. .3 u 1 luv»! “th“: I. a VOIOAOIDDIIQ! '5 \ I .420 It‘ll. II w o ’15.!!! .1 t: 00 EVIL!) I l‘ . . finahhfiqc‘tdtfi. to..- uni}. nua".bl.| Q ‘uli. ‘l‘v‘l‘lv. .olob‘ I! taou.101 “II1 Liv -l‘.‘ 3.00. ioIQlfl‘Xt‘kl’it-i . a- .-"‘ I.‘\OCr.\.M.NO.I0.I\-‘tlll...l 'luxl 1!...nn)|...l\¢xll ‘llxvtzmnm . i .' I till] i. A . .Illlll' p 'f 55443 .IIIHWh... VEHIAWWMJIHH ‘h‘t‘y‘t‘ - ('0 Al ‘ v I ultk ‘va'l 5"!- avl: . {$3. . II . . rte t .sitdw. ... . JI.»P.2.I.»J!:.J.. oV‘O-onnllln o1... . .. . .53.. .. .4 . a. y. 2 I... ’ o.‘1»)ofltl!\a . 06;. XI I: :13: .I....A . .1. .1... .av‘"- v&$v . v 11".WvIOo-vx.‘ ¢ :47 u.: y. n at it’lbv'ool: ‘a..¢.l i313... - Hal! - xiii]. I 3410an n v: a} .. a. .. . ‘11! I 1: Lo‘ It‘ll.lv a vva'ul II ; K-Xl‘l. u‘ t. . 99.49 .\....I .htflntol!‘ ‘ 4 Iotuuzo‘“ ’ v ‘ .3;v . . . o1 . I. 301.: I: lit.‘ . tic-ll! n bntrlv; ‘nv‘bvv‘n‘ I ! \Illl n v no .Ihuufi “to I... unfinitta fibrin)“ GL1... . 1 Q. .0 :q‘s :1“ ‘ ' a". Elfin}, 'v ‘ I l.‘» '. to” .n ‘ A A- ‘3 ' I r-i-n fl ‘ r:1;.'! '0 . o- - .- “ F‘—-fi-‘ ‘- i .10 I o F ' ‘- W‘éa'dfi «1-3.1». fi-bId'fi;--aV—- -r“-‘-¢ Universit- This is to certify that the dissertation entitled Stopped-Flow Studies of the Reduction of Cytochrome g_0xidase and Magnetic Susceptibility Studies of Cytochrome g_0xidase and Some of Its Derivatives presented by Zexia Kay Barnes has been accepted towards fulfillment of the requirements for Ph . D - degree in Chemi slam ajor professor Date 7/21/86 MS U is an Aflirmatiw Action/Equal Opportunity Institution , 0-12771 MSU LIBRARIES M v RETURNING MATERIALS: Place in book drop to remove this checkout from your record.‘ FINES will be charged if book is returned after the date stamped below. STOPPED-FLO' STUDIES OF THE REDUCTION OF CYTOCEEONE g,OXIDASE AND NAONETIC SUSCEPTIBILITY STUDIES OF CTTOCEEONE g_OIIDASE AND SONE OF ITS DERIVATIVES By Zeria lay Barnes A DISSERTATION Snbnitted to Michigan State University in partial fulfillnent of the reqnirenente for the degree of DOCTOR OF PHILOSOPHY Department of Cheniatry 1986 T‘ J .2 .’ “ .‘O "I I -- I). V t l- , 9 .Li ‘ #07739? ABSTRACT STOPPED-FLOW STUDIES OF THE REDUCTION OF CTTOCERONE §,OXIDASE AND NAGNETIC SUSCEPTIBILITY STUDIES OF CYTOCERONE g,OXIDASE AND SONE OF ITS DERIVATIVES By Zeria Kay Barnes Stopped-flow spectrophotcnetry was used to study the anaerobic reduction of cytochrone g oridase by 5. 10- dihydro S-nethylphenaaine (KPH). CuA reduction was conplete after 400 ns. The 830 an absorbance decay lagged the 605 nn absorbance growth. Analyses of data collected at various NPR concentrations shoved that CuA was reduced via cytochrone g, not directly by IPE. The sane reaction was followed at 444 nn in the presence of Treen-zo and lauryl naltoside. Both reactions had a second-order phase followed by two first-order phases. The rate constants were not affected by the detergent used. Aerobic reduction of oxygenated ensyne by [PR was followed at 444 nn. There was a binolecular phase, a steady-state phase, and two phases that were first order in enzyne. 0f the last two, one contributed only slightly. indicating alnost conplete ensyne honogeneity. The reaction of [PH and cytochrone ; was rate-liniting: the rate constant of the intranolecular 1 reaction was greater than 4.9 s- . The temperature dependence of the nagnetic ;.1 ' 7 -‘Rf; J. ‘0‘; :u \ .‘ 'I so . A " .\ ' v u '3 v‘ ‘ ‘A' I - ‘-I 9 ~' " .— 4 Q ~ 1‘ ' : J " S .' .x R . ' . ‘ E1 “”4", I g? ‘ ‘T T J '1 . II' II ' w' "’ ‘ . . ' o , ' .. -‘ i '3 U 3 ‘ . ' I.“ s . f' .3 V. v . .c v A i ' 3 J "' ‘ . 4 x. 1 ~ ~' ‘11, J 5 J ‘ susceptibility was neasured for resting. cyanide-bound and fornate-bound cytochrone 9_ oxidase. The a3 center of the resting and fornate-bound enzyne showed antiferronagnetic coupling. 'The cyanide-bound ensyne was nagnetically heterogeneous with 20% of the enxyne having a -J value of 30 on"1 and 80% having a I value of 'nagnitude 1-2 OI-l. The type of coupling in the latter is unknown. The nagnetic behavior of the resting and cyanide-bound enxyne was unaffected by the ensyne isolation technique. Use of Tween-20 instead of lauryl naltoside did not affect the susceptibility of the resting enxyne or that of the najor conponent of the cyanide-bound ensyne. The coupling factor of the ninor conponent of the cyanide-bound ensyne was doubled in lanryl naltoside. Use of glucose oxidase and glucose to renove oxygen affected the susceptibilities of resting and fornate-bound enayne, presunably by causing peroxide binding. The spin state of cytochrone Q was found to be 3/2 in peroxide-bound cytochrone oxidase. ACKNO'LEDGNENTS I would like to thank Dr. Janes L. Dye for suggesting this project. for the nany helpful discussions. and for encouraging ne when I needed it. I would also like to thank Dr. Gerald T. Babcock for his help with this project and for the financial support. I also wish to thank the Departnent of Chenistry. the National Science Foundation, and the National Institute of Health for financial support. I would like to thank Dr. Tho-as V. Atkinson. lartin Babb, and Ianfred Langer for the technical assistance always cheerfully given. "Without then there is no science.” I would also like to thank the nany friends at Iichigan State and at Purdue, as well as elsewhere,-and ny fanily, for encouraging no as I did experinents and wrote this nanuscript. 11 LIST OF TABLES LIST OF FIGURES Chapter TABLE OF CONTENTS I. IntrOductiou O O O O O O O O O O O O O O A. Cytochrome Oxidase A.1 wwww wax >>>>> e e e Gu‘tnlo QM‘N NH. Structure . . . . . . . . . . . Metal Centers . . . . . . . . . Optical Spectra . . . . . . . . Forms of Cytochrome Oxidase . . Kinetics . . . . . . . . . Magnetic State of the a3-Cu3 Center . . . . . . . . . . . netic Studies of Ensynes . . . . Types of Magnetic Behavior . Dianagnetic and Background Corrections . . . . . . Curie Law . . . . . . . Zero-Field Splitting . Van Vleck Equation . . Intramolecular Effects II. Materials and Methods . . . . . . . . . A. Materials . . . . . . . Stopped-Flow Experiments C. 3.1 2 3 4 5 S Stopped-Flow System Data Handling . . . . . . . . . Deoxygenation . . . . . . . . . MPH Preparation . . . . . . . . Enzyme Preparation . . . . . . netic Susceptibility Experiments 111 H NQMNHH 17 24 26 27 30 32 33 35 35 36 36 36 37 38 41 41 III. Stopped-Flow Kinetics of Cytochrome Oxidase 43 A. Concentration Dependence of the CuA Reaction . . . . . . . . . . . . . . 43 B. Detergent Effects on Kinetics . . . . 50 B.1 Spectral Changes and the Reaction 53 B.2 Data Analysis . . . . . . . . . . 55 B.3 Discussion . . . . . . . . . . . 60 C. The Reduction of Oxygenated Cytochrome Oxidase . . . . . . . . . . . . . . . 60 C. 1 Procedure . . . . . . . . . . . 60 C. 2 Reaction and Spectral Changes . . 61 C. 3 Data Analyses . . . . . . . . . . 63 C.4 Discussion . . . . . . . . . . . 73 IV. The Magnetic State of the a3 Center in Cytochrome c Oxidase and Some of Its Derivatives . . . . . . . . . . . . . . . 75 A. Survey of the Literature . . . . . . 75 B. Susceptibility Experiments . . . . . 78 B.1 Resting Cytochrome Oxidase . . . 78 B.2 Peroxide-bound Cytochrome Oxidase 83 B.3 Oxidized Formate-bound Cytochrome Oxidase . . . . . . . . . . . . . 87 B.4 Oxidized Cyanide-bound Cytochrome Oxidase . . . . . . . . . . . . . 92 B.5 Cyanide-bound Cytochrome Oxidase in Ethylene Glycol . . . . . . 97 B.6 Isolation and Detergent Effects . lOl B.7 Discussion . . . . . . . . . . . 103 v. Futut. 'ork O O O O O O O O O 0 O O O O O O 110 REFERENCES 0 O O O O O O O O O O O O O O O O O O O 1 13 iv LIST OF TABLES Page III.1 Rate Constants for the Initial Phase of the Anaerobic Reaction of MPH and Cytochrome Oxidase at Varying MPH Concentrations . . . . 49 111.2 Rate Constants for the Anaerobic Reaction of MPH and Cytochrome Oxidase in the Presense of Lauryl Maltoside and of Tween-20 . . . . . . 58 111.3 Rate Constants for the Post Steady-state Reduction of Oxygenated Cytochrome Oxidase by "PH 0 O O O O O O O O I O O O O O O I O O 6 8 IV.1 Squares of Effective Magnetic Moments for the Possible Magnetic States for the a3 Center Of CytochrOIO Oxidlto e e e e e e e e e e e e 81 IV.2 Coupling Factors for Cyanide-bound Cytochrome oxid.’° O O O O O O I O O I O O O O O O O O O 96 IV.3 Curie Law Constants for Cyanide-bound Cytochrome Oxidase . . . . . . . . . . . . . 98 IV.4 Effective Magnetic Moments Squared For Resting Cytochrome Oxidase . . . . . . . . . 102 11.1 11.2 111.1 111.2 111.3 111.4 111.5 111.6 LIST OF FIGURES The Arrangement of the Subunits in Cytochrome Oxidase in the Membrane . . . . . The Structure of Home g . . . . . . . . . . The Optical Spectra of Oxidized and Reduced Cytochrome Oxidase . . . . . . . . . . . . . The Optical Spectra of Oxidized and Cyanide-bound Cytochrome Oxidase . . . . . . A Schematic of the Bottle Used for the Anaerobic Titration of MPMS with NADH . . . The Spectra of an Anaerobic Titration of "PIS '1th "Ann O O O O O O O O O O O O O O O The Absorption Change at 830 nm During the Reduction of Cytochrome Oxidase with MPH . . The Calculated and Experimental Absorptions at 830 nm for the Reduction of Cytochrome Oxidase with MPH . . . . . . . . . . . . . . The Residuals of the Calculated Absorptions at 830 nm for the Reduction of Cytochrome oxid". 'ith [PH O O O O O O O O O O O O O O The Calculated and Experimental Absorptions at 605 nm for the Reduction of Cytochrome ox‘d“. 'ith “Pa O O O O O O O O O O O O O O Residuals of the Calculated Absorptions at 605 nm for the Reduction of Cytochrome Oxidase with MPH . . . . . . . . . . . . . . The Spectra from the Partial Reduction of Cytochrone Oxidase with MPH in the Prescence of Lauryl Maltoside . . . . . . . vi 39 40 45 47 48 51 52 54 111.7 111.8 111.9 111.10 111.11 111.12 111.13 IV.1 1V.2 IV.3 IV.4 The Calculated and Experimental Absorbances at 444 nm for the First Phase of Reduction of Cytochrome Oxidase with MPH in the Presence of Lauryl Maltoside . . . . . . . . The Calculated and Experimental Absorbances at 444 nm for the Final Phases of the Reduction of Cytochrome Oxidase with MPH in the Presence of Lauryl Maltoside . . . . . . The Change in Absorbance at 444 nm During the Reduction of Oxygenated Cytochrome Olid‘Oo O O O O O O O O O O O O O O O O O O The Calculated and Experimental Absorbances at 444 nm for the Presteady-state Reduction of Oxygenated Cytochrome Oxidase with MPH: Empirical One Exponential Equation . . . . . Residuals and Calculated and Experimental Absorbances at 444 nm for the Poststeady- state Reduction of Oxygenated Cytochrome Oxidase with MPH: Empirical Two Exponential Equation . . . . . . . . . . . . . . . . . . The MPH Concentration Dependence of the Pseudo First-order Rate Constant for the Poststeady-state Reduction of Cytochrome Oxidase . . . . . . . . . . . . . . . . . . The Residuals and Calculated and Experimental Absorbances at 444 nm for the Poststeady-state Reduction of Oxygenated Cytochrome Oxidase with MPH: Mechanistic Equation . . . . . . . Magnetic Susceptibility vs Reciprocal Temperature for Resting Cytochrome Oxidase . Experimental and Calculated Magnetic Susceptibilities vs Reciprocal Temperature for Resting Cytochrome Oxidase . . . . . . . Magnetic Susceptibility vs Reciprocal Temperature for Peroxide-bound Cytochrome 011d.'. O O O O O O O O O O O O O O O O O O Magnetic Susceptibility vs Reciprocal Temperature for Formate-bound Cytochrome oxid‘.° O O O O O O O O O O O O O O O O O O vii 57 59 62 65 67 69 72 79 84 86 1V.5 1V.6 1V.7 IV.8 1V.9 Experimental and Calculated Magnetic Susceptibilities vs Reciprocal Temperature for Formate-bound Cytochrome Oxidase . . . . 89 Possible Structures for the Cytochrome .3 Oxygen Reducing Site . . . . . . . . . . . . 91 Magnetic Susceptibility vs Reciprocal Temperature for Cyanide-bound Cytochrome 0x1.“ '. O O O O O O O O O O O O O O O O O O 94 Magnetic Susceptibility vs Reciprocal Temperature for Cyanide-bound Cytochrome Oxidase in 50:50 Ethylene Olycol/HEPES . . 100 A Diagram of the Energy Levels for Ferromagnetically and Antiferromagnetically Coupled Systems of Two 8-1/2 Ions . . . . 107 viii CHAPTER 1 W Winn Cytochrome g oxidase (ferrocytochrome 1:02 oxidoreductase:E.C. 1.9.3.1) catalyzes the transfer of electrons from cytochrome q,to oxygen: 4 cytochrome 92+ +02+4H+ 5 4 cytochrome g?+ + 2H20. More than 90% of the oxygen consumption by living organisms on earth occurs through this reaction (Vikstrom et al. 1981). The free energy change is about -l92 kJ per four electrons transferred. This energy is stored as an electrochemical proton gradient and is used by the cell in subsequent ATP synthesis. Wm Cytochrome oxidase is a Y-shaped protein which spans the inner mitochondrial membrane. Twelve or thirteen subunits copurify in stoichiometric amounts with the metal centers (Kadenbach and Merle. 1981). but only eight of these subunits are required for electron transport activity 1 2 and for generation of a transmembrane proton gradient (Azzi. 1980: Downer. et a1. 1976). The arrangement of the subunits has been investigated by using cross-linking and chemical binding (Briggs and Capaldi. 1977: Fuller. et al. 1981). A model consistent with the data from these studies is shown in Figure 1.1 (Capaldi et.al.. 1983). Wm There are two iron and two copper ions in each cytochrome oxidase monomer. Although both iron species are isolated from the enzyme as heme g,(Figure 1.2), they have different structural and functional properties in the enzyme. Uhen there. they are called cytochrome ; and cytochrome g3. The copper complexes also differ from each other. Because of an assumed structural and functional association with the cytochromes. the copper species are known as CuA and Cup. The traditional distinction between the cytochrome moieties is that cytochrome g3 will bind ligands such as CO. HCN. and NO while cytochrome ; will not. In addition. cytochrome g3 exhibits no epr signal in the oxidized enzyme while cytochrome g_has signals of a low-spin heme at g-3.0. 2.2. 1.5. Evidence suggests that cytochrome 53 hp; one histidine ligand (Chan et.al.. 1982) and that cytochrome E has two histidine ligands (Stevens et.al.. 1982) 3 Figure 1.1 The Arrangement of the Subunits in Cytochrome Oxidase in the Membrane (from Capaldi et.a1., 1983) heme 9 Figure 1.2 The Structure of Heme a. 5 Different epr and optical properties show that the environments of the two copper ions differ. No epr signal is exhibited by Guy vhile Cug has a 3-2 signal. Isotopic substitution studies (Capaldi et.al.. 1983) indicate two histidine ligands for CuB. ENDOR and epr studies (Stevens et.al..l982) on yeast cytochrome oxidase identify one cysteine and one histidine as ligands of CuA. Because there are no epr signals from Cun or fro. cytochro-e £3 in the resting or oxidized cyanide-bound forms of the enzyme, it is believed that these two metals are magnetically coupled in those enzyme forms. The pair may then be refered to as the 33 site. Wu The cptical spectra of the fully oxidized and fully reduced enzyme are shown in Figure 1.3. The bands in the near-UV ('Soret") and visible ("0.") regions are due to the porphyrin ring of the homes. A discussion of the origin of porphyrin spectra is given by Gouterman (1959). The two cytochromes contribute about equally to the absorbance at 444 nm. Cytochrome L contributes about 80% of the absorbance at 605 nm. with most of the remainder due to cytochrome 13. The broad band at 330 nm is at least 851. due to CIA (Wharton and Tzagoloff. 1964: Boelens and Vever.1980: Beinert et.al.. 1980). 200 ISO I20 80 4O 0 'e _u '2 E M 01111111111141111111 ' 500 550 600 650 700 8 6 4 2 “~-—----.. 0 lllLiLilllJl 11141111 iJ-i~~ 650 me 750 800 850 Wavelength, nm Figure 1.3 The Optical Spectra of Oxidized and Reduced Cytochrome Oxidase: .... , oxidized, and -— , fully reduced. Molar absorptivities are expressed per unit containing two hemes and two copper ions. Fromllalaka (1981) 7 The optical spectrum of the oxidized cyanide-bound enzyme is shown in Figure 1.4. (Nicholls and Chance. 1974). It is similar to that of the oxidized resting enzyme. with the most important difference being the lack of a shoulder at 650 nm. The absence of this shoulder is used as a check for complete cyanide binding of the enzyme. WW Cytochrome oxidase as isolated is said to be in its "resting" state. 'Ihen reduced and recxidized it is converted to a form called "oxygenated" (Okunuki et.al.. 1959). The oxygenated form. so called because it was originally believed to be an oxygen adduct. is more active than the resting enzyme (Antonini et.al.. 1977: Brunori et.al.. 1979: Peterson and Cox. 1980). Because its greater activity may mean it is the form present during catalytic cycles. the oxygenated enzyme is of great interest. The traditional method used to make the oxygenated form involves reduction of the enzyme with an excess of sodium dithionite and subsequent reoxidation with oxygen. This procedure yields the more active form with a near-UV absorpton at 428 nm (vs 418-424 nm for the resting form) and a slightly enhanced absorption in the visible region (Lemberg and Stanbury. 1967). This enzyme will decay in a two-step process to a form with an activity and spectrum very similar to those of the resting oxidase. The shift in Absorbance 0J8 012 (108 0134 400 500 600 Wavelength (nm) Figure 1.4 The Optical Spectra of Oxidized and Cyanide-Bound Cytochrome Oxidase. —-oxidized, -—- cyanide-bound aouoqiosqv 9 Soret maximum from 428 nm to shorter wavelengths has been correlated with the loss of enhanced activity (Brunori et.al.. 1981) and the decay in the visible region has been shown to follow the disappearances of a g-S. 1.78. 1.69 epr resonance (Armstrong et.al.. 1983). Kumar et.al. (1984) have recently shown that the more active form of the enzyme can also be produced with its near-UV maximum at 420 nm. This is accomplished by using catalase to prevent formation of peroxide (due to the presence of dithionite) during enzyme reoxidation with 02. Addition of peroxide to this 420 nm form of the more active enzyme produces the 428 nm form of the oxygenated enzyme. This indicates that the latter is a peroxide derivative or reaction product. Oxidized cytochrome oxidase is believed to exist in more than one conformation because the resting oxidase is heterogeneous in its behavior with several ligands. 'ork done at Michigan State University by F. Halaka showed that the resting enzyme isolated by the method of Hartzell and Beinert (1975) is heterogeneous in its reduction by sodium dithionite. From 15 to 30 percent of the cytochrome ;3 is reduced directky in an initial fast phase. The remaining enzyme is reduced at the cytochrome g site. with its cytochrome g3 reduced through intramolecular electron transfer. 'Using enzyme isolated by the Yonetani method. Jones et.al. (1983) found that all of the enzyme is reduced 10 by dithionite via cytochrome g... Brunori et.al. (1985) found no difference in the dithionite reduction of cytochrome oxidase isolated by using either the Ionetani or the Hartzell-Beinert techniques. These results may indicate a difference in the properties of the enzyme 1 soula.teld iii di.ff'erreiit l.al>oi:at:or'ie s on: a preparation-to-preparation variation with the same isolation technique. Bickar et.al. (1982) reported that the enzyme they isolated by the Yonetani method always bound at least some peroxide. but that the extinction coefficient varied. This indicates that the proportion of enzyme that bound the peroxide changed from one isolation to the next. even though the same procedure was used each time. Brudvig et.al. (1981) report that three conformations mty be present in the enzyme as isolated. Addition of NO induces a high-spin cytochrome epr signal in one fraction of the enzyme. presumably by uncoupling the cytochrome g3 and CuB. Another fraction exhibits a g'-12 epr signal and is further distinguished from the first fraction by the slowness with which it binds cyanide. A third fraction does exhibit the high-spin cytochrome signal in the presence of N0 alone. but does when both NO and F" are present. Brudvig et.al. (1981) also looked at the oxygenated enzyme. They found that the presence of NO by itself does 11 not induce a high-spin oytchrome epr signal. but addition of F“ or of F" and NO induces the high-spin cytochrome epr signal. The oxygenated form does not show the g'-12 signal. Its behavior thus corresponds to that of the third fraction of the resting enzyme. A g'-12 signal appears as the 428 nm band shifts to 420 nm when the oxygenated form decays. ‘The heterogeneity seen by Brudvig et.al. was confirmed by Vilson et.al. (1982). who found that the differences disappeared during turnover. The kinetics of cyanide binding has been used as a probe of the g3 site in both the resting and the oxygenated enzyme. The number of phases and their rate constants have been reported for the omygenated enzyme and for resting cytochrome oxidase isolated by several different procedures. Van Buuren et.al. (1972) reported that the binding of cyanide to the resting enzyme consists of more than two components with the initial second-order phase having a rate constant of 1.8 M-1 s-l. Brittain and Greenwood (1976) prepared oxygenated enzyme and found that cyanide binds in one phase with a rate constant of 22 M-1 s-l. Naqui et.al. (1984) found that the resting enzyme bound cyanide multiphasically. The phases present and the percent contribution of each phase vary not only with the method of enzyme isolation, but from preparation to preparation when the same isolation procedure is used. For 12 all the enzyme except one of the two Ionetani preparations, the initial phase was second-order. The average rate constant was 2.5 M-1 s-l. Other phases were first-order in 1 enzyme and had rate constants of either 1.1 s- or 0.06 s-l. Naqui et.al. suggest that the first-order phases are due to conversion of enzyme forms incapable of binding cyanide to a form that is capable. The fast phase would then be due to the binding of cyanide by the capable form initially present. Like Brittain and Greenwood. Naqui et.al. found monophasic binding of cyanide to the oxygenated enzyme. but with an average second-order rate constant of 2.3 M.1 s-l. All preparations of oxygenated enzyme showed the same behavior. regardless of the method used to isolate them from the membrane. Naqui et.al. noted the similarity of the second-order rate constants in the resting and oxygen- ated cases and suggested that the portion of the resting enzyme capable of binding cyanide may actually be in the oxygenated form. While the rate constants for cyanide binding to the oxygenated enzyme vary considerably from one investigation to another. it is plain that the oxygenated enzyme does not show the heterogeneity that the resting enzyme presents. 13 mum.“ LLWW Since cyto- chrome oxidase molecules span the inner mitochondrial membrane. at least some of their in_1iyg environment must be hydrophobic. 'hen removed from the lipid environment of the membrane. the enzyme requires phospholipids or nonde- naturing detergents to form a solution and to maintain electron transport activity (Awasthi et.al.. 1971: Brierly et.al.. 1962). The steady-state activity of the soluble enzyme is affected by the dispersing detergent used. A higher activity results from the use of nonionic detergents such as Triton I-lOO or the Tween series than from the use of ionic bile salts such as cholate. Robinson and Capaldi (1977) report that the higher activity is due to the greater fluidity of the fatty acid chain of the nonionic detergents. The identity of the detergent headgroup also has an effect since cytochrome oxidase has a higher steady—state activity in solutions of lauryl maltoside than in solutions of Tween-20. though these two detergents have identical fatty acid chains (Rosevear et.al.. 1980). Robinson et.al. (1985). using several types of detergents. found that the rates of electron transport were dependent upon the structure of the head group and the length of the hydrocarbon tail. 14 The variation of activity with the nonionic detergent used may be related to the aggregation state of the enzyme. In lauryl maltoside the enzyme is monodispersed and is 2-10 times more active than in Tween-20. where it is oligoneric and polydispersed (Rosevear et.al.. 1980). WWW 300"!” reduced cytochrome g.is the natural substrate of cytochrome oxidase. the reaction between them has been studied in several laboratories. Two of the difficulties associated with using cytochrome g as a reductant in transient-state kinetic studies should be noted: (1) the reaction is very fast (the second-order rate constant is about 8 x 107 M"1 s-l) and much of the reaction is over within the deadtime of many of the stopped-flow systems used. and. (2) under anaerobic conditions. only two electrons are transferred from cytochrome q to an oxidase molecule. so complete four-electron. anaerobic reduction cannot be studied. The most comprehensive anaerobic study was done by Antalis and Pmlmer (1982). who varied the ionic strength and cytochrome q:oxidase ratio. Both changes affect the number of phases observed. Low ionic strength or reductant:oxidase ratios of 1:1 or 2:1 lead to monophasic kinetics. High ionic strength or high cytochrome g:oxidase ratios result in a biphasic reaction. In additon. the ionic strength affects the amplitudes of the two phases. 15 This explained the disagreements in the literature about the number of electrons transferred in the fast initial phase (Andreasson et.al.. 1975: Iilson et.al.. 1975: Van Buuren et.al.. 1974) and in the slower second phase (Gibson et.al.. 1965). Antalis and Palmer also found that the total number of electrons transferred did not depend on the ionic strength. but varied form 0.8 to 2.0 as the reductant:onzyme ration went from 1 to 8. This explained the differing reports of the number of electrons tranr— ferred (Andreasson et.al.. 1975: 'ilson et.al.. 1975). WWW Reduction of oxygenated cytochrome oxidase by cytochrome g has been studied in several laboratories. Comparisons with the reduction of the resting enzyme indicate a difference in the kinetic behavior of the two forms. Antonini et.al. (1977) found that the reduction of oxygenated cytochrome oxidase by cytochrome 1 had a steady state velocity 4-5 times greater than that of the resting oxidase. This difference disappeared at very low oxidase:cytochrome q,ratios. The difference in reaction rates observed was due to faster reduction of cytochrome £3 during a phase that was first-order in enzyme. rather than the faster reaction between cytochrome g and the enzyme. Peterson and Cox (1980) extended these studies by using both cytochrome g and reduced methyl and benzyl viologens 16 as reductants. They reported a rate constant of 7.5 s-1 for cytochrome g3 reduction in the oxygenated enzyme. Gibson et.al. (1965) reported a rate constant of 0.5 s"1 for this reduction in the resting enzyme. Thus the kinetic difference between resting and oxygenated enzyme is due to enhanced intramolecular electron transfer in the latter form. The disappearance of this kinetic difference at low oxidase:reductant ratios was confirmed by Antonini et.al. (1985) and is believed to be due to conversion of the resting form to the oxygenated form by multiple turnovers. AIi1A_BrJJa1iaa_1x_1:Ma1hI1_2hanazininm_lu£hxlrnliaia Anaerobic stopped-flow studies of the reduction of cytochrome oxidase by 5-methyl phenazinium methylsulfate (MPH) were done by F. Halaka (1981). Comparison of initial and final spectra with those of the oxidized and reduced MPH and the oxidized and reduced enzyme showed that the reaction has the overall stoichiometry: 2MPH + cytochrome oxidase (oxidized) -——-9 2MP+ + cytochrome oxidase (reduced). The changes of absorbance in the difference spectra showed that the species that absorbed at 430 nm (cytochrome g) was reduced more quickly than the species that absorbed at 410 nm (cytochrome g3). l7 Nonlinear least-squares fitting of the data showed that the reaction had a fast phase that was first-order in each reactant and was followed by two slow phases that were both first-order in enzyme concentration. The absorbance change at 830 nm. due mainly to the reduction of CuA. lagged the change at 605 nm at the one set of concentrations used. The model proposed to account for this was 3+ 2+ MPH + [cyt _a_ Cu A J : MPHT[cyt 32+ 012+] ; MP+ + [cyt 52+ CuZ] Electron transfer to cytochrome g,was proposed to be fast relative to electron transfer from cytochrome g to CuA: thus the cytochrome would always appear reduced. The two phases that were first-order in enzyme must represent intramolecular electron transfer. The presence of two first-order phases may be due to heterogeneity of the enzyme. Both of these phases are too slow to be significant in the catalytic cycles of the enzyme. 6 Me n tic tate o e to hrome a :g13_g£3£1£ Because neither the iron ion nor the copper ion of the g3 center exhibits an epr signal in the resting or oxidized cyanide-bound enzyme. it has been proposed that the two metals are magnetically coupled (Van Gelder and Beinert. 18 1969) in these forms of the enzyme. It is generally believed that the coupling is a superexchange interaction operating through the orbitals of a bridging ligand. The magnetic states of the ;3 center of the cyanide-bound and resting forms of the enzyme have been studied by using Massbauer and magnetic-circular dichroism (MCD) spectroscopies and magnetic susceptibility measurements. The conclusions drawn from these studies are conflicting. with some studies indicating ferromagnetic coupling and some indicating antiferromagnetic coupling. Wm Because of the low (25) natural abundance of 57 Fe. Mossbauer studies of cytochrome oxidase were done with bacterial enzyme isolated from 11:31}; thnxngnh111;,(Kent et.al.. 1982). Comparative studies were later done with bovine cytochrome oxidase (Kent et.al.. 1983). For the L3 center of the oxidized cyanide-bound bacterial enzyme. Kent et.al. report a zero-field quadrupole splitting and an isomer shift typical of a low spin ferric heme. Application of a magnetic field broad— ened the spectrum. indicating a ground state of integer spin. 811. These data were explained by ferromagnetic coupling of an 881/2 iron ion with an 8-1/2 copper ion. The bovine cyanide-bound cytochrome oxidase had zero-field parameters similar to those of the bacterial l9 enzyme. The presence of a quadrupole doublet evinced a :zero or integer spin ground state. Application of a magnetic field broadened the spectrum. indicating an integer spin. The similarity of parameters and the effect of the field suggest ferromagnetic coupling to yield an 8-1 ground state in the bovine enzyme as in the bacterial enzyme. It must be noted. however. that the data for the bovine enzyme are extremely noisy and these conclusions are based mostly on comparison with the bacterial data. The zero-field parameters of the g3 center in the resting oxidized bacterial enzyme were typical of a high-spin ferric heme. The lack of zero-field hyperfine structure indicated that the ground state spin was zero or an integer. Surprisingly. application of a magnetic field gave results that were preparation dependent. The lack of change in the spectrum of one preparation indicated an 8-0 ground spin state. The spectrum of a second preparation broadened when the field was applied. indicating an 811 spin ground state. Kent et.al. were unable to provide an explanation for this heterogeneity. or a mechanism by which a high-spin iron ion could couple with an 8-1/2 copper ion to yield a diamagnetic state. At zero field the g3 center of the resting oxidizmd bovine cytochrome oxidase exhibited a quadrupole doublet with parameters similar to those of the bacterial enzyme. Applicatimn of a 60 mT field did not broaden the spectrum 20 within the limits of their signal-to-noise ratio. This ratio was poor enough that Kent et.al. do not believe the lack of broadening conclusively shows an 8-0 ground state. If they are correct. then a zero or integer spin ground state is indicated with clear-cut conclusions about the spin and the nature of the coupling being impossible. WWW- Tho-son ot-tl- (1981). used mcd spectroscopy to study the magnetic properties of cytochrome 53 of the oxidized cyanide-bound bovine enzyme. Their enzyme was isolated using the Yonetani technique. The magnetization curve of the ;3 site of the cyanide-bound enzyme was typical of a ground state elec- tronic doublet. The sigmoidal nature of the curve indi- cated a low-lying excited state. These curves and the lack of an epr signal were explained by Thomson in terms of an 8-1 ground state with an axial distortion such that the M. - I 1 components were at least 10 on"1 lower in energy than the M,-0 component. An epr signal would not be expected because M, - r 2 transitions are forbidden. This model implies ferromagnetic coupling between two 8-1/2 centers. The coupling constant was greater than 10 cm-l. LWW Tvoodlo ot-nl- (1978) measured the magnetic susceptibilities of the oxidized resting and oxidized cyanide-bound forms of bovine 21 cytochrome oxidase isolated with the Hartzell-Beinert technique. Their temperature range was 7 K to 200 K. Moss et.al. (1978) extended the temperature range to 1.4 K for the oxidized resting form. Tweedle et.al. found that. except at the lowest temperatures. the magnetic susceptibility of the resting oxidized bovine enzyme conformed to the Curie Law. (a)2 N 82 S(S+1) 2 N 82 u: x = - 3kT 3kT ff The slope of the X vs T-1 plot corresponds to a nsz - 31.5. The contribution to this by cytochrome a3+ and CS: was calculated from their epr parameters and found to be 7.2. Thus the a3 site contribution was 24.3. The most plausible coupling scheme that could produce this value for ":ff is antiferromagnetic coupling of an 8-5/2 cytochrome a§+and the 8-1/2 CuB. The linearity of the X vs T-1 plots at temperatures as high as 200 K led to the conclusion that -2J > 200 cm—1. The nonlinearity and the decrease in slope at low temperatures obtained by Moss et.al. were accounted for by reasonable values for the zero-field splitting and the rhombic zero-field parameter 1 (D-9 cm"1 and E-0.1 to 1 cm- ). 22 The plot of susceptibility vs reciprocal temperature for the oxidized cyanide-bound enzyme was curved. The limiting slope at low and high temperatures corresponded to a :1er of 7.8 and 15.1 respectively. These values were expected to be 7.2 and 15.2 if an antiferromagnetically coupled center composed of two 8-1/2 ions was present and if the absolute value of 2J was about kT. The curvature of the plot would arise as the excited state became more extensively populated at higher temperatures. The equation for the classical model of such a two-spin interaction is x - Lgfig-fl- [ 1 +31; exp (-2J/1 8 J 2 exp (-nMJ) -J where Tl =gBH/kT. Now. J 2 exp (-nMJ) = [sinh (J+%)n lsinh(n/2) -J and J a J 2 (4538) exp (-nMJ) = m 3H [ 2 exp<-MJn)J -J 2 -J Consequently. z 3H J <11 > a RT gr {ln[ 2 exp('MJnj} 2 -J Combining these equations gives =36 {(J+ls) coth [(J+*5)n] - heath (Tl/2)} (DZ) 8 gBJBJ‘n 29 where 133(0) is the Brillouin function. The macroscopic magnetic moment for N noninteracting ions is then M = 8J5 BJ(n) For large ‘n (high fields or low temperatures). B 2 1 J and M x H - NgBJ/H Note that this susceptibility does not depend on the temperature. but is a constant value known as the " saturation moment" . This physically corresponds to all ions being in the ground magnetic state. For small n (low fields and high temperatures). BJ(n) = (J+k) n/3 and x = Hg2 82 J(J+1)/3k'r = 'SI'; (1.2) where C-Ng282J(J+1)/3k.Equation 1.2 is known as the Curie Law. Another common expression of the Curie Law is called the spin-only formula for the susceptibility: 2 x = N2 82 S(S+1)/3kT = N 82 niff/Bk'r where “iff = g2 S(S+1) 30 This equation applys for many complexes of ions of the iron series. such as home proteins. because their orbital angular momentum is quenched (L-O) by the ligand field. ILJLJUI£:EiaLiJhflJJJins Spin-orbit interactions may partially lift the degeneracy of the ground state. leaving a set of spin states. This splitting is commonly called zero-field splitting because it occurs in the absence of an external magnetic field. It is characterized by the axial and rhombic zero-field splitting factors. D and E. respec- tively: li-l ,- 8 / “E M ._1 S-1 -—-( s x‘ D \ -..- M =0 8 The effect of this splitting on the susceptibility becomes most apparent as kT approaches the energy difference between the split states. For many heme proteins D~1o °.-l (Tasaki et.al.. 1966: Uenoyama et.al.. 1968: Behere et.al.. 1979). Since the symmetry of a heme approaches axial. the rhombic splitting factor. E tends to be much smaller and the spin Hamiltonian H - D 8% is sufficient to represent the fine structure of the electronic ground state of many heme proteins. 31 Although the average susceptibility is affected by the zero-field splitting. it is not the best parameter to use to determine the zero-field splitting. If crystals of the material in question are aw1ilable. measurements of the magnetic anisotropy provide a better value. Some heme proteins have been found to have magnetic behavior that deviates from the Curie Law at high temperatures. The magnetic susceptibility of cytochrome c peroxidase. :for example. deviates form the Curie Law behavior above 170 K. Its ":ff was found to vary from 5 to 35 (Izuka. 1968). These are the values expected from pure low- and high-spin behavior d5 ions. The behavior of the system matches that expected if there is a thermal equilibrium between low and high spin states. W An equation for the susceptibility which includes field dependent effects on the energies can be derived by returning to Equation 1.1. Assume that the energy can be expanded as a power series in the field: E . E(o) + H E(1) + H E<2) + ... n n n n Then exp (-En/kT) 8 (l-Eé1)/kT) exp (-E§o)) and. 1111 - “BE/3H I _Eril) _ ZHEISZ) 32 Substitution of these into Equation I.1. expansion of the exponentials. and elimination of terms involving powers of H larger than one (except for the exponential involving En) gives the Van Vleck equation: (1) (2) (1) u a N 5 (-En - 2HEn ) (l-HEn /kT) n (0) (1) § exp (~En / kT) (l-HEn / kT) (0) (1) (2) where En is the zero-field energy and En and En are the first and second order Zeeman terms. respectively. W While the size of enzyme molecules often prevents intermolecular interaction between metal ions. the ions within a molecule may be close enough for their spins to interact. One way that spin-spin coupling can occur is by direct contact of the orbitals on adjacent metal ions. The Pauli principle then requires that the spins be aligned antiparallel. so the coupling is antiferromagnetica Another mechanism of spin-spin coupling is that the orbitals of the metal. which have unpaired electrons. overlap with filled orbitals of bridging atoms and the unpaired electrons are thus delocalized. This is referred to as superexchange. and the resultant coupling can be either ferromagnetic or antiferromagnetic. 33 The spin interaction between two metal ions with spins 81 and 82 can be described by the Hamiltoniai. Q - -2J EIOEZ where J' in energy units is called the exchange coupling constant and is different than the quantum number I. With this Hamiltonian an antiferromagnetic interaction leads to a negative I and a ferromagnetic interaction leads to a positive I. Using the relationships 31.§2-;5(52-si-s§) .nd 82=S(S+1) a system with 31 " 52 ‘31 and antiferromagnetic coupling has an energy of 3J/2 for the 8-0 level and and energy of -J/2 for the 8-1 level. The energy difference between the levels is thus 21. Substituting these energies into the Van Vleck equation gives. upon rearrangement. 2N 232 1 -1 x = 31.1: [1 +-3-exp (-2J/kT)] 34 This equation does not consider the effects of any zero-field splitting which may be present. CHAPTER 2 W Mauls The beef hearts used to isolate cytochrome oxidase were obtained fresh from Van Alstine's Packing House in Okemos. Michigan. Phenazine methosulfate (MPMS). B-nicotinamide adenine dinucleotide (NADH). cholic acid. N-2-hydroxyethyl piperazine N-2-ethane sulfonic acid (HEPES). polycry- ethylene sorbitan monolaurate (Tween-20) and Triton X-114 were obtained from Sigma Chemical Company. Argon was purified by passing it through a 50 cm BASF R3-11 catalyst column at 100 °C. Cholic acid was purified by recrystallization from 95% ethanol. The crystals were mixed with equivalent amounts of potassium hydroxide to give a 20% (w/v) solution in cholate ion. Hydrochloric acid was used to adjust the pH to 8.0. Tween-20 and Triton 1-114 were kept refrigerated as 20% (v/v) solutions. All other reagents were of analytical grade and were used without further purification. 35 36 W W The stopped-flow system used a double-beam. vacuum- tight. rapid-scanning stopped-flow spectrophotometer (Papadakis et.al.. 1975: Coolen et.al.. 1975). Up to three solution bottles were attached to each side of the system at the top of burets. Liquids flowed from the burets into a reservoir containing a magnetic stir bar. Thus. the concentrations of the solutions mixed in the cell could be varied in the reservoir before their introduction into the push syringes. The scanning collection mode was used to follow absorption changes at several wavelengths during a single mix of solutions within the cell. When better time resolution was desired. data were collected in the fixed wavelength mode. Detailed descriptions of the stopped-flow system are given in the Ph.D. dissertations of N. Papadakis and R.B. Coolen. Details of the periodic checks run on the system are given in the Ph.D. dissertation of F. Halaka. W Data were collected with a PDP 8/1 or PDP 8/E computer interfaced to the stopped-flow system. then transferred to the MSU (whet 750 for analyses. Correction was made for finite scan times. channel numbers were converted to 37 wavelength. and voltages were changed to absorption values by using programs described in Apendix E of R. Cochran's P h .l). d iiss eirtza t ic»n.. 13h e are s'ul.t iiig time-absorption-wavelength data were fit by using the nonlinear least-squares fitting program KINFIT4. a modified version of KINFIT (Dye and Nicely. 1971). Win All solutions were deoxygenated by alternately evacuating their bottles and allowing purified argon to equilibrate with the solutions. This was repeated six to ten times and then 3 psig of argon was introduced. The sealed bottles of buffer. MPH. and water were attached to the stopped-flow apparatus with 5 mm Fhscher-Porter sclv-seal joints. After the bottles of solutions (except the enzyme) were attached. the stopped-flow system was made anaerobic by evacuation. ‘to 100 microns pressure four times. with each evacuation followed by an equilibration with purified argon at 3 psig. The system was then evacuated to less than 1 micron pressure and water was allowed to enter the cell. Then 1-2 psig of argon was introduced to prevent foaming of the detergent-containing solutions when they were allowed to flow into the burets. The anaerobicity of the stopped-flow and the buffers was checked by monitoring the MPH spectrum for several 38 minutes. Any oxygen present would have reacted with this species. oxidized it. and produced an increase in the absorbance at 388 nm. After anaerobicity checks and calibrations. the liquid-containing parts of the system were isolated with Kontes valves and the enzyme bottle was attached to the system. The part of the system not isolated was then evacuated and filled with argon several times. The enzyme bottle was not hung until this point of the experiment so it could be kept cold as long as poss- ible. Wanna MPH wms prepared by anaerobically titrating MPMS with NADH by using the bottle shown in Figure 2.1. Because MPH showed an instability when kept in HEPES buffer overnight. no buffer was used in its solutions. NADH was added with a calibrated Hamilton gas-tight syringe. The cell on the bottle allowed the reduction of MPMS to be monitored with a Cary 17 spectrophotometer. NADH was added until only a shoulder remained at 388 nm: this shoulder indicated that a small amount of MPMS remained. ensuring that no excess NADH was present. .A typical set of titration spectra is shown in Figure 2.2. The concentration of MPH was calcu- lated by using A5388 -21.0 mM-1 cm.1 (Halaka. unpublished). The MPH was allowed to stand overnight before the final spectrum was taken. Figure II.1 39 Titrom 1 l2 cm A Schematic of the Bottle Used for the Anaerobic Titration of MPMS with NADH. (a) Kontes valve (b) Fischer-Porter solv-seal joint (c) one centimeter quartz cell (from Halaka, 1981) 40 oonfivewnrwwrwfi II) 350 400 450 5CD Woulenqtmnm Figure II.2 The Spectra of an Anaerobic Titration of MPMS with NADH (from Halaka, 1981) 41 All bottles and burets used to handle MPH or MPMS were wrapped in aluminum foil due to the photosensitivity of these compounds. Room lights were turned off during the experiment. There are no windows in the room. Wen The cytochrome oxidase used in stopped-flow experiments ‘was isolated from the membrane by the method of Hartzell and Beinert (1975). During all steps the enzyme was kept between 0 °C and 10 °C. The final pellet was solubilized in 1-2 mL of 50 mM HEPES buffer (pH-7.4) containing 0.5% Tween-20. ‘The enzyme solution was stored in liquid nitrogen until the day of the experiment. when it was diluted with more of the same buffer. Fhr the reduced enzyme. the ratio of the absorbance of the reduced enzyme at 444 nm to that at 420 nm was at least 2.4 (Gibson et.al.. 1965). W Magnetic susceptibilities were measured with an 8.H.E. Corporation Variable Temperature Susceptometer. which utilizes superconducting quantum interference devices (SQUID) as a basis for its measuring system. Data were taken at 7.00 kG over a temperature range of 2 to 200 K. Sample holders were made from either an aluminum- silicon alloy obtained from 8.H.E. Corporation or from 42 poly(monochlorotrifluoroethylene). whose trade name is Kel-F. 'The Kel-F sample holders were cleaned in boiling nitric acid before use. ‘The enzyme used was prepared either by the method of Hartzell and Beinert (1975) or Ionetani (1966). The latter was a gift from P. Moroney. Hartzell and Beinert enzyme was in 50 mM HBPES solution containing 0.5% Tween-20. Yonetani enzyme was in 100 mM phosphate buffer containing Tween-80 or in Tween-20. Because oxygen is paramagnetic. the enzyme and buffer samples were made anaerobic before they were frozen. This was done by one of two methods: (1) a series of evacua- tions of the sample bottle. each followed by argon equili- bration. or (2) addition of small amounts of glucose and glucose oxidase to the sample while it was in a helium-flushed glovebag. After the oxygen had been removed. the samples were frozen and stored in liquid nitrogen. CHAPTER 111 STOPPED-FLOW KINETICS OF CITOCHROME OXIDASE WAW Several laboratories (Wilson et.al.. 1975: Gibson et.al.. 1965: Antalis and Palmer. 1982) have demonstrated that electrons are donated to at least two sites during the rapid initial phase of the reaction between oxidized cytochrome oxidase and ferrous cytochrome 9, Wilson et.al. (1975) showed that a decay of the 830 nm band that is assigned to CuA occurs during the initial rapid phase of that reaction. The 605 and 444 nm bands also change during this phase. an indication of cytochrome a3+ reduction. The decay at 830 am either lagged or was simultaneous with formation of the 605 nm band. It has been commonly postulated that electrons enter through the cytochrome g, which is then in rapid equilibrium with CuA. It has also been suggested several times (destr m and Casey. 1985: Capaldi. 1983: Azzi. 1980) that CuA may be the site where some or all of the electrons enter the enzyme. When the enzyme is reduced with MPH the decay at 830 nm lags the growth of absorbance at 605 nm: this led Halaka (1981) to propose that CuA receives its electrons via cytochrome ;. 43 44 Halaka demonstrated this lag in the decay at 830 nm at only one MPH concentration. so the experiment described here was designed to do two things: (1) to see if this lag is reproducible at other MPH concentrations. and (2) to see whether the rate constant obtained from the absorbance change at 830 nm depends on the reductant concentration. The existence of such a dependence would indicate that the MPH wms reacting at the CuA site rather than only at the cytochrome g, site. The absorbance change at 605 nm was also followed. The typical absorbance decrease at 830 nm during the first 300 ms is shown in Figure 111.1. The lag in absorbance is evident at the two MPH concentrations used. showing that the presence of the lag does not depend on reductant concentration. The scheme that Halaka proposed to account for the lag at 830 nm was MPH + [cyt 33+ CuA] ——k-L—-v PIP-i-[cyt 32+ C1122] (A) (B) k2 ‘1 MP+ + [cyt a2+ CuZ] (C) The concentration of B+C at time t is given by Lb <1-exp(-kt>) <[B]+[cl) = R'exP('kt) (111.1) 45 x: 22.18“ cowuuavom onu wmfiumn a: omw um uwmmno moquHOmn< may :1 93 A55 .zmz mugs omovfimo each:00umo mo H.HHH seamed 35.25» emu . owm . com . 9.21 om. . o.» 1 we . o~_o.o O J O . ooo 40—00 00 00000 i . v 0000 JCNOO w. o o o O o 000 4 J o m o oo oo ..¢N0.0 U 00 o oo o o 1 an o 0.8 my w... as $8.0 ooofluw i . waned O mnod 46 'where R is [L]°/[M]°. [L10 and [M10 are the initial concentrations of MPH and cytochrome oxidase respectively. and k is k1([L]°-[M]°). The rate of change of absorbance at 830 nm is proportional to the production of C. dC/dt - k2(([BJ-+'[CJ) - [03) The right side of Equation 111.1 was substituted for ([B] + [0]) above and the resulting equation was solved numerically with the value of k2 being adjusted. A typical fit of the equation to the data is given in Figure 111.2. ‘The residuals from this fit are given in Figure 111.3. Their randomness indicates that the equation fits the data and that the above mechanism. where CuA receives its electrons via cytochrome g, is sufficient to explain the data of this experiment. The values found for k2 are given in Table 111.1 Examination of these values shows that the rate constants exhibit no dependence on the concentration of MPH. again indicating that CuA does not receive its electrons directly from the reductant. The initial part of the reaction was also followed through the absorbance change at 605 nm. The first step in the above reaction scheme was used to develop the following second-order rate equation that related absorbances and the initial concentrations of reactants: A. - (1,-10) i[l-exp(-k't)]/[1-exp(-k't)/R]l + A. 47 I Absorbance 0.055 b I 0.050 0.045 0 Figure 111.2 Time (sec) The Calculated and Experimental Absorbances at 830 nm for the Reduction of Cytochrome Oxidase with.MPH. 0 calculated, A experimental, [aa3]-16.5 um, [MPH]- 80.2 mi 48 0.50 0.30 - 0.l0- ° 0 -0.|0 r O 0 Residual X |00 -o.3o - J _ 0.50 l I ‘ i l 0 Lo 2.0 3.0 4.0 5.0 Time (seconds) Figure 111.3 The Residuals of the Calculated Absorptions at 830 nm for the Reduction of Cytochrome Oxidase with MPH. [aa3]=16.5 uM, [MPH]=80.2 uM 49 Table 111.1 Rate Constants For the Initial Phase of the Reaction of MPH and Cytochrome Oxidase at Varying MPH Concentrations [KPH]. vi [443]. pl k1 x 105 u'1 s-1 ‘ k2 .‘1" 27.8 8.2 1.10:0.04 40.1 8.2 1.07:0.02 40.1 8.2 1.10:0.05 11.2 30.0 2.8x0.3‘ee 40.1 8.2 11.810.5 40.1 8.2 13.4:2.3 40.1 ‘ 8.2 13.4:2.s 27.8 8.2 17.0:4.9 27.8 8.2 12.2:0.4 10.8 27 17.8:i0.s’u 9 from data at 605 nm 9‘ from data at 830 nm 9“ from Halaka (1981) 50 where At, A0, and A. are the absorbances at tine t, zero. and at the end of the phase. respectively. k' is k1(Lo-lo) and I is Lollo. Since this second-order equation fit the initial absorbance changes. the [PE is reacting at cytochrome I.- The fit of this equation to the data is given in Figure 111.4 . and the residuals of this fit are given in Figure 111.5. The rate constants found were about one-third of those found by Halaka, apparently indicating a difference in the enzyne preparations. W The steady-state activity of soluble cytochrone oxidase varies with the type of nonionic detergent present. The identity of the detergent head group affects the activity. Rosevear et.al.(l980) found that cytochrone oxidase was two to ten tines more active in lauryl naltoside than in Tween-20. though the two detergents have identical fatty acid chains. Robinson et.al. (1985) used a variety of types of detergents and found that the electron transport activity of the enzyne depended on both the head group and the fatty acid chain of the detergent present. Cytochrone oxidase as isolated is heterogeneous. None of its soluble forns are as active as the membrane-bound enzyle. The nore active forns of the solubilized ensyle nay be structurally nore similar to the enzyne in the nenbrane. Absorbance 51 (148 I (146 (L44 I (142- CL40- (138- (136 I 5 0.34? Figure 111.4 l 1 so (20 Iéo Time (ms) 1 L 200 1 240 The Calculated and Experimental Absorbances at 605 um for the Reduction of Cytochrome Oxidase with MPH. [aa3]=16.5 11M, [MPH]-55.0 11M 52 m :1 o.nmunmng .2: m.oHum mag .mmz no“: ommefixo mEOHAUOuho mo aowuuavom wnu you a: new um mou=m390mn< wouoasoawo ecu mo mamsvfimcm 2.5 2.: m.HHH muamfim 0v 8N q OWN 4 emu u ow— d om- J Cum d 1 ON;|. .umAVI Aucxvl nXHo owgu omAu 001 x snonpyseu 53 If the steady-state activity of an enzyme form is higher. then one or nore steps in its turnover reaction must occur at a faster rate. Determination of which step is :faster would give clues about the enzyme structure of the more active form. The stopped-flow experiment described here was designed to see mech step in the reaction is accelerated when the enzyme is reduced in the presence of lauryl maltoside. The enzyme solutions contained either 0.5i lauryl maltoside or 0.5i Tween-20. ‘The kinetics of the anaerobic reduction of cytochrome oxidase by 5-nethylphenazinium methylsulfate (KPH) was studied. IPF was used as the reductant because its oxidation is easily followed through changes in its optical spectrum and its reduction potential is low enough that it will completely reduce cytochrone oxidase under anaerobic conditions. W Spectral changes in the 340-500 nn region were followed while the enzyne was reduced by KPH. Oxidation of an led to an increase in absorbance at 388 nm, while reduction of cytochrone oxidase led to a decrease in absorbance at 420 nm and an increase in absorbance at 444 nm. The first and last spectra collected are shown in Figure 111.6. 54 L00 0.90 - ° o 0.80 » . 0.70 - . e . ‘ 0.60 . . 050- - V . 040- Absorbance 030- 020- 0.10- . 2 0'00 340 3&04 3180 L 400 A 420 J 4404 4éo A 480 560 Wavelength (nm) Figure III.6 The Spectra from the Partial Reduction ~ of Cytochrome Oxidase with MPH in the Presence of Lauryl Maltoside. 0 first spectrum, A last spectrum ' 55 It was shown by Halaka (1981) that the overall stoichiometry of the reaction of NPR and oxidized cytochrome oxidase. in the prescence of Tween-20. is 2 NPR + cytochrome oxidase (oxidized) ---+ 2 IP+ + cytochrome oxidase (reduced). Halaka also found that the reaction was triphasic when carried out in Tween-20. The initial, fast phase was first order in both reactants. The two phases that followed were both slower. first order in enzyme. and zero order in MPH. The same triphasic behavior was observed when the enzyme was reduced by IPH in the prescence of lauryl maltoside instead of Tween-20. W A second-order rate equation that related absorbances and the initial concentrations of reactants was used to do the nonlinear least-squares fitting. The equation was derived for the second-order reaction k L + M —-D products If x is the fraction of the reaction remaining, then the rate of the reaction may be expressed by dxldtI-klo(R-l+x) where 1.0 and lo are initial concentrations of L and l and l-Lollo. Integration of this equation gives 1n [x/(R-l+x)] I -k't + q 56 so x ' I(R-1)exp(-k't+q)l/[l-exp(-k't+q)] (111.2) where k' - k(R-l)lo and q is the product of (R-1) and the integration constant. The fraction of reaction left. x. may also be expressed in terms of absorbances by x I 1 - (At-Ao)/(Ao-A°). where At. A0. and A. are the absorbances at times t. 0. and at the end of the phase. Setting this equal to the right- hand side of Equation 111.2 and solving for At gives At - (A.-A°)I[l-exp(-k't)l/[l-exp(-k't)/R]1 + Ao (after the integration constant is evaluated by using At-Ao at t-O). Figure 111.7 shows a typical fit of this second order equation to the data collected during the initial phase of the reduction. The randomness of the residuals indicates that the equation used fits the data. The second-order rate constant obtained was 5.74 x 105 [-1 s”1 (Table 111.2). The last two phases were fit by the equation representing two first order processes: At - A. - AlexP(-k1t)-A2exp(-k2t). The fit of this equation to the data is given in Figure 111.8. Again, the residuals are small and random. indi- cating the equation fits the data. The two rate constants 1 1 obtained were 0.031:t.002 s- and 0.0021i.0003 a. (Table 111.2). 57 .02-° .Ol- o -.Ol~c90 ° (0) C16C>~ Olfiih C152.~ Absorbance L 0.48 + ° 0.44 P 0.40 -g A Figure 111.7 0.0 .20 .40 o. l l l .80 .60 I.'00' l.20 Time (seconds) The Calculated and Experimental Absorbances at 444 nm for the First Phase of Reduction of Cytochrome Oxidase with.MPH in the Presence of Lauryl Maltoside. [as ]-3.0 pH [MPH]=2. 7 11M 3 58 Table 111.2 Rate Constants for the Anaerobic Reaction of HPR and Cytochrome Oxidase in the Prescence of Lauryl Maltoside and Tween-20. Second Order Phase First Order Phases k x 105 I-1 s—1 k1 s -1 k2 s -1 5.710.5 . 0.03610.002 a 0.002830.0003' 7.3io.8 4 0.031:0.002 t 0.0021:0.0003¢ s.s:0.s 4. 0.031to.004 .. 0.004110.0007u 7.32. .4 0.05:0.01 .. 0.00l4I0.000S" 4.910.: 8* 0.030:0.002 .4 0.002830.0004u ' in lauryl maltoside ‘0 in Tween-2O 59 GAO- Residuol X l03 _0 .5 O (c) 01%!- (18C)- (178'- Absorbance (176 (174 I O (b) l 1 l l Figure 111.8 80 I20 Iéo ' 200 240 l 250 Time (seconds) The Calculated and Experimental Absorbances at 444 nm for the Final Phases of the Reduction of Cytochrome Oxidase with MPH in the Presence of Lauryl Maltoside. [aa3]=2.8 1.1M, [MPH]-3.2 11M 60 The corresponding rate constants found for the three phases of the reaction between this preparation of the enzyme and lPR. carried out in Tween-20, are also given in Table 111.2. We}; There is no significant difference in the rate constants obtained when the enzyme is in Tween-20 and when the enzyme is in lauryl maltoside. This indicates that either (1) the step that is faster during turnover in lauryl maltoside is not one of the steps that can be observed at 444 or 605 nm. or (2) the enzyme converts to the oxygenated form during turnover and it is this form which has a higher activity in lauryl maltoside. W W Anaerobic solutions of 104 nl lPH and 23.1 nl oxidized cytochrome oxidase were mixed in a 2:1 volume ratio in the reservoir on one side of the stopped-flow instrument. Aerobic buffer was prepared in the reservoir on the other side of the instrument by mixing anaerobic and oxygen-saturated 50 ml HEPES buffer. All solutions contained 0.5% Tween-20. The enzyme solution was 150 ml in HEPES and the lPR solution had no buffer. 0.5 units of catalase were present per milliliter of enzyme (after 61 mixing) to prevent formation of peroxide. The proportion of anaerobic and oxygen-saturated buffers was varied during the experiment. resulting in.different oxygen concentrations. Approximately ten minutes after the lPR and enzyme were mixed. the resulting solution of reduced enzyme and excess lPR was mixed with the aerobic buffer and the reaction was observed. W A typical absorption-time curve at 444 nm is shown in Figure 111.9. At the beginning of observation. about 75% of the enzyme was recxidized. which produced the oxygenated enzyme. An increase in absorption at 444 nm is due to the reduction of cytochromes g_and g3. Adter observation was begun. the absorption initially increased. a plateau was reached. and then the absorption increase continued. Before the plateau was established. electrons were entering the enzyme faster than they were leaving it. The absorbance increase after the plateau is due to continued reduction of the enzyme upon deletion of the oxygen. For the plateau to exist, the change in the sum of the concentrations of reduced cytochromes ; and ;3 must be small. This implies that, during the plateau period, electrons entered the cytochrome L site at approximately the same rate as that at which they left the enzyme. The steps that occur after cytochrome ; reduction are electron 62 L30 0 O O O O O O 120- ° 0 IJO- ° 0 o g 100- I! .0 5 § 090 ° <1 . 030- ° 0 0° saw 0.70 ° 1 l 4 l 4 l a l 1 l L _l J l 0 L0 20 30 4D 50 60 10 Time (s) Figure III.9 The Change in Absorbance at 444 um During the Reduction of Oxygenated Cytochrome Oxidase. [aa3]-3.8 1.1M, [Haj-13.0 1.1M 63 transfer through the enzyme, in one or several steps. and the transfer of electrons to oxygen. During the plateau period, one of these steps was occurring at the same rate at which electrons were entering the enzyme, resulting in the establishment of an approximate steady state. If the oxygen reaction had been slower than the intramolecular electron transfer steps, then the level of oxidation of the enzyme during the steady state would have depended on the oxygen concentration. No such dependence ‘waa observed. so an intramolecular electron transfer reaction must be slower than the oxygen reaction. If that step directly involved cytochrome ;?+ or a species that it transfers electrons to quickly, then d(Absorbance)/dt - d(;?+)/dt - kl [lPB][§?+]-k2[g?+] ~ 0. Then h2~k1 [lPH][13+]/[;2+]. Because there was oxygen present during the plateau period. the concentration of reduced cytochrome g3 was small. Since the absorbance level indicates the enzyme was very close to 50% reduced. most of the cytochrame g_must have been reduced. Thus [g?+]l[g?+] would have been small and k2 would be small compared to the pseudo order rate constant for the lPR reaction. C.1_nixi_AuLLxxei The initial data analyses were done in an empirical manner. Although most of the data were collected in scanning mode, fixed wavelength files were used to analyze 64 the increase in absorbance before the plateau. This was due to the increased time resolution of the fixed-wavelength data and the short time before the plateau. The following one exponential equation was found to fit this initial part of the absorbance increase: At - A0 + (A. - A0) [1 - oxp(-xt)] At, A0. and A. are the absorbances at time t. time zero, and at the end of the phase. The calculated and observed absorbances and the residuals are shown in Figure 111.10. The rate constants found, from two data files. were 1212 s"1 and 14:2 s-l. This region would include the changes due to a second-order reaction between lPH and cytochrome ;, The rate constants are thus pseudo first-order constants. They correspond to the bimolecular rate constants 4.4:0.6 x 105 u'1 s'1 and 5.1:0.7 x 105 u'1 s-l, which agree well with the value 3.2tO.5 x 105 l".1 s"1 found for this phase of the anaerobic reduction of resting cytochrome oxidase by lPR (Halaka. 1981). It should be noted that the data used do not include those points just prior to the plateau since the oxygen reaction would be expected to be increasingly important as the steady-state phase is approached. Also. the initial absorbance found by fitting the equation indicates that about 20% of the enzyme was not recxidized within the mixing time. This may be due to the competition between the oxygen and lPR reactions. 65 f0 9. 2. x O o o ‘5 ° ° o :3 0.0 0000000 0000000000 E o 0 00 o (a) m -2.0 0.70 0.68 6 o O 02 0 ° 0 s 5 00" e 0.66 boo O 6 .2 9.8“ <1 be 0.64 66 8° ‘60 30 (b) 0 20 40 60 Time (ms) Figure 111.10 The Calculated (A) and Experimental (0) Absorbances at 444 nm for the Presteady- State Reduction of Oxygenated Cytochrome Oxidase with MPH: Empirical One Exponential Equation. [aa3]=3.8 11M, [MPH]=26.9 1.1M 66 .A two exponential equation was fit to the data after the plateau: At - A... -A1 exp(-k1t)- Azexp(-k2t) where the absorbances are as defined above. A1 and A2 are parameters proportional to the fraction of the reaction described by each exponential. and k1 and k; are rate constants. A typical fit to data collected in the scanning mode is shown in Figure 111.11. The values of Aq and A2 indicate that the reaction described by the second exponential (k2) accounts for only one percent of the total absorbance change that occurs after the plateau” ‘The rate constant found for this minor component of the reaction varies from 0.1410.08 a"1 to 0.4:o.1 8"1 (Table 111.3). The average value of 0.28 s"1 is 1 found by Halaka (1981) for the close to the value of 0.2 s- first-order phase which accounted for the reduction of about fifteen percent of the cytochrome g3 in the resting enzyme. Since one phase accounts for essentially all of the absorbance change and the rate constant for the minor component is similar to that found for a minor component of the resting enzyme indicate that essentially all of the heterogeneity characteristic of the resting enzyme is absent in the oxygenated enzyme of this experiment. The rate constant for the predominant phase was found to vary linearly with the concentration of lPR at the end of the plateau. This is shown in Figure 111.12. This implies 67 O 9 X 0.4 —_. . ° 0 o O loo 0° 0 .8 0.0” O o O O 0:5 "'0 O O 0 ‘03-04“- (a) IZO 0063 9 o o 6 R o s 8 . ’ o |.l6b _6 3 c |.|2- o g -. 8 LOB- <1 |.04- L00; (b) ‘04 s 8 I40 :2 Figure III.11 Time (s) (a) Residuals and (b) Calculated and Experimental Absorbances at 444 nm for the Poststeady-state Reduction of Oxygenated Cytochrome Oxidase with MPH: Empirical Two Exponential Equation. [aa3]=3.8 uM 68 Table 111.3 Rate Constants for the Reduction of Oxygenated Cytochrome Oxidase with lPH. Post Steady-State. [urn]‘. an 11 .“1 12 s-1 21.4 4.2:0.2 0.27:0.05 20.4 3.910.: 0.14:0.08 18.4 3.310.2 0.3:0.1 13.4 2.6*0.1 0.37:0.06 14.8 2.610.1 0.2:0.2 13.5 2.03:0.09 0.4:0.1 13.0 1.98:0.05 0.29*0.05 ' at the end of the plateau 69 5.0 ‘13? .4? (- I.O ' O 1 I l 1 l 1 l l I l2.0 l4.0 l6.0 I8.0 20.0 22.0 MPH Concentration (pM) Figure 111.12 The MPH Concentration Dependence of the Pseudo First-Order Rate Constant for the Poststeady- state Reduction of Cytochrome Oxidase 70 that the reaction is pseudo first-order in lPH. The slope of the graph corresponds to the bimolecular rate constant 2.610.1 x 105 u‘1 3'1. In order for the absorbance change after the plateau to 'be due to the reaction between lPH and cytochrome oxidase. this must be the rate-limiting step. If the intramolecular electron transfer were rate limiting the rate constant would not show an lPH concentration dependence. The rate constant for the intramolecular reaction must thus be greater than the largest pseudo firt-order rate constant found. and so is larger than 4.2 s-l. This is substantially greater than the value 0.002 s-1 found for the intramolecular electron transfer for the dominant species of the resting enzyme (Halaka. 1981). An attempt was made to find a more accurate value for the intramolecular rate constant. The equations were developed from ‘2+ 3+ 1‘2 .3+ C(R-lH urn + .3*c(“1’* _fiq‘ucu-n'r + 111* It was assumed that the concentrations of completely oxidized enzyme and of reduced C were negligible at the end of the plateau and that the bimolecular reaction was pseudo first-order. The fraction of completely reduced enzyme at time t was defined as x: x - I.“ dbl“ 1 ./ [PP]. 71 Then x is given by x - l-ki exp(-k2t)l(ki-h2) - k; exp(-kit)l(ki-kz) (111.3) where hi I kIIlPB]. It may also be written in terms of absorbances: x I (At - A°)I(Ao - A0) where At, A0 and A. are the absorbances at time t. time zero (defined as the end of the plateau) and at the end of the reaction, respectively. Then At - (A. - A0); + A0, Equation 111.3 was substituted for x and the result was fit to the data after the plateau. Because the lPB reaction was pseudo order, the fixed wavelength files could be used. This equation did not fit the data as well as the empirical two exponential equation as the residuals were quite systematic (Figure 111.13). The data were not sufficient to fit the intramolecular rate constant well. The program would converge for only one data file. The rate constants 1 and 1935 s-l. Although the obtained were 2.42 10.04 s- equation is symmetric in the rate constants and the oxygen concentration is not known for this file, the smaller of the two rate constants should be that for the lPR reaction: not only is the lPH reaction expected to be rate-limiting, but the smaller rate constant is in the range of that expected for the lPH reaction based on the results from the data analyses of the scanning files. .22 (3 :3 2 (D O) 0: -.02 l.25 us 0D (J 5 |.05 .13 L- C) 3 q 0.95 0.85 Figure 111.13 0 LC 2.0 3.0 4.0 72 l l l l l 1 Time (s) The (a) Residuals and (b) Calculated III-and Experimental (0) Absorbances at 444 nm for the Poststeady-state Reduction of Oxygenated Cytochrome Oxidase with MPH: Mechanistic Equation. [aa3]=3.8 uM 73 W In order for the post steady state reaction to be limited by the reductant concentration. the intramolecular electron transfer must occur with a rate constant larger than the highest pseudo-order rate constant found for the post steady-state lPB reaction. That is. the intramolecular electron transfer must have a rate constant larger than 4.2 s-1 (Table 111.3). The average number of electrons transferred can be calculated from the average rate of oxygen reduction during the plateau. 14 ul/s. The corresponding average electron transfer rate is 14 s’l. The limiting intramolecular electron transfer thus has a rate constant between 4.2 and 14 s-l. This rate constant cannot be definitively assigned to a specific intramolecular step. If it corresponds to the reduction of cytochrome gg+. then it confirms the literature reports of an acceleration of that rate in the oxygenated enzyme. Peterson and Cox (1980) report a comparable value, 7.5 s-l. for that reaction in the absence of oxygen. Hill and Greenwood (1984) and Greenwood (1967) report a value of 300 s"1 for the reduction of cytochrome 23 in the presence of oxygen. This large difference in rate constants indicates that the presence of oxygen somehow changes the nature of the 53 site. If the intramolecular rate constant found in this experiment is for the reduction of CuA, then it is less than 74 1 found for that reaction in the resting the value 17.8 s- enzyme (Halaka. 1981). This experiment also confirms that the rate constant for the reduction of cytochrome g_in the oxygenated enzyme is not very different than that of the resting enzyme. Another interesting note is that the enzyme was oxidized more than 50% during the flow time, in spite of the presence of excess reductant: yet. by the beginning of the observation time, when the oxygen concentration had been lowered by a larger fraction than had the reductant concen- tration. the enzyme species present was being reduced by the lPH. This suggests that when the resting enzyme is reduced. an enzyme species is produced which loses its electrons more readily than does the oxygenated form of the enzyme. CHAPTER IV MN TI TAT F THE 43w W W The current literature concerning the magnetic state of the ‘3 center in the resting and oxidized cyanide-bound forms of cytochrome oxidase is conflicting. lhile it is relatively uniform in stating that the iron and copper ions are magnetically coupled. there is disagreement over the nature of that coupling. lagnetic circular dichroism (lCD) data (Thomson et.al.. 1981) indicate that the ions of the :3 center of the oxidized cyanide-bound bovine enzyme are ferromagnetic- ally coupled with J)10 cm'l' loaabauer data for both bacterial and bovine cytochrome oxidase (Kent et.al.. 1982) also indicate ferromagnetically coupled ions in the :3 center of the cyanide-bound enzyme. Thomson could not comment on the magnetic state of the g3 center of the resting enzyme as it is lCD silent. Kent found preparation-dependent results for the oxidized bacterial enzyme: the ground state sometimes appeared to be SIO and sometimes SIl. lcssbauer data for the bovine resting enzyme are somewhat ambiguous due to the small signal-to-noise ratio. 75 76 By using magnetic susceptibility measurements. Tweedle et.al. (1978) found that the iron and copper ions of the g3 center are antiferromagnetically coupled in both the oxidized cyanide-bound and the resting enzyme. There are several problems with trying to compare the results of these lCD. ldssbauer. and magnetic susceptibility experiments: (1) Each type of measurement was performed on cytochrome oxidase isolated by a different technique. Yonetani. Toshikawa and Hartzell-Beinert isolation procedures were used for the enzyme for lCD, loss- ‘bauer. and magnetic quceptibility measurements. respectively. In some cases enzyme isolated by different techniques has shown different reduction (Halaka. 1981: Jones et.al.. 1983) and ligand binding properties (Naqui et.al.. 1984) at the ;3 center. It is not known if the magnetic properties are affected by the isolation procedure used. (2) lhile both Thomson and Tweedle used the dispersing detergent Tween-20. lent used either Triton 1-100 or deoxycholate with the bacterial enzyme and does not report what detergent was used with the bovine oxidase. lCD samples must be optical glasses. so the lCD sample was in a 50% ethylene glycol solvent. Enzyme activity differs with the detergent used (Robinson and Capaldi, 1977: Rosevear et.al.. 1980: 77 Robinson et.al. 1985). For simpler transition metal compounds. small changes in bond angle can strongly influence the nature of the coupling between the metals (Hoard. et.al.. 1967). Also. the coupling of the ions in a compound prepared as a model for the ;3 center depended on the solvent used (Gunter et.al.. 1984). ‘Thus it can be seen that the environment provided 1J1 solution may change the way that the enzyme behaves. Ihether the changes in the environ- ment caused by different detergents or by different solvents change the magnetic properties of the enzyme is not known. (3) To make susceptibility measurements, dissolved oxygen must be removed from the sample. Tweedle et.al. removed it by adding the oxygen scavengers glucose oxidase and glucose to the sample. These scavengers produce hydrogen peroxide. which is known to bind to the enzyme (Bickar et.al.. 1982). It is not known if the susceptibility samples prepared in this way are peroxide bound or whether peroxide binding affects the magnetic susceptibility of the enzyme. The experiments described here were designed to see if the magnetic susceptibility was influenced by the procedural differences described above. lagnetic susceptibility measurements were made on enzyme isolated by 78 two techniques (Hartzell-Beinert and Yonetani) and in different detergents (Tween-20 and lauryl maltoside). as well as in a 50% ethylene glycol solvent. To check the reproducibility. samples were exchanged with L. Wilson at Rice University. Houston, Texas. WW WW9. The magnetic susceptibility of resting cytochrome oxidase was measured at temperatures from about 2 to 180 1. Above 20 I the susceptibility is inversely dependent on the temperature: that is, it conforms to the Curie Law 2 2 2 2 2 x - N g B S(S+1) g N.B ”eff =.ueff 3kT 3kT 7.95 T where 8 is the Bohr magneton number and flgff is the square of the effective magnetic moment. given by [1sz .. (2)2 8 (8+1) . The .x vs 1/1' plot of resting cytochrome oxidase in maltoside. shown in Figure 1V.l. has a slope at higher temperatures which corresponds to a ufiff of 32.0 i 0.5. Each of the four metal ions in resting cytochrome oxidase contributes additively to the susceptibility, and 79 03.6.35 maou500u>o wafiummm you ouaumuoqea. Hooch—«00m w> huflaqflunnomam 0:0:me H .>H 0.3me .75 7232380... ON. 0.. 2. I. N _. 0.. no. mo. #0. NO. 050. oo\ . 0.. 0000 o I. O o 1 ON. 0 4 x O ._ on. d 8 L J O . w o . ow m. . O o D ._ on. C. O o . ow. . Oh. 80 so to the square of the effective magnetic moment. The contribution of cytochrome g?+ and CuA can be calculated from their spin states (SIl/2 for both) and their g values: 82' gy, and gx are 3.03. 2.21. and 1.45 for cytochrome g?+ and 2.18. 2.02. and 1.99 for on;2+ (Aasa,et.al..1976). The 2'i- g: ). The mean g values are then (E)2 I (1/3)(‘i + 3y effective magnetic moments squared that are found from these values are 4.0 for cytochrome a3+ and 3.2 for CuA. Thus the contribution of the ;3 center to flgff is 7.2 and the “if; for the ;3 center is then 32.0-7.2 - 24.8. Comparing this with the values of the “Eff for the possible magnetic states of the ;3 center (Table IV.1) permits a determination of the actual magnetic state. Assuming the iron ion is in the +3 state. there are only two states which give a ngff value of about 24 for the g3 center. One is an intermediate spin cytochrome coupled ferromagnetically to the copper, the other is a high spin cytochrome coupled antiferromagnetically to the copper. The iron of cytochrome g3 is generally believed to be high spin (Babcock et.al.. 1976). so the ions of the ;3 center must be antiferromagnetically coupled. The fact that the X vs l/T plot remains linear at high temperature shows that the coupling factor, I. must be large enough that the higher energy levels are not substantially populated at high temperature, that is, -2J ) 126 cm-1. 81 Table 1V.l Squares of Effective lagnetic loments for the Possible lmgnetic States for the ;3 Center of Cytochrome Oxidase. Cyt g3 CuB Coupling usz‘ SI1/2 SIl/2 antiferromagnetic 0(strong).6(weak) SI1/2 SI1/2 none 6 SI1l2 SIl/2 ferromagnetic 8(strons).6(weak) SIS/2 SIl/2 antiferromagnetic 8(strons).18(weak) SIS/2 SIl/2 none 18 SI3/2 SIl/2 ferromagnetic 24(strong),l8(weak) SI5/2 SI1I2 antiferromagnetic 24(strong).38(weak) SI5/2 SI1/2 none 38 SI5/2 SIl/2 ferromagnetic 48(strong).38(weak) 120,, = (g>2 s (3+1). (1)2 - 2.0 82 At lower temperatures the susceptibility no longer follows the Curie Law. This may be explained by assuming that the spin degeneracy of the ground state is lifted by spin-orbit interaction. This phenomenon is commonly called zero-field splitting because it occurs in the absence of an applied magnetic field. This splitting is generally small enough that it does not influence the susceptibility at high temperatures, but it can influence the susceptibility when the temperature is low enough for kT to approximate the energy difference between the split levels. Griffith (1971) has derived the equation that describes the temperature behavior of the square of the effective magnetic moment of a coupled ion pair consisting of an SI5I2 ion and an SIl/2 ion. This equation includes the axial zero-field splitting parameter, D, but assumes that the rhombic parameter, E. is zero. This assumptimn is reasonable for cytochrome oxidase because for heme proteins E is often small compared to D (Unenoyama et.al.. 1968; Nakano et.al.. 1972). Griffith's equation for flgff may, after modification, be used to find the temperature dependence of the suscepti- bility of all the metal centers in the enzyme. The contri- bution of cytochrome g?+ and CuA is accounted for by adding 7.2 to Griffith's flgff: u2 __. 02(2) +1023) exp <-x> eff 5 + 783190-10 (1v.1 ) 83 where “2(2) , 4L9+(2y-7)expHy/3)+‘(8y—2)ex:(-161/311 y[1+2exp(-4y/3)+2exp(-16y/31]_ u2(3):§__[90+(5y-65)exp('2113)+’(20Y-16)BXP(_-5Y)+(4§Y'9)EXP(-§Y)J 5y[1+2exp(-2y/3)+2exp(-8y/3)+2exp(-6y)]' and x I 3J/kT and y = D/kT. Theoretical values of the magnetic susceptibility were calculated by using a large value of J and various values of D. The results. shown along with the experimental data in Figure 1V.2. indicate that a 0 value between 5 cm-1 and 10 cm-1 is adequate to explain the deviation from the Curie Law behavior that occurs at lower temperatures. This is within the range of D's that has been found for other heme proteins (Tasaki et.al.. 1966: Uenoyama et.al.. 1968: Behere et.al.. 1979). W 'hen resting enzyme samples were made anaerobic by the use of glucose and glucose oxidase. the susceptibilities found were different than when the oxygen was removed from the samples by evacuations followed by equilibrations with argon. Typical data for the samples made with the oxygen 84 b ‘ l 06 D = 5cm“ '- Q 0.98 b c: .- Q . actual 0.90 ~ m w: _ ‘ D c: t! ., 0.82 . o ‘e’ r . ° L- . O a 0.74 i’ . D 0"0cm_' x r A O Q . 4 0 0.66 . D o . e o ‘ D O 0.58- a” DD 0 "no 00° 0.50 ~ 0 o ° 00 1—‘_l——L-—h—-_.d—h 4 ‘ - ‘—a—i——h—_1 0.20 025 0.30 0.35 0.40 0.45 0.50 0.55 l/T (K") Figure IV.2 Experimental and Calculated Magnetic Susceptibilities vs Reciprocal Temperature for Resting Cytochrome Oxidase 85 scavengers are shown in Figure IV.3. Above 15 1 the susceptibility shows a Curie Law behavior. The high temperature slope corresponds to a “6ff of 24.8. and thus an g3 center contribution of 17.6. Comparing this with the calculated values in Table 17.1 indicated three possible magnetic states for the Q center: an SIS/2 cytochrome weakly coupled. either ferromagnetically or antiferromagne- tically, to the SI1/2 CuB, or an isolated SI3I2 cytochrome and SIl/2 C33. An epr spectrum of the sample did not indicate the presence of an isolated SI3/2 center. so the ions must remain coupled in some way. An optical spectrum of the sample was taken after the susceptibility determination had been made. A near-UV peak at 428 nm indicated that the enzyme was peroxide-bound (Iumar et.al.. 1984). IIt is postulated that the oxygen scrubbers produced enough peroxide to lead to this peroxide binding. To ensure that all of the enzyme capable of binding peroxide did so, twenty equivalents of hydrogmn peroxide were added to another resting enzyme sample which was then made anaerobic with the oxygen scrubbers. The susceptibility was then measured. The resulting data correspond to a "6ff of 26.0 (ge3). indicating no increase in the concentration of peroxide-bound enzyme. SUSCEPTIBILITY (mole 00:.) UI O J) O 01 O N O 5 Figure IV.3 86 L O .— £0 E? 1 . . . . . . . . . 0 .04 .08 .l2 .I6 .20 TEMPERATURE-'(K-I) Magnetic Susceptibility vs Reciprocal Temperature for Peroxide-bound Cytochrome Oxidase 87 LWLWM Formate-bound cytochrome oxidase was prepared by incubating the enzyme overnight with a 60- to lOO-fold excess of sodium formate. Binding was complete as indicated by the shift of the near-UV optical band to 417 nm (Nicholls. 1976). Then the oxygen in the sample was removed by cycles of evacuations and equilibrations with argon, the suscepti- bility behavior was very similar to that of the resting enzyme. Figure IV.4 shows that at higher temperatures the susceptibility followed the Curie 1" ; the tiff was 31.2 (24.0 for the g3 center). Since the iron of formate-bound cytochrome ‘3‘ is high spin (Nicholls.l976: Babcock et.al.. 1976). an antiferromagnetically coupled cytochrome 53-01% center is indicated. The linearity of the plot to 180 K indicates that -21 > 126 cm-1. .At low temperatures the susceptibility was less than predicted by the Curie Ltw. As with the resting enzyme. this deviation can be explained by zero-field splitting. Equation 1V.l can be used to calculate theoretical values of the susceptibility at various D values. As shown in 1 and 10 cm.—1 Figure 1V.5. a D value between 5 cm- adequately explains the lower than predicted susceptibility values. It has been suggested (Babcock et.al, 1981) that the magnetic behavior of the oxidized formate-bound enzyme could 88 .56 I .48 - .40 - ° .24 x per mole 003 .I6- 000- c9 l l l l l l 7 l 0 0.4 .08 .I2 IS .20 Temperature’l ( K"') Figure IV.4 Magnetic Susceptibility vs Reciprocal Temperature for Formate-bound Cytochrome Oxidase L38 L30 l.22 |.l4 .0 m m x per male 003 .0 (D O .0 Q N 0.74 0.66 0.58 0.50 Figure 89 ' 083cm" ’ O . a .- 0 e (experimental) o a . 4 a . ° D=5cm O .- A o D t‘ O L . ° ° _ O D A 0 a . ° 0 c A D A I- o D A A ,_ o 22 _' A Q . Dg l0¢m r 8 a - p- A A " A A II ‘ ‘ .“ . . A A 1 1 ‘ l 1 1 L l 1* 1 1 1 1 1 1 1 1 1 2.0 .24 .28 .32 I.36 ' .40 .44 .48 .52 Temperature (K ) 1V.5 Experimental and Calculated Magnetic Susceptibilities vs Reciprocal Temperature for Formate-bound Cytochrome Oxidase 90 help distinguish between two modes by which exchange coupling may occur. The models for these. illustrated in Figure 1V.6. may be classified as "back-side” bridging or as "front-side" ‘bridging. lore specific models. also shown. have been proposed by Palmer et.al. (1976) and by Blumberg and Peisach (1976). Formation of the formate-enzyme complex ’ would perturb the structure shown in (c) only slightly and likely cause little change in the susceptibility. Displacement of the u-oxo bridge in the structure shown in (41) would likely cause a large change in susceptibility. Since the susceptibility of the formate bound enzyme is essentially the same as that of the resting enzyme. the back-side bridging model is indicated. The above discussion assumes that formate binds to the iron of cytochrome g3. If it binds elsewhere, this discussion.of the geometry'of the binding does not apply. When the oxygen was removed from the formate-bound enzyme sample by use of the oxygen scavengers glucose oxidase and glucose. a different susceptibility behavior was observed. though the near-UV peak remained at 417 nm. indicating that the formate remained bound. The value of flgff was then 45.3. so the 13 center contribution to it was 38.1. Comparing this with the calculated values of ugff for the possible magnetic states of the g3 center (Table IV.1) seemed to indicate that the coupling was either weakened or broken altogether when the enzyme sample was prepared in this way. Since the resting enzyme was shown to bind peroxide when 91 Reslifl Cytochrome gsipossible Structures 8 8 2+ ' 3 5 Cu -8— 58 ’\—L L—F.e”-8—c.f’ : ‘s : \s [L so: NK “02 a) “back- side“ . b) ”front-side' . / r 't‘ I : Coat—his —- e‘“ —l~l20 his -F.e" Cu” I g \ / : : 0 I I N . N c) Palmer :1 g! 0976) d) Blumberg and Peisach “979) fi Figure IV.6 Possible Structures for the Cytochrome 8 Oxygen Reducing Site (from Babcock et.al., 1981 92 oxygen scavengers were used. it is reasonable to suppose that it may have also occurred'with the formate-bound enzyme. though this requires that the peroxide and formate bind at different locations. To test whether the enzyme was completely peroxide bound. four equivalents of peroxide were added to this formate-bound enzyme. the oxygen was removed with scavengers. and measurements were repeated. The 11%“- was then found to be 51.9. indicating an 53 center contribution of 44.7. This does not match any of the values on Table 1V.1. If it is assumed that the enzyme was closer to completely peroxide bound than it was when no peroxide was added. but not yet completely bound. ferromagnetic coupling between the high-spin cytochrome and the copper is indicated. In any case it is clear that removing the oxygen by the use of glucose oxidase perturbs the magnetic state of the formate-bound enzyme in some way. presumably by peroxide binding. W 'hen the oxidized enzyme binds cyanide. cytochrome 53 goes from high to low spin (Babcock et.al..l978). The 53 center still has no epr signal. so the metals remain coupled. lost of the disagreement in the literature about the magnetic state of the g3 center concerns the nature of the coupling in the cyanide-bound enzyme. 1f antiferromagnetic coupling with a large 21 value is retained then the g3 center would not contribute to flgff 93 since the populated spin state would have SIO. The contribution of cytochrome ; and CuA would remain at 7.2 since their spin states are unchanged from those in the resting enzyme (Babcock et.al.. 1976). The susceptibility would follow the Curie Law. If 21 is small compared to kT then the g3 center would be a thermally randomized SIl manifold whose contribution to gift would be 8.0. lith the value of 7.2 from the other metal ions. the total tiff would be 15.2. and the susceptibility would follow the Curie Law. If 21 has an intermediate value. then at high temperatures the susceptibility would show behavior typical of 21kT. Thus. flgff would be 15.2 at high temperatures and 7.2 at low temperatures and the susceptibility would not follow the Curie Law. The observed behavior is that expected from an intermediate value of I (Figure 17.7). Below is 3 I the value of "6ff is 7.4 while at m 180 K it is 16.0. Tweedle et.al. (1978) also observed limiting values of Biff, but they reached the values of 7.8 below 50 I and 15.1 at high temperatures. The equation that describes the temperature dependence of the susceptibility of two antiferromagnetically coupled spins is 2 2 gLsNB .1. _ -1 X 3kT [1 + 3 exp ( 2J/kT)] 94 o 0.60- ” o o 050- 0° 0 a" - o° 0 0° .9- 040" oo o E - a 5;. 0.30 - ° C. o - o 2" o +- o E 0.20” o :9 c +5. " o 8 m 010;] :3 (O - 00 0.10 0.20 0.30 0.40 0.50 Temperature" (K-') Figure IV.7 Magnetic Susceptibility vs Reciprocal Temperature for Cyanide-bound Cytochrome Oxidase 95 where 2.1 is the energy difference between the two spin states. In order to fit this equation to the cytochrome oxidase data. a Curie Law term. CIT. and a temperature independent term were added for the contribution of the a-center and the diamagnetism of the protein. respectively. Then the modified equation was fit to the data collected between 5 and 180 I the value of -1 that was found ranged 1 to 56.5 i 5.7 cm-1 from 29.6 t 11.6 cm— (Table IV.2). These values bracket the -1 value of 38.5 1 1.3 cm”1 found by Tweedle et.al.. but have a much larger standard deviation. Substantially different 1's were found when the equation was fit to data collected between 2 and 180 I. The I value 1 to 1.31 i 0.11 cr’l. then ranges from 1.26 i 0.13 cm- The differmnt 1 values can be explained if the enzyme is heterogeneous with two forms of the enzyme present. One form has a moderate value of J and the other form has a low value of I. ‘lhen high temperature data are used. the moderate I value is found because the enzyme with a low I value contributes to the signal in a Curie manner. and so does not contribute to the fit of the coupling term in the equation. At low temperatures the signal of the enzyme with a moderate :1 value is zero because saturation to the SIO level has occurred: thus the low I value is obtained when the low temperature data are used to fit the equation. This is supported by the fact that the Curie term found from high 96 Table IV.2 Coupling Factors for Cyanide-Bound Cytochrome Oxidase 22121211121 Tsanerstuxs_xanxa 1.11111 Yonetani 3-190 1 -29.6111.6 c.“1 (in Tween-20) 2—190 1 -1.2610.13 cm-l Hartzell-Beinert 5-180 x -30.116.4 cu."1 (in Tween-20) Hartzell-Beinert 20-201 1 -56.5ts.7 cm-1 (in Lauryl laltoside) 2-201 1 -1.31:0.11 cm-l 97 temperature data is larger than the value 0.90 expected from the cytochrome g,and the CuA (Table IV.3). An equation with two coupling terms. a Curie term. and a diamagnetic term was fit to the data collected between 2 and 180 11 The coefficients of the coupling terms were allowed to vary and the Curie constant was fixed at 0.9 to account for the contributions of the cytochrome ; and CuA. The coefficients thus represented the fraction of enzyme present in each form. Irt was found that 7811 percent of the enzyme had the I value of lower magnitude. It must be noted that when the equation is fit in this way. the accuracy of one I value is increased at the expense of the accuracy of the other 1 value. that is the J values 1 and -44I28 cm-l. This is because found were —1.7OI0.03 cm- the equation used does not describe the system perfectly. particularly at low temperature where zero-field splitting would have an effect. Because the errors in measurement of the high temperatures are greater. the program assumes most of the error is in the high temperature data. As a result. the equation is fit such that the I value obtained from the high temperatures is most affected by the deviations from the equation. MW In order to reproduce the enzyme environment in the sample used for the lCD studies done by Thomson (1981).a 98 Table IV.3 Curie Law Constants for Cyanide-Bound Cytochrome Oxidase 21121111121 1.81112 9 Yonetani 5-190 K 1.6130.04 (in Tween-20) 2-190 K 0.8110.4 Hartzell-Beinert 5-180 1 1.4910.11 (in Tween-20) Hartzell-Beinert 20-201 1 1.60:0.04 (in Lauryl laltoside) 2-201 1 0.6910.03 99 sample of cyanide-bound enzyme was dissolved in a solution that was fifty percent ethylene glycol and fifty percent HEPES buffer. The detergent used was lauryl maltoside. Direct addition of pure ethylene glycol to an enzyme sample previously dissolved in a buffer caused precipitation of the enzyme due to local concentration effects. Consequently. the cyanide-bound enzyme was prepared in the buffer. precip- itated with ammonium sulfate. and then redissolved in the solution of ethylene glycol. buffer and detergent. This method was chosen as ammonium sulfate precipitation does not harm the resting enzyme (note. however. that it was the cyanide-bound enzyme that was precipitated). It has the added benefit of avoiding the dilution of the enzyme that occurs with direct addition of ethylene glycol. The susceptibility data for the cyanide-bound enzyme in the ethylene glycol/buffer solution are shown in Figure 1V.8. A feature immediately apparent is the peak in the susceptibility values. which indicates an antiferromagnetic signal. What is surprising is that this peak shows up in the susceptibility data before the contributions of the cytochrome ; and CuA are subtracted. The equation fit to this data indicates that the Curie contributions of cytochrome g and CuA that were present in the other forms of the enzyme are not present in this sample. What has caused their loss is not known. but precipitation of the cyanide-bound enzyme 100 L45- L43 - |.4| - X per mole 003 0 L39 - l m 4 1 L l 0 40 80 Tao ‘1 I60 200 TEMPERATURE (K) Figure IV.8 Magnetic Susceptibility vs Reciprocal Temperature for Cyanide-bound Cytochrome Oxidase in 50:50 Ethylene Glycol/HEPES 101 has not been tried previously and it may affect the enzyme in some way. The most likely occurance would be auto-reduction. Wm T3919 IV-4 3h0'8 th' ”6ft values obtained for resting cytochrome oxidase that was prepared by two isolation procedures (Hartzell-Beinert and Yonetanti) and that was dissolved in two types of detergent (lauryl maltoside and Tween). The constancy of the effective magnetic moments indicates that neither the isolation procedure nor the detergent influences the magnetic state of the g3 center of the resting enzyme. The possibility of isolation procedure or detergent effects on the magnetic state of the g3 center of the low I and intermediate 1 forms of the cyanide-bound enzyme may be considered by examining their I values (Table IV.2). The isolation procedure did not change the I value of either form of the enzyme. Thus the choice of isolation procedure did not influence the ;3 center magnetic state in the -cyanide-bound enzyme. The choice of detergent did not influence the antiferromagnetic coupling factor of the low 1 form of the enzyme. again indicating no change in the magnetic state of its g3 center. However. the intermediate .1 form of the cyanide-bound enzyme had a larger coupling factor when in lauryl maltoside than it had when in Tween. This indicates a change in the bond overlap or bond angles at the ;3 center 102 Table IV.4 Effective lagnetic laments Squared for Resting Cytochrome Oxidase 2 txsnirstiea Tannaratuxe_xzaze Eeff Yonetani 12-169 K 33.112.5 (in Tween-20) Hartzell-Beinert 20-201 K 30.211.1 (in Tween-20) Hartzell-Beinert 30-180 1 32.010.s (in Lauryl laltoside) 103 of the enzyme. This change is probably not dramatic since the 1 value is still moderate enough to result in non-Curie Law behavior when the enzyme is in a lauryl maltoside solution. W The magnetic susceptibility behavior of resting oxidized cytochrome oxidase clearly indicates that the SI5/2 cytochrome g3 is antiferromagnetically coupled to the SI1I2 CuB ion. The Curie Law behavior that extends to temperatures as high as 180 K means that the energy separation between the two lowest energy levels exceeds 126 cm-1. The deviation from Curie Law behavior at low temperatures is readily explained by a zero-field splitting whose magnitude is typical for that of heme proteins. ‘These results confirm the conclusions reached by Tweedle et.al. (1978). In addition. this work shows that the magnetic state of the g3 center of the resting oxidized enzyme does not depend on the isolation procedure or on the detergent usede This eliminates one of the possible reasons for the conflicting reports of the type of coupling present in resting cytochrome oxidase. The susceptibility behavior of the peroxide-bound enzyme indicates that an SI3I2 cytochrome 53 is present. Such a spin state is extremely unusual in a heme protein. Peroxide binding studies done by Bickar et.al. (1982) indicated that some preparations of cytochrome oxidase may not bind peroxide completely. even in the presence of a large excess of peroxide. 104 but the 428 nm near-UV band of these samples means all of the enzyme in this sample was peroxide bound. The linearity of the susceptibility vs reciprocal temperature graphs also indicates that there is no shift in equilibrium occuring between spin states and that the coupling factor remains large. It thus appears that the peroxide-bound enzyme actually has an SIS/2 iron ion. The lack of a cytochrome g3 lCD signal from what was formerly believed to be the oxygenated enzyme (but is now believed to be peroxide-bound) (Babcock et.al.. 1976) confirms that the cytochrome 53 is not low spin. This work shows that the formate bound enzyme has an ;3 center which is composed of an SI5/2 cytochrome g3 coupled to an SIl/2 Cun. The adherence to the Curie Law at temperatures as high as 180 1 indicates that the energy difference between the two lowest energy levels is at least 126 cm-1. As with the resting enzyme. the deviation from the Curie Law behavior that occurs at low temperatures can be explained with reasonable values of a zero-field splitting. The relative deviations from the Curie Law of the resting and formate-bound enzyme suggest that the zero-field splitting of the formate-bound enzyme may be less than that of the resting enzyme. It must be remembered. however. that the average susceptibility is not very sensitive to the zero-field splitting. Indeed. the values of the zero-field splitting obtained from the average susceptibilities have sometimes been found to be in significant error when compared to the values obtained later 105 from the more dependable anisotropy measurements (litra. 1977). The calculations of the zero-field splitting shown here were undertaken to explain the low temperature deviation from the Curie Law. not to pinpoint the value of D. The magnetic susceptibility of oxidized cyanide-bound cytochrome oxidase indicates that this form of the enzyme is magnetically heterogeneous. Twenty-two percent of the cyanide-bound enzyme was in a form with a moderate I value while the rest was in a form with a low 1 value. The percent heterogeneity matches that found by Jensen et.al. (1984). who found a heterogeneity in partially reduced cyanide—bound cytochrome oxidase with twenty percent of the molecules showing an epr signal from the cytochrome ;§*-ncn complex. The behavior of the enzyme form with a moderate I value matches that found by Tweedle et.al. (1978). The magnitude of the coupling factor found in these experiments did not depend on the enzyme isolation procedure used. 1]:did depend somewhat on which detergent was used. but the variation was not sufficient to explain the conflicts in the magnetic susceptibility literature. lhile the form of the cyanide-enzyme with a moderate I value clearly shows antiferromagnetic coupling behavior. there remains some doubt about the behavior shown by the low 1 form of the cyanide-bound enzyme. Equation fitting to the data gave a small negative coupling factor with a magnitude of one to two reciprocal centimeters. but this 106 does not mandate true antiferromagnetic coupling. Another possibility is weak ferromagnetic coupling with a zero-field splitting that results in a lowest energy level with l. of zero. As shown in Figure 1V.9. neither ferro- nor antiferromagnetic coupling can result in an ordering of energy levels that would give a unique susceptibility behavior if the magnitude of the zero-field splitting is unknown. Hence the determination of the field dependence of the saturation temperature cannot permit the determination of the type of coupling. To check the reproducibility. samples of the resting and cyanide-bound enzyme were exchanged with L. Idlson of Rice University. Houston. Texas. The enzyme received from Rice was isolated with the Hartzell-Beinert technique and the oxygen had been removed by using glucose and glucose oxidase and catalase. This sample of resting enzyme gave the same susceptibility behavior as those we made that were deoxygenated by evacuations and argon equilibrations. that is. the use of the oxygen scrubbers had no effect on the susceptibility. When glucose and glucose oxidase were added to a sample of their enzyme (that had no catalase) in the presence of air. the near-UV band shifted to 428 nm. but it did so over the course of several minutes. In contrast. when the oxygen scrubbers were added to our enzyme. also with no added catalase. its near-UV band shifted to 428 nm 107 Ms=+l Ms=+| 3:0 M511 2.)! r’ S" \ D l ‘ M =0 M =-l 1‘ s 8 23:3 ti ‘\\\\‘\\_ Figure 1V.9 A.Diagram.of the Energy Levels for Ferromagnetically and Antiferromagnetically Coupled Systems Composed of Two S=k Ions. 108 within 15 seconds. 1t is suggested that the samples from Rice were frozen quickly enough after addition of oxygen scrubbers to prevent the binding of peroxide. or their added catalase kept peroxide from reaching a high concentration when the glucose oxidase was added in the helium-flushed glovebag. The results obtained for'the cyanide-bound enzyme sent from Rice were consistent with those obtained for the enzyme made in our laboratory. This is important because the peeple in the laboratory at Rice obtained results consistent with those reported by Tweedle et.al. (1978). that is. antiferromagnetic coupling with a -J vmlue around 40 cm-1 and no heterogeneity present. 'hile they can collect data only to 20 K. the standard deviations of the I values they obtained by data fitting are much smaller than those obtained in our laboratory. indicating closer adherence to the coupling equation with one I factor. However. the Curie factors they obtain are approximately 1.2 instead of the 0.9 that is expected from the a center. The deviation from 0.9 indicates the presence of a species (besides the ; center) that contributes in a Curie fashion in the temperature range used“ 'The value consistently obtained in our laboratory for the high temperature (5-180 I) Curie factor was approximately 1.5. Since all of these results were obtained consistently in both laboratories. it appears that there is a difference. as yet unknown. in the way the enzyme is handled 109 in the two laboratories which affects the susceptibility. 'hether there could have been a smaller amount of heterogeneity in their enzyme that was hard to detect above 20 K is open to question. but even if there was. the amount of heterogeneity changed with sample handling. This sensitivty to handling procedure is important when interpreting the results obtained tn different laboratories. 'hen these results differ. it may be due to different procedures. The effect of the presence of peroxide on the magnetic susceptibilities of the resting and formate-bomnd enzyme raised some questions. It appears that the peroxide binding changes the spin states of the iron ion to 3/2. The lack of heterogeneity of peroxide binding could be confirmed by finding the spin state of the peroxide-bound oxygenated cytochrome oxidase. which shows no peroxide-binding heterogeneity. It also appears that the formate binds in a different place than peroxide does. but where is not known. ‘This needs to be known before the susceptibility for the formate-bound enzyme can be interpreted in terms of a back- or front-binding model. It has been suggested (Palmer et.al.. 1976) that there is a structural change that occurs upon reduction of the enzyme which makes cytochrome g3 more accessible. For instance. the reaction of resting oxidase with sodium azide is rapid and the changes in the spectrum are small (lever et.al.. 1973). making it unlikely that the azide is binding at the iron ion. However. the partially reduced enzyme reacts with aside in a way which causes the conversion of a high-spin epr signal to a low-spin signal. demonstrating 110 111 that under those conditions the azide is reacting with the iron ion” ‘The magnetic susceptibility behavior of these vac azide derivatives would test the hypothesis that the azide binds in a different place in the resting and partially reduced enzyme forms. W The experiment detailed above indicated that if the resting enzyme is more active under anaerobic conditions. it is due to a step which cannot be observed at 444 nm. ‘This would suggest that the reaction should be followed. under the same conditions. by omservation of the 830 nm band to see if a difference in the CuA reaction is evident. Another possibility is that the detergent effect on the activity occurs with the oxygenated enzyme rather than the resting enzyme. This suggests that a comparison of the reactions of the oxygenated enzyme in different detergents needs to be carried out. with observation in the near-1R as well as the near-UV and visible spectral regions. Given previous suggestions that different aggregation states may be responsible for the difference in enzyme activities. the question of whether the aggregation states of the enzyme in the two detergents is the same for the oxygenated and resting enzyme becomes important. ‘The megnetic susceptibility results indicate that the formate ion and peroxide bind at different places on the 112 enzyme molecule. This could be confirmed if formate-bound enzyme is able to use peroxide as an electon acceptor. as the resting enzyme is able to do. L1 81‘ 0F REFERENCES LIST OF REFERENCES Aasa.R.. Albracht. S.P.J.. Falk. l.-E.. Lanne. B.. and Vanngard.T. (1976) Biochim. Biophys. Acta 122, 260-272. Andreasson. L.E. (1975) Eur. J. Biochem. 11: 591-597. Antalis. T.l. and Palmer. G. (1982) J. Biol. Chem. 211. 6194-6206. Antonini. E.. Brunori. l.. Colosimo. A. Greenwood. C.. and 'ilson. l.T. (1977) Proc. Natl. Acad. Sci. USA 11. 3128-3132. Antonini. G.. Brunori. l.. Colosimo. A.. lalatesta.. F.. and Sarti. P. (1985) J. Inorg. Biochem. 21. 289-293. Armstrong. F. Shaw. R.'.. and Beinert. H. (1983) Biochim. Biophys. Acta 111. 61-71. Awasthiu '!.C.. Chuang. T.F.. Keenan. T.'. and Crane. F.L. (1971) Biochim. Biophys. Acta 111. 42-48. Azzi. A.. (1980) Biochim. Biophys. Acta 121. 231-252. Babcock. G.T.. Vickery. L.E.. and Palmer. G. (1976) J. Biol. Chem. 111. 7907-7919. Babcock. G.T.. Vickery. L.E.. and Palmer. G. (1978) J. Biol. Chem. 151. 2400-2411. Babcock. G.T.. Callahan. P.l.. Ondrias. l.R. and Salmeen. I. (1981) Biochemistry 29; 959-966. Beetlestone. I. and George. P. (1964) Biochemistry a, 707- 714. Behere. D.V.. Date. S.K.. and litra. S. (1979) Chem. Phys. Letters 61, 544-547. Beinert. H.. Shaw. R.'.. Hansen. R.B.. and Hartzell. C.R. (1980) Biochim. Biophys. Acta 121. 458-470. 113 114 Bickar. D. Bonaventnra. 1.. and Bonaventnra. C. (1982) Biochemistry 11. 2661-2666. Blumberg. 1.8.. and Peisach. J. (1979) Dev. Bioch. 5, 153- 159. Boelens. B.. and 'eyer. 8.0. (1980) FEBS Lett. L11. 223-226. Brierly. G.P. and lerola. A. (1962) Biochim. Biophys. Acta 11: 205-211. Briggs. l.l. and Capaldi. R.A. (1977) Biochemistry 11.73-77. Brittain. T. and Greenwood. C. (1976) Biochem. J. 111, 453- 455. Brndvig. G.'.. Stevens. ‘l‘.B.. Horse. R.B.. and Chan. 8.1. (1981) Biochemistry 22 3912-3921. Brunori. l.. Colosimo. A. Rainoni. G.. Iilson. l.T.. and Antonini. B. (1979) J. Biol. Chem. 111. 10769-10775. Brunori. l.. Colosimo. A.. Sarti. P.. Antonini. B.. and Iilson. l.T. (1981) FEBS Lett. 111 195-198. Brunori. l.. Bickar. D.. Bonaventnra. I. and Bonaventnra. C. (1985) J. Biol. Chem. Zifl- 7165-7167. Capaldi. R.A.. lalatesta. F.. Darley-Usmar. V.l. (1983) Biochim. Biophys. Acta 111. 135-148. Canghey. 8.8.. Iallacc. 8.1.. Volpe. J>A. and Yoshikava. 8. (1976). in thg_figgzgg; (Boyer. P.D. ed.) Volume 13. Academic Press. New York. Chan. 8.1.. Brudvig. G.'.. Martin. C.T.. and Stevens. T.B. (1982) in (Ho. C. ed.) pp. 171-177 Blseyier. Amsterdam. Chuang. T.F. and Crane. F.L. (1973) Biochim. BioPhys. Acta 2&1; 563-570. Coolen. R.B.. Papadakis. N.. Avery. 1.. Bnke. C.G.. and Dye. J.L. (1975) Anal. Chem. 41. 1649-1655. Downer. N.'.. Robinson. N.C. and Capaldi. R.A. (1976) Biochemistry 11, 2930-2936. Dye. J.L. and Nicely. V.A. (1971) J. Chem. Ed. 18. 443-448. 115 Fuller. 8.D.. Darley-Usmar. V.M.. and Capaldi. B.A. (1981) Biochemistry 29» 7046-7053. Gibson. 0.8.. Greenwood. C.. 'harton. D.C.. and Palmer. G. (1965) J. Biol. Chem. ZAQ. 888-894. Gibson. C.R.. Palmer. G.. and Iharton. D.C.. (1965) J. Biol. Ch... “9.. 915-922e Gouterman. M. (1959) J. Chem. Phys. 19. 1139-1145. Griffith. 1.8. (1971) Molecular Physics 21, 141-143. Gunter. M.J.. Berry. 1.1.. and Murray. 1.8. (1984) J. Am. Chem. Soc. 191. 4227-4235. Halaka. FR.. (1981) Ph.D. dissertation. Michigan State University. East Lansing. MI. Hartzell. C.R.. and Beinert. B. (1974) Biochim. Biophys. Acta 1&1. 318-338. May. P.J.. Thibeault. J.C.. and Boffmann. R. (1975) J. Am. Chem. Soc. 21. 4884-4899. Board. J.L.. CODOI. C.R. and Click. I.D. (1967) J. Al. Chem. 300. up 1992-1996e Iiruka. T.. Iotani. M. and Yonetani. 'l‘.. (1968) Biochim. Biophys. Acta 111, 257-267. Jensen. P.. 'ilson. M.T.. Aasa. R. and Malmstrom. B.G. (1984) Biochem. J. 114. 829-837. Jones. G.D.. Jones. M.G.. Iilson. M.T.. Brunori. M.. Colosimo. A. and Sarti. P. (1983) Biochem. J. 122. 175-182. Iadenbach. B. and Merle. P. (1981) FEBS Lett. 111. 1-11. Kent. T.A.. Munch. B.. Dunham. 8.8.. Filter. 8.8.. Findling. K.L.. Yoshida. T. and Fee. J.A. (1982) J. Biol. Chem. 251. 12489-12492. Kent. T.A.. Young. L.J. Palmer. G.. Fer. J.A. and Munch. E. (1983) J. Biol. Chem. 111. 8543-8546. Iumar. C.. Naqui. A.. and Chance. B. (1984) J. Biol. Chem. 212, 2073-2076. Lemburg. B.. and 8tanbury. 1.8. (1967) Biochim. Biophys. AOt. mp 37-51. 116 Mitra.8. (1977) Progr. Inorg. Chem. 11. 309-408. Moss. 1.8.. Shapiro. 8.. King. 1.8.. Beinert. B.. and Bartrcll. C. (1978) 1. Biol. Chem. 111. 8072-8073. Munch. B. (1978) Methods anymol. 11. 346-379. Nahano. H.. Otsuha. 1.. and Tasahe. A. (1972) Biochim. Biophys. Acta 116. 355-371. Naqui. A.. Inumar. C.. Ching. 1.. Powers. L. and Chance. B. (1984) Biochemistry 21, 6222-6227. Nicholls. P. and Chance. B. (1974) in W W. (D. Bayashi. ed.) pp.479-534. Academic Press. New York. Nicholls. P. (1976) Biochim. Biophys. Acta. 410. 13-29. Ohunuhi. I. and 8ehuru. 1.. an. Chem.. Tohyo and Kyoto. (Maruren. Tokyo. 1958). Orii. Y. and King. T.E. (1972) FEBS Lett. 21. 199-201. Orii. 1. and King. 1.8. (1976) 1. Biol. Chem. 111. 7487-7493. Palmer. G.. Babcoch. G.T.. and Vichery. L.E.. (1976) PNAS 11. 2206-2210. Papadakis. N.. Coolen. R.B. and Dye. 1.L. (1975) Anal. Chem. £1. 1644-1649. Peterson. L.C. and Cox. R.P. (1980) Biochim. Biophys. Acta 121. 128-137. Robinson. N.C.. and Capaldi. R.A. (1977) Biochemistry L6, 375-381. IRobinson. N,C.. Neumann. 1.. and 'iginton. D.. (1985) Biochemistry. 11. 6298-6304. Rosevear. P.. VanAhen. 1.. Barter. 1.. Ferguson-Miller. S. (1980) Biochemistry 12. 4108-4115. Shaw. R.'.. Bansen. R.B.. and Beiner. B. (1978) 1. Biol. Chem. 111. 6637-6640. Stevens. 1.8.. Martin. C.T.. Wang. H.. Brudvig. G.'.. Scholes. C.P.. and Chan. 8.1. (1982) 1. Biol. Chem. 2.5.1: 12106-12113. 117 Stevens. 1.3.. Brudvig. G.'.. Bocian. D.F. and Chan. 8.1. (1979) Proc. Natl. Acad. Sci. USA 16. 3320-3324. Tasahi. A.. Otsuha. 1.. and Kotani. M. (1967) Biochim. Biophys. Acta Lil. 284-290. Thomson. A.1.. Johnson. M.K.. Greenwood. C.. Gooding. P.E. (1981) Biochem. J. 121. 687-697. Tweedle. M.F.. 'ilron. L.J.. Garcia-Iniguez. L.. Babcoch. G.T.. and Palmer. G. (1978) J. Biol. Chem. 211, 8065-8071. Uenoyama. 1.. Iizuha. T.. Morimoto. H. and Kotani. M. (1968) Biochim. Biophys. Acta 119. 159-166. VanBuuren. K.1.B.. Nichollr. P. and Van Gelder. B.F. (1972) Biochim. Biophys. Acta 111. 258-276. Van Buuren. K.J.B.. Van Geldcr. B.F.. Wilting. J.. and Braams.R. (1974) Biochim. Biophys. Acta 313, 421-429. Van Gelder. B.F. and Beinert. H. (1969) Biochim. Biophys. Acta 182. 1-24. Vih. 8.3.. and Capaldi. R.A. Biochemistry (1977) Li. 5755- 5759. 'ever. 8.. Muijers. A.0.. Van Gelder. B.F. Bahher. E.P. and Van Buren. K.1.H. (1973) Biochim. Biophys. Acta 111 1-7. 'harton. D.C. and Tragoloff. A. (1964) J. Biol. Chem. 212, 2036-2041. Wikstron. 11.. Xrab. L. Saruto. 11- WA Synthggil. Academic Press. New Yorh. 1981. 'ihstrom. M. and Casey. R.P. (1985) J. Inorg. Bioch. 11,327-334. Wilson. M.T. Greenwood. C.. Brunori. M. and Antonini. E. (1975) Biochem J. 111. 145-153. Wilson. M.T.. Jensen. P.. Aasa. B.. Malmrtrom. B.G.. Vanngard. T. (1982) Biochem. J. 221, 483-492. Yonetani. T. (1966) Biochem. Prep. L1. 14-20. Yu. C.. In. L. and King. T.E. (1975) J. Biol. Chem. 210. 1383-1392 . “WIVES 4 WTIWHWIWITUITH “Wilvllfilmfil'filw I“ 3 1 2 9 3 0 3 0 8 2 7 24