphate-Inhibited Forms of the Enzyme . . . . . . . . . . . . . . . . . . 77 EPR Spectrosc0pic Studies of the Purple Acid Phosphatase from Bovine Spleen. . . . 77 Discussion. . . . . . . . . . . . . . . . . . 108 CHAPTER 4 - METAL ION SUBSTITUTION STUDIES WITH BOVINE SPLEEN PURPLE ACID PHOSPHATASE . . . . . . . . . . . . . 123 Materials and Methods . . . . . . . . . . . . 125 Results . . . . . . . . . . . . . . . . . . . 126 Metal Analyses . . . . . . . . . . . . . . 126 Kinetics Data. . . . . . . . . . . . . . . 128 Electronic Spectra . . . . . . . . . . . . 128 EPR Spectroscopy . . . . . . . . . . . . . 131 Magnetic Measurements on the Fe-Zn Purple Phosphatase . . . . . . . . . . . . 139 Discussion. . . . . . . . . . . . . . . . . . 139 CHAPTER 5 - PREPARATION AND CHARACTERIZATION OF TWO MODELS FOR THE IRON CENTER IN PURPLE ACID PHOSPHATASE. . . . . . 148 Materials and Methods . . . . . . . . . . . . 149 Results . . . . . . . . . . . . . . . . . . . 152 vii Chapter Page Discussion. . . . . . . . . . . . . . . . . . 160 SUMARY O O O O O O O O . O 0 O O O O O O O O O O l 71 PART II - ISOLATION FROM HUMAN AND BOVINE SPLEEN OF A GREEN HEME PROTEIN WITH PROPERTIES OF MYELOPEROXIDASE LITERATURE REVIEW . . . . . . . . . . . . . . . . 176 CHAPTER 1 - A MYELOPEROXIDASE-LIKE GREEN HEME PROTEIN FROM BOVINE SPLEEN. . . . . . 179 Materials and Methods . . . . . . . . . . . . 179 Extraction and Purification. . . . . . . . 179 Analytical Methods . . . . . . . . . . . . 181 Spectral Characterization. . . . . . . . . 182 Results . . . . . . . . . . . . . . . . . . . 183 Isolation. . . . . . . . . . . . . . . . . 183 Visible Spectral Data. . . . . . . . . . . 186 Electron Paramagnetic Resonance Spectrosc0py . . . . . . . . . . . . . . . 194 Molecular Weight Studies . . . . . . . . . 194 Iron Content . . . . . . . . . . . . . . . 204 Substrate Specificity. . . . . . . . . . . 204 Attempts at Heme Removal . . . . . . . . . 206 Resonance Raman Spectrum . . . . . . . . . 207 Discussion. . . . . . . . . . . . . . . . . . 207 CHAPTER 2 - A MYELOPEROXIDASE-LIKE GREEN HEME PEROXIDASE ISOLATED FROM HUMAN SPLEEN. . . . . . . . . . . . . . . . 216 Materials and Methods . . . . . . . . . . . . 216 Results . . . . . . . . . . . . . . . . . . . 218 Chapter Discussion. SUMMARY . . APPENDIX A. REFERENCES. ix Page 227 229 230 232 Table LIST OF TABLES Page Sources, Type of Metal, and pH Optimum of Purple Phosphatases. . . . . . . 6 Purification of Purple Acid Phos- phatase from Beef Spleen. . . . . . . . . . 26 Inhibition of Beef Spleen Purple Acid Phosphatase. . . . . . . . . . . . . . 51 Inorganic and Covalently Bound Phos- phatate in Purple Acid Phosphatases . . . . 55 Double Integrated Values for the EPR Spectra of Reductive Titration of Purple Acid Phosphatase with Shift in Absorption Maxima Shown. . . . . . . . . 93 Major Resonance Raman Bands Exhibited by the Native, Reduced and Phosphate Inhibited Purple Acid Phosphatase from Bovine Spleen and the Corresponding Bands in the Porcine and Sweet Potato Enzymes . . . . . . . . . . . . . . . . . 113 Double Integration of the EPR Spectra of the Oxidized (Native), Reduced (Pink), and Reduced Plus Phosphate Forms of the Table 10 11 12 13 14 15 Page Purple Acid Phosphatases from Bovine Spleen and Porcine Uterus . . . . . . . . . 119 Metal Content and Enzymatic Activity of Metal-Substituted Purple Acid Phosphatase . . . . . . . . . . . . . . . . 127 Visible Absorption Maxima and Molar Absorptivities of Several Metal Substituted Forms of Purple Acid Phosphatase from Bovine Spleen. . . . . . . 134 Elemental Analysis of Fe(sa1his)2PF - 6 CH CH OH and Fe(sa1hisH BPh 151 3 2 2)2 4. . . . . . Summary of Crystallographic Data for Fe(sa1his)2PF6. . . . . . . . . . . . . . . 153 Representative Fe-O and Fe-N Distances for Fe(sa1his)2PF6 and Similar Com- pounds. . . . . . . . . . . . . . . . . . . 167 Some of the Fe(salhis)2PF6 Bond Angles of Ligating Atoms with the Central Iron Atom . . . . . . . . . . . . . 168 Purification of Green Heme Protein from Beef Spleen. . . . . . . . . . . . . . 189 Summary of Visible Absorption Data for Green Heme Peroxidase and Myeloperoxidase . . . . . . . . . . . . . . 197 xi Table 16 17 Page Substrate Specificity of the Green Heme Peroxidase from Bovine Spleen . . . . . . . 205 Summary of Visible Absorption Data for Green Heme Peroxidase from Human and Bovine Spleen . . . . . . . . . . . . . . . 226 xii Figure LIST OF FIGURES Carboxymethyl cellulose (CM-52) ion ex- change chromatography of the been spleen purple acid phosphatase . . . . . Chromatography of the purple fractions from CM-52 on Sephadex G-75 with ab- sorption at 550 and 280 nm . . . . . . . Gel permeation molecular weight estima— tion on Sephadex G-75. . . . . . . . . . SDS-gel electrophoresis estimation of the molecular weight of the pur- ple phosphatase from beef spleen . . . . Lineweaver-Burk plots of kinetics data with p-nitrophenyl phosphate as sub- strate (phosphate, cyanide, and azide inhibition shown). . . . . . . . Lineweaver-Burk plots of kinetics data with p-nitrophenyl phosphate as sub- strate (F- and PABP inhibition shown). Lineweaver-Burk plots of kinetics data with p-nitrophenyl phosphate as substrate (shown are native enzyme alone xiii Page 28 3O 34 36 45 47 Figure 10 11 12 and with MoOi-, W0 and voi'). . . . . 2- 4 Gel electrophoresis of purple acid phosphatase (FePase) before (A) and after (B) treatment with neuramini- dase (N) . . . . . Proposed mechanism for the reduction, binding of substrate, ducts and inhibition by fluoride of the release of pro— purple acid phosphatase from beef spleen . . . . . . Visible absorption spectra of the pur— ple acid phosphatase. Shown are the native enzyme and the reduced enzyme with 10 mM F- and 10 mM phosphate. Visible absorption spectra of the pur- ple acid phosphatase. time course of dithioerythritol reduc- Shown are the tion and the reduction with ferrous ion and ascorbate. Visible absorption spectra of the pur- ple acid phosphatase. native enzyme, enzyme reduced with fer- rous ion and ascorbate, and enzyme re- Shown are the duced with ferrous ion and ascorbate with 1 mM vo3’ added . 4 xiv Page 50 53 59 69 71 74 Figure 13 14 15 16 Visible absorption spectra of the reductive titration with methyl viologen radical cation of bovine spleen purple acid phosphatase . Resonance Raman spectra (1 = 514.5 nm) of purple acid phosphatase from bovine spleen. The native enzyme excitation (spectrum A), reduced form (spectrum B), and the reduced form plus 10 mM phosphate (spectrum C) are shown EPR spectra at 4.5 K of beef spleen purple acid phosphatase. Spectrum A, oxidized (purple) form; spectrum B, reduced (pink) form produced by treating oxidized enzyme with 1.0 mM (NH4)2Fe(SO4)2 and 100 mM ascorbate, pH 5.0; spectrum C, sample prepared as for B followed by addition of 10 mM sodium phosphate, pH 5.0 . . . . EPR spectra at 4.2 K. Spectrum A I pink form of beef spleen acid phos- phatase produced by Fe2+/ascorbate treatment (same sample as in Figure 15, Spectrum B); Spectrum B, oxidized enzyme (0.23 mM) plus 1 equivalent of Na in 0.073 M Tris-HCl buffer, pH 8 XV 2 820 4 Page 76 79 81 84 Figure 17 18 19 20 21 22 Page Power saturation curves for the 4.2 K EPR spectrum of reduced (pink) purple acid phosphatase. Shown are curves for the g = 1.95 peak and for the g = 1.77 peak. . . . . . . . . . . . . 81 Effect of temperature upon the EPR spectrum of reduced (pink) enzyme. . . . . 89 EPR spectra of the reductive titration of purple acid phosphatase from beef spleen . . . . . . . . . . . . . . . . . . 92 Effect of pH upon the EPR spectrum of the purple acid phosphatase from bovine spleen. . . . . . . . . . . . . . . 95 EPR spectra of the purple acid phos- phatase from bovine spleen prepared without the use of hydroxylapatite. Shown are the native enzyme (spectrum A), the redued (pink) form (spectrum B), and the reduced form plus 10 mM phosphate (spectrum C) . . . . . . . . . . 98 EPR spectra of the purple acid phos- phatase from porcine uterine fluids. Shown are the native enzyme (Spectrum A), the reduced (pink) form (Spectrum B), and the reduced form plus 10 mM phosphate (spectrum C) . . . . . . . . . . 100 xvi Figure 23 24 25 26 Page EPR spectrum of the reduced form of the purple acid phosphatase from 3- 4 Mossbauer spectrum of purple acid bovine spleen with 1 mM VO added . . . . 102 phosphatase from bovine spleen at 4.2 K and 0 field. . . . . . . . . . . . . 104 Magnetic susceptibility data as a function of temperature T for various forms of beef spleen purple acid phos- phatase. (A) Purple (oxidized) enzyme at 1.5 mM. The solid line shown is calculated for 2.5% high-spin Fe3+. (B) Pink (reduced) enzyme at 0.75 mM, prepared by reduction with 100 mM dithiothreitol for 200 min. The solid line is calculated for a spin-only S = 1/2 paramagnet at 0.75 mM (using 9 = 1.77) . . . . . . . . . 107 Magnetic susceptibility data for native purple acid phosphatase as a function of temperature. Shown are three computer generated curves for a Spin 5/2-spin 5/2-coup1ed system with values of 2J equal to: -134, -300 and -200 cm-1. . . . . . . . . . . . . . . 110 xvii Figure 27 28 29 30 31 32 Page Lineweaver-Burk plots of kinetic data for the Fe-Zn enzyme with PNPP as substrate. Shown are the native 3.. 4 I mM F‘, 10 mM F‘, and 10 mM Pi. . . . . . . 13o enzyme alone and with 1 mM AsO 2.5 Visible absorption spectra of sev- eral metal substituted forms of bovine spleen purple acid phosphatase. Shown are the Fe-Zn, Fe-Cd, and Fe-Cu forms of the enzyme. . . . . . . . . . . . 133 EPR spectra of the Fe-Zn purple acid phosphatase as isolated (spectrum A), reduced with ferrous ion and ascorbate (Spectrum B) and reduced plus 10 mM phOSphate (spec- trum C). . . . . . . . . . . . . . . . . . 136 EPR spectra at 77 K of Fe—Zn form of beef Spleen purple acid phos- phatase in the absence (spectrum A) and presence (spectrum B) of 10 mA phosphate and 100 mM ascorbate . . . . . . 138 EPR spectrum of the Fe-Cd form of purple acid phosphatase (3.4 mg/mL). . . . 141 Magnetic behavior of the Fe-Zn form of bovine spleen purple acid xviii Figure 33 34 35 36 37 38 39 phosphatase (1.1 mM). The solid line is calculated for an S = 5/2 paramagnet (1.1 mM). . . . . . . . . . . Visible absorption spectra of Fe(sa1his)2PF and Fe(sa1his)2BO 6 4° ' EPR spectra of the Fe(sa1hisH2)2BO4 (spectrum A), conalbumin (spectrum B) and Fe(sa1his)2PF6 (spectrum C) . . . Magnetic behavior of Fe(sa1his)2PF6 (1.1 mM). The solid line is cal— culated for an S = 5/2 paramagnet (1.1 mM) . . . . . . . . . . . . . . . . Unit cell of Fe(sa1his)2PF6 (stereo view). . . . . . . . . . . . . ORTEP diagram for Fe(sa1his)2PF6 (50% probability) with hydrogens and PF6 omitted for clarity. . . . . . Stereo view of Fe(sa1his); - hydrogens omitted for clarity. . . . . . Proposed structures of the two iron center in bovine spleen purple acid phosphatase. Shown are the oxidized (purple) and reduced (pink) forms 0 O O O O O O O O O O O O O O O O O xix Page 143 155 157 159 162 164 166 175 Figure 40 41 42 43 44 45 Page Ion exchange chromatograph of the green heme peroxidase on carboxy— methyl cellulose (CM-52—Whatman) . . . . . 185 Chromatography of the green frac- tions from CM-52 on Sephadex G-75. . . . . 188 Disc-gel electrophoresis of green heme peroxidase on 7.5% polyacryl- amide. . . . . . . . . . . . . . . . . . . 191 Visible absorption spectra of the green heme peroxidase. Shown are the Fe(III) native enzyme, the Fe(III) cyanide adduct, the oxidized pyridine hemochromogen, and the re- duced pyridine hemochromogen . . . . . . . 193 Visible absorption spectra of the reduced Fe(II) forms of the green heme peroxidase. Shown are the reduced native form, the CN- adduct, the N3 adduct, the NO adduct, and the CO adduct. . . . . . . . . . . . . . . . . 196 EPR Spectrum of the green heme 6 (com- peroxidase at gain 1 x 10 pilation of 10 scans on signal averaging package) . . . . . . . . . . . . 199 XX Figure 46 47 48 49 50 51 Page Gel permeation molecular weight estimation on G-75. Standards were: A, B-lactoglobulin; B, a-chymotrypsino- gen; C, ovalbumin; D, bovine serum albumin. . . . . . . . . . . . . . . . . . 201 SDS-gel electrophoresis estimate of the molecular weight of the green heme peroxidase. Standards were: A, lysozyme; B, B-lactoglobulin; C, d-chymotrypsinogen; D, ovalbumin; E, bovine albumin . . . . . . . . . . . . . . 203 Resonance Raman Spectrum (Aexcitation = 413.1 nm) of bovine spleen green 3+) . . . . . . . . . . 209 heme peroxidase (Fe Carboxymethyl-cellulose KC1 gradient chromatography of the human green heme peroxidase from spleen. . . . . . . . 221 Sephadex G-75 chromatography of the human green heme peroxidase from spleen . . . . . . . . . . . . . . . . . . 223 Visible absorption spectrum of the native (Fe3+) human green heme per- oxidase from spleen. . . . . . . . . . . . 225 xxi ASC CM-52 DTE DTH EDTA EPR MES 2-MET P-ll PABP PMSF PNP PNPP SDS TEMED ABBREVIATIONS Ascorbic Acid Carboxymethyl cellulose CM-52 Dithioerythritol Sodium Dithionite Ethylenediaminetetracetic Acid Electron Paramagnetic Resonance 2-(N-morpholino)ethanesulfonic Acid B-Mercaptoethanol Methyl Viologen Radical Cation Methyl Viologen Dication Inorganic Phosphate Cellulose Phosphate P-ll [p-(Aminoacetyl)benzylj-phosphonate Phenylmethanesulfonyl fluoride p-Nitrophenol p-NitrOphenyl phosphate Sodium Dodecylsulfate N,N',N"JV”-Tetramethylethylenediamine xxii PART I EVIDENCE FOR A SPIN-COUPLED BINUCLEAR IRON UNIT AT THE ACTIVE SITE OF THE PURPLE ACID PHOSPHATASE FROM BOVINE SPLEEN INTRODUCTION The detailed study of metalloenzymes is becoming more and more the domain of the inorganic chemist. Many of the physical methods utilized to study transition metal complexes are equally applicable, given the sensitivity of modern analytical and spectroscopic instrumentation, to the study of metal ions in biological systems. Thus, electron para- magnetic resonance spectroscopy can provide information on the environment of metal ions even when present in only micromolar quantities. In a similar fashion, recent ad- vances in liquid helium cooled superconducting magnets and electronics have provided superconducting quantum inter- ference susceptometers (SQUID) capable of detailed magnetic measurements on samples containing as little as one to two micrograms of metal - ideal for the study of the magnetic- ally dilute metal ions found in metalloenzymes. Improved metal analysis methods such as plasma emission spectroscopy provide accurate metal analyses in the parts-per-billion range, again ideal for the detection of trace metals in biological systems. A principal technique of inorganic chemists has always been electronic absorption spectro- photometry since many transition metal chelates are vividly colored. Often the initial clue to the presence of a metallo- enzyme in a particular preparation is the visual observa- tion of a brightly colored band on a chromatography column. The appearance of such bands typically creates an irresis— tible temptation (at least to a bioinorganic chemist) to explore the nature of such a chromophore. Reports of the widespread occurrence of a class of intensely purple phos- phatases (some of which are reported to contain iron) represent an opportunity to utilize traditional, as well as new, inorganic chemical techniques for the study of the metal chromophore in these enzymes. The following chapters describe the isolation of the purple acid phosphatase from bovine spleen and the physico- chemical study of the iron at the active site. Part I, chapter 1 details the development of a new purification procedure for the enzyme along with preliminary biochemical characterization. Chapter 2 provides insight into the bio— chemical nature of the enzyme via kinetics studies, while chapter 3 describes e1ectronic,EPR, Mdssbauer, and reson- ance Raman spectroscopic and magnetic studies of the enzyme. Chapter 4 is a description of the properties of several metal substituted forms of the purple phosphatase; chapter 5 details the preparation, properties, and structure of model compounds synthesized to provide comparative spectral and magnetic data. Part II describes the discovery and characterization of a new green heme peroxidase fortuitously discovered during the purification of the purple acid phos- phatase from beef spleen (chapter 1) and its isolation from human spleen (chapter 2). '6 ' J Dc! : LITERATURE REVIEW Phosphatases Enzymes that catalyze the hydrolysis of phosphate esters are termed phosphatases and are found in a wide variety of organisms. These enzymes have important meta- bolic roles in controlling levels of phosphate and phos- phorylated metabolites and in regulating the activity of other enzymes. Several subdivisions of phosphatases exist, perhaps the most common being between acid and alkaline phos- phatases. This distinction arises from the existence of enzymes with optimal activity at either acid or at alkaline pH (but not both)l. Certainly of more interest to inorganic or bioinorganic chemists is the division of the phOSphatases according to prosthetic group. A large number of zinc-containing alka- line phosphatases is known; they have been extensively studied and are the subject of numerous reviews2-4. Many acid phosphatases are thought not to contain transition metal ions and have been exhaustively researcheds. In many cases, however, the quantities of enzyme available have precluded metal analyses. Both the acid and alkaline phosphatases are generally colorless even in concentrated solutions, but a small (yet growing) number of intensely purple (€500_550 e 2-3000) enzymes has been reported. These purple phosphatases are in general poorly charac- terized and as a rule are difficult to obtain. Table 1 lists the sources of known purple phosphatases; all except the M. sodenensis enzyme are basic glycoproteins. The metal prosthetic group (if known), and whether their range of maximum activity is at acid or alkaline pH are also given in the table. Only the enzymes from beef spleen, porcine uterine fluids, and sweet potato will be discussed in de- tail, as the others have not yet been extensively studied. The kidney bean enzymes’7 has been Shown to contain iron and to have hemaglutinin activity, but no additional infor- mation has been published. Of the plant sources, only the sweep potato enzymel6-18 has been extensively studied with most of the characterization done by Sugiura and co-workersl4-l6. The M. sodenensis purple phosphatase has been studied by 24 Glew and Heath , who found eight Ca+2 ions per molecule of glycoprotein; this is the only known example of an alka- line purple phosphatase. Acid phosphatases from potato, wheat germ, and human prostate and alkaline phosphatase from E. ggli_were all shown to proceed via a phosphoenzyme intermediatezs. The rate of reaction is independent of product alcohol concen- tration and they all covalently bind phosphate (Pi) upon incubation with excess inorganic Pi and catalyze 180 ex- change between water and Pi' All involve Pi binding to a . H A-“ Table 1. Sources, Type of Metal, and pH Optimum of Purple Phosphatases. Prosthetic pH Source Group Optimum Ref. Kidney bean Fe acid 6,7 Bovine spleen Fe acid 8-10 Porcine uterine fluids Fe acid 11-13 Sweet potato Mn acid 14-18 Soybean Mn acid 19,20 Spinach leaves Mn acid 21 Cultured rice plant cells Mn acid 22 Neurospora crassa -- acid 23 Micrococcus sodenensis Ca alkaline 24 serine residue and many of the enzymes that catalyze hydrolysis of phosphoric esters require a divalent cation +2 2). Phosphoprotein phosphatases from heart, (Ca or Zn+ liver, and muscle tissue are also known which, like the bovine spleen and porcine uterine purple phosphatases, are activated by addition of dithioerythritol (DTE), ascorbate (ASC), and B-mercaptoethanol (2-MET)26-30. These enzymes are, however, difficult to obtain and are not well charac- terized in terms of metal content. They do bear a remark- able similarity to the bovine spleen purple phosphatase in terms of molecular weight, pH optimum, inhibition charac- teristics and hydrolysis mechanism. 31 have isolated an acid phos- DiPietro and Zengerle phatase from human placenta which has properties much like the porcine uterine purple phosphatase. The two enzymes have similar molecular weight (35,000), are inhibited by phos- phate but not tartrate, and show marked increase in activity upon treatment with mild reductants. The human enzyme does not, however, exhibit either molybdate or fluoride inhibi- tion. Tartrate resistant acid phosphatase activity has also been found in human spleens of patients with leukemic 32 and Gaucher's disease33’34. Both reticuloendotheliosis enzymes are inhibited by Pi and molybdate, but not by tar- trate, are highly basic, and are not active catalysts for hydrolysis of glycerophosphate. In addition, the Gaucher's Spleen enzyme has a molecular weight of 33,000 and upon sodi sodium dodecylsulfate (SDS) gel electrophoresis exhibits two subunits of molecular weight 16,000 and 20,000. Kinetics data for this enzyme are remarkably consistent with the mammalian purple phosphatases; for example, molybdate (Ki 3 2 mM), fluoride (Ki 3 1 mM), and ferrous (Ki 2 1.5 mM) ions are all inhibitors. Unfortunately only microgram quan— tities of the Gaucher's Spleen enzyme were isolated, so no electronic spectra nor metal assays were attempted. Purification and Properties of Bovine Spleen Purple Acid Phosphatase Intrigued by the reports of Harris in 1946 of a reducing agent—activated phosphatase in frog's eggs35 and of Fein- 36 stein and Volk in 1949 , who reported finding a similar enzyme in rat spleens, Sundararajan and Sarma described in 1954 a partial purification of ox spleen acid phosphatase37. They obtained a 200-fold purified enzyme fraction strongly activated by ascorbate, cysteine, and thioglycolic acid. (Note the similarity to the activation of phosphoprotein 26-30 phosphatases from other tissues .) Their principal means of purification was successive precipitation with ammonium sulfate and acetone. Singer and Fruton38, Hof- 40 fman39, and Sundararajan and Sarma later reported modi- fications of the original method including butanol extrac- 39 tion, but with little if any improvement in yield and no substantial increase in purity. 42 41 and Revel and Racker in 1960 were Glomset in 1959 the first to use modern liquid chromatographic methods for purifying the purple phosphatase from beef spleen. The latter obtained a 240-fold purification and were the first to report stabilization of the reduced form of the enzyme by ferrous ion. Glomset, however, has the distinction of first reporting an electrophoretically pure protein having an intense purple color. A year later, in collaboration with Porath43 , he published the first electronic absorption spec- trum of the purified enzyme, as well as the first metal assays. Unfortunately, the sensitivity of the assay method apparently precluded the detection of iron in the protein. They also reported the first evidence for a shift in Amax to lower wavelength upon treatment with mild reducing agents. Campbell and Zerner8 reported a new purification method in 1973 and found one iron atom per molecule of enzyme. This value was later corrected to two iron atoms per mole- culeg; the original error resulted from the use of impure enzyme. In the latter paperg, Zerner's group reported the molecular weight to be 38,000 with the enzyme composed of two subunits of 15,000 and 26,000 molecular weight. They also found the equivalent weight per iron to be 19,400, corresponding to two iron atoms per molecule. The molar absorptivity per iron at 550 nm was found to be 2040 L— nmde-l cm-l, while the absorbance at 280 nm of a 1 mg per lnL solution was 1.59. Their preparation was reported as ‘3 95% pure by e1ectr0phoresis, but unfortunately no details 10 of the purification method have been published. In the same paper, Zerner's group compared the bovine spleen enzyme to the purple acid phosphatase from porcine allantoic fluid. Their data show a remarkable similarity between the two enzymes in terms of molecular weight, iron content, electronic spectra, molar absorptivity, and activa- tion by reducing agents. Buhi, 33 21-11'44 have prOposed a role for the porcine uterine enzyme in iron transport in uterine fluids. They have termed this enzyme "uteroferrin" presumably from its similarity in this function to that of transferrin. Zerner's results are not, however, in complete agreement with the data presented by Chen, 33 31.45, Schols— nagle, 33 31.12 46 , and Roberts, gt 31. who have reported a molecular weight of 32,000 and only one iron per molecule. Even though Zerner's group pointed out in their 1978 paper that serious errors may have been made by Robert's group in terms of molecular weight estimation and in determination of protein concentration (and hence in iron content), the iron content and molecular weight of the porcine purple acid phosphatase are still not generally agreed upon. Similarly, the sweep potato enzyme, reported to contain one atom of manganese per 110,000 molecular weight, has not yet been shown to be free of iron, yet its visible absorption spectra is much like that of the reduced form of the iron containing 1 -1 l6 enzymes (Amax = 515 nm, 8515 = 2460 L-mole cm ) . H Ki 11 Kinetics,ypH Activity, and Specificity of Bovine Spleen Purple Acid Phosphatase Several workers have reported on the substrate specif- icity of the beef spleen enzyme and shown that the enzyme will hydrolyze a broad range of substrates, including phos- 37139140 38,42 phosphoramidates , adenosine tri- phosphate42, and phenyl phosphates42. Similar broad 12,13 phoproteins, Specificity has been reported for the porcine and sweet potato16 enzymes. While both mammalian enzymes fail to catalyze the hydrolysis of B-glycerophosphate39'12, it is a good substrate for the sweet potato enzymel6. Detailed kinetics studies have been performed on the por- cine enzyme, including determination of the Michaelis con- stant (Km = 2.0 mM) for p-nitrophenylphosphate (PNPP), pH optimum (4.9), and increase in activity in the presence of mild reductants46. These results are identical to those found for the bovine spleen enzyme37'10'39 (Km = 2.0 mM, pH optimum = 4.9). The effects of inhibitors on the two enzymes are also very similar. Singer and Fruton report fluoride inhibition of the bovine enzvme. This has been con- firmed by Revel and Racker42, who also found inhibition by iron chelators such as phenanthroline and 8-hydroxyquinoline. Inhibition by molybdate, Pi' and (p-(aminoacety1)benzy1)- phosphonate (PABP), a substrate-like molecule containing a C-C-P bond instead of a hydrolyzable C-O-P linkage, have also been shown for the bovine enzyme37'42. The porcine 12 enzyme exhibits similar inhibition behavior12'46, but no detailed studies have been published for that enzyme. Kinetics and inhibition studies have also been published for the sweet potato enzymels. The binding constant, Km, was reported to be 0.17 mM for (PNPP), but conditions of assay were substantially different than for the mammalian enzymes. Fluoride, Pi' arsenate, and molybdate were potent inhibitors with Ki values of 0.3, 2.3, 0.05, and 0.0011 mM, respectively. The effect of phosphate as a product inhibitor along with the lack of inhibition by p-nitrophenol (PNP) is con- sistent with the mechanism associated with many acid and alkaline phosphatases and phosphoprotein phosphataseszs'26. This coupled with the activation of many of these enzymes by ASC, DTE, and 2-MET suggests a singular mechanism and intermediate for all known phosphatases. spectral Characteristics of the Bovine Spleen Pugple Acid Phosphatase The electronic absorption spectrum of the native bovine purple acid phosphatase is dominated by a large peak at 280 nm, a shoulder at 320 nm, and a broad visible absorption band at 550 nm43. Although a shift to shorter wavelength upon mild reduction has been discussed in other publica- 8,9,43 10 tions , until this work a spectrum of the pink, reduced form of the enzyme had not been presented. 13 Schlosnagle, st 31,12, have published spectra of both the native and reduced forms of the porcine enzyme that are very similar to those of the bovine spleen enzyme. It should be noted that the porcine enzyme may be isolated as a mixture of the oxidized and reduced formslz; Antanaitis and Aisen47 have shown that addition of two equivalents of ferrous ion to native porcine purple phosphatase results in a shift of Xmax from 545 nm to 525 nm. It seems probable that oxygen in the air space above the sample, however, may have oxidized much of the Fe(II) to Fe(III), precluding stoichiometric reduction of the purple to the pink form (particularly since the samples were shaken overnight at 4°C). No evidence has been presented for a similar color change upon mild reduction of the sweet potato enzyme, nor has any attempt to oxidize the pink (Amax = 515 nm) sweet potato enzyme to a purple form been reported. Glomset and Porath43, were unable to detect an electron paramagnetic resonance (EPR) absorption spectrum for the bovine spleen enzyme; the temperature of the sample and the sensitivity of the instrument were not reported. No other reference to EPR spectra of the bovine spleen enzyme is avail- able except for this work10’48. EPR spectra of the porcine material have been reported, however. Schlosnagle 2E 31.49 described a strong 9 = 4.3 EPR signal (due presumably to high spin Fe(III) at 77 K for the native and reduced forms of the porcine enzyme, as did Nochumson for the kidney bean 14 enzyme7. Later Roberts and others reported the g = 4.3 ab— sorption to account for less than 8% of the total iron presentso. They also reported that an unusual g' = 1.74 spectrum accounted for over 90% of the iron (i.e., 0.9 spins/Fe). The high field EPR spectra of both the native and reduced porcine enzyme are much like that of the reduced form of the bovine Spleen enzyme (g = 1.77), both in line- shape and g-values48’51. Results similar to those reported for the porcine enzyme were obtained on beef spleen enzyme that had not been exposed to hydroxylapatite (HAP) during purificationSI. This may lend strength to the argument that, like many phosphatases, the beef spleen enzyme may form an internal phosphate ester as part of the hydrolysis mechanism. The loss of EPR signal in the native (oxidized) form when purified on HAP was attributed to antiferromag- netic coupling of the two iron atoms by covalently bound 51 P. . 1 No EPR spectrum could be detected at 20°C for 1 mM sam- ples of the native sweet potato enzymelG. Upon denaturation with HCl and heat, however, a Mn(II) EPR signal was ob- served at 20°C. Of itself this is not proof of the absence of an iron EPR spectrum as the high field absorption of the reduced mammalian enzymes cannot be detected above 40 K48'50. It is also not clear that the manganese found in the sweet.potato enzyme is responsible for the chromophore or that it is involved in catalysis; manganese is common in plants and is likely to be found as nonspecifically 15 bound metal in plant proteins. Several resonance Raman spectroscopic studies have been reported on purple acid phosphatases. Sugiura and cowork- 14-16 1 ers, attribute bands in the 1200 to 1650 cm- region upon excitation at 514.5 nm to vibrations associated with a metal-bound tyrosyl moeity. A sharp band at 370 cm-1 was tentatively assigned as a Mn(III)-S stretch, although a Mn(III)-O (tyrosyl) stretch (or indeed an Fe(III)-O stretch) could not be ruled out. Resonance Raman Spectra of the porcine enzyme have been reported by Bazer et al.52 and Antanaitis st 31.53'54 (A excitation = 514.5 and 647.1 nm). Their spectra are essentially identical and are very similar to those of Sugiura's sweet potato enzymel6. Antanaitis st 31.54 also used circular dichroic spectroscopy which implicated tyro- sine ligation to the iron in the chromOphore. No resonance Raman data on the bovine purple acid phosphatase have been published to date. Magnetic Prgperties of Purple Acid Phosphatases Antanaitis and coworkers have reported on the magnetic properties of the porcine enzymeso. Using a Faraday balance with the sample cooled to liquid helium temperatures, they found a value of s = l/2 per iron; it should be noted that the iron concentration used in the calculations may be in error. The iron concentration used is referred to as 16 "chromophoric" iron; if this means that the absorption at 550 nm was used to calculate the iron content, then the assay may be low by a factor of m1.5. Antanaitis st al.54 —_ quoted a molar absorptivity at 550 nm of 3100 L-mole-l cm-l per iron, significantly different from the value of 2000 1 cm"1 found on a dry weight basis by Campbell gt 31.9. 0 (40,300 L-mole-l cm-l) L-mole- If Antanaitis' reported value of 528 was used to determine protein concentration, a similar er- ror could result as Campbell's dry weight measurement yields a value of €280 = 56,800 L-mole-l cm-l. Use of either molar absorptivity reported by Antanaitis st 31. could result in an error in iron concentration of 1.55 or 1.43 for c and 8280’ respectively, if Campbell's values are 550 correct. Data has been presented on the oxidized (purple) and reduced (pink) forms of the bovine spleen purple phos- phatase48. These data show diamagnetic behavior for the purple form and paramagnetic behavior (3 = 1/2 per two irons) for the pink form. This is consistent with an s = 0 (high spin Fe(III)-high spin Fe(III)) antiferromagnetically coupled two iron system in the native purple form of the enzyme and a one electron reduced 5 = 1/2 (high spin Fe(III)-high spin Fe(II)) coupled unit in the reduced form. .-u “U. 81; 17 Metal Substitution Studies of Purple Acid Phosphatases The only published experiments regarding the substitu- tion of other transition metals for iron in these enzymes are those of Keough et al.55 for the porcine enzyme. They illustrated a rather exciting property of this enzyme, whereby upon reduction to the colorless (all ferrous) form one of the two iron atoms diffuses out of the protein matrix much more slowly than the other. This property allowed the preparation of a "half-apo" form of the enzyme and subsequent reconstitution with Zn(II) to form an iron- zinc enzyme that has 100% of the enzymatic activity of the native enzyme. This group also reported the reconstitu- tion of the apo-enzyme with two Zn(II) ions as well as partial reconstitution with Ni(II) (1.40 atoms per mole- cule), but activity of the Ni2 and Zn2 forms was less than 2% that of the native enzyme. Similar reconstitution re- sults have been published for the bovine spleen enzyme48. Replacement of one iron by zinc gives a visible spectrum Similar to that of the two iron form with nearly the same molar absorptivity (6550 = 2100 vs. 6550 = 2000 L-mole-1 cm-l, respectively) and the same absorption maximum (550 nm). Unlike the two iron form, the iron-zinc form is paramagnetic (s = 5/2, “eff = 6.0) and has a g = 4.3 EPR signal characteristic of high spin Fe(III)48. The kinetics parameters of the iron-zinc form are identical to those of the two iron form (Vmax’ Km for PNPP, and Ki for Pi’ PABP, 18 and fluoride)10. No comparative data are available for the Fe-Zn form of the porcine enzyme. From the preceding literature review, it is clear that much controversy exists regarding the nature of the catalytic site (or sites) of the purple acid phosphatases. Certainly, the involvement of iron and the number of iron atoms per molecule in the porcine and bovine enzymes are still in question. The possibility of the existence of similar phos- phoprotein phosphatases in other tissues such as liver, heart, and muscle, as well as the previously identified purple phosphatases from a wide variety of sources, is a strong incentive for the study of these systems. The wide- spread requirement for divalent cations for phosphatase activity coupled with the probable formation of Fe(II) in the active form of the bovine (and probably the porcine) enzyme and the activity of the iron-zinc form of the beef and porcine enzymes is particularly exciting to bioinorganic chemists. It is the purpose of the following chapters of this dissertation to describe some properties of the bovine spleen purple acid phosphatase and to clarify the role of the metal as well as the number and oxidation state of the irons in the enzyme. :J 101 n V t CHAPTER 1 PURIFICATION AND CHARACTERIZATION OF BOVINE SPLEEN PURPLE ACID PHOSPHATASE Since the discovery of spleen acid phosphatase in 37 1954 , attempts to obtain the homogeneous enzyme have been many and varied. Most noteworthy were the efforts of Glom- 41’43, who reported the first homogeneous set and coworkers preparation and also the first indication of a violet colora- tion. Their method involved the pH 3.0 extraction and am- monium sulfate precipitation steps of Singer and Fruton, followed by TEAE—cellulose chromatography; preparative electrophoresis was used as the final purification step. No yield data were reported. Further refinements were not reported until 1973 when 8 described the use of carboxymethyl- Campbell and Zerner cellulose in place of the TEAE-cellulose in the first chromatography step following ammonium sulfate precipitation. A final chromatography step using Sephadex G-75 gave a pro- duct that was 40% pure in an 8.3% overall yield. The same group reported an improved method in 19789 that gave 5-10% overall yield using a batch extraction onto carboxymethyl- cellulose followed by cellulose phosphate and G-75 chroma- tography. This method gave electrophoretically pure enzyme 19 20 similar to the material obtained by Glomset43. A major drawback to all of the previous methods is the poor overall yield of enzyme (typically less than 1 mg per kg spleen)9. By combining conditions and improving upon those techniques, as well as introducing new purifica- tion steps, a significantly improved purification procedure has been developed. Yields consistently approaching 50% overall and 8 mg per kg spleen have been obtained using the method described below. Materials and Methods Chemicals Cellulose phosphate (P-11) and carboxymethylcellulose (CM-52) were manufactured by Whatman and obtained through the Anspec Co., Inc. Gel permeation media (Sephadex G-75 and G-100) were obtained from Pharmacia and hydroxylapatite (BIO—GEL HTP) was purchased from Bio-Rad Laboratories. Ultrafiltration of CM-52 eluate was done on an Amicon 8MC unit using a 25 mm PM—10 membrane filter, while dialysis was done with cellophane dialysis tubing obtained from VWR Scientific, Inc. Bovine albumin, egg albumin, d-chymotrypsi- nogen, B-lactoglobulin, and lysozyme were obtained from Sigma as were 2-(N-morpholino)ethanesulfonic acid (MES), phenylmethylsulfonyl fluoride (PMSF), and PNPP. Reagents for gel electrophoresis were purchased from Bio-Rad 21 Laboratories as a kit. Other, more common reagents were obtained from a variety of sources and were all ACS reagent grade or better. Spleens were obtained from the local abattoir within three hours of slaughtering. Analytical Methods and Instrumentation The assay method of Campbell, s; sl.9, was modified by substitution of MES for cacodylate buffer and used for all activity measurements. Protein concentrations were deter- 56 mined by the method of Lowry st a1. using the coefficient of 1.59 reported by Campbell 33 31.9. Activity assays were done in duplicate. Fresh spleens were assayed by excising l g from the thickest portion of the organ followed by grinding in 10 mL of 0.25 M KC1 in a 15 mL tissue grinder. The samples were then adjusted to pH 3.5 with 6 M HCl, centrifuged for 5 min in a bench tOp Waco Separator and assayed for phosphatase activity. Normally spleens with less than 20 units per g were discarded. Assays were performed and electronic spectra obtained on a Cary 219 or on a Beckman DU spectrophotometer equipped with a Gilford Model 252-D accessory. Gel scans were performed on the DU using a Gilford gel scanning accessory. Gel electrophoresis experiments were done using a Bio-Rad Model 220 cell with Coomassie blue development of protein bandsS7. Native protein gels were 7.5% polyacrylamide with 22 3% of the monomer asbfisacrylamide crosslinking agent. These gels were topped by a 1 cm stacking gel of the same composition but containing 5% total polyacrylamide. Stack- ing gels were cast in a pH 7.0 acetate buffer and running gels were cast in a pH 4.3 acetate buffer. Ammonium per- sulfate and N,N',N",N"'-tetramethylethylenediamine (TEMED) were used as co-catalysts for gel formation. Immediately after casting, n-butanol was layered over the surface of the forming gel to insure a sharp, square interface. Sam- ples contained 0.25 M KC1 and 0.01 M acetate buffer, pH 5.0; typically 2.0 mA per well were applied until the sam- ple had entered the gel and then the current was increased to 4 mA per well for the balance of the run. Voltage ap— plied was 150 volts initially and rose up to 200 volts during the experiment which was run at constant current. The running buffer was B-alanine (0.5 M) adjusted to pH 4.5 with glacial acetic acid. SDS gel electrophoresis was performed in the same manner as the native protein electrophoresis except that running gels were 15% polyacrylamide, gel buffer was 1.4 M Tris- HCl (pH 8.8 - 0.4% SDS), and running buffer was 0.025 M Tris, 0.192 M glycine (pH 8.3 - 0.1% SDS). Protein samples and standards were prepared by making them 2.5% in SDS and 1.0% in B-mercaptoethanol followed by heating for 5 min in a boiling water bath. 23 Extraction and Purification of the Enzyme Beef spleen strips (1 cm x 5 cm x 15 cm) were suspended in 0.25 M KC1 (2 mL/kg spleen) at 4°C and homogenized in a Waring blender at medium speed for one minute and high speed for two minutes. The resulting homogenate was ad- justed to pH 3.5 with 6 M HCl and stirred at room tempera- ture for a minimum of 3 h and a maximum of 20 h. The ex- tract was then centrifuged at 9000 g for 10 min at 20°C. The supernate was filtered through glass wool, made 0.1 M in ASC, and adjusted to pH 5.5 with 12% NaOH. The solution was then treated with l g of P-ll per 2000 units of enzyme. After stirring for 15 min, the P-11 was filtered, washed with 2 x 100 mL of H20, resuspended in a minimum of 2.0 M KC1 (usually 200-300 mL), and stirred for 2 h. The P-ll was filtered and washed with 50 mL of 2.0 M KC1, resulting in a pale yellow-green solution. The combined filtrate and wash were dialyzed overnight against distilled water, cen- trifuged at 9000 g for 20 min at 20°C, and adjusted by dilu- tion to 0.1 M KC1 or less. All subsequent operations were carried out at 5°C in a cold cabinet. The clarified solution was loaded onto a carboxymethyl- cellulose column (2.5 x 20 cm) at 6.0 mL/min. The column, which exhibited an intense purple band, was washed with 100 mL of 0.15 M KC1 and 0.05 M sodium acetate, pH 5.0, and eluted with a 0.15 to 1.0 M KC1 gradient in 0.05 M sodium acetate buffer, pH 5.0. The KC1 concentration was 24 monitored by measuring the conductivity of the sample and comparing the results to a set of KC1/buffer standards. Fractions with A280/A550 < 40 were pooled, concentrated to 5 mL by ultrafiltration, loaded onto a Sephadex G-75 column (1.5 x 85 cm), and eluted with 0.2 M KC1 in 0.05 M sodium acetate buffer, pH 5.0. The KC1/buffer used for this column and the subsequent hydroxylapatite column was run through a 1.5 x 15 cm bed of Chelex ion exchange resin (Bio-Rad Laboratories) to remove any transition metal con— taminants. The purple fractions with A280/A < 25 were 550 combined and pumped through a 1.5 x 3 cm bed of hydroxyl- apatite previously equilibrated with the G-75 eluent. A greenish-yellow band appeared, usually occluding the tOp 50% of the column. Residual purple acid phosphatase was recovered from the column by washing with 10 mL of 0.75 M KC1 in 0.05 M acetate buffer, pH 5.0. The resulting product was normally concentrated to %8 mL and typically had an A280/A550 of 15-17. Molecular Weight Determinations Molecular weight data were obtained by gel permeation chromatography on Sephadex G-75 with bovine and egg albumin, a-chymotrypsinogen, and 8-1actoglobulin as standards. Pur- ple acid phosphatase obtained by the usual method as well as a sample purified in the presence of 1 mM PMSF (to determine whether or not proteases in the crude extract 25 were affecting the molecular weight) were run on the G-75 column. The G-75 results were confirmed by SDS gel elec- trOphoresis data using the same standards with the addition of lysozyme. Results Purification This improved purification method for beef spleen purple acid phosphatase results in a 50% recovery of enzyme puri- fied to homogeneity in only five steps. Table 2 displays typical yield and recovery data for each step. The use of a preassay to screen individual spleens for activity com- bined with improved acid extraction conditions increases the initial amount of activity obtained. Spleens ranged from 1 unit/g to over 100 units/g with an average of 20-25 units/g. It should be noted that dairy cow spleens appeared to have 3 to 5 times the amount of enzyme per g that was found in steers, bulls or calves. Figure 1 shows the elution profile of the purple phos- phatase activity profile. The large 430 nm peak is due to the green heme peroxidase reported in Part II of the dis— sertation. It copurifies with the purple phosphatase prior to CM-52 chromatography. Sephadex G-75 chromatography is shown in Figure 2; the absorption at both 550 nm and 280 nm is plotted. The small 26 Table 2. Purification of Purple Acid Phosphatase From Beef Spleen. Specific Yield % Activity Step (Units) Recovery Units/mg Acid Extraction (1000 g Spleen) 20,000 100 0.24 Cellulose Phosphate (P-ll) Batch Adsorption 17,000 85 32 Carboxymethyl Cellulose (CM-52) Column 13,600 68 480 Sephadex G-75 Column 11,600 58 1110 Hydroxylapatite Column 10,000 50 1240 27 96% An: 3939... $32893 eat 8V a: omm and £9 one :2 0mm um :03 Imuomnm nuw3 A cozam e .o. x 02.4 . O 8N4 . op. - I . a ll.'|lll .IQd .lmd .IQO SS. 38; . IION .IO€ 29 . AOL E: omm paw AIV omm um coHumHOmnm nufl3 mnnu xmpmsmmm co mmlzu Eonm mcoHuomuw OHQHSQ may mo xzmmumoumfiounu .N wusmwm 30 no.0 0004 0.6 N musmflm 4E .mE:_o> cozam Om. ON. a fl ‘ _\ Om O.— OmN< O.N 31 peak at 105 mL elution volume is due to the green heme protein impurity which has a significant absorbance at 550 nm. Hydroxylapatite chromatography of the purple acid phos- phatase obtained from the G-75 step was successful in puri- fication of the enzyme, but activity losses were unaccept- able. As an alternative, a small bed of HAP was used to "polish" the purple phosphatase, successfully removing re- maining impurities with m90% recovery of phosphatase activity. Protein Assay Protein assay by the Lowry method56 , using bovine serum albumin as a standard, consistently gave values of protein concentration (mg/mL) 1.6 times the value obtained using the molar absorptivity at 280 nm or 550 nm reported by Camp— bellg. This is probably due to the difference in tyrosine content between the two proteins and is mirrored in the dif— ference in the absorption at 280 nm of 1 mg/mL solutions of the two proteins. The gravimetrically determined 8 found 9 550 by Campbell provides a better estimate of protein concentra- tion and has the advantage of being relatively specific for the purple acid phOSphatase (as was used for most estimates). 50 Molecular Weight Data and Electrophoretic Purity Two methods were used to estimate the molecular weight of the purple acid phOSphatase. Gel permeation chromatography 32 on Sephadex G—75 gave a molecular weight of 40,000 for the native enzyme. Figure 3 is a plot of log molecular weight * VS. K 58 av for four molecular weight standard proteins and the corresponding Kav values for purple acid phosphatase prepared in the presence and absence of PMSF. Figure 4 shows SDS gel e1ectr0phoretic data plotted in a similar fashion, where dS/df is the ratio of sample migra- tion distance to the distance of the front from the origin. The presence of two principal bands for the purple phospha- tase at positions equivalent to molecular weights of 24,000 and 15,000 in roughly equal proportions is denoted by the 0 marks on the plot. This is consistent with information published by Campbell, st s1.9. Native protein gel electro- phoresis of the purple acid phosphatase followed by Coomas- sie blue staining showed a single protein band although the band was somewhat broader than the standard proteins - pos- sibly indicating the presence of isoenzymes. A 600 nm scan using a Gilford gel scanner followed by peak integration showed the purple phosphatase represented >95% of the protein present. * A number related to elution from a column by the equation: K = ve - V0 av Vt - VO where Vt is the total column volume, V is the column void volume, and Ve is the elution volume 0 the leading edge of a peak at 1/2 height. 33 .xmmm comm mo mmpm mcflwmma mnu mo macaw one mo ammoumucfi >2 pecaeumuwp mwz AmEsHo> cofluSHmv > .cHEBQHm_Eusm mcw>on .0 kneadnam>o .O “cmmocammwuuoe>£ola .m “ceasnoamouomalm .6 ”mnm3 mpumpcmum .Aao meadow mo outmoded .ue\me .mom wa.o .xm.m mac mahosHm z mmH.o .mehe z mmo.o .hdmmsn scented “mom we.o .Am.m mac Hum Image 2 v.H .ummmsn Hmm «Hem mcflxomum wm .me prEmenom wma “Oov “Hmccmno \«E v “mmmcx0flnu How 68 m.H "muwz mcofluflpcoo oaumuonmouuowam .cHESn lam Esuwm mcfl>on .m ucflESQHm>o .Q “cwmocflmm>HMOE>£OId .U “pHHSQOHmouOMH um .m “mewnomma .< "mH03 mpumpcmum .An: comamm moon Eoum mmwumnmmonm mamuzm mzu mo unmflmz HMHSOOHOE may mo coflumeflumw mammuonmouuowam melQO .v musmflm 36 v musmfim To. x 50.63 8.3862 mm m. _ - O. 0.0 II the .x 37 Discussion The improved purification for spleen purple acid phos— phatase has a number of advantages over previous methodology. By far the most important is the improved yield (8 mg of homogeneous enzyme per kg of spleen) over that of the pre- vious procedure (<1 mg/kg)9. This remarkable improvement is not due to a single feature but derives from three separate modifications. First, using both 0.25 M KC137 and pH 3.538 (thus combining conditions used by previous researchers) increases the yield at the extraction step by up to three-fold. These conditions also result in faster extraction although stirring for up to 20 h does not af- fect the yield. By preassaying the spleens and using only those with average or above enzyme content, it is possible to obtain an additional two to three-fold increase in ex- traction stage yield, due primarily to the large variation in enzyme content from spleen to spleen. Finally, the use of cellulose phosphate as a batch adsorbent, replacing the more conventional ammonium sulfate precipitation, not only is more convenient but also increases yield slightly at the second step. Interestingly, cellulose phosphate P-l, a more fibrous product also manufactured by Whatman, is totally ineffective as a batch adsorbent even though the phosphate loading and chemical structures are similar. It is not known if this behavior extends to other cellulose phosphate pro- ducts. A batch extraction method using carboxymethylcellulose 38 was alluded to by Campbell 23.31f9 but has apparently never been published. The 130 fold purification achieved through the use of cellulose phosphate P-ll is surprisingly good for such a crude method. It is possible that a form of affinity adsorption of the protein involving the phosphate groups is reasponsible for the specificity in addition to the highly basic nature of the protein. The loss in ef- fectiveness of P-ll with repeated use as an adsorbent for the purple phosphatase supports the affinity absorption theory. The dialysis following the batch adsorption step serves two functions. First, dialysis yields a much lower volume than direct dilution and allows a more convenient means of achieving the desired low KC1 concentration. (The lower the salt concentration the sharper the bands on the CM-52 column and hence the better the purification upon elution.) The second advantage is the precipitation of large amounts of contaminating protein during dialysis, resulting in some degree of purification. It is not clear whether or not a rapid dialysis system such as a hollow fiber dialyzer would be an advantage as the protein precipitation may be a time dependent phenomenon. The final purification steps are common chromatographic procedures and give the expected degree of purification. The recovery in the final hydroxylapatite step is dependent upon the size of the sample, however. Small quantities of 39 protein (15-20 mg) will usually result in recoveries of about 85% while larger lots (50-60 mg) give better than 90% recovery. This is probably an indication that the losses are primarily due to handling with some minor loss due to adsorption or denaturation. The molecular weight of the beef spleen acid phospha- tase has been shown by this work to be m40,000 daltons. The protein is composed of two non-identical subunits of molecu- lar weight 24,000 and 15,000. No significant changes in these values are observed when 1 mM PMSF (a serine protease inhibitor) is maintained throughout the preparation, sug— gesting that enzymatic degradation of the protein during purification is not occurring. The enzyme is unusually stable during purification (pH 3.5, 24°C for up to 20 h) and is resistant to damage during repeated freeze-thaw cycles. This evidence does not, however, rule out the possibility that the presence of two subunits is due to acid hydrolysis of a sensitive peptide during the extrac- tion step. These data substantiate previous findings9 where es- sentially identical values for A280/A550 (15.1), molecular weight of holoenzyme (38000), and subunit molecular weights (15,000, 26,000) were reported. The improved purification provides a ready supply of enzyme for further studies in- cluding kinetics, spectroscopic and magnetic measurements. CHAPTER 2 MOLECULAR PROPERTIES OF BOVINE SPLEEN PURPLE ACID PHOSPHATASE Kinetics studies are commonly used to determine the mechanism of enzymatic action as well as to elucidate the nature of active sites of enzymessg. Other studies that are necessary to understand the properties of the active site of a metalloenzyme are determination of the metal content and presence of bound moieties such as phosphate. This chapter describes kinetics studies with several inhibitors of purple acid phosphatase, detailed metal assays, determination of bound phosphate, and an examination of the glycoprotein nature of the enzyme. As mentioned in the literature survey, Campbell s; 31.9, have reported that both the bovine spleen and porcine uterine enzymes contain two iron atoms per 40,000 molecular weight protein. Schlosnagle et al.49, and Antanaitis and Aisen50 , however, have found only one iron per 35,000 molecular weight porcine protein. Certainly careful metal assays are required for the beef spleen enzyme to validate the results found by Zerner's group. The role of the iron in catalysis has not yet been determined for the beef spleen enzyme and is also in doubt for the porcine 40 41 enzyme. Clearly, detailed kinetics experiments can show the probable mechanism of the hydrolysis and may, by use of appropriate inhibitors, implicate the iron in catalysis. Some kinetics data is available for comparison from studies on the porcine enzyme. Materials and Methods Reagents for phosphate analysis were obtained as a phosphate assay kit from Sigma Chemical Co. The assay is based on the method of Fisk and SubbaRow60 61 as modified by Bartlett , except that the ashing method of Van de Bogart and Beinert62 was used. Gel electrophoresis was done as in Chapter 1 but a BioRad Model 155 disc gel electrophoresis unit was used and gel concentrations were as noted in Figure 8. [p-(Aminoacetyl)benzle-phosphonate (PABP) was prepared 63 1 by S. Lin and characterized by H NMR spectrosc0py on a Varian T-60 instrument. p-Nitrophenol was obtained from Eastman Chemical Co. and recrystallized by S. Lin63. Kinetics data were obtained by using the standard p-nitro- phenylphosphate substrate method of Campbell s; 31.9, modified as noted in Chapter 1 of this dissertation. Data were taken on a Beckman DU spectrOphotometer equipped with a Gilford Model 252D accessory. Some kinetics data were obtained on a Cary 219 Spectrophotometer. All inhibitor solutions (except sodium cyanide) were adjusted to pH = 6.0 immediately prior to use. 42 Iron assays were run according to the method of Van de Bogart and Beinert62 or by a modification of that method using 1,10-phenanthroline and eliminating the isoamyl- alcohol extraction step. Analyses for iron, copper, man- ganese, molybdenum, magnesium, zinc, nickel, and cobalt were done by inductively coupled plasma emission spectros- copy on a Jarrell-Ash Model 955 plasma atomcomp, using the G-75 and hydroxylapatite elution buffer as a blank solution. Sialic acid glycoprotein linkages were hydrolyzed using neuraminidase from Clostridium perfringens. The enzyme (grade IX) was purchased from Sigma Chemical Co. Gel electro- phoresis molecular weight distribution studies of the native enzyme were done with a neuraminidase-treated sample (24 h, 25°C, pH 5.0) and a blank enzyme sample containing all buffers and reagents except neuraminidase. Results Metal Content Duplicate samples of the purified enzyme were used to determine the metal content. The chemical method gave a value of 2.2 iron atoms per 40,000 molecular weight, while the plasma emission method gave 2.0 iron atoms per mole. Magnesium was undetected at a detection limit of 0.4 mole per mole enzyme, while copper, manganese, molybdenum, co- balt, and zinc were undetected at 0.1 mole per mole of 43 enzyme. Nickel was also undetected at that level, but poor linearity for that metal assay prohibited a closer ap- proximation of the nickel content. However, substantial quantities (>0.1 mole per mole enzyme) of nickel are not likely to be present. Kinetics Data Figure 5 shows Michaelis-Menten behavior for the purple acid phosphatase using p-nitrophenylphosphate (PNPP) as substrate. No inhibition by 10 mM cyanide or azide was ob- served; the cyanide data are corrected (by apparent extinc- tion coefficient of PNPP) for a pH shift of 0.3 pH units caused by high concentrations of cyanide. (No significant change in activity is observed over that pH range.)39 Values of Km and Vfiax of 3.0 mM and 580 sec-1 were cal- culated for PNPP. Data for p-nitrophenol do not show in- hibition even at 5 mM (limit of concentration due to high absorbance at 410 nm); however, phosphate does show strong product inhibition with Ki = 3.6 mM. The data in Figure 6 show that PABP, a substrate-like molecule with a carbon-phosphorus bond instead of the hydrolyzable oxygen-phosphorus bond in the substrate, is a potent competitive inhibitor with a Ki of 1.9 mM. Addi- tional data in Figure 6 show that fluoride exhibits complex inhibition behavior, changing from apparently competitive inhibition at fluoride concentrations below 1 mM to more 44 .21 632.. odd .1! 63:88 .on mumcmmocm SE oH nufl3 ppm macaw Axv me>wcw m>flumc mum c3osm .mumuum Insm mm mumnmmonm HmcmzmouuHCIm cue; pump mOADmcflx mo mpoHQ xusmlum>mmzwcflq .m musmflm 45 m musmflm m“.0. x 295.. no. O. m 0 m- _ _ _ , 1 . M. . . .1 ON a 1 CV 2061?? 1 cm b. .ucm @282 . _ a“ E w“ .. I O -nvod as o. o o I om l 00. 46 . HOV SE OH um coHananH mmm ppm ADV SE H paw . 2V m.~ . A3 o . A! 0H um coHUHQanH opHuosHm mum :3onm .mumuquSm mm mumnmmonm Hhcmnmouucha nuHB dump mOHuwcHx mo muon xusmlnm>mw3mcHH .m musmHm 47 w wusmHm ».O_ x o_oE\._ % N T _ d . 00. \ 22: 1\c_E I On. kw: u... .28 _ a nu 2E 0N4 :1 SE 0 o I L. SE 0. I CON mmdm SE 0. o 48 complex inhibition behavior at concentrations of 2.5 mM and higher. Figure 7 shows the inhibition behavior for molybdate, arsenate, vanadate, and tungstate. Molybdate and tungstate show similar highly potent competitive inhibition with Ki = 5 uM and 6 uM respectively. Arsenate is a weaker com- petitive inhibitor (Ki = 0.24 mM), but is still stronger than phosphate (Ki = 3.6 mM). Vanadate also shows com- petitive inhibition with Ki = 3.2 mM. Inhibition constants, substrate and inhibitor concentrations, and percent inhibi- tion data are summarized in Table 3. Reactivity of Purple Acid Phosphatase with Neuraminidase The gel scan in Figure 8A shows two major peaks at ap- proximately 39,000 and 40,000 molecular weight. Prior to Coomassie blue staining, two faintly purple bands were ob- served in the same positions, indicating that both bands are due to the enzyme and not to an impurity. Since the enzyme is known to be a glycoprotein9'4l, the effect of neuraminidase (an enzyme that cleaves sugars from proteins at sialic acid linkages64)on the molecular weight and on the distribution of the two species was examined. Figure BB is the gel scan of a sample identical to that of Figure 8A, but treated for 24 h at 25°C with one unit per mL of neuraminidase. The shift of the bulk of the higher molecular weight peak to the same position as 49 e . Adv umo> :5 o4 to? out .1; umoma och . Os smog . ADV Imooz S: OH “~qu ppm mcon ADV 953cm w>Humc who c395 .wpmuumnsm mm mumnmmosm chwnmouuHCIo SHHB mumo mOHpmcHx mo muoHo xusmnnm>mw3mcHH .b musmHm 50 0. Nb. x was) I h OHSon m _ 22 .103 IMOmd. o 2250.: -mo> e :2 105M015 o Eiocmodz a 9:35. $.62 b O? om 39: SSE: > Om 00_ ON. 51 Table 3. Inhibition of Beef Spleen Purple Acid Phosphatase. % Inhibition Inhibitor [I] [s] = 1 mM Ki poi” 10 mM 74 3.6 mM PABPZ- 10 mM 75 1.9 mM F- 10 mM 82 3.0 mM woi' 10 uM 73 6 uM Mooi- 10 uM 76 5 uM A302- 10 uM 31 0.24 mM voi' 1 mM 37 3.2 mM 502', CN', N3 10 mM 0 -—— php 5 mM 0 --- D,L-Tartrate 40 mM 0 -__ 52 .m.e mo .mcHdemum z m.o ummmsn mCHccsu “ochcouImumumom m.q mm .oconmumeumumom o.h mm "whom Imsn Hmm “Hwo mconmum wm .Hmo prEmH>uom wm.h “Oov «Hmccm20\HHmc mnu .mmoumnmmoso pHom wHouso one mo muuowom :oHumuomnm mHQHmH> .OH muson 69 8» OOO OH wedded :5: a ooh OOV OOm _ _.o No MO v.0 70 oumnuoomm ppm :oH moonumm cuH3 :oHuoopmu one was ASHE om . oH ..... mum SBOSm .TIIIV o u u pm 1...... SEE .IIIIV :oHuoopwu HouHHnumumoHnqu mo mmusoo wEHu may .ommumnmmoso pHom wHonsm on» mo muuowom :oHumHomnm anHmH> “CHE o .HH musmHm HS modded f . . s M 1.... .... . ,, J). I... . I .1 3, 5....» .1. u... a) w) TIC {Knock}... MCfQ IKCHU (CACH- {\OM . :loS x T... _ III, I, 9: x........ O_\_ X II.I. 71 72 2 is also shown native form with 0.4 M ASC and 0.08 M Fe+ in Figure 11; reduction under these conditions is complete within one minute. When Pi (10 mM) is added to the pink form of the enzyme, an immediate shift to the purple form is observed. However, under anaerobic conditions, addi- tion of phOSphate does not result in this shift. Reduction 2 as before, of the enzyme to the pink form with ASC/Fe+ followed by addition of 1.0 mM vanadate results in the spectrum shown in Figure 12. Unlike Pi’ the vanadate does not result in a shift to the purple form under aerobic con- ditions. Figure 13 shows the anaerobic titration of the native enzyme, with MV+ as reductant, monitored over the range 400 to 700 nm. The visible absorption peak shifts slowly from 550 nm for the native enzyme to 505 nm for the active reduced pink form; full conversion to the pink form is achieved at 1.1 electrons per two iron atoms. Addition of more than 1.2 electrons per two irons resulted in tur- bidity, caused by precipitation of protein, which obscured the spectra; however, the pink color (as determined visually) was almost entirely eliminated at 2 electrons per two iron atoms. At 2.2 electrons per two irons a blue color appeared, presumably due to a slight excess of reduced methyl violo- gen radical cation. 73 Figure 12. Visible absorption spectra of the purple acid phosphatase. Shown are the native enzyme ( ----- ), enzyme reduced with ferrous ion and ascorbate ( ), and enzyme reduced with fer- rous ion and ascorbate with 1 mM VOi‘ added (----) o 74 0.2 - ().. 4oo ' 1 500 600 Mnm) Figure 12 75 Figure 13. Visible absorption spectra of the reductive titration with methyl viologen radical ca- ion of bovine spleen purple acid phosphatase. 76 FePose reductive titration 0a 02 l ‘L 400 500 600 kJm1 Figure 13 700 77 Resonance Raman Spectroscopy of the Oxidized (Purple), Reduced (Pink), and Phosphate-Inhibited Forms of the Enz e Laser excitation (514.5 nm) of the enzyme results in a resonance Raman spectrum; this has been measured for the oxidized, reduced, and phosphate-inhibited forms of the en- zyme. Figure 14 shows the resonance Raman Spectra of the three forms of the purple phosphatase; the four intense bands between 1100 and 1620 cm.1 are typical of tyrosyl vibration mode516'52'71. The shift to slightly lower 1 band is the only wavelength and broadening of the 1286 cm— major change in the Spectrum upon reduction; addition of phosphate does not alter these effects and does not lead to the appearance of any new features. Two strong bands are also seen in the 500 to 600 cm.1 range. EPR Spectroscopic Studies of the Pugple Acid Phospha- tase from Bovine Spleen The oxidized (purple) form of the beef spleen enzyme is EPR silent even at liquid helium temperatures and rela- tively high power settings (10 mW). The spectrum obtained under those conditions is shown in Figure 15 (spectrum A). Also included in that figure is the spectrum of the pink (reduced) form of the enzyme (spectrum B), which shows a complex Signal centered at g = 1.77 and a sharp g = 4.3 signal. The 9 = 1.77 signal is highly temperature sensitive, 78 .n3onm ouo AU Esuuoooov oumnomono SE 0H mSHm EHOM pooooon onu pno .Am Eduuoommv Euom poOSpoH .nm Eduuooomv oESNno o>Huon one .nooHom onH>on Eoum omouonm Imono pHom oHouso mo AES m.VHm n coHuouHoonV ouuooom noEom oonon0mom .VH ouoon 79 Vom _ OOV _ ooh owo _ _ _km _ . _ _ mmm mkm OON 00m 00m _ . mom who SH dhsoem OOO. _ oo__ oom_ oop 83 com; ooo. oot _ So: now _d SEQ 83 868310 _ . . _ . . _ so: mow. _ . hom_ mow see amused £2 888m .m _ . . . . _ _ so: \KJPXS homo moo. ohm. . 8353 o>zoz .4 80 .Ao.mv moa x m.m ho lac men x m .dth unmadhumde “Dom OON .oEHu noon .0 OOOO .oonou nooo “O OOm.~ .uom pHon “Dom N.O .ucoumnoo oEHu “O OH .opsuHHmEm coHuoHDOOE “me OOH .xonosoonm COHuoH lopOE “3E OH .uoBOm o>o30HOHE ammo bv.m .Sonosooum o>o30HOHE "mnoHqu Inoo OCH30HH0m onu Hops: HouoEouuooom DOONmm Hoxsum o no pocHouno ouo3 ouuooom .O.m mo .Hommsn ououooo EsHpom SH AHE\oE my nHououo SE m~.O ponHounoo moHoEom HH< .O.m mo .ouonooono ESHpoo SE OH HO EoHqupo Sn poBOHHOM m Hem mo pouoooum oHoEom .O Eouuoomm “O.m mm .ouonuoomo SE OOH out Aeomvmomnvmzc :5 e.H tuna meshtm odhooexo oeeummuu an tweet Iouo EHOM Aanov poospou .m Esuuoomm “EACH AonuomO poNHpon .< Eduu Iooom .ooouonmmonm pHoo oHoqu noonm moon mo M m.v um ouuoomm mom .mH ousmHm 81 mH ouson 82 broadening above 12 K and unobservable above 30 K. The 9 = 4.3 signal varies in intensity (but never accounts for more than 5% of the total iron) and is absent in samples prepared by slow reduction with DTE alone; this signal may be due to slight air oxidation of the ferrous ion in the ASC/Fe+2 mixture used to reduce the enzyme rapidly. Addition of 10 mM phosphate to a sample identical to that of spectrum B results in loss of the g = 1.77 signal (spec- trum C), as does addition of phosphate to enzyme reduced with DTE alone. This is entirely consistent with the optical spectral changes reported above. Figure 16 shows the high field portion of Figure 15, spectrum B(spectrum 16 A); double integration of the spectrum gives 0.9i0.l Spins per molecule (i.e., one spin per two iron atoms). This signal is very similar to that reported for the purple and pink 50 and for the beef spleen enzyme prepared without the use of HAPSl. forms of the porcine uterine fluid enzyme Both the overall appearance and the g values of the prin— cipal features of spectrum 16 A are in close agreement with those reported for the porcine and beef spleen enzymes. Careful examination suggests that both the g = 1.77 and g' = 1.74 (porcine enzyme) spectra are composed of at least two overlapping specieSSI. Because the purple to pink conversion in the porcine enzyme has been attributed to reduction of one or more disulfide links resulting in a conformational change about Figure 16. 83 EPR spectra at 4.2 K. Spectrum A, pink form of beef spleen acid phosphatase produced by Fe2+/ascorbate treatment (same sample as in Figure 15,Spectrum B); spectrum B, oxidized enzyme (0.23 mM) plus 1 equivalent of NaZSZO4 in 0.073 M Tris-HCl buffer, pH 8. Conditions of EPR spectrometry as in Figure 15, except microwave power, 5 mW (A) or 2.5 mW (B), field set, 4,000 G; scan range, 2,000 G, instrument gain, 4 x 105 (A) or 6.4 x 105 (B). :('J‘fi[\\, 84 L92 I77 LS5 ' Figure 16 ZOO G ASC/Fe2+ ZOO G DTH MNJJI'W LOO 85 the iron, and because the purple form of the bovine enzyme was EPR silent in our studies, it was necessary to investi- gate the stoichiometry of the reduction carefully. Ac- cordingly, an aliquot of carefully standardized DTH solu- tion containing 0.5 equivalents of DTH (i.e., 1.0 electron per two iron atoms) was added anaerobically to the purple form of the enzyme, giving the EPR spectrum shown in Figure 16, spectrum B. Although spectrum B is not identical to that in spectrum A, a remarkable similarity is apparent; minor differences in g values and in the relative intensities of the peaks can be attributed to the higher pH required for DTH reduction (8.0 vs. 5.5), as shown below in the pH dependence studies of the spleen purple phosphatase EPR spectrum. Double integration of the signal in spectrum 16 B gives 0.8i0.1 spins per two iron atoms. Both the ASC/Fe+2 reduced and the DTH reduced enzyme spectra are consistent with formation of a new paramagnetic species upon one electron reduction of the two iron unit. The power saturation curves for the two major peaks (9 = 1.92 and 1.77) in Figure 16 spectrum A are shown in Figure 17; the different saturation curves indicate a contribution to one of the peaks from another species (see Figure 20). Figure 18 illustrates the effect of temperature on the g = 1.77 complex signal at pH 5.5. An aerobic titration similar to that depicted in Figure 13 was conducted and followed by EPR spectra. Thus EPR 86 .Oom N.O .unoumnoo oEHu “O OH .opSuHHQEo noHuoHSpOE “me OOH .Sonosooum :oHuoHSCOE uwmo hv.m .mocoso Ioum o>o30uoHE "mnoHqunoo Houuooom .nuv xoom hh.H n m on» now new Ans noon mm.H n m onu How mo>uso ono n3onm .ommuonmmonm pHoo oHQHSQ AxcHoO poospou mo Esuuoooo mom M N.v onu HON mo>uso :oHuousuom Hozom .hH ousmHm 87 SH ouson :2: .mmeodc m_oo_oo_u .cozoacoth .5on 88 8: 3; 8O 3 3.9 Nov , o. 9 ON mm on . . _ _ a _ _ O . .. m Elna. a n. 5. 8.. no 6 II. H In Elli... .V xxi elull 88 Figure 18. Effect of temperature upon the EPR spectrum of reduced (pink) enzyme. Power and gain are as shown, other spectral conditions are as in Figure 15. . 89 A. 4.4 K 5mW gain 8 x I05 2000 l ' WW I L76 |.9l B. I4 K g I 5mW ‘ L7 goin 2 x l05 4 I | l.7| '6' 3800 0 Figure 18 90 samples of enzyme reduced by 0.0, 0.3, 0.6, 1.0, 1.3, and 2.0 electrons were prepared. Results are shown in Figure 19, where maximal absorption at g = 1.77 occurs upon ad- dition of m1.0 electron, thus confirming the visible spec- tral titration study and showing the g = 1.77 EPR signal and the pink color to be associated with a one-electron reduced two iron form of the enzyme. Table 5 lists the results of double integration of each sample, as well as the approximate wavelength of the visible absorption maxima from Figure 13. It is clear that changes in Amax of the Optical Spectra upon reduction parallel the increases in intensity of the EPR Spectra. The effect of pH on the g = 1.77 EPR signal is illustrated in Figure 20 for pH 3.5, 4.5, 5.5, 6.5, 7.5, 8.5, and 9.6. At the higher pH values protein precipitation occurs and samples, even when frozen in less than 10 seconds in an isopentane (113 K) slush, show lower integrated intensities than expected. Of particular interest, however, is the apparent shift in population of the EPR signals with pH. The dominant form at low pH is a rhombic signal with g values of'gX = 1.92, gy = 1.77, and g2 = 1.63. The EPR signal that strengthens at high pH appears to be axial with 9|) t 1.62 and gi = 1.74. It is not clear what the under- lying cause of this shift in population may be, but de- protonation of a metal ligand may be responsible. Antanaitis and Aisen51 have reported that the purple 91 .mH ousmHm CH mo ouo mnoHUHpnoo Houuooom .noonm moon Eoum omouonomonm pHoo onuso mo :oHuouuHu o>Hu05poH onu mo muuoomm mom .mH ousmHm 92 OH ouson O Comm _ -oONm we... 6 m._ o -a o._ a -a ooo .23 .m g to O.< T11 0 00m . cozotfi 9:881. omedoh. 93 Table 5. Double Integrated Values for the EPR Spectra of Reductive Titration of Purple Acid Phosphatase with Shift in Absorption Maxima Shown. Sample Integration Amax 0 e- 0 spins 550 nm 0.3 e‘ 0.26 spins 538 nm 0.6 e’ 0.56 spins 527 nm 1.0 e- 0.97 spins 510 nm 1.3 e- 0.74 spins --- 2.0 e' 0.0 spins --- Figure 20. 94 Effect of pH upon the EPR spectrum of the purple acid phosphatase from bovine spleen. Spectral conditions as in Figure 15. Samples were prepared by reduction with dithioeryth- ritol and Fe+2 (pH<7) and with dithioerythri- tol alone (pH>7) followed by freezing in liquid N2 after one hour at room temperature. mi35 pH45 md55 MiSS deS mi85 L92 I78 95 l LBS | 40856 Figure 20 2006 96 phosphatase from beef spleen exhibits a g = 1.74 EPR signal in both the oxidized and reduced forms. This is similar to the results from the same workers on the porcine en- zymeso; however, the beef spleen enzyme used was purified without the use of hydroxylapatite. Omission of the hydroxylapatite step in the preparation gives enzyme that exhibits EPR spectra similar to those reported by Antanaitis and Aisen49 (Figure 21). For comparison, EPR spectra of the porcine enzyme were measured (using a sample provided by Phillip Aisen); about 0.2 spins per two iron atoms were observed (Figure 22) both in the sample as received and after reduction and treatment with Pi' Upon anaerobic re- duction of the beef spleen enzyme by one electron and sub- sequent anaerobic addition of Pi' the enzyme does not under— go the expected shift from the pink to the purple form (Figure 21). EPR experiments corroborate the visible spec- tral observations, showing that more than 75% of the g = 1.77 signal is retained upon anaerobic Pi addition. Addi- 3- 4 results in little change in the optical spectrum (Figure tion of 1.0 mM V0 to the reduced pink form of the enzyme 12). As with anaerobic addition of Pi the EPR signal is retained, but in this case it is dramatically perturbed (Figure 23). Several features are evident and may be due to the 51 V nuclear spin of 7/2. Figure 24 shows the 4.2 K MOssbauer (MB) spectrum of the native purple acid phosphatase using natural abundance 97 .mH ousmHm :H mm mnoHqunoo Houuooom .AO Esuuoooov ouonomonm SE OH mSHo EHOM poospon on» new .Am Esuuooomv Euom Aanmv Godunon onu .AS Esau Iooomv oESNno o>Huon onu ono ozonm .oquooonxoupwn mo om: onu poonuH3 pouomoum nooHQm onH>on Eoum omouonomonm pHoo oHousm onu mo ouuooom mom .HN ouson Hm oHSOHm O mmov _d .250. + 8688. .o . mm. m: ow. _ . _ t o 8% . NH x now heath/a 6886mm 6.262 .4 368358th 505:5 notoooi omooou 99 .mH oHSOHm SH mo mnoHqunoo Hmuuoomm .AU Esuuooomv oumnowono SE OH mSHo EACH poospou onu pno .Am ESE» Iooomv EHOM Aanov po05pou on» .no Esuuooomv oESNSo o>Huon onu oum nzonm .mOHSHm onHHou: onHonom Eoum ooouonmmono pHoo oHouso onu mo oupooom mom .mm onsoan 100 mm ouoon O mmov to oom . .o .289 + ooozpom .O @182 sedated .m 6262 .< mmanomoco manna oESod 101 .mH ousmHm CH mo oum mcoHqucoo Houuooom .poppo Imo> SE H nuH3 noonm onH>on Eoum omouonomono OHoo oHouso on» mo EHOH poospou onu mo Esuuooom mom .mm ousmHm 102 mm ouoon o 8% _ mm.— _ 3 ooom. t ...o> SE 3.86.1885 68861 added 103 .pHon O poo M N.v no cooHom ocH>on Eoum omouonmmono pHoo oHonso mo Esnuoooo HosmnommS .vm ousmHM 104 ON oHSOHm A.omn\EEO E 5_oo_o> m m o _. m- .m- _ . . H . . . _ . . H o: 88.0 .m. who 08.. 6.54 m . .: l NO _.0 IUBOJOd u! 109143 105 57Fe. Quadrupole splittings for the two doublets are 0.56 and 1.16 mm/sec while the corresponding isomer shifts are 0.16 and 0.46 mm/sec, respectively (relative to metallic iron at room temperature). Application of a 320 gauss magnetic field to the sample (4.2 K) does not cause magnetic splitting of the spectrum and increasing the temperature to 146 K causes very little broadening of the absorption peaks (data not shown). Figures 25 A and B Show the magnetic susceptibility of the purple (oxidized) and pink (DTE reduced) forms, respec- tively, as a function of temperature. The purple form is effectively diamagnetic, with only a small residual para- magnetism equivalent to less than 3% of the iron present (assuming high-Spin Fe(III), (S = 5/2). The one-electron reduced form of the enzyme exhibits Curie law paramagne- tism with values somewhat higher than expected for an S = 1/2 spin only paramagnet per two iron atoms. A plot of Xm (corrected for 5% adventitious high-spin Fe(II)) gives a straight line with slope equivalent to a value of “eff = 2.2 “B (Bohr magneton) per molecule of protein containing two iron atoms. No deviation from linearity was observed even at 180 K, suggesting strong antiferromagnetic coupling in both forms of the enzyme. In an attempt to estimate the value of J (the anti- ferromagnetic coupling constant), a 17 mg sample of freeze- dried native protein was run at temperatures up to 300 K on the SQUID susceptometer. The results for the temperature Figure 25. 106 Magnetic susceptibility data as a function of temperature T for various forms of beef spleen purple acid phosphatase. (A) Purple (oxidized) enzyme at 1.5 mM. The solid line shown is cal- culated for 2.5% high-spin Fe3+. (B) Pink (reduced) enzyme at 0.75 mM, prepared by re- duction with 100 mM dithiothreitol for 200 min. The data have been corrected for the presence of 5% adventitious iron (analysis shows 2.1 Fe per molecule), assumed to be high— spin Fe2+. The solid line is calculated for a spin-only S = 1/2 paramagnet at 0.75 mM (using 9 = 1.77). 107 0.l-' A FeFe(ox) 0.05 O- i + 2 L Xp B 0.3- FeFe(red) ; _ ”L I i -l l ' l O 50 . IOO , ISO T(K) '* Figure 25 108 range of 75 to 300 K are Shown in Figure 26, along with plots of susceptibilities expected for -2J = 134, 200, and 300 cm-1. The theoretical values for S = 5/2, S = 5/2 coupling were obtained from the appropriate form of the Van Vleck equation, using a TRS-80-l6K-Leve1 II computer. Equations and computer programs are described in detail in Appendix A. The value of -2J for the native purple enzyme is estimated to be Z 250 cm-1. The reduced (pink) form of the enzyme is estimated to have a -2J Z 100 cm-1; the lower limit for the reduced form is smaller than that for the oxidized, purple form simply because of the increased error associated with measurements on the more dilute frozen liquid samples employed for the reduced enzyme. Discussion The first visible absorption spectrum of the beef spleen purple acid phosphatase was published in 196043, but re- versible reduction to a pink form similar to the porcine enzyme has previously only been mentioned in passing. The visible and ultraviolet spectra Of the bovine spleen purple acid phosphatase shown in Figures 10 and 11 demonstrate that the enzyme can be reversibly reduced to a pink form upon treatment with mild reducing agents. Addition of excess sodium dithionite causes immediate bleaching and complete loss of color and activity. When mild reductants plus ferrous ion are used, conversion to the pink form is 109 .HHHHOoM cHom ann msoHuHuno>pm wm.m mo oonomouo onu MOM pouoouuoo ohm dump 39:8 4! Ted com- one .101 com- . 2: em? "8. H38 2 mo 8365 no? Eoumwm OoHQSOOIm\m :HQmI~\m SHQm o MOM mo>uoo pououonoo HousoEoo oounu ohm ozonw .oEMNno poHHpIoNoon mo OE OH mm3 oHoEom .ousuouooEou mo noHuOQSM o mo AoO omouoanonm pHoo onuso o>Huon How ouop SDHHHnHumoomsm OHuocmoS .ON ousmHm 110 ON ousmHm 316:. CON _ 09 OO_ _ Boo 6.5.5.3wa 7E0 CON- _-Eo 00m . .qu Mum... dim. 111 rapid, but without ferrous ion conversion requires at least one hour at room temperature with either ASC or DTE. 2 . . acceleration 15 not es- Although the mechanism for Fe+ tablished, its role may be that of a one-electron outer sphere transfer agent. Both ASC and DTE are poor one- electron reductants, are higher molecular weight than ferrous ion, and both are negatively charged in the pH range normally used to reduce the enzyme. Ferrous ion, on the other hand, is positively charged, an excellent one- electron reducing agent, and is a small, mobile ion. The fact that reduction occurs at all in the absence of addi- tional Fe+2 may well be due to trace quantities of adven- titious iron (i.e., that responsible for the g = 4.3 EPR signal in the oxidized form of the enzyme). The reversion to the purple form, even in the presence of excess reductant, upon addition of 10 mM Pi is par- ticularly interesting; similar behavior is observed upon addition of PABP. With PABP, however, conversion is in- complete as the absorption maximum is 535 nm rather than 550 nm. Neither fluoride nor vanadate perturb the visible spectrum of the pink form significantly. The stepwise re- duction of the enzyme shown in Figure 13 is strong evidence that the reduced (pink) form of the enzyme is, in fact, a one-electron (per two iron atoms) reduced species. This is in direct conflict with the hypothesis that, in the porcine enzyme, the purple to pink shift is due to a pro- tein conformation change caused by reduction of a disulfide 112 bond (or bonds - in any case this would require a minimum two-electron reduction). This evidence suggests that a reduction of one Fe(III) to Fe(II) may be the cause of the color shift in the beef spleen enzyme. Data from EPR and magnetic measurements to be discussed below support this conclusion. The resonance Raman Spectra of the oxidized and reduced forms of the beef spleen enzyme are virtually identical to those reported for the corresponding porcine enzymesz. Both are similar to the sweet potato enzyme spectruml6: data for all three enzymes are compiled in Table 6. The high frequency bands for uteroferrin originate from a tyrosyl residue(s) and have been assigned as follows: C-H 1), C-O stretching (1285 cm-l), first ring 1) bending (1168 cm- stretching (1503 cm- cm-1)52'54. These assignments are similar to those made for , and second ring stretching (1603 transferrin71 and have been Shown to be relatively inde- 52 pendent of the type of coordinated metal ion . Frier sp_sl.72 and Sjoberg sp_s1,73, have reported hemerythrin resonance Raman absorptions in the 450 to 510 cm.1 region and, using the isotOpic shift of 180 enriched samples, have assigned the absorptions to an Fe-O stretch in a u-oxo bridged iron dimer. A weak band appears at 523 cm-1 in the resonance Raman spectra of purple acid phosphatase from beef spleen; however, it is not clear that this (or the stronger absorption arising at 572 cm-1) represents an 113 .noHoou mHnu CH oHnoHHo>o uo: ouuoomm I .o.n ONOH «OOH mOOH OOOH OOOH mOOH momH vomH momH momH momH momH mmmH mmmH mmmH mmNH mmNH OONH ommH mmHH OOHH OOHH EOHH OOHH .o.n III mum mum mum Hum .o.n OOO mom mom OOO mom .o.n mum Hum Hum mum mum .o.n .o.c .o.n mmm mmm Omm Nnm III III III III III o>Huoz pooopom OoNHpHxO poananH Ho pooopom poNHpon OHououoo uooBm mm monouo onHouoo nooHom onH>om "a HIEO .numnoHo>o3 .moEMNnm ouopoo noozm ppm ocHouom onu EH mpcom mchnoomoHHoo on» cam nooHom ocH>om Eoum omouonmmonm pHOS oHouom poanHnnH oponmmonm pno poospom .o>Huoz on» Sn poananm monom noEoM oonon0mom HofloS .O oHnoB 114 1 band is Fe-O stretch of a bridging oxygen. The 520 cm— not reported for either the porcine or the sweet potato enzyme, as the published spectra do not cover that region. 1 A band at 370 cm- in the sweet potato enzyme Spectrum has been tentatively assigned as a Mn(III)-S stretchl6; however, neither the porcine nor the beef spleen enzyme exhibits such an absorption. These results are solid evidence for assignment of the Visible absorption to a tyrosyl ligand to metal charge transfer transition. It is unclear whether one or both iron sites have tyrosyl ligation. The reduction of one Fe(III) to Fe(II) does not decrease the molar absorptivity of the charge transfer band, suggesting that only one iron is ligated by tyrosine. However, as will be shown in Chapter 4, removal of one iron (and replacement with Zn(II)) results in the loss of 50% of the absorption at 550 nm, implying that both iron sites have tyrosyl ligands. It is probably more likely that both iron sites contain a tyrosyl ligand and that the change in wavelength (shifting of the Am toward the larger, ax tailing UV absorption) upon reduction, as well as changes in the electronic configuration of the chromophore are responsible for the similar molar absorptivity of the reduced and oxidized forms. The EPR spectra in Figure 15 show that, for bovine en— zyme isolated using HAP, the oxidized and phosphate-in- hibited forms are EPR silent while the reduced form has a 115 complex high field absorption similar to that observed for the porcine enzyme, uteroferrinso. The signal for the bovine spleen enzyme represents 0.91.1 spins per two iron atoms. The results of careful titration studies (Figures 16 B and 19) using DTH or MV+ show that the maximal g = 1.77 signal occurs at approximately one electron per two iron atoms, and are entirely consistent with the parallel optical titration (Figure 13). Both results are indicative of a spin-coupled [Fe(III)]2 unit in the oxidized form and a spin-coupled Fe(III), Fe(II) unit in the pink form of the enzyme. These EPR results are in contrast to the published 50, where double integration results on the porcine enzyme of the EPR signal yielded one spin per iron atom and the color shift upon mild reduction was attributed to reduc- tion of a disulfide bondso. This is not at all consistent with the one electron reduction (per two iron atoms) and loss of covalently bound phosphate seen in the beef spleen enzyme. It is difficult to reconcile these differences since the two enzymes are so similar in other respects. Recent results Show the porcine enzyme to contain a two iron coupled unit that also undergoes a one electron reduc- tion65. EPR studies on hemerythrin lend strong support to the Fe(III)-Fe(II) hypothesis, as this protein is known to contain a spin-coupled two-iron unit. Two different 116 "semi-met" mixed valence Fe(III)-Fe(II) forms of hemerythrin have been observed, and exhibit EPR spectra similar to those Observed for the reduced purple phosphatases74. One-electron reduction of methemerythrin gives a species with a rhombic (g = 1.95, 1.87, 1.67) EPR Spectrum, while the one-electron oxidation of deoxyhemerythrin yields a species with an axial spectrum (9 = 1.95, 1.72). The overall shapes, 9- values and temperature dependence of these semi-met heme- rythrin spectra closely resemble those of the bovine and porcine purple phosphatase signals, strongly suggesting the existence of a related binuclear iron unit in the purple phosphatases. The apparent existence of the bovine and porcine enzymes as a mixture of two species (as suggested by the EPR spectra) may be due to the simultaneous exist- ence of two spectral forms of the enzyme (such as the two forms seen for semi-methemerythrin). The spectra in Figure 20 may explain the observed heterogeneity of the EPR spectra, as the ratio of the two species shifts with pH. Ligation to the iron atoms in hemerythrin is probably substantially different from that in the purple phosphatases (there is no evidence for tyro- 75), in which case the pH sine ligation in hemerythrin sensitivity of the hemerythrin may be substantially less. Also, if methemerythrin and deoxyhemerythrin have dif- ferent ligation, and electron transfer upon oxidation or reduction is rapid compared to rearrangement of ligands, 117 then the rapid freezing (to prevent disproportionation) of the hemerythrin samples after preparation may account for the different Spectra observed. The pH dependence of the EPR spectrum of the purple phosphatase implies that deprotonation of an iron ligand is responsible for the appearance of the axial EPR signal observed at higher pH. The ligand that is probably de- protonated is most likely H20, which is known to have a pKa of ~7.5 when bound to Fe(III)76, although deprotona- tion of an imidazole ligand cannot be ruled out (Fe(III)- imidazole is reported to deprotonate with pKa's of 8 to 77). As 11, depending upon substituents on the imidazole shall be seen later in this chapter, the magnetic behavior of the bovine purple phosphatase and hemerythrin also sup- ports the existence of a spin-coupled iron dimer in the bovine enzyme. Antanaitis and Aisen51 showed that bovine spleen purple acid phosphatase isolated without the use of HAP gave an EPR signal at g = 1.77 for the purple form of the enzyme. Figure 21 demonstrates that the bovine enzyme prepared in a similar fashion gives similar results. Double integra- tion of the g = 1.77 EPR signal of the native (as isolated) form shows that the signal is only m10% of the intensity of the Signal observed for the fully reduced form of the enzyme; it is probably due to the presence of a small amount of reduced enzyme in the enzyme as isolated, as has been 118 reported for the porcine enzymelz. The behavior of the por- cine enzyme (furnished by Professor Phillip Aisen) shown in Figure 22 is nearly identical to that of the bovine en- zyme. Table 7 gives results of double integration of the EPR spectra of the oxidized, reduced, and phosphate inhibited forms of both enzymes. Evidently only 10-20% of either enzyme is EPR active in the native (as isolated) form, while upon reduction both show maximal EPR activity. Addition of inorganic phosphate to the reduced enzymes restores the original spectra, indicating that the residual signal is not affected by simple phosphate binding. The results in Chapter 2 regarding the binding Of phos- phate to the purple form, but not to the reduced (pink) form of the beef spleen enzyme, are supported by the EPR results discussed earlier. Evidently in the absence of oxidant (02 or oxidized ASC or DTE), addition of phosphate does not cause a loss of the EPR signal nor does it cause rever- sion to the purple form. These results strongly suggest that phosphate is associated with the purple form of the enzyme, but not the pink, and are consistent with those phosphatases which covalently bind Pi to a serine residue26. The somewhat broader EPR spectrum observed upon the anaero- bic addition of Pi to the pink form (spectrum not shown) may be due to a 31 P nuclear hyperfine interaction, but this is not clear at this time. In direct contrast to phosphate, vanadate apparently 119 Table 7. Double Integration of the EPR Spectra of the Oxidized (Native), Reduced (Pink), and Reduced Plus Phosphate Forms of the Purple Acid Phos- phatases From Bovine Spleen and Porcine Uterus. Enzyme Native Reduced Reduced + Pi Bovine Spleen 0.04 spins 0.80 spins 0.04 spins Porcine Uterine 0.075 spins 1.06 spins 0.075 spins 120 binds to the pink form of the bovine purple phosphatase, strongly perturbing the g = 1.77 EPR signal (Figure 23). The EPR signal observed only accounts for 56% of the ex- pected signal, but it is not clear whether this is due to an equilibrium between reduced, inhibited enzyme and the oxidized form, to a loss of protein due to denaturation at the relatively high pH level (7.0) necessary to keep the V04 in solution, or to other factors. In any case, if the vanadate is truly bound to the iron, or at least nearby, as the EPR suggests, ENDOR or pulsed EPR spin echo studies may Show a clear interaction between the iron and the van- adate. Such detailed studies do not, however, fall within the scope of this dissertation. The MOssbauer spectrum of the native purple acid phos- phatase from bovine spleen suggests that the enzyme contains two non-equivalent iron sites in a 1:1 ratio. The quad- rupole splittings and isomer shifts are similar to high- spin iron (III) values (typically 0.2 to 0.8 mm/sec and 0.4 to 0.85 mm/sec, respectively). However the isomer shift of 0.16 nm/sec of one doublet is lower than is typical for high-spin ferric iron. The quadrupole splitting Of 1.16 57Fe enrichment mm/sec is also out of the usual range; of the enzyme will enable detailed MOssbauer studies which may clarify the relationship of the two iron sites and pro- vide information regarding the antiferromagnetic coupling of the two iron atoms. It is clear, however, that the iron 121 present in the purple phosphatase is not adventitious iron, as the effect of an applied magnetic field is minimal as is the broadening of the absorption upon increasing the sample temperature to 146 K. Paramagnetic Fe(III) should be magnetically split and Show temperature broadening of the Mdssbauer Spectrum. The MOssbauer spectrum of beef spleen purple phos- phatase is unlike those of metaquohemerythrin or (Fe(salen))20, both of which exhibit only a single quadrupole doublet78. The quadrupole splittings of those two spectra are 1.57 and 0.85 mm/sec, respectively, not unlike those of the purple enzyme. The isomer shifts are both 0.46 mm/sec, the same value observed for one of the phosphatase doublets. Another binuclear iron enzyme, ribonucleotide reductase, has a Mdssbauer spectrum consisting of two quadrupole doub- lets (quadrupole splittings 1.65 and 2.45 mm/sec and isomer 79). Magnetic splitting shifts 0.53 and 0.45, reSpectively of the doublets of the reductase is observed at 30 kilo- gauss, indicating strong antiferromagnetic coupling. Further Mossbauer studies of the purple acid phosphatase will be re- quired to adequately compare the enzyme to available models and other iron containing enzymes. The results of magnetic measurements shown in Figures 25 and 26 represent unequivocal evidence for the existence of a spin-coupled binuclear iron center in bovine spleen purple phosphatase. The oxidized (purple) form is 122 essentially diamagnetic over the range 4.5 to 300 K. The lack of a g = 4.3 EPR signal or deviation from linearity of a Xm vs l/T plot for the purple form indicate a strong antiferromagnetic interaction between the two iron atoms. 1 The magnitude of -2J is estimated to be 3 250 cm- for this form of the enzyme and is similar to that reported for 1)80 metaquohemerythrin (-2J = 268 cm- This high value for -2J is not unusual for u-oxo bridged iron dimersBl. The magnetic behavior of the one-electron reduced (pink) form of the enzyme is consistent with the S = 1/2 EPR spec- trum observed. It has slightly higher magnetic suscepti- bility than would be expected for a Spin only Curie law S = 1/2 paramagnet (for g = 1.77, u = 1.36 uB). Anal- eff ysis of a Xm vs. l/T plot results in a value of “eff 2.2 “B per two iron unit. The reason for the higher than ex- 2+ from pected value is not known, but it may be due to Fe enzyme degraded upon reduction and freezing. The visible, EPR, and MOssbauer data presented above, when combined with the magnetic behavior of the oxidized and reduced forms of the purple acid phosphatase, Show that the enzyme contains a spin-coupled binuclear iron center. Direct comparison with the porcine uterine enzyme shows an indubitable relationship between the enzymes from spleen and uterine fluids. Whether or not this two iron cluster is present in other purple phosphatases remains as an urgent question to be resolved by other researchers. CHAPTER 4 METAL ION SUBSTITUTION STUDIES WITH BOVINE SPLEEN PURPLE ACID PHOSPHATASE Metal ion substitution experiments are commonly used by bioinorganic chemists to examine the active sites of metalloenzymes. The effect of such substitutions upon the catalytic activity and upon spectral and magnetic prOperties of the metalloenzyme can often provide important insight into the mechanism, ligation, coordination geometry, and metal oxidation state of the native enzyme. Keough and 55 have demonstrated that selective removal and coworkers replacement of one of the two iron atoms is possible for the porcine purple acid phosphatase. Upon reduction of the enzyme with excess sodium dithionite, one iron atom is ob- served to diffuse out of the protein rapidly (tl/2 < 15 sec), while the second iron atom is removed much more slowly (tl/2 W 30 min). By quickly removing excess reductant and the liberated iron via G-25 chromatography, a "half-apo" enzyme containing a single iron atom per molecule can be prepared. Exposure of the "half-apo" form of the enzyme to Zn(II) results in the formation of a one iron, one zinc form of the porcine enzyme. The zinc-substituted enzyme 123 124 thus formed retains 100% of the catalytic activity of the native (Fez) form toward PNPP. Similar resultstr replacement of one iron atom by zinc have been reported for the bovine spleen purple acid phosphatase48. The zinc-iron bovine enzyme thus prepared also shows full retention of enzymatic activity; Vfiax and Km for PNPP and the Ki for phosphate have also been shown to be the same for the Fe-Zn as for the native form of the enzyme. NO metal substitution experiments have been described to date in which one iron atom is replaced by a metal ion other than Zn(II), although Keough s3 31.55 do report replacement of both iron atoms with Zn(II) and partial replacement of both with Ni(II). Attempts by us to prepare bovine spleen enzyme with both iron atoms replaced by 57Fe enriched iron met with little success, Zn(II) and by usually resulting in precipitation of large quantities of colorless apOprotein. The results presented in this chapter Show that the Fe(III)-Zn(II) and Fe(III)-Cu(II) forms of the bovine en- zyme can be prepared and that the Fe(III)-Cd(II) and Fe(III)- Ni(II) forms are probably formed, although they have not been isolated in pure form. Metal analyses, kinetics data specific activities, electronic and EPR spectra, and mag— netic measurements will be presented. 125 Materials and Methods Metal analyses were done using the plasma emission method described in Chapter 2; specific activities were determined and kinetics experiments performed with PNPP as substrate using the modified assay method described in Chapter 1. Electronic spectra were obtained using a Cary 219 spectrophotometer, while the liquid helium temperature EPR spectra were obtained using a Bruker ER200D spectrom- eter equipped with an Oxford liquid helium cryostat and a Nicolet Model 1180 computer. EPR experiments at 77 K were performed using a Varian E-4 spectrometer equipped with a liquid nitrogen dewar as a sample holder. The 77 K EPR spectrum of the Fe-Zn enzyme was quantitated by double integration versus a 1.0 mM Cu(II)-EDTA standard using the g-value corrections of Aasa and Vanngard66, and versus 0.5 mM bis-salicylidenehistamine iron(III) hexafluorophosphate ([Fe(III)balhis)2]PF6) solution in ethanol. Magnetic susceptibility data over the range 4.5-250 K were obtained on an SHE Corporation SQUID susceptometer using frozen liquid samples (40 pk, 1.1 mM FeZn enzyme). The raw data were treated as described in Chapter 3. All reagents used were ACS grade or better and were Obtained from the usual supply houses. The metal substitution experiments were performed ac- 55 cording to the method of Keough s3 31. with the following 126 modification. Prior to loading the fully reduced enzyme onto the G-25 column to remove liberated iron, 1,10-phen- anthroline, and excess DTH, the column was equilibrated with a 10 mM solution of a salt of the metal ion to be substi- tuted (CuC12,Zn (acetate)2, N°(NO3)2,Or CdClz). This im- mediately exposed the "half—apo" form of the enzyme to the desired metal ion and minimized disproportionation of the half-apo form to equal parts native two-iron enzyme and apoenzyme. Results Metal Analyses The results of metal analyses on the Fe-Zn, Fe-Cu, Fe- Cd, and Fe-Ni preparations are shown in Table 8. For com- parison, the catalytic activity of each form toward PNPP is given as the percent of native enzyme activity, and the absorbance ratios A280/A4max are also included. The Fe-Zn and Fe-Cu preparations appear to contain one mole of Fe and one Of Zn or Cu per mole enzyme, with both metal sites fully occupied. The Fe-Cd preparation contains about one excess cadmium atom per mole of enzyme, while the Fe-Ni preparation contains about 0.4 nickel and 1.16 iron atoms per molecule. Substitution experiments were also attempted with Mn(II) and Co(II), but incorporation of less than 0.1 metal atoms per molecule by either metal ion was achieved under the 127 Table 8. Metal Content and Enzymatic Activity of Metal- Substituted Purple Acid Phosphatase. Fe-M Fe/mol M/mol Activity A280/A550 Fe-Zn 0.97 1.05 100% 28 Fe-Cu 1.0 0.98 18% 28 Fe-Cd 0.86 2.2 7% 27 Fe-Ni 1.16 0.4 15% 38 Fe-Fe 2.0 --- 100% 15.4 128 conditions of the experiments. Attempts to make the apo- enzyme and reconstitute with two Zn(II) atoms led to pre- cipitation of much of the protein and no apparent forma- tion of the an form. Similar attempts to prepare 57 Fe enriched enzyme from apoenzyme also met with failure, even though the jporcine enzyme iron reconstitution conditions described by Keough, s3 31.55 Kinetics Data , were rigorously followed. Detailed kinetics data were obtained only on the FeZn form of the enzyme, as the other Fe-M forms obtained have less than 20% of the catalytic activity of the native en- zyme at pH 6.0. Figure 27 shows the kinetics behavior of the Fe—Zn enzyme in the presence and absence of 10 mM and 3- 4. tions were adjusted to pH 6.0 prior to addition. The 2.5 mM F“, 10 mM Pi, and 1.0 mM A50 A11 inhibitor culated values for Km and Vmax (PNPP) were 3.8 mM and -1 3- sec , respectively. Ki values for A304 and Pi were solu- cal- 570 4.0 mM and 3.6 mM respectively; the same anomalous fluoride inhi- bition behavior observed for the native (Fez) form of the enzyme is seen for the Fe-Zn form. Electronic Spectra The visible absorption spectra of the Fe-Zn, Fe-Cu, Fe-Cd and Fe-Ni forms Of the bovine purple phosphatase 129 Figure 27. Lineweaver-Burk plots of kinetics data for the Fe-Zn enzyme with PNPP as substrate. Shown are the native enzyme (0) alone and with 1 mM Asoi' (A), 2.5 mM F' (O), 10 mM F’ (A), and 10 mM Pi ([3) . 130 '200_ ~ ' 0 Native Enzyme / A A lmM A502 / / o IOm Pi / o 2.5mM F'/ A lOmM F-‘ ISO '- l _ / ,- V IOO / / min/[uncle / / so-l / éL/m‘ole x (02. Figure 27. 131 are shown in Figure 28. Table 9 lists the visible absorption maximum for each form and the molar absorptivity per iron atom and per mole of protein. All four forms have a shoulder at 330 nm and a large peak at 280 nm, as does the native enzyme. EPR SpectrOSCOpy Several EPR experiments were performed with the FeZn form of the enzyme. Figure 29 illustrates EPR spectra of the Fe-Zn enzyme at 4 K. Spectrum A shows the EPR of the FeZn form as isolated, while spectra B and C Show the EPR 2 and ASC/Fe+2 plus phosphate, after addition of ASC/Fe+ respectively. The observed increase in intensity and sharpening of the signal seen in Spectrum C also occurred upon addition of phosphate without the inclusion of the ASC/Fe+2 reductant. The low temperature (T < 30 K) EPR spectra with a broad g = 4.3 signal (Figure 29, spectrum A) is easily saturated (above 0.25 mW at 13 K) and broadens rapidly above 30 K. This is qualitatively characteristic of high-spin Fe(III) in a rhombic environment, but the peculiar relaxation properties make quantitation by double integration impossible. Addition of saturating concentra- tions of Pi results in sharpening of the signal and an apparent increase in the spin-lattice relaxation time, such that the signal is easily observed at 77 K (see Figure 30). Thus, Figure 30 shows the EPR spectrum at 77 K with 132 Figure 28. Visible absorption spectra of several metal substituted forms of bovine spleen purple acid phosphatase. Shown are the Fe-Zn (———), Fe-Cd (----), and Fe-Cu (---) forms of the enzyme. Spectra for Fe-Zn and Fe—Cu are offset vertically by 0.06 and 0.05 A, re- Spectively. 0.05 133 \. \ '\ .\. ‘\ ’\ \. l I l 1 400 500 600 700 1, nm Figure 28 134 Table 9. Visible Absorption Maxima and Molar Absorptivities of Several Metal Substituted Forms of Purple Acid PhOSphatase from Bovine Spleen. Form Amax e (per mole enzyme) 5 (per mole Fe) FeFe 550 nm 4000 m’1 cm’1 2000 m‘1 cm‘1 FeZn 545 2100 2160 FeCd 500 2270 2640 FeCu 560 2210 2210 FeNi 550 1680 1450 Figure 29. 135 EPR spectra of the Fe-Zn purple acid phospha- tase as isolated (spectrum A), reduced with ferrous ion and ascorbate (spectrum B) and reduced plus 10 mM phosphate (spectrum C). Conditions were: T = 4.2 K; 9 mg/mL enzyme; microwave frequency, 9.46 GHz; microwave power, 50 uW; modulation frequency, 100 kHz; modulation amplitude, 10 G; time con- stant, 0.2 sec; field set, 2050 G, scan range, 4000 G; scan time 200 sec; instru- ment gain, A = 2 x 106, B = 1 x 106, C = 5 x 105. C FePose FEZn + P. 1 A KJfMNdvaSI-~m~vvnAvaVhArw1:;:;:“~ FeZn 9=4.3 l l l l J 1 06 20006 40000 Figure 29 137 .2; Omm no Amv OOOH .nHoo unoEsuumnH “CHE v .oEHu cooo .0 OOOH .oonou noow “O ommH .pom pHon “oom m.O .unouonoo oEHu “O OH .opsuHHoEo coHuoHSOOE unmx OOH .Socoooouw noHuoHSOOE uno30uoHE “NmO OH.m .Monosooum o>o3ouOHE "mnoHuchoo onHonH0m on» Hops: Houo IEouuooom vim coHuo> o no ponHouno oHoB ouuooom .O.m Mm .Hommsn onouooo EoHpom CH nHououo SE mm.O pocHopnoo moHoEow nuom .oponuoomo SE OOH Ono ouonomono SE OH HO no Efiuuooomv oonooouo poo Am Esuuooomv oonomno onu EH omouonomono pHoo oHoHoo nooHom moon mo EHOM SNIom mo M no no ouuooow Mom .Om ouson 138 o 8.1 N0? om oHDmHm 139 (spectrum A) and without (spectrum B) addition of Pi' Double integration of the sharp g = 4.32 EPR signal at 77 K obtained in the presence of phOSphate yielded a value of 0.9i0.l spins per Fe atom. EPR spectra of the Fe-Cu and Fe-Ni forms of the enzyme show only weak signals (g m 2.0, i 0.1 spins/mol); the Fe-Cd form shows a g z 4.3 absorption similar to that of the Fe—Zn form (Figure 31). Magnetic Measurements on the Fe-Zn Purple Phosphatase Low temperature magnetic susceptibility data on the Fe-Zn protein are shown in Figure 32; simple Curie law behavior consistent with high spin Fe(III) is observed. The slope of a plot of Xm vs l/T (data treated as in Chapter 3) yields a temperature-independent magnetic moment of “eff = 6.0:0.2 0B. Discussion The results presented above demonstrate that one of the iron atoms of the spin-coupled binuclear iron center of the purple acid phosphatase can be replaced by any of several divalent transition metal cations. Only the Fe-Zn form of the enzyme retains substantial enzymatic activity at pH 6.0; this form shows kinetics and inhibition behavior similar to that of the native two iron protein; the only 140 .OOH x N n :Hoo pno 3E m n uo3oo o>o30HOHE umooxo mm ouson EOM mo oEom onu ouo onoHqunOo Houuooom .AHE\OE v.mv omouonomonm pHoo oHouoo mo EEOm pOIoM onu mo Esupooom Mom .Hm ouson 141 O OOOV Hm ousmHm OO O OOON H ng 142 .ASE H.HO uonmoE Iouoo N\m n m no MOM pouoHooHoo mH onHH pHHOm one .ASE H.HO ooouonm Imono pHoo oHouso nooHom onH>on mo Epow nNIoM onu mo MOH>onon oHponooS .mm onsono 143 mm ouoon CO... nXu_ Aum . nu _ cNom 144 discrepancy is the Ki for arsenate, 4.0 mM, which is substantially larger than the value of 0.24 mM found for arsenate inhibition of the native enzyme. Due to the cat- alytic activity and apparent purity of the Fe-Zn form of the enzyme, the bulk of additional studies were concentrated on that form. In early experiments, using the method of Keough, as 531-55 , to prepare the Fe—Zn form of the enzyme, de- tectable disproportionation of the "half-apo" enzyme to yield small amounts (m10%) of holOprotein containing two iron atoms occurred during reconstitution. As a result the technique described in "Materials and Methods" was used to quickly expose the "half-apo" enzyme to the substituting metal ion while still on the G-25 column. In the case of Zn(II) substitution, this method minimized the formation of the Fe2 form of the enzyme. The rate of disproportiona- tion was not measured but probably has a tl/2 > one hour based on the observed formation of the Fe2 form during the m20 min required for the G-25 separation and subsequent Zn(II) addition. Using the technique of adding Zn2+ to the G—25 column eluent, formation of the Fe2 form is typically <5% as determined by the increased enzymatic activity with DT/Fe+2. The EPR spectra of the Fe-Zn form of the bovine spleen purple acid phosphatase, which under appropriate conditions (77 K, 10 mM Pi), shows a strong g = 4.3 signal equivalent to approximately one spin per iron atom, together with the 145 magnetic behavior (“eff = 6.0:.2 uB), demonstrates that replacement of one Of the iron atoms by Zn(II) abolishes the magnetic coupling, as expected. Phosphate, a competi— tive inhibitor of phOSphate ester hydrolysis, strongly perturbs the EPR spectrum of the Fe-Zn protein, suggesting that phosphate (and by inference, substrate) binds at or near the bimetallic site. At this point, there is no evi- dence that can differentiate between binding of phosphate and substrate in close proximity to either the exchangeable (Fe2+ or Zn2+) or the non-exchangeable (Fe+3) metal ions. A number of techniques are available which, in the future, may differentiate between these two possible binding sites: ENDOR (electron nuclear double resonance) and pulsed EPR experiments with 67Zn-substituted or Fe2 holoenzyme and 31 75 3 AsO4-; and Fe and Zn EXAFS (extended x-ray absorption fine structure) spectroscopy on phosphate-, P-phOSphate or arsenate- and tungstate-inhibited Fe2 and Fe-Zn enzyme. The data in Tables 8 and 9 are interesting in that the three metal substituted forms of the enzyme that are most readily formed (Cd, Cu, Zn) all have similar molar ab- sorptivities and absorption ratios (A280/AA ). It does max appear, however, that some Fe-Ni form is obtained, as significant amounts of Ni are associated with the enzyme. The relatively high absorption ratio and low molar ab- sorptivity associated with the Fe-Ni form suggest that the amount of Fe-Ni enzyme formed is small. Since only 0.4 atoms of nickel bind to the "half-apo" bovine spleen 146 enzyme upon attempts at substitution, it is interesting to note that Keough, 22'sl.55 , report that reconstitution of the apo-enzyme with Ni(II) results in only 1.4 atoms of nickel binding per mole of enzyme. It is not clear whether the second binding site is somehow restricted to 0.4 Ni atoms per mole due to an equilibrium phenomenon, or if a change in pH, ionic strength or temperature of the re- constitution conditions will be sufficient to allow stoi- chiometric replacement of iron by nickel at the exchangeable site. Similar changes in reconstitution conditions may allow formation of the Fe-Mn, Fe-Co, and other forms of the enzyme. The absence of a definitive EPR signal for the 5Fe(III) Fe-Ni enzyme is peculiar in that a combination of d and d8Ni(II) should result in half-integral spin species and hence in an EPR signal whether or not antiferromagnetic coupling occurs. This conspicuous absence could imply that the nickel may either be present in a Ni-Ni form of the enzyme (antiferromagnetically coupled) or as non-specifically bound metal, although magnetic broadening or unusual relaxa— tion prOperties may also be responsible for the lack of an observable EPR absorption. Clearly, the ability of this enzyme to bind a variety of metal ions at one if not both sites represents a real Opportunity for inorganic chemists to study this type of bimetallic complex. Although the Fe-Cu and Fe-Cd forms of the enzyme are inactive at the pH used, studies of their 147 magnetic properties, EPR (for the Fe-Cd form), EXAFS, MOssbauer, and resonance Raman Spectra may provide infor- mation on the structure of the catalytic site of the holo— enzyme. The relationship of the bovine spleen enzyme to the enzyme from porcine uterine fluid is considerably clarified by these metal substitution experiments. Certainly, the existence of the Fe-Zn forms of both enzymes clearly demon- strates the bimetallic nature of the active site in each. The combination of full enzymatic activity of the Fe-Zn forms, the metal assay results of Keough 3; s1,55 , and the stepwise removal of the two iron atoms observed for both enzymes are convincing evidence of the similarity between the bovine spleen and porcine uterine enzymes. CHAPTER 5 PREPARATION AND CHARACTERIZATION OF TWO MODELS FOR THE IRON CENTER IN PURPLE ACID PHOSPHATASE Tyrosyl ligated Fe(III) is known to be present in non- heme iron-containing proteins such as dioxygenases,82 transferrin83, and porcine uterine purple phosphatasesz, and has been implicated in the iron chromophore of purple acid phosphatase from beef spleen (Chapter 3). The di- oxygenases and transferrin contain discrete mononuclear iron centers, while the purple phosphatases from beef Spleen and porcine uterus contain a spin-coupled binuclear iron center at the active site. All of these proteins exhibit a strong ligand—to-metal charge transfer band in their visible spectra. It is often convenient to synthesize model compounds which mimic the spectral properties of the chromophore of a metalloenzyme to provide insight into the more compli- cated protein metal centers. Que sp‘sl.82 , have prepared a salicylaldehyde-histamine Schiff base chelate of Fe(III) (bis-salicylidenehistamine iron(III) perchlorate-Fe(sal- his)2 C104), which has similar resonance Raman and visible absorption spectra to the beef spleen purple acid 148 149 phosphatase. The compound was isolated (by trituration) as a dark purple powder. In this chapter we report the preparation and electronic and EPR spectra of the Fe(sal- his)2 PF 6 salt and the reduced Schiff base chelate Fe(sal— hisHZ)zBPh4. The crystal structure of the former is also presented. Both compounds are useful as models, particularly for the Fe-Zn form of the purple acid phosphatase. Materials and Methods Histamine was Obtained from Sigma and used as received; salicylaldehyde (as the bisulfate) was purchased from East- man. ACS reagent grade anhydrous ferric chloride was used to form the chelates. Sodium tetraphenylborate and am- monium hexafluorOphosphate were purchased from Alfa Pro- ducts. Sodium borohydride was a Matheson Coleman and Bell product. All other reagents were ACS reagent grade or better and were obtained from a variety of sources. Electronic spectra were obtained using a Cary 219 spectrophotometer. EPR spectra were obtained using either a Varian E-4 (77 K) with a liquid nitrogen dewar as a sample holder or a Bruker ER200D (4-30 K) equipped with an Oxford liquid helium cryostat. Fe(salhis)2PF6 was prepared in the same manner as Fe(salhis)2ClO482 except that ferric chloride (anhydrous, 0.75 g) was substituted for ferrous perchlorate and the resulting solution of chelate in methanol was treated with 150 4.6nmmfl.(0.49 g) of ammonium hexafluorOphosphate, yielding a purple precipitate. The precipitate was filtered, washed 2x with 10 mL diethyl ether and air dried. The product was crystallized from ethanol by slow cooling to -20°C. The reduced form of the Schiff base chelate was prepared by mixing 1.03 g of histamine with 1.13 g of salicylaldehyde in 50 mL of absolute methanol. After the intense yellow color of the Schiff base was fully developed (NS min), small quantities (m10 mg) of sodium borohydride were slowly added with stirring until the yellow color was no longer visible. The solution was cooled to -20°C, filtered, and 0.75 g anhydrous FeCl3 added creating an intense purple solu— tion. Sodium tetraphenylborate (1.57 g) was added, and the resulting purple solid filtered and washed 2x with 10 mL of diethyl ether. The resulting product was air dried and redissolved in hot isopropyl alcohol. Although the product precipitated upon cooling, only microcrystals were obtained (even with cooling in a controlled temperature bath with the temperature lowered at 1°C per h). Elemental analysis for Fe(salhis)2PF (C, H, N, Fe, P, F) and Fe(salhisH2)2BPh4 6 (C, H, N, Fe, B) were performed by Galbraith Laboratories and are given in Table 10. x-ray crystallographic data were obtained by mounting a suitable crystal of Fe(salhis)2PF6 in a glass capillary sealed in an argon atmosphere. The crystal data, collection and reduction of the x-ray diffraction data, and solution and 151 Table 10. Elemental Analysis of Fe(salhis)2PF and Fe(salhisH2)2BPh 6 CH3CHZOH 4. Fe(salhis)2PF -CH3CH20H Fe(salhis H2)2BPh4 Element Expected 6 Found Expected Found C 46.26 45.58 71.23 70.08 0 7.11 7.11 3.95 4.31 (by difference) N 12.45 12.39 10.37 7.80 H 4.49 4.42 6.22 6.25 Fe 8.44 8.44 6.89 7.42 P 4.59 4.59 ---------- F 16.89 16.94 ---------- B ----- ----- 1.33 4.14 152 refinement of the structure are provided in Table 11. Results The results of elemental analyses of Fe(salhis)2PF6 and Fe(salhisH2)2 shown in Table 10 indicate that the crystalline Fe(salhis)2PF is a relatively pure compound 6 with one ethanol of crystallization. The results for Fe(salhisH2)2PF6 are less clear, suggesting that some im- purities may be coprecipitating with the desired compound. Reduction of the Schiff base C=N will result in an optically active secondary amine; a mixture of diasteriomers will thus be produced, upon formation of the bischelate. Either of the above possibilities might explain the difficulty in preparing x-ray quality crystals from Fe(salhisH2)BPh4. The electronic spectra of both model compounds are shown in Figure 33, Fe(salhis)2PF has a Am = 535 nm, 6 ax BPh4 has Am = 525 nm; the molar while the Fe(salhisH ax 1 2’2 absorptivities are 3800 and 2190 L-M- , respectively. Figure 34 depicts the EPR Spectrum of the two models, showing a g = 4.3 high-spin Fe(III) EPR spectrum for both. The spectrum of Fe(III) conalbumin is included for com- parison. The magnetic data for Fe(salhis)2PF6 shown in Figure 35 indicate that the model compound exhibits Curie law behavior; a plot of Xm vs l/T yields a value for the 153 Table 11. Summary of Crystallographic Data for Fe(salhis)2 PF6. Formula Molecular Weight Space Group a,$. b, A c, A o, deg 8. deg Yr deg g/cm3 Dcalc. Radiation Used max (sine/1) 0. cm"1 no. of reflections R Diffractometer Method of Solution Method of Refinement C24H24N602Fe-PF6-C2HSOH 675.38 Pbca 16.434 (11) 18.208 (11) 19.853 (13) 90 90 90 5940.6 (2) 8 1.510 Mo Ko (A = 0.70926 3) 0.60 5.814 2160 0.072 Picker FACS—I V.D.O.D.S.84 Patterson Full Matrix Least Squares 154 v sommxmasasmvon one A O O .< m.O uowmwo AIIIIO momannHomvom mo ouuooom noHuoHowno oHnHmH> .mm ousmHm 155 20 A 300 I 400 1 l 500 600 )(mm . Figure 33 700 800 4 m-.— 156 Figure 34. EPR spectra of the Fe(salhisH2)2BPh4 (spec- trum A), conalbumin (spectrum B) and Fe(sal- his)2PF6 (spectrum C). Spectral conditions were: microwave frequency, 9.090 GHz; micro- wave power, 10 mW; modulation amplitude, 5 G; modulation frequency, 100 kHz; field set, 2000 G; scan range, 4000 G; time constant, 0.3 sec; scan time, 4 min; T = 8.5 K; gain, 250 (A,C) and 2,500 (B). 157 58 427 Figure 34 158 .H66 on osea oeaom one . ASE H.Hv .ASE H.HO m umnooEouoo N\m n m no MOM pouoHsO mm AmHnHomOom mo uoH>onon OHuonmoS .mm anaana 159 On. mm ouson A v. V... 00— OD _ __ (0i 9 160 magnetic moment (peff) of 6.0:.1 “B - near the expected spin-Only value for high spin Fe(III) (5.92 0B). X-ray crystallographic data were obtained using a crystal of the Fe(salhis)2PF6-EtOH. The unit cell is shown in stereo view in Figure 36; eight formula units (C24H24N602Fe° C H OH) are contained within the unit cell. The PF anion 2 5 6 (not shown) has octahedral symmetry, while the six ligating atoms about the Fe(III) central atom are arranged in near octahedral symmetry as shown in Figure 37 and in stereo view in Figure 38. Hydrogen atoms have been omitted for clarity. Table 12 contains the Fe—O andlkroidistances and also shows Fe-O distances in several similar chelates found in the literature. Table 13 contains L-Fe-L bond angles for the Fe(salhis)2PF model. 6 Discussion The visible absorption spectra of the two model compounds have absorption maxima (535 and 525 nm) which lie between the maxima observed for the pink and purple forms of the purple acid phosphatase (505 and 550 nm, respectively). It is not clear what electronic or distortion effect is responsible for the differences in wavelength between the two models; however, the molar absorptivity (3800 L-M-l) of the better characterized model, Fe(salhis)2PF is nearly 6' twice that of the enzyme (per iron atom). This may be due to the presence of two phenolate ligands per iron atom in 161 Figure 36. Unit cell of Fe(salhis)2PF (stereo 6 view). 162 Figure 36 163 Figure 37. ORTEP diagram for Fe(salhis)2PF6 (50% probability) with hydrogens and PF6 omitted for clarity. ' 164 Figure 37 165 Figure 38. Stereo view of Fe(salhis):-hydrogens omitted for clarity. 166 Figure 38 167 Table 12. Representative Fe-O and Fe-N Distances for Fe(salhis)2PF6 and Similar Compounds. Fe(salhis)2PF Fe(saloph)CatHa 6 Atom Atom Distance, A Atom Atom Distance, A Fe 01 1.917 (6) Fe 01(sal) 1.897 (5) Fe 02 1.910 (6) Fe 02(sal) 1.912 (5) Fe N1 2.127 (8) Fe 03(cat) 1.828 (4) Fe N2 2.138 (7) Fe N1 2.104 (6) Fe N4 2.119 (8) Fe N2 2.090 (7) Fe N5 2.156 (7) [Fe(salen)]2hqb NaFe(Meso-ehpg)C Atom Atom Distance, A Atom Atom Distance, A Fe 01(sal) 1.898 (2) Fe 01(sal) 1.922 (7) Fe 02(sal) 1.912 (2) Fe 02(sal) 1.893 (7) Fe O3(hq) 1.861((2) Fe 03(carb) 2.011 (8) Fe N1 2.089 (3) Fe 04(carb) 2.087 (8) Fe N2 2.103 (2) Fe N1 2.157 (8) Fe N2 2.140 (9) a[N,N'-(l,2-Phenylene)bis(salicylideniminato)J(catecholato- O)iron(III), Reference 89. bu-(l,4-Benzenediolato-O,O')bis[N,N'-ethylenebis(salicyliden- iminatoiron(III)] Reference 89. CSodium salt of the iron(III) chelate of meso-N,N-ethylene- bis(O-hydroxyphenylglycine), Reference 90. 168 Table 13. Some of the Fe(salhis)2PF6 Bond Angles of Ligating Atoms with the Central Iron Atom. Atoms Angles (Degrees) 01-Fe-02 94.58 (.31) 02-Fe—Nl 90.79 (.32) Ol-Fe-Nl 86.86 (.34) 02—Fe-N2 90.00 (.32) 01-Fe—N2 174.12 (.37) 02-Fe-N4 51.58 (.37) 01-Fe-N4 90.40 (.31) 01-Fe-N5 91.13 (.32) 02-Fe-N5 172.47 (.43) 169 the model versus only one per iron atom in the enzyme. This does not explain the relatively low intensity of the visible absorption found for the Fe(salhisH BPh 2)2 4 model, although the purity and structure of this compound are in question. It is clear, however, that non-heme iron enzymes with tyrosyl ligation and model compounds with phenolate ligand have similar visible absorption spectra and molar absorptivities.88 The EPR spectra Shown in Figure 34 are typical of high spin Fe(III) with rhombic symmetry; the signals are nearly isotropic with only small differences in lineshapes evident. (These absorption arise from the middle Kramer's doublet of systems with large values of D.) The spectrum of Fe(salhis)2 PF6 is entirely consistent with the crystal structure Ob- tained. The magnetic behavior (Figure 35) represents con- clusive evidence for the high-spin Fe(III) state of the metal in Fe(salhis)2PF6. The crystal structure of Fe(salhis)2PF6 shows similar phenolate oxygen-iron bond lengths to those observed in other octahedral and square pyramidal Fe(III) salicylaldehyde Schiff base chelates. Such complexes have been reported to have Fe-O (sal) bond distances of 1.893 to 1.922 A89’9O, in good agreement with the Fe(salhis)2PF6 values of 1.917 and 1.910 A. Iron-nitrogen distances are also similar, with values ranging from 2.089 to 2.157 A for typical 89,90 complexes and 2.119 to 2.156 A for Fe(salhis)2PF 6' 170 Bis-salicylate oxygen-iron bond angles for other cis- oriented Fe(III) salicylates are also in good agreement with the Fe(salhis)2PF6 structure (ranges of 90.1 to 95.1 degrees versus 94.6 for this structure). Phenolate oxygen- iron-nitrogen angles in other complexes are typically 81.2 to 96.2 degrees, similar to the values for this model, although distortion from 90° seems to be minimal in the case of Fe(salhis)2PF6. The well characterized Fe(salhis)2PF6 appears to be a good initial model for the iron center in the Fe-Zn purple acid phosphatase. The reduced form of the chelate, Fe (salhisH2)2BPh4,is probably a more likely model because of the secondary amine rather than imine iron coordination, which provides a ligation more typical of a biological system; however, until an x-ray structure is obtained, confirming the proposed structure, other models must suffice. S UMMARY A new purification method for bovine spleen purple phosphatase: has been developed, which provides up to ten- fold greater yield than previous methods and requires only five steps. Using this method sufficient enzyme has been obtained to perform significant physical and chemical studies of the enzyme. The protein contains two iron atoms per 40,000 molecular weight molecule, and is composed of two non-identical subunits Of 14,000 and 26,000 molecular weight. The bovine spleen purple phosphatase exhibits similar kinetics behavior to the purple phosphatases found in sweet potato and porcine uterine fluids. It binds a single phos- phate per molecule in the purple (resting) form, but not in the pink (active) form. This behavior is consistent with that of the phosphatases that utilize a serine resi- due to covalently bind phosphate during catalysis. Electronic and EPR spectroscopic studies Of the pink (reduced), purple (nativeh and phosphate-inhibited forms of the purple acid phosphatase show that the purple form is EPR silent, the pink form is a one-electron reduced EPR active species with principal absorbance at g = 1.77, and the phosphate inhibited form is EPR silent. MOssbauer 171 i.” E Eli-m. 172 spectroscopy shows that the enzyme has two discrete iron sites in a ratio of 1:1 and that the iron atoms are anti- ferromagnetically coupled. This coupling is corroborated by magnetic measurements, which show the purple form to be diamagnetic and the pink (one electron reduced) form to l for have 8 = 1/2. The coupling is strong (2J Z -250 cm- the purple form) supporting the presence of a u-oxo bridged iron dimer. Resonance Raman spectroscopy shows the pres- ence of tyrosyl ligation to the iron, indicating that the visible absorption is due to a tyrosyl ligand-to-metal charge transfer transition. Metal substitution experiments show that one of the two iron atoms can be readily replaced by a number of di- valent transition metal cations. The Fe-Zn form of the purple phosphatase retains full activity and exhibits similar kinetics behavior to the native (Fe-Fe) enzyme. This evidence, along with the spectroscopic evidence that the active species is a one electron reduced Fe(III)-Fe(II) form of the native enzyme, suggests that an Fe(III)-M(II) unit is the actual catalytic form. The metal substitution evidence clearly establishes the enzyme as a bimetallic species. Synthetic models with phenoxide ligation to Fe(III) establish the visible absorption band as a ligand-to-metal charge transfer absorption. These models are useful for understanding the properties of the Fe-Zn enzyme and sup— port the presence of tyrosyl ligation in both the Fe-Zn 173 form and in the native enzyme. Figure 39 illustrates the probable structural features of the purple and pink forms of the purple acid phosphatase from bovine spleen. The covalently bound phosphate stabilizes the purple form of the enzyme and thus is responsible for the rapid reversion of the pink form to purple upon addi- tion of saturating concentrations of phosphate. Both iron atoms are shown as containing tyrosyl ligands, as one-half the visible absorption is lost upon replacement of one iron with zinc. The Fe-O-Fe bridge angle is shown as W1800, although the unit may be bent to some extent. This minimal model is consistent with the bulk of the data obtained dur- ing the research presented in this dissertation. The model may not be valid for all other purple phosphatases; but the strong similarity between the bovine spleen and porcine uterine fluid purple acid phosphatases is unlikely to be fortuitous. 174 .mEHOm Axcflmv pwospmu cam Amamusdv Ummflpfixo wnu mum cBosm .mmmumcmmond pflum madman :mmHmm wcfl>on :fl Hmucwo cone o3u on» NO mwusuosuum pmmomoum .mm musmflm 175 mm musmflm _n_ + xEd Lmm ... OI emu lo lawn. @ . \ O @ i .3 28:1 tmmo lama \amulol am“. lo 0 © @ PART I I LITERATURE REVIEW Enzymes that catalyze the oxidation of a substrate us- ing H202 (donor: H202 oxidoreductases), commonly called peroxidases, are unusual in that individual enzymes exhibit broad specificity. Substrates for a given enzyme may range from inorganic ions such as C1- and I- to complex organic species such as ascorbate. Peroxidases are found in plants, bacteria, fungi and animalsgl; some of the well charac- terized enzymes are the peroxidases from horseradishgz, baker's yeast93, turnip94, Japanese radishgs, cow's milk96, 97, and infected dog uterigB. All of these enzymes thyroid contain heme iron and are typically characterized by an intense visible absorption spectrum. Usually the most in- tense band (called the Soret or V band) occurs in the 380 to 415 nm region; however, some hemoproteins (not all of which are peroxidases) exhibit a red-shifted Soret with Amax in the 415 to 435 nm region. This peak, together with an intense « band at 580-620 nm, gives the enzymes a dis- tinctive green hue and consequently, they are referred to as green heme proteins. 99,100 Jacob and Orme-Johnson have reported a green catalase from Neurospgra crassa, De Filippi and Hult- 101,102 quist have isolated a green heme protein of unknown 176 177 function from bovine erythrocytes, and Stelmaszyn'ska and 103 Zgliczyn'ski discovered a green peroxidase in hog in- testinal mucosa. The enzyme isolated from infected dog .98 . . . uteri is also a green heme perOX1dase termed verdOperOXi- dase by its discoverer, K. Agner. Later the same enzyme was renamed myeloperoxidase (EC 1.11.1.7); this green heme peroxidase is characterized by unusually long wavelength «, B, and y (Soret) heme bands in the visible spectrum. Myeloperoxidase has since been isolated from porcine 104 105,106 leucocytes , human leucocytes , and human mono- cytele7. The green heme peroxidase from bovine spleen has visible absorption spectra similar to those of myeloperoxidase for a number of adductsgB’lOB'log’llo, has similar substrate 109-112, has a molecular weight similar to the large subunit of myeloperoxidasegl’108’113'114, and has a 115,116 specificity nearly identical EPR spectrum (for the native enzyme) 117, by reductive cleavage of mye10per- Andrews and Krinski oxidase, have produced an active hemi-myeloperoxidase with one light and one heavy subunit. Isolation of such a protein suggests that the function, structure, and heme moeity of spleen green hemeperoxidase may be closely related to that of myeloperoxidase. As a result, Optical and EPR spectra, metal assays, molecular weight (estimated both by gel permeation and disc gel electrophoresis), and substrate specificity studies of the bovine and human green heme 178 peroxidases from spleen have been performed and are pre- sented in chapters one and two of part two of this disser- tation. CHAPTER 1 A MYELOPEROXIDASE-LIKE GREEN HEME PROTEIN FROM BOVINE SPLEEN While develOping the purification procedure for beef spleen purple acid phosphatase10 presented in Part I of this dissertation, a dark green band was observed upon carboxymethyl cellulose chromatography of beef spleen ex— tracts; the electronic spectrum was characteristic of a heme protein. The isolation, purification and characteriza- tion of this protein, a peroxidase with properties similar to myeloperoxidase (but with significantly lower molecular weight and unusual catalytic properties), are described in this chapter. Materials and Methods Extraction and Purification The green heme protein was extracted from beef spleen strips (spleens obtained as in Part I, Chapter 1) suspended in 0.25 M KC1 (2 ml/g of spleen) at 4°C, and homogenized in a Waring blender at medium speed for 1 min and at high speed for two min. The homogenate was adjusted to pH 3.5 179 180 with 6 M HC1 and stirred at room temperature for 3-20 h, followed by centrifugation at 9000 x g for 10 min at 20°C. The clarified SUpernatant was filtered through glass wool to remove small waxy particles, and the resulting filtrate was made 0.1 M in ascorbic acid and adjusted to pH 5.5 with 12% NaOH. The solution was then treated with 4 g/L of cellulose phosphate (P-11, Whatman). The P-11 was fil— tered, washed twice with 100 mL of H20, resuspended in 200 mL of 2.0 M KC1, and stirred for 2 h. The P-1l was filtered and washed with 50 mL of 2.0 M KC1, resulting in a pale yellow-green filtrate. At this point, slightly cloudy solutions were centrifuged at 9000 x g for 20 min at 20°C for clarification to avoid plugging the chromatography column in the succeeding step. (All subsequent steps were carried out at 5-10°C in a cold cabinet.) The clear filtrate was diluted to 0.15 M KC1 and loaded onto a carboxymethyl-cellulose (CM-52, Whatman) column (2.5 x 20 cm) at 6.0 mL/min. The column (with a dark green band at the top) was washed with 100 mL of 0.2 M KC1, 0.05 M sodium acetate, pH 5.0, and eluted with a 0.15 to 1.0 M KC1 gradient in 0.05 M sodium acetate buffer, pH 5.0. The KC1 concentration was monitored by measuring the conductivity of the samples and comparing them to a set of KC1/buffer standards. The fractions with absorbance Z 0.05 at 434 nm were pooled, concentrated to 5 mL by ultrafiltration (Ami- con 8MC, PM—lO membrane) loaded onto a Sephadex G—75 column 181 (1.5 x 85 cm), and eluted with 0.2 M KC1. The green frac- tions with A434/A280 3 0.6 were combined and concentrated to 10 mL as before. The final purification step was rechromatography on CM-52 carboxymethyl-cellulose. After loading the green heme protein as before and eluting with 100 mL of 0.2 M KC1, 0.05 M sodium acetate, pH 5.0, the column was sec- tioned, and the first 5 mm were placed in 5.0 mL of 2.0 M KC1 and allowed to stand for 20 min. The suspension was then centrifuged briefly and the purified green heme pro- tein was decanted: A434/A280 was 0.8 for this fraction. Analytical Methods Molecular weight data were obtained by gel permeation chromatography using Sephadex G-75, with bovine albumin, egg albumin, a-chymotrypsinogen, and B—lactoglobulin (all from Sigma Chemical Co.) as standards. Polyacrylamide gel e1ectr0phoresis (both native and SDS) was done on a BioRad model 220 electrophoresis cell with Coomassie blue develop- ment of protein bandsS7. For SDS gels, the same molecular weight standards were used as in the Sephadex G-75 experi- ments, with the addition of lysozyme (Sigma). Disc gels for electrophoresis of native enzyme were composed of stacking (5% acrylamide) and separating (7.5% acrylamide) sections. Stacking gels were prepared in 0.064 M sodium acetate buffer (pH 6.0) and 5.0 mM TEMED; 182 separating gels were prepared in 0.36 M sodium acetate buffer (pH 4.3) and 0.046 M TEMED. Iron content was de- termined as previously describedlo. Chlorination of substrate by the green heme peroxidase (chloroperoxidase activity) was determined according to the 118, using 1,1-dimethy1-4-chloro- method of Morris and Hager 3,5-cyclohexanedione (monochlorodimidon) as substrate. One unit of chloroperoxidase activity is defined by the forma- tion of l umole of dichlorodimedon per minute under the standard assay conditions. Spectral Characterization Spectral data for protein estimation (A280) and for determination of enzyme activity were obtained on a Beck- man DU equipped with a Gilford model 252-D accessory. Elec— tronic spectra were obtained on a Cary 219 spectrometer. The EPR spectrum was run on a Bruker ER-ZOOD spectrometer equipped with an Oxford liquid helium cryostat. All reduced heme optical spectra were run using an argon flushed cell; all reduced samples were stored under argon. Reduction was done with 170 mM sodium dithionite (DTH). The NO adduct was formed by reduction of the heme iron with DTH, fol- lowed by addition of 0.05 M sodium nitrite in a 100 mM solution of DTE. The pyridine hemochromogens were prepared by making the enzyme solution 30% (v/v) in pyridine. Some of the optical spectral data were obtained on samples of 183 protein prior to the column sectioning on CM—52; electro- phoresis results indicated that these samples were m80% pure with m20% contamination by the copurifying purple acid phosphatase. The latter has been purified to homogeniety (Part I) and is EPR inactive, has zero peroxidase activity and makes only an insignificant contribution (i 5%) to the visible absorption under the conditions examined here. Resonance Raman data were obtained using the 413.1 nm line of a Spectra Physics 164-ll krypton laser and a Spex Ramalog Spectrometer67. The Spectrum was generated by 90° transverse excitation with the sample contained in a 3 mm x 5 mm square flat bottom quartz (Precision Scientific) cell. Optical spectra were taken before and after the ex- periment to ensure that protein degradation did not occur. The sample cuvette was maintained at 4°C using a stream of cold nitrogen. Results Isolation The isolation of the green heme protein is detailed above under "Materials and Methods". Figure 40 shows a typical carboxymethyl-cellulose chromatogram with the KC1 gradient superimposed. The green band elutes at a KC1 concentration of 0.65 M, suggesting that the protein is 184 .5593 on E: omv um ppm ACV E: 0mm um :oHuQHOQO nuns lav ucmflomnm Hos An :oflusflm .AcmEum;3uNmuzov mmoHsHHmo assume Ixxonumo co mmmpflxoumm mam: cwmum may mo >£Qmuooumeouzo mmcmnoxm :oH .ov ousmflm 185 ow masons ._E .mE=_o> cozam noon. CON OO_ l v.0 I 0.0 I QC o. x 83 o 3.9.. .38: .1 om 186 highly basic. Figure 41 shows the elution profile of the enzyme on Sephadex G-75; a small amount of a lower molecu- lar weight protein elutes shortly after the green heme protein. Typical purification data are shown in Table 14 for the cellulose phosphate, carboxymethyl-cellulose, and Sephadex G-75 steps. A convenient criterion of purity is the ratio of the absorbance at 434 nm to that at 280 nm; this ratio rises throughout the purification to a final value of 0.80. Analysis of the gel electrophoresis results (see below) suggests that the protein is Z 95% pure. Figure 42 shows densitometer traces of three native enzyme electro- phoresis gels run simultaneously; plot 1 is the scan of an unstained gel at 430 nm and illustrates the position of the green protein; plot 2 is the scan of an identical gel, stained with iodide and H202 at pH 7.0 to show the location of peroxidase activity; plot 3 is a scan of another identi— cal gel, stained with Coomassie blue to show the position of protein bands. The coincidence of these peaks strongly suggests that the visible absorption and peroxidase activity are due to the same apparently homogeneous protein. Visible Spectral Data Figure 43 shows the spectrum of the native (Fe+3) form of the protein along with those of the CN- adduct and the oxidized and reduced hemochromogens. Of particular interest are the remarkably long wavelength a, B, and Y bands in all 187 .c3onm Auuluv E: owm new A v E: omv um mcoHUQHOmnm nuHB mhuw xwpmnmmm co mmlzu Eouw mcoflpownw compo on» mo xzdmumoumaounu .Hv musmflm 188 Hv wusmflm JE\i_O> 20:.3m 00. _.o No 8.3.834 Md v.0 189 Table 14. Purification of Green Heme Protein From Beef Spleen. Total Total Specific Step Proteina Activity Activity Recovery mg units units/mg % Acid extractc 111,000 49,000d 0.44 100 p-11 1,760 34,000d 19 70 CM-52 60 7,000 117 14 G-75 21 5,760 275 11 CM-SZ-column sectioning 12 4,800 400 9 aEstimated by assuming that an absorption at 280 nm of 1.0 is equivalent to 1.0 mg of protein/m1. bDetermined by oxidation of p-aminobenzoate (0.04 M) in 0.1 M phosphate buffer, pH 7.0. with 0-5 mM H202: AA625 0.010/min is one unit of enzyme activity. of CFrom 100 g of beef spleen. dMay include non—green heme peroxidases. 190 Figure 42. Disc-gel electrophoresis of green heme peroxi- dase on 7.5% polyacrylamide. Plot 1, 1:430 nm, no stain; plot 2, IT H202 activity stain, A=350 nm; plot 3, Coomassie blue stain, 1:625 nm. 1131 I) 007111 «11...... 1..” 2) R 0151A Absorbonce 3x22, Kl 1:330 3) ( 0.3oIA l J llj 3 2 I o (O 00L. ‘4*' . m:— r) (fib- 3 A- Figure 42 192 .A...v comoeouno IOEwn mcflpfluaa pmospmu wnu can .A v CODOEOHLUOEOQ mafipflumm pmuflpflxo may .Al I Iv uosppm wpflcmwo AHHHth mnu .AI.I.IVOE>Ncm w>flumc AHHHVwm on» mum CBOQm .mmmpflxoumm mam: cwmum mcu mo muuomdm COAuQHOmnm wanflmfl> .mv musqflm 193 ms mucous Etc: Nd v.0 m0 194 four spectra. Treatment with N; (75 mM) caused no discern- ible changes in the native spectrum. Figure 44 shows the spectra of the reduced heme and the CN-, N-, NO, and CO adducts of the reduced (Fe+2) form of the enzyme. Table 15 summarizes the spectral data for the a, B, and y bands of the various forms of the protein; values reported for derivatives of myeloperoxidase are included for comparison. Electron Paramagnetic Resonance Spectroscopy Figure 45 shows the X-band EPR spectrum of the green heme protein. The g values of 6.81, 4.99, and 1.94 are characteristic of high spin ferric heme (for example, the fluoride adduct of catalase)109. Molecular Weight Studies The similarity of the above visible spectra to those of the corresponding derivatives of myeloperoxidase91'108' 113'114 suggested that the green heme protein from beef spleen might be myeloperoxidase. Figure 46 shows the mol- ecular weight estimation for the green green heme protein on a Sephadex G-75 column (1.5 x 85 cm) using four marker proteins of known molecular weight. A value of 57,000 is obtained for the molecular weight of the green heme protein from beef spleen. Similar results were obtained using SDS- polyacrylamide gel electrophoresis; the data are shown in Figure 47. Comparison of the SDS and native gel data Figure 44. 195 Visible absorption spectra of the reduced Fe(II) forms of the green heme peroxidase. Shown are the reduced native form ( ), the CN“ adduct (------), the N3 adduct (...), the NO adduct (-°-°), and the CO adduct (----). 196 700' 600 x(nm) Figure 44 500 400 197 Table 15. Summary of Visible Absorption Data for Green Heme Peroxidase and MyelOperoxidase. Absorption Maxima of Following Bands (nm): a Form a B y Native (Fe3+) GHPb 574 500(sh) 434 Native (Fe3+) MP0C 580 500(sh) 433 GHP (Fe3+) + CN'b 632 452 GHP (Fe3+)PyHCb 580 434 MP0 (Fe3+) PyHcd 590 434 GHP (Fe2+) PyHCb 590 438 MP0 (Fe2+) PyHcd 597 438 GHP (Fe2+)b 636 590 473 MP0 (Fe2+)c 636 472 GHP (Fe2+) + cu'b 613 522 461 MP0 (Fe2+) + CN-e 615 463 GHP (Fe2+) + Mgb 614 462 MP0 (Fe2+) + Ngb 615 460 GHP (Fe2+) + cob 629 459 MP0 (Fe2+) + cof 634 468 GHP (Fe2+) + Nob 626 454 aGHP, green heme protein described in this work; MPO, myelo- peroxidase; PyHC, pyridine hemochromogen; sh, poorly resolved shoulder. bThis work c Reference Q: Reference e Reference HI Reference 98. 113. 111. 126. 198 .x H.0H .ousuwumemu “me QOH .aocwsvmnm coHumH 1:605 “m ~.o .ucmumcoo 08H» “HICHE o ooom .wumu mcflccmom “0 m .wpsuflHQEm :ofiumaspoe “muum3HHHHE 0H .Hm3om w>m3ouoHE ammo mov.m .wocmsqmum ”mumz wmoomouuommm mam MOM mcofluflpcou .Ammmxomm mowmmum>m chmflm :0 mcmom OH HO coflumafldeoov moa x H :Hmm pm mmmpflxoumd mam: comma wnu mo Esuuowmm mam .ma mnsaam 199 mv musmfim 0 000? 0 OOON 00 _ mm? 080m 200 . Emmamm :mEDmnI .cmmadm mat/om" ‘v xwmd comm mo mmpw unflpmma 93 mo macaw may mo ummoumucfl we meflEuwump mmB AwEsHo> coflusHmv w> .GHEDQHM Ezuwm wcw>on .o “CHEDQHM>O .U “cwmocflmm>uu08>£ond .m “GHHDQOHmouUMHIm .d “meB mpumpcmum .mhuw co coHumEfluww unmflm3 unasowaoe cofiummEHmm How .mv wudmflm 201 00. we musmwm Po. x 2963 6.8662 mm Nn m. _ O. (°/\-"/\ ) / (°/\ -a/\) 202 .cflmfluo Eoum ucoum mo wocmumep >3 Umpfl>flp meEMm mo wocmumflp .moxmt .mam wa.o .im.m mac meaosam z NaH.o .meae z mmo.o .uwmmsn oceans“ umom wv.o .Am.m may Humlwflue z v.a .Hmwmsn Ham “Hem mcflxomum wm .me mpwfimamuom wma “Dov “chcmno\<fi w ammmcxoflcu Add as m.H “mumB mCOHuflpcoo ofiumuozmouuomam .cHESQHm mcH>on .m “CHEDQHm>o .o “cmmocflmmxuuoewnolc .U “seashonouomalm .m umE>N0m>H .d "wumz mpumpcmum .mmmpfixouwm mew: cmmum mcu mo ucmfiwz umanomaoe may Mo wpweflumm memmnonmouuomam Hmmlmom .nq musmflm 203 mm .1 111.1... m. . .uflufiuallvu L be wusmflm To. x E063 8336.2 mm m. _ 0. 0.0 I the .x 204 indicate the presence of a single polypeptide chain of molecular weight 56,000. Scanning of the stained native gel on the Gilford DU showed the green heme protein to be 3 95% pure. Iron Content Colorimetric iron analyses suggest an iron content of 1.2 g atoms of Fe/57,000 molecular weight protein. The protein concentration used to obtain this value was deter- mined from the absorbance at 280 nm and the assumption that the molar absorptivitity at that wavelength is the same as that reported by Agner for crystalline myeloperoxidasell4. This result must thus be regarded as approximate, until suf- ficient quantities of the green heme peroxidase are ob- tained to permit an accurate gravimetric determination of the molar absorptivity. Substrate Specificity Several typical substrates for peroxidases were tested at pH 5.0 and 8.5; the rate of reaction was monitored by visible or UV absorption change at the appropriate wave— length. Results are given in Table 16. The green heme peroxidase was also tested for catalase activity by addi- tion of enzyme to 5 mM H202 solution at pH 8.5; no catalase activity, as evidenced by evolution of gas or decrease in H...- .F— Table 16. The conditions were: 205 Substrate Specificity of the Green Heme Per- oxidase from Bovine Spleen. 0.05 M acetate buffer (pH 5.0) or 0.05 M sodium bicarbonate (pH 8.5); [H202]=2.2 mM; [enzyme]=15 ug/ml; pathlength, 1 cm. Concen- Substrate tration pH 1 Result M nm I'a 0.04 8.5 350 0.5 unit/mg I'a,b 0.04 5.0 350 400 units/mg AscorbateC 0.005 8.5 280 No reaction AscorbateC 0.005 5.0 280 No reaction Pyrogallolc 0.005 8.5 430 0.13 A/min Pyrogallol 0.005 5.0 430 m3.0 A/min l-Naphthold 0.035 8.5 550 No reaction l-Naphthold 0.035 5.0 550 No reaction p-AminobenzoateC 0.005 8.5 500 0.2 A/min p-AminobenzoateC 0.005 5.0 500 2.2 A/min Pyrogallolc'f 0.005 8.5 430 0.02 A/min aMethod of Reference 109. bConditions the same as in the legend except [enzyme]=l.5 ug/ml. CMethod of Reference d eMethod of Reference f of H202. Method of Reference Conditions the same 119. 120. 121. as in the legend except for the absence 206 absorption at 240 nm, was observed. Addition of the green heme protein to pyrogallol at pH 8.5 in the absence of H202 resulted in the slow formation of a red color, suggest- ing weak oxidase activity. Using the same conditions as 111 for iodide oxidation and assuming that an Hosoya, pp 31. absorption of 1.0 at 434 nm corresponds to a protein con- centration of 1.0 mg/mL, we find that the Specific activity of the green heme peroxidase for iodide oxidation is 30 units/mg. This is below the value reported for lacto- peroxidase (108 units/mg), but much higher than that found for myeloperoxidase (5 units/mg)112. Results of the chlorodimidon assay for chloroperoxidase activity show that the green peroxidase is capable of oxidizing chloride ion. Under the stated conditions of the assay, one mole of enzyme catalyzes the formation of 36,000 moles of dichloro- dimedon per minute from monochlorodimedon, chloride ion, and hydrogen peroxide. Attempts at Heme Removal Two attempts were made to remove the heme moiety from the protein. First the classical HCl/acetone method of extraction was attempted. One mL of enzyme (%2 mg/mL) was mixed with l M HCl in acetone (2 mL); precipitation of the protein occurred and the green color due to the heme remained associated with the precipitate. The method of 122 Paul for cleaving thioethers in cytochrome c (using 207 AgNO3 in glacial acetic acid) was also attempted. Again the green chromophore remained with the protein fraction, although significant bleaching of the green color was ob— served. Similar bleaching was also noted upon addition of a large excess of HCl (final concentration ml M). Resonance Raman Spectrum Figure 48 shows the resonance Raman spectrum from 1 1 1100 cm- to 2000 cm- of the green heme peroxidase. Typi- cal strong metalloporphyrin absorption bands appear in the region 1100 to 1650 cm-1; the large number of bands and the strong doublet at ml600 cm.1 are particularly unusual features. Discussion The isolation of the green heme protein described above relies on a convenient batch absorption to and elution from cellulose phosphate (P—ll). This results in at least an 80- fold purification with very good recovery (m70%). At this point, however, there is still too much residual absorbance in the 400 to 450 nm region for the A434/A280 ratio to give any indication of the actual amount of green heme protein present. The subsequent step, chromatography on carboxymethyl-cellulose, consistently results in only a 20% recovery of peroxidase activity. This, together with the r_:nfi13 208 cmmum cmmamm mCH>OQ mo A6: H.MH¢H coflummfloxm .A+mmmv mmmpflxoumd mam: «v Esuuomdm cmEmm mocmc0mmm .mv musmflm 209 GO: OON_ . VON. 00m. _ . hog moat mmm. 00v. .92 9%. we mnsmaa com. 000. 00.: com. d a q u . : 63 l. . Own. _ mvm. . mam. mm. OOm_ OOON #1 210 observation that a dark band appears at the top of the CM- 52 column during loading and does not migrate even with 2.0 M KC1, suggests that at least a portion of the peroxidase activity in the crude extract may be associated with other colored proteins. The actual recovery of green heme per- oxidase may thus be higher than indicated in Table 14. The protein has been purified at least 900-fold with Z 10% recovery in only four steps. The major impurity in the preparation after Sephadex G-75 chromatography is the purple acid phosphatase discussed in Part I of this dissertationi23-lzq Absorption of the green protein onto a CM-52 column followed by column sectioning removes the purple acid phosphatase impurity. A comparison of Spectroscopic properties suggests that the chromophores of the green heme peroxidase and of myelo- peroxidase are Similar, if not identical. Thus the position of the Soret (y) peaks in the optical spectra of the two proteins agree within 1-2 nm for the native (ferric), oxidized and reduced pyridine hemochromogens, and ferrous cyanide and azide derivatives, while the positions of the long wavelength (0) peaks are also generally in agreement. The observed differences in the 0 peaks are greatest for the pyridine hemochromogens (approximately 7-10 nm) and may reflect the use of different concentrations of pyridine in preparing the samples. The only serious discrepancy is in the Spectra of the CO complex. We observe Soret and 0 bands at 459 and 629 nm, respectively, compared to 486 and 634 nm 211 for myeloperoxidase126. In addition to the position of the peaks, the observed intensity ratios of the long wavelength features to that of the Soret band are also essentially the same as in myelOperoxidase. These Spectral features are all substantially red-Shifted compared to the Spectra of analogous derivatives of common heme proteins, such as cytochromes b, a, and even P-450127 , suggesting an unusual structure of the heme prosthetic group. The resonance Raman Spectrum of the native green heme protein also supports an unusual structure for the heme moiety. The number of bands and the positions of the major absorptions are unlike those of heme a (in any of its spin states or coordination numbers), metmyoglobin (F-, H 2 or N; coordinated), or typical protohemeslzg. It is pos- 0: sible that the large red-shift of the Spectra of the green heme peroxidase may be due to the presence of a keto or formyl group on the ring periphery; however, the absence of a strong Raman peak above 1620 cm.1 argues against thislzg. If, however, the formyl moiety is distorted out of the heme plane such that the resonance enhancement is minimized, then the C=O stretch may not be observed under the conditions of the experiment; in such a case, however, the C=O will not be conjugated to the porphyrin and cannot be responsible for the large red-Shift. In addition, the EPR spectrum of the native form of the green heme peroxidase is nearly identical to that of 212 myeloperoxidase (g values of 6.81, 4.99, 1.94 for the former and 6.82, 5.05, and 1.91 for the latter)115. These Spectra are again atypical of normal heme proteins; for myeloper- oxidase, the spectrum has been attributed to a quantum mechanical mixed Spin StatelBO. The Similarities between the green heme peroxidase and myeloperoxidase are not limited to Spectroscopic properties. Both are highly basic proteins with moderate peroxidase activity toward a variety of substrates, both appear to 2 contain heme prosthetic groups tightly bound via covalent bonds involving other than ester or thioether linkages. There are, however, Significant differences between the two proteins with respect to apparent molecular weight, sub- strate Specificity and specific activity, and distribution in tissues, which suggest that the two are not in fact identical. Certainly the major difference between the two enzymes is the discrepancy in observed molecular weight. Our gel permeation and acrylamide gel e1ectr0phoresis data clearly Show that the green heme peroxidase from beef Spleen is a relatively low molecular weight molecule (W57,000), probably consisting of a Single polypeptide chain. In contrast, myeloperoxidase is reported to be a tetrameric protein (molecular weight m140,000) with an 0282 subunit Structure; the a and 8 subunits are estimated to have molecular 131 weights of W57,000 and 10,500, respectively It is worth 213 noting that one heavy subunit of myeloperoxidase apparently contains the covalently bound heme moietyl3l. There are thus extensive similarities between the green heme peroxidase from beef spleen and the heavy subunit of myeloperoxidase. The work of Andrews and Krinski on reductive cleavage of myeloperoxidase to form hemi-myeloperoxidase may have Significant bearing on the isolation of the beef spleen peroxidasell7. Reductants (such as 40 mM DTE) are ap- parently capable of cleaving the disulfide bond(s) link- ing the two large subunits of myeloperoxidase. Since 0.1 M ascorbate iS routinely used in the bovine green heme peroxidase purification (at the P-11 step), it is possible that the product isolated is actually a reductively cleaved fragment of the holoenzyme myeloperoxidase. The absence of a 10,500 molecular weight subunit upon SDS gel electrophoresis of the bovine green heme protein, however, makes a Simple cleavage (such as that of myeloperoxidase to hemi-myeloperoxidase) unlikely. The retention of full specific activity by the hemi-myeloperoxidase does not illuminate the Situation Significantly; only full Specificity studies of hemi-myeloperoxidase including substrates such as ascorbate, iodide, and l—naphthol will suffice to Show a strong relationship between hemi-myeloperoxidase and the green heme peroxidase from bovine spleen. Significant differences exist in catalytic properties of the green heme protein and the holoenzyme myeloperoxi- dase. Although both are able to oxidize a relatively 214 broad Spectrum of substrates, the green heme protein is unusual in that it cannot oxidize ascorbate (a good sub- strate for myeloperoxidase), despite oxidizing iodide rapidly. This selectivity cannot have a thermodynamic basis, inasmuch as ascorbate (E6 = +0.06‘V)132 is much more easily oxidized than iodide (E6 = +0.54 VIEIB. The dif- ference must lie in the substrate binding site of the two proteins: that of myeloperoxidase must be able to accommo- date ascorbate as well as iodide, while that of the green heme peroxidase cannot tolerate the relatively bulky and highly charged ascorbate molecule. Despite the difficulty of comparing Specific activities under precisely the same conditions, it is clear that the green heme peroxidase has a significantly higher Specific activity than myelOperoxi- dase for iodide oxidation (30 versus 5 units/mg). The value for the green heme peroxidase approaches that of lactoperoxidase, an enzyme that is clearly differentiated from green peroxidases by its Spectroscopic properties. The green heme peroxidase also Shows chloroperoxidase activity: one mole of enzyme produces 36,000 moles of product per minute. This value is in the same range as that of chloroperoxidase from Caldariomyces fumago (67,000 moles of product per minute per mole of enzyme)135. The fact that the Spleen green heme peroxidase has the ability to chlorinate substrate using readily available chloride is not totally surprising. The generation of active chlorine 215 at the Site of infection is a normal function of mammalian immune systems and the Spleen is known to function as an integral component in immune functions. Finally, myeloperoxidase has not been isolated to date from reticuloendothelial tissue such as spleen. A recent report of the isolation of myeloperoxidase from human mono- cytes suggests that it should be present in spleen, which contains significant amounts of monocytes in addition to lymphocytes and erythrocytes. The precise relationship of the green heme protein described here to myeloperoxidase remains unclear. The possibility that it corresponds to the heavy subunit of myeloperoxidase and arises from de- gradation of the latter during the acid extraction and subsequent steps cannot be ruled out. However, varying the temperature 4-200C) and the duration of the acid extraction step has no Significant effect on the purification, nor does inclusion of 1 mM PMSF (a potent inactivator of serine protease5134). ‘I CHAPTER 2 A MYELOPEROXIDASE-LIKE GREEN HEME PEROXIDASE ISOLATED FROM HUMAN SPLEEN An unusual green heme peroxidase has been isolated from bovine Spleen110 and Shown to be a 57,000 molecular weight protein with properties similar to those of myelo- peroxidase98. A cleavage product of myeloperoxidase, hemi- myelOperoxidase, has been reported to have peroxidase ac- tivity, to have a molecular weight of m70,000, and to contain one heavy and one light subunit of myeloperoxi- dase117 (the native enzyme contains two of each). Green heme proteins have not previously been reported in human reticuloendothelial tissue, although human monocytes (found in spleen) have been shown to contain large amounts of myeloperoxidase. This chapter describes the isolation and preliminary characterization of a green heme peroxidase from human Spleen. Materials and Methods A 1.5 kg human spleen, provided courtesy of N. Dimitrov, M.D. was removed from an adult female with malignant lymphoma and kept on ice for two hours prior to extraction with 216 —- 217 0.25 N KC1, pH 3.5 (2.0 L). Only 1.0 kg of Spleen was used in the extraction as ~0.5 kg was utilized for pathological studies. After homogenation in a Waring Blendor at medium Speed for one minute and at high speed for two minutes, the homogenate was adjusted to pH 3.5 with 6 M HCl and stirred at room temperature for 20 hours. The extract was cen- trifuged at 9000 X g for 10 minutes at 20°C and the super- natant filtered through glass wool to remove small waxy particles. The resulting filtrate was made 0.1 M in as- corbic acid and adjusted to pH 5.5 with 12% NaOH. The solution was then treated with 4 g/L of cellulose phOSphate (P-ll, Whatman). The P-ll was filtered, washed twice with 50 mL of H O, resuspended in 100 mL of 2.0 M KC1, and stirred 2 for one hour. The P-ll was filtered and washed with 25 m1 of 2.0 M KC1 resulting in a pale yellow green filtrate. The cloudy solution was centrifuged at 9000 x g for 20 min at 20°C for clarification to avoid plugging the chromato- graphy column in the succeeding step. (All subsequent operations were carried out at 5-10°C in a cold cabinet.) The clear filtrate was diluted to 0.15 M KC1 and loaded onto a carboxymethyl-cellulose (CM-52-Whatman) column (1.0 x 10 cm) at 1.5 mL/min. The column was washed with 30 mL of 0.2 M KC1, 0.05 M sodium acetate, pH 5.0 and eluted with a 0.15-1.0 M KC1 gradient in 0.05 M sodium acetate buffer, pH 5.0. The KC1 concentration was monitored by 218 measuring the conductivity of the samples and comparing to a set of KC1/buffer standards. The green fractions A434/A280 Z 0.10 were pooled, concentrated to 2.0 mL by ultrafiltration, loaded onto a Sephadex G-75 column (1.5 x 85 cm), and eluted Wlth 0.2 M KC1. The green fractions With A434/A280 3 0.4 were combined and loaded onto the CM-52 column used earlier. After elution with 30 mL of 0.2 M KC1, 0.05 M sodium acetate, pH 5.0, the column was sectioned, the top 5 mm extracted with 5.0 mL of 2.0 M KC1 by Shaking, and the extract de- canted after brief centrifugation. The ratio A434/A280 was 0.67 for this fraction. Electronic spectra were obtained using a Cary 219 Spectrophotometer; activity data using I- as substrate were obtained using a Beckman DU spectrophotometer equipped with a Gilford 252D accessory. Experimental conditions were identical to those for the bovine spleen enzyme activity. Reduction and adduct formation with CN- and NO was per- formed as described in Chapter 1IO7. Molecular weight estimation was done using the same standards and the same Sephadex G-75 column used for the bovine Spleen enzyme, also detailed in Chapter 1. Results The human green heme protein was purified in the same manner as the enzyme from bovine Spleen; in all stages of the purification the human enzyme behaved in a manner 219 similar to that of the bovine enzyme. The final purity achieved with the human enzyme was, however, lower than that of the similar bovine spleen enzyme with R2 (A434/ A280) of 0.67 versus 0.80, respectively. Figure 49 illustrates the CM-52 chromatographic step; shown are absorbance at 280 and 430 nm and the KC1 grad- ient used. The green enzyme elutes with 0.65 M KC1. Figure 50 Shows the G-75 chromatography step, with a broad protein ( A280) peak and a narrower peak due to the green protein. Figure 51 Shows the visible absorption Spectrum for the native (as isolated) human enzyme; principal features include a large peak at 434 nm (Y band) with shoulders at 370 nm and at 500 nm (8 band) and a peak at 572 nm (a band). Both the relative Size and positions of the peaks in the human enzyme spectrum are essentially identical to those of the bovine Spleen enzyme. Table 17 summarizes the visible ab- sorption data for the beef Spleen and human spleen enzymes and several adducts. Except for the reduced forms, all Spectra of the two enzymes have maxima agreeing within 3 nm. The spectra of the reduced forms of the two enzymes are Similar, but the Soret (y) band of the human enzyme is at 5 nm Shorter wavelength than the corresponding band of the bovine enzyme. Experiments using I- as substrate Show that, like the bovine enzyme, the human Spleen green heme protein exhibits peroxidase (activity, utilizing hydrogen peroxide to form I3 220 . HOV E: one an :oHumHOMnHm man can . HOV Es omm um :oHumHOMnHm mnH . H3 unmSHm HUM Ho coHumuucmocoo mnu mum czozm .cmmHmm Eoum mmmponumm mam: cmmum amass mcu mo >£QmumoumEouno ucmemum HUM mmoHsHHmuonnHmE>xonumu .mv musmHm 221 cm ma wusmam 1:: . ._o> cozam om 00. 0? ON i O 4 . nMwHu {v.0 6.0 O_ x 034 o . o._ 034 . 3358: 100. 1 0.. ON 222 .AIIIIV E: omv um can A v EC omm um coHumu0mnm mzv mum czonm .cmdem Eoum mmmponumm mam: :mmum amen: mnu mo >£dwumoumeouno menu xmpmcdmm .om mHDmHm 223 On. 1 4E .._o> cozam 09 on _ 1,1 \ . _.O . No 224 .cmmHmm Eoum mmmpon Inmm mam: cmmum 268:: A+mmmv m>Hpmc may wo Enuuommm coHumu0mnm mHnHmH> .Hm musmHm Hm mHDmHm E: .4 CON com com 00¢ H H 225 226 va omv III III mmm mmm 02 + H+Nmmv UOODUOm Hea owe mmm omm MHe CH6 .20 + A+N0MS twosomm mnv mow omm mmm mmm mmm A+Nmmv pmospmm mme omv III III mmm omm :20 + +mmmv m>Humz vmv «me com com «um mum H+mmhv m>Humz mcH>om :mEsm mcH>om :mEdm mcH>om amasm Show “Ecv mEmez :oHuQHOm24 .cmmHmm mcH>om paw :mEsm Eoum mmmponumm mEmm :mmuw MOM mumo coHumu0mQ< mHQHmH> mo mumEESm .hH mHnme 227 from I-. The Specific activity of the enzyme is difficult to ascertain since the purity is low; however, the Specific activity is in the same range as the bovine enzyme. Figure 46 (Chapter 1) shows the results of Sephadex G-75 chromatography used to determine the molecular weight 58 of the enzyme. The method of Porath was used to estimate the molecular weight, which resulted in a value of 58,000 g per mol. This is in good agreement with the values of 56,000 and 57,000 g per mol determined for the molecular weight of the bovine enzyme using gel electrophoresis and G-75 gel permeation chromatography. Discussion The isolation of human green heme peroxidase described above is identical to the procedure for the purification of green heme peroxidase from bovine spleenllo. The yield of human enzyme is substantially lower than that typically achieved with bovine spleen (about 2 mg/kg vs. 6 mg/kg), possibly accounting for the lower degree of purification obtained. The visible absorption Spectrum shown in Figure 51 is identical in every respect with that of the bovine Spleen 110 green heme peroxidase and with that of myeloperoxi- dase98. As can be seen in Table 17, the human and bovine enzymes are also much alike in terms of the visible absorption spectra of their oxidized (Fe+3) cyano and 228 reduced (Fe+2), reduced cyano, and reduced NO forms. In no case do the maxima of the spectra differ by more than 5 nm. Both enzymes also have similar molecular weights as determined by gel permeation chromatography (bovine Spleen N57,000 g per mole and human spleen m58,000 g per mol). The Spectral similarity of both enzymes to the hemi-myelo- peroxidase produced by reductive cleavage of myeloxeroxi— dase117 suggests a common chromophore on these enzymes. If femur-111 the heavy subunit of myeloperoxidase from human monocytes is indeed similar to the green heme protein from human spleen, then immunochemical techniques may Show that Similarity by demonstrating cross-reactivity between antibodies prepared using the two enzymes. Certainly much additional information in terms of resonance Raman, EPR and electronic Spectroscopy, as well as detailed substrate specificity studies, will be required to establish an unequivocal relationship between the bovine and human enzymes and myeloperoxidase. These will, however, be limited by the availability of sufficient enzyme. SUMMARY Novel green heme proteins from human and bovine spleen have been purified to near (human) and apparent (bovine) homogeneity. Both enzymes have Similar visible spectra with unusually long wavelength absorptions due to a, B, and y porphyrin bands; the EPR values of the bovine protein are 6.81, 4.99 and 1.94 g; (EPR values for myeloperoxidase are similar to those found for the bovine enzyme). Both proteins have peroxidase activity and a visible Spectra Similar to myeloperoxidase (EC 1.11.1.7). The observed molecular weights of the green heme proteins from spleen are Similar (bovine m57,000 and human m58,000 daltons) much lower than that found for myeloperoxidase, clearly distinguishing the two spleen proteins from myeloperoxidase. Many of the common derivatives of heme proteins, such as the reduced form and the CN-, CO, N0, N3 adducts have similar optical spectra to myeloperoxidase but Show minor differences. Substrate Specificity is also different for the bovine Spleen enzyme, ascorbate being a good substrate for myeloperoxidase but a poor one for the green heme per- oxidase. It is possible that the human and bovine spleen enzymes are Similar to hemimyeloperoxidase, a protein prepared by reductive cleavage of myeloperoxidase, but the available data are not sufficient to establish this conclusively. 229 APPENDIX 230 .mmmm OOHBOHHOM map so cm>Hm MH paw momsmcmH UHmflm CH cmHHHu3 mmB Emummm m\m u m .m\m n m map MOM x mo mmsHm> mumumcmm Op Omms Hammond amusmeoo mne OH+iaxm\eauvmx0m+iex\emncaxme+Aex~\eamucaxma+ieaxanuvax0m ex 2 I x mxmea+iexm\eaacmx0~a+iex\emucawm~\mm+iex~\eHmuvmxwm+iexxeNHucaxmmxH memo "OOH msqm m2H3O o m mMH .u . HH m nu mchs pmchuno mHmB Emumwm Hm u m .m\m n my map How mmHuHHHQHummomsm OHumcmmE HOHHEHm .ex\eomucawaH+iax\Somuvaxma+exxe~Hucaxme+leakeeuaxmm+iex\nmuvaxmm+H ex u x iexxeomuvmxmoHH+iexxeomuvmxmoe+iax\e~Hucaxmmm+iex\eeuwmmeH+iex\nmuvmxmm Nemmz mMHucOHumsqm OCHBOHHOM mnu OH OSHOMOOOM pmcHEumump was Emumwm chm zouuomHm pmHmsoo Am\m u m .N\m n my map mo >UHHHnHummOmsm OHumcmmz < xHozmmmfl 10 15 20 30 40 50 2010 2020 2030 231 '5/2 MAG' INPUT "ENTER A VALUE FOR J"; J INPUT "ENTER A VALUE FOR TEMPERATURE"; T GOSUB 2010 PRINT "VALUE OF x IS" x "FOR J = "J" AND T = "T (30 TO 15 Y = J/(.695-T) x = .251/T-((2-EXP(-2-Y)) + (10-EXP(-6-Y)) + (28-EXP(-12-Y)) + (60-EXP(-20-Y)) + (110oBXP (-30-Y)))/(1+(3-EXP(-2-Y)) + (5-EXP(-6-Y)) + (7-EXP(-12-Y)) + (9-EXP(-20~Y)) + (ll-EXP (-30 Y)))-3 RETURN E3 _________------ 11:...i- I REFERENCES 10. 11. 12. 13. REFERENCES Boyer, P. D. "The Enzymes", Vol. IV, 3rd Ed., Aca- demic Press, New York, 1971, p. 337-746. Chlebowski, J., and Coleman, J. E. (1976) in "Metal Ions in Biological Systems", V. 6, H. Sigel, ed., 1-140, M. Dekker, New York. Spiro, T. G. (1973) in "Inorganic Biochemistry," V. 1, G. L. Eichorn, ed., Elsevier, New York, p. 549. Reid, T. W., and Wilson, I. B. (1971) in "The Enzymes" V. IV, p. 373, P. D. Boyer, ed., Academic Press, New York, NY. Hollander, V. P. (1971) in "The Enzymes", V. IV, p. 450, P. D. Boyer, ed., Academic Press, New York, NY. Nochumson, S., O'Rangers, J. J., and Dimitrov, V. (1974) Fed. Proc., 33, 1378. Nochumson, S., Ph.D. Dissertation, Hahnemann Medical College (1975). Campbell, H. D., and Zerner, B., (1973) Biochem. Bio- phys. Res. Comm., 54, 1498. Campbell, H. D., DionysiuS, D. A., Keough, D. T., Wilson, B. B., deJersey, J., and Zerner, B. (1978) Biochem. Biophys. Res. Comm., 82, 615. Davis, J. C., Lin, S. S. and Averill, B. A. (1981) Biochemistry 20, 4062. Buhi,W., Bazer, F. W., Ducsay, C.,Chun, P. W., and Roberts, R. M. (1979) Fed. Proc., 38, 733. Schlosnagle, D. C., Bazer, F. W., Tsibris, J. C. M., and Roberts, R. M. (1974) J. Biol. Chem., 249, 7574. Schlosnagle, D. C., Sander, E. G., Bazer, F. W., and Roberts, R. M. (1976) J. Biol. Chem., 251, 4680. 232 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 233 Sugiura, Y., Kawabe, H., and Tanaka, H. (1980) J; Am. Chem. Soc., 102, 6581. Sugiura, Y., Kawabe, H., Tanaka, H., Fujimoto, S. and Ohara, A. (1981)J3 Am. Chem. Soc., 103, 963. Sugiura, Y., Kawabe, H., Tanaka, H., Fujimoto, S., and Ohara, A. (1981) J. Biol. Chem., 256, 10664. Uehara, K., Fujimoto, S., and Taniguchi, T. (1971) J. Biochem. (Tokyo), 12, 183. Uehara, K., Fujimoto, S., and Taniguchi, T. (1970. J. Biochem. (Tokyo), 15, 627. Fujimoto, S., Nakagawa, T., and Ohara, A. (1977). Agri. Biol. Chem., 41, 599. Fujimoto, S., Nakagawa, T., and Ohara, A. (1977). Chem. Pharm. Bull., 25, 3283. Fujimoto, S., Nakagawa, T., Ishimitsu, S., and Ohara, A. (1977) Chem. Pharm. Bull., 2_5, 1459. Igaue, I., Watabe, H., Takahashi, K., Takekoshi, M., and Morota, A. (1976) Agri. Biol. Chem., 40, 823. Jacobs, M. M., Nyc, J. F., and Brown, D. M. (1971) J. Biol. Chem., 246, 1419. Glew, R. H. and Heath, E. C. (1971) J. Biol. Chem., 246, 1556. Knowles, J. R. (1980) Ann. Rev. Biochem., 49, 877. Krebs, E. G. and Beavo, J. A. (1979) Ann. Rev. Bio- chem., 48, 933. Gratecos, D., Detwiler, T. C., Hurd, S. and Fischer, E. H. (1977) Biochemistry, 16, 4812. Ray, K. P. and England, P. J. (1976) Biochem. J., 157, 369. Chou, C. K., Alfano, J. and Rosen, O. M. (1977) J. Biol. Chem., 252, 2855. Khandelwal, R. L., Vandenheede, J. R., Krebs, E. G. (1976) J. Biol. Chem., 251, 4850. DiPietrO, D. L. and Zengerle, F. S. (1967) J. Biol. Chem., 242, 3391. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 234 L81“, K-Wo' Lai, L. C., Burkart' Po To, Yam, L. To (1977) J. Biol. Chem., 252, 3371. Robinson, D. B. and Glew, R. H. (1980) Arch. Bio- chem. BiophyS., 205, 156. Robinson, D. B. and Glew, R. H. (1980) J. Biol. Chem., 255, 5864. Harris, D. L. (1946) J. Biol. Chem., 165, 541. Feinstein, R. N. and Volk, M. E. (1949) J. Biol. Chem., 177, 339. Sundararajan, T. A., and Sarma, P. S. (1954) Biochem. 92. 5_6. 125- ‘— Singer, M. F. and Fruton, J. S. (1957) J. Biol. Chem., 229, 111. Hoffman, T. (1958) Biochem J., 9, 135. Sundararajan, T. A. and Sarma, P. S. (1959) Biochem. g... 7_1. 537- " Glomset, J. (1959) Biochim. Biophys. Acta, 32, 349. Revel, H. R. and Racker, E. (1960), Biochim. Biophys. Acta, 43, 465. Glomset, J. and Porath, J. (1960) Biochim. Biophys. Acta, 32, l. Buhi, W., Ducsay, C. A., Bazer, F. W., and Roberts, R. M. (1982) J. Biol. Chem., 257, 1712. Chen, T. T., Bazer, F. W., Cetorelli, J. J., Pollard, W. E., and Roberts, R. M. (1973) J. Biol. Chem., 248, 8560. Bazer, F. W., Chen, T. T., Knight, J. S., Schlosnagle, D. C., Baldwin, N. J., and Roberts, R. M. (1975) J. Anim. Sci., 41, 1112. Antanaitis, B. C. and Aisen, P. (1982) J. Biol. Chem., 257, 1855. Davis, J. C. and Averill, B. A. (1982) Proc. Natl. Acad. Sci. USA, 79, 4623. Schlosnagle, D. C., Sander, E. G., Bazer, F. W., and Roberts, R. M. (1976) J. Biol. Chem., 251, 4680. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 235 Antanaitis, B. C., Aisen, P., Lilienthal, H. R., Roberts, R. M., and Bazer, F. W. (1980) J. Biol Chem., 255, 11204. Antanaitis, B. C. and Aisen, P. (1982) J. Biol. Chem., 257, 5330. Gaber, B. P., Sheridan, J. P., Bazer, F. W., and Roberts, R. M. (1979) J. Biol. Chem., 254, 8340. Antanaitis, B. C., Aisen, P., Roberts, R. M., Bazer, F. W., and Strekas, T. C. (1981) Fed. Proc., 40, 1665. _ Antanaitis, B. C., Strekas, T., and Aisen, P. (1982) J. Biol. Chem., 257, 3766. Keough, D. T., Dionysius, D. A., deJersey, J. and Zerner, B. (1980) Biochem., Biophys. Res. Commun., ‘33, 600. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem., 193, 265. O'Farrell, P. H. (1975) J. Biol. Chem., 250, 4007. Porath, J. (1963) J. Pure & Appl. Chem., 6, 233. Cleland, W. W., The Enzymes, Vol. II, 3rd Ed., P. D. Boyer Ed., Academic Press, New York, 1970, p. 1. Fisk, C. H. and SubbaRow, Y. (1925) J. Biol. Chem., 33, 375. Bartlett, G. R. (1959) J. Biol. Chem., 234, 466. Van de Bogart, M. and Beinert, H. (1967) Anal. Bio- chem., 39, 325. Lin, S. S., M. S. Thesis, Michigan State University (1979). Hatton, M. W. C. and Regoeczi, E. (1973) Biochim. Biophys. Acta, 327, 114. Aisen, P. (private communication). Aasa, R. and Vanngard, T. (1975) J. Magn. Reson., (£2. 308. Ondrias, M. R. and Babcock, G. T. (1980) Biochem. Biophys. Res. Commun., 33, 29. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 236 Kankeleit, E. (1964) Rev. Sci. Instrum., 33, 194. Averill, B. A., Bale, J. R., and Orme-Johnson, W. H. (1978) J. Am. Chem. Soc., 100, 3034. Erbes, D. L. and Burris, R. H. (1978) Biochim. Bio- phys. Acta, 525, 45. Gaber, B. P., Miskowski, V., and Spiro, T. G. (1979) J. Am. Chem. Soc., 33, 6868. Freier, R. R. (1980) Arch. Biochim. Biophys., 205 499. SjOberg, B—M., Loehr, T. M., and Sanders-Loehr, J. (1982) Biochemistry, 33, 96. Muhoberac, B. B., Wharton, D. C., Babcock, L. M., Harrington, P. C., and Wilkins, R. G. (1980) Biochim. Biophys. Acta, 626, 337. Steukamp, R. B., Siecker, L. C., Jensen, L. H., and Sanders-Loehr, J. (1981) Nature, 291, 263. Mahler, H. R. and Cordes, E. H. in "Biological Chem- istry", 2nd ed. (1971), Harper, New York, p. 665. Johnson, C. R., Shepherd, R. B., Marr, B., O'Donnell, S. and Dressick, W. (1980), J. Am. Chem. Soc., 102, 6227. Okmura, M. Y. V., Klotz, I. M., Johnson, C. E., Winter, M. R. C., and Williams, J. P. (1969) Bio- chemistpy, 3, 1951. Atkin, C. L., Thelander, L., and Reichard, P. (1973) J. Biol. Chem., 248, 7464. Dawson, J. W., Gray, H. B., Hoening, N. B., Rossman, G. R., Schreder, J. J., and Wang, R. H. (1972) Bio- chemistry, 33, 461. Murray, K. S. (1974) Coord. Chem. Rev., 33, 1. Que, L., Jr., Heistand, R. H., II, Mayer, R., and Roe, A. L. (1980) Biochemistry, 33, 2588. Tomimatsu, Y., Kint, S., and Scherer, J. R. (1976) Biochemistry, 33, 4918. Lenhert, P. G. (1975) J. Appl. Cryst., 3, 568. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 237 Doyle, P. A. and Turner, P. S. (1968) Acta Crystallogr., A24, 390. All least squares refinements were based on the mini- mization Ovawi||Fo| = FCII2 with the individual wi = 1/0(Fo)2 . All calculations were performed on a CDC Cyber 170 model 750 computer using a variety Of programs which have evolved from, or been assimilated into, the system Of Allan Zalkin (Lawrence Berkeley Laboratory), private communication (1974). Other programs used in- clude ORTEP by C. K. Johnson, Report ORNL-3794, Oak Ridge National Laboratory, Oak Ridge, Tenn., 1965. Keyes, W. E., Loehr, T. M., and Taylor, M. L. (1978) Biochem. Biophys. Res. Commun., 33, 420. Heistand, R. H., II, Roe, A. L., and Que, L.,Jr. (1982) Inorg. Chem., 33, 676. Bailey, N. A., Cummins, D., McKenzie, E. D., and Worthington, J. M. (1981) Inorg. Chim. Acta, 39, 111. Saunders, B. C., Holmes-Siedle, A. G., and Stark, B. P., "Peroxidase," Butterworths, Washington, 1964, p. 234. Keilin, D. and Hartree,E. F. (1951) Biochem J., 49, 88. '—_ Altschul, A. M., Abrams, R., Hogness, T. R. (1940) J. Biol. Chem., 136, 777. Yamazaki, I., Fujinaga, K., Takehara, I., and Takahashi, I. (1956) J. Biochem. (Tokyo), 33, 377. Morita, Y. and Kameda, K. (1957) Mem. Res. Inst. FOOd Sci., Kyoto Univ., 33, 1. Theorell, H. and Akeson, A., Ark. Kemi, Mineral Geol. 17B, 7. Hosoya, T. and Morrison, M. (1967) Biochemistry, 3, 1021. Agner, K. (1941) Acta Physiol. Scand. 2, Suppl. VIII, 1. __ Jacob, G. S. and Orme-Johnson, W. H. (1979) Biochemistry, 33, 2967. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 238 Jacob, G. S. and Orme-Johnson, W. H. (1979) Biochem— istry lg, 2975. DeFillippi, L. J., and Hultquist, D. E. (1978) J. Biol. Chem., 253, 2946. DeFillippi, L. J., and Hultquist, D. E. (1978) 3; Biol. Chem., 253, 2954. Stelmaszyn'ska, T. and Zgliczyn'ski, J. M. (1971) Eur. J. Biochem., 33, 56. Odajima, T. and Yamazaki, I. (1972) Biochem. Biophys. Acta, 284, 360. Bakkenist, A. R. J., Wever, R., Vulsma, T., Plat, H. and Van Gelder, B. F. (1978) Biochim. Biophys. Acta 524, 45. Matheson, N. R., Wong, P. S., and Travis, J. (1981) Biochemistry, 39, 325. BOS, A., Wever, R. and R005, D. (1978) Biochim. Bio- phys. Acta 525, 37. Advances in Enzymology, Agner, K. V3, 137—148 (1943). Morrison, M. (1970) Meth. in Enzym. 17A, 653. Davis, J. C. and Averill, B. A. (1981) J. Biol. Chem. 256, 5992. Hosoya, T., KondO, Y., Ui, N. (1962) J. Biochem. (Tokyo) 33, 180. Hosoya, T. and Morrison, M. (1967) Biochem., 6, 1021. — Schultz, J., and Shmuckler, H. W. (1964) Biochem., 3, 1234. Agner, K. (1958) Acta Chem. Scand. 33, 89-100. Wever, R. and Bakkenist, A. R. J. (1980) Biochim. Biophys. Acta, 612, 178. Blum, H., Chanceu’B., Gunson, D. E., and Litchfield, W. J. (1978) FEBS Lett., 33, 37. Andrews, P. C. and Krinski, N. I. (1981) J. Biol. Chem., 256, 4211. I 239 118. Morris, D. R. and Hager, L. P. (1966) J. Biol. Chem., 241, 1763. 119. Chance, B. (1950) J. Am. Chem. Soc., 72, 1577. 120. Falk, K. G., McGuire, G., and Blount, E. (1919) J. Biol. Chem., 33, 229. 121. Lipmann, H. (1949) J. Biol. Chem., 139, 977. 122. Paul, K. G., (1955) Acta Chem. Scand.,4, 239. 123. Glomset, J. and Porath. J. (1960) Biochim. Biophys. Acta, 33, 1. 124. Campbell, H. D. and Zerner, B. (1973) Biochem. Bio- phys. Res. Comm., 33, 1498. 125. Campbell, H. D., Dionysius, D. A., Keough, D. T., Wilson, B. B., deJersey, J., Zerner, B. (1978) Bio- chem. BiOphyS. Res. Comm., 33, 615- 126. Odajima, T. and Yamazaki, I. (1970) Biochim. Biophys. Acta, 206, 71. 127. Boyer, P. D., "The Enzymes" 3rd Ed., Academic Press, New York, San Francisco and London, 1976 Vol. XI. p. 399 and Vol. XIII p. 171. 128. Callahan, P. M. and Babcock, G. T. (1981) Biochemistry, 32, 952. 129. Spiro, T. G. and Strekas, T. C. (1974) J. Am. Chem. Soc., 33, 338. 130. Maltempo, M. M. (1974) J. Chem. PhyS., 33, 2540. 131. Harrison, J. E. and Schultz, J. (1978) Biochim. Bio- phys. Acta, 536, 341- 132. Boyer, P. D., Lardy, H., and Myrback, K. eds. "The Enzymes" 2nd ed. Academic Press, New York and London, 1960, Vol. 3, p. 389. 133. Latimer, W. M., "The Oxidation States Of the Elements and Their Potentials in Aqueous Solutions", Prentice- Hall, Inc., New York, 1952, p. 67. 134. Fahrney, D. E., and Gold, A. M. (1963) J. Am. Chem. Soc., 33, 997. 135. Morris, D. R. and Hager, L. P. (1966) J. Biol. Chem., 241, 1769. 240 136. Carlin, R. L. (1970) Transition Metal Chemistry, Vol. 6, p. 24, Marcel Dekker, Inc., New York.