UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Role of heme propionates in myoglobin electron transfer Lim, Anthony Richard 1990

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1990_A1 L55.pdf [ 6.11MB ]
JSON: 831-1.0098750.json
JSON-LD: 831-1.0098750-ld.json
RDF/XML (Pretty): 831-1.0098750-rdf.xml
RDF/JSON: 831-1.0098750-rdf.json
Turtle: 831-1.0098750-turtle.txt
N-Triples: 831-1.0098750-rdf-ntriples.txt
Original Record: 831-1.0098750-source.json
Full Text

Full Text

ROLE OF HEME PROPIONATES IN MYOGLOBIN ELECTRON TRANSFER by ANTHONY RICHARD LIM B.Sc, The University of British Columbia, 1983  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES THE DEPARTMENT OF BIOCHEMISTRY  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA March 1990 © Anthony Richard Lim, 1990  In  presenting this thesis in partial fulfilment of the  requirements  for an  advanced  degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department  or  by  his  or  her  representatives.  It  is  understood  that  copying  or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department The University of British Columbia Vancouver, Canada  DE-6 (2/88)  ABSTRACT Myoglobin (Mb) is a well characterized hemeprotein found in skeletal muscle. dimethylester evaluate  heme-substituted  the involvement  reactions of Mb.  derivative of equine  of the heme  Mb  propionate  (DME-Mb)  groups  The  was prepared to  in the electron transfer  To achieve this goal, an efficient procedure to reconstitute and purify  DME-Mb in high yield was developed.  The near UV-visible absorption spectra of DME-  Mb in various states of ligation and oxidation did not change significantly relative to those of native Mb.  The *H NMR  oxidation state) and metDME-Mb  spectra obtained  for native metMb (heme Fe(III)  showed differences in the electromagnetic  environment  of their respective heme groups. The  reactivity of DME-Mb was different from that of native Mb. (Fe-H 0) has a lower pK  water ligand of metDME-Mb  2  a  For example the  than that of native metMb as  determined by spectroscopic pH titrations.  The autoxidation rate of oxyDME-Mb (Fe(II)-  0 ) is faster than that of native oxyMb.  MetDME-Mb apparently has a binding affinity  2  for ferricyanide not evident in native metMb.  Compared to native Mb, DME-Mb has  decreased susceptibility to the oxidant hydrogen peroxide. The  oxidation-reduction equilibrium of DME-Mb has been studied under a variety of  solution conditions.  At standard conditions (pH 7, 7=0.1 M  and 25°C) the midpoint  reduction potential (E ) of DME-Mb is 100.0(2) mV vs. SHE, which is 39 mV higher than m  the E  m  of native Mb.  identity or ptf  a  Analysis of the pH dependence of E  m  showed differences in the  between titratable groups found in native and DME-Mb.  strength dependence of E  m  showed a higher net positive charge estimate  than native Mb consistent with the nature of the chemical temperature dependence of E  m  The ionic  for DME-Mb  modification involved.  The  showed that DME-Mb has a greater difference in stability  between oxidation states than native Mb.  ii  The  kinetics of metDME-Mb  variety of conditions.  reduction  by Fe(EDTA) " were also studied under a 2  At standard conditions, metDME-Mb reacted  with the reductant  Fe(EDTA) " at a second order rate constant (Jfc ) two orders of magnitude greater than 2  12  that  of native  metMb.  After  between reactants, metDME-Mb  correcting still  for the differences  reacted  at a significantly  metMb, indicating differences in their reaction mechanisms. ionic strength dependences of Jfc  12  had  in reduction  potential  faster rate than native  The pH, temperature and  for DME-Mb and Fe(EDTA) " showed that DME-Mb 2  electrostatic and thermodynamic properties significantly different from that of native  Mb. The  functional differences between DME-Mb and native Mb can be attributed to the  structural and electrostatics properties of the heme propionate groups. of  these  discussed studies.  groups  within  with reference  the surrounding  protein  to the structure of Mb  and the external  The interactions environment are  available from x-ray crystallographic  As a result, it is concluded that the heme propionate groups are involved in the  structural stability, electron transfer specificity and reactivity of Mb.  iii  TABLE OF CONTENTS ABSTRACT  ii  TABLE OF CONTENTS  iv viii  LIST OF TABLES LIST OF FIGURES  ix  ABBREVIATIONS  x  ACKNOWLEDGEMENTS  xi  INTRODUCTION A. Historical Perspective  1  B. Biology of Myoglobin  1  1. Physiology  3  2. Genetics  4  C. Molecular Structure of Myoglobin  5  1. Amino Acid Composition and Sequence of Myoglobin  6  2. X-Ray Crystallography of Myoglobin  6  3. Characteristics of Myoglobin in Solution  10  D. Functional Studies of Myoglobin  11  1. Heme Replacement  11  2. Ligand Binding  13  3. Oxidation State Changes  15  E. Biological Electron Transfer  16  1. Oxidation-Reduction Equilibrium  16  2. Spectroelectrochemical Determination of Reduction Potentials  20  3. Electron Transfer Theory  22  4. Marcus Theory  26  5. Approaches to Studying Electron Transfer Reactions  27  iv  F. Outline and Purpose of Thesis  29  EXPERIMENTAL PROCEDURES A. General Procedures  31  B. Purification of Myoglobin  32  C. Preparation of Apo-Mb  32  D. Reconstitution of Apo-Mb  33  E. Spectroscopic Characterization of DME-Myoglobin  35  1. Visible Region  35  2. 'H NMR  36  3. EPR  36  F. Determination of the Heme pK  37  a  G. Measurement of Autoxidation Rates  37  H. Synthesis of Pentaammineimidazoleruthenium (III) trichloride  38  I. Spectroelectrochemical Experiments  39  J. Reduction Kinetics with Fe(EDTA) "  41  2  RESULTS A. Preparation and Isolation of DME-Myoglobin  44  B. Spectroscopic Characterization of DME-Myoglobin  45  1. Visible Region  45  2. *H NMR  47  3. EPR  50  C. Spectrophotometric pH Titration  50  D. Autoxidation Rates  55  E. Synthesis of [Ru(NH ) Im]Cl 3  5  59  3  v  F. Reduction Potential Measurements  59  1. pH Dependence of Reduction Potential  63  2. Temperature Dependence of Reduction Potential  67  3. Ionic Strength Dependence of Reduction Potential  70  G. Reduction Kinetics Measurement  73  1. pH Dependence of Reduction Rate  76  2. Temperature Dependence of Reduction Rate  80  3. Ionic Strength Dependence of Reduction Rate  80  H. Marcus Analysis of Kinetics  85  DISCUSSION A. Preparation of DME-Myoglobin  86  B. Spectroscopic Properties  87  1. Near UV-Visible  87  2. 'H NMR  89  3. EPR  90  C. Autoxidation Rate  90  D. Electrochemical Studies  93  1. pH Dependence  96  2. Thermodynamics  100  3. Electrostatics  102  E. Reduction Kinetics  104  1. pH Dependence  107  2. Thermodynamics  108  3. Electrostatics  109  F. Marcus Theory Analysis  113  vi  CONCLUSIONS A. Emerging Role of Heme Propionate Groups  115  B. Myoglobin  115  C. Cytochrome b  116  D. Cytochrome c  118  E. Cytochrome c Peroxidase  119  F. Further Studies  119  BIBLIOGRAPHY  121  5  APPENDICES A. Midpoint Reduction Potentials (vs. NHE)  132  B. Second Order Rate Constants  135  C. Second Order Rate Constants Adjusted for Driving Force  137  D. Derivation of Autoxidation Equation 17  138  E. Marcus Theory Calculation of kff  140  vii  LIST OF TABLES I  Proposed Assignments for 'H NMR  Signals  49  II Parameters from Ionic Strength Dependence of Reduction Potential  73  III Parameters from the Ionic Strength Dependence of Reduction Rate  83  viii  LIST O F FIGURES  1. Heme Structure  2  2. Amino Acid Sequence of Myoglobin  7  3. Tertiary Structure of Myoglobin  9  4. Potential Energy Diagram  24  5. Visible Spectra of DME-Mb  46  6. *H NMR Spectra of DME-Mb and Mb  48  7. EPR spectrum of DME-Myoglobin  51  8. pH Titration of DME-Mb Visible Spectrum  52  9. Linear Plot of pH Titration Data  54  10. Autoxidation Spectrum of DME-Mb  56  11. Linear Plot of Autoxidation Data  58  12. OTTLE cell Spectra  61  13. NernstPlot  62  14. pH Dependence of E with one titratable group  65  15. pH Dependence of E with three titratable groups  66  16. Temperature Dependence of E  69  m  m  m  17. Ionic Strength Dependence of E  72  m  18. AA vs. Time ± Ferricyanide  74  19. 2  75  n d  Order Plot of AA vs. Time  20. pH Dependence of Jt  77  12  21. pH Dependence of jfc (adj)  78  12  22. Temperature Dependence of k  81  12  23. Ionic Strength Dependence of it j2  84  24. Heme Region of Mb  99  25. Dipole Moment of Mb  112  ix  ABBREVIATIONS CyDTA  trans-1,2, -diaminocy clohexanetetraacetate  dipic  pyridine-1,2-dicarboxylate  DME  dimethylester  DMSO  dimethyl sulfoxide  E  midpoint reduction potential  EDTA  emylenediamine tetraacetate  EPR  electron paramagnetic resonance  Hb  hemoglobin  His  histidine  I  ionic strength  m  k  second order rate constant  ^obs  observed pseudo-first order rate constant  Lys  lysine  Mb  myoglobin  n  Mb0 , oxyMb  oxymyoglobin  metMb  metmyoglobin  NaCl  sodium chloride  NaPi  sodium phosphate  NMR  nuclear magnetic resonance  OLIS  On-Line Instrument Systems  OTTLE  optically transparent thin-layered electrode  SCE  standard calomel electrode  SHE  standard hydrogen electrode  Tris  tris(hydroxylmethyl)aminoethane  UV  ultraviolet  2  ACKNOWLEDGEMENTS  I would like to acknowledge  the support of my family and the invaluable help I've  received during my Graduate Studies from past and present members of the Mauk lab.  In  particular I would like to thank Dr. Bhavini Sishta for her help in developing the D M E Mb preparation.  In addition, I would like to thank Dr. Colin Tilcock for his help with  the NMR spectroscopy. and information on the  Thanks also go to members of the Brayer lab for their assistance three-dimensional  structure of  myoglobin,  especially  Dr. Steve  Evans for his work on the dipole moment calculations and Gordon Louie for his help with the Mb radius calculation and molecular graphics program.  xi  INTRODUCTION A. Historical Perspective  Myoglobin  (Mb) is a heme containing  protein found  muscle cells of a wide variety of animal species.  in the skeletal and cardiac  Until the beginning  of this century,  the reddish colour of muscle was believed by many to be due to the circulation of blood containing hemoglobin (Hb) through the muscle tissue.  Morner (1897; Kagen, 1973)  extracted a pigment from dog muscle and compared its visible absorption spectrum with that of hemoglobin.  Slight differences in the wavelengths of absorption were found, but  the results were thought to be artifacts.  Theorell eventually established that myoglobin  is a distinct protein (1932; 1934; Kagen, 1973). heart  by crystallization  and determined  He subsequently purified Mb from horse  its molecular  weight  by sedimentation  centri-  fugation to be ~ 17,500, which was much less than that of Hb.  He also measured its  oxygen (0 ) affinity and showed it to be greater than that of Hb.  Hill (1933), who used  2  the term "muscle hemoglobin," demonstrated that Mb exhibits a hyperbolic oxygen binding curve that is clearly different from the characteristic sigmoidal curve of Hb.  B. Biology of Myoglobin  Mb consists of two separable components, a globin or apoMb protein component, and a  heme prosthetic group.  The heme in Mb  is more accurately  identified  protoporphyrin LX (Fuhrop and Smith, 1975), which is depicted in Figure 1. iron (Fe) atom of heme is coordinated by the nitrogens of the pyrrole rings. the site of 0 0. 2  2  and other ligand binding.  as iron  The central The Fe is  Only the Fe(II) (ferro-) oxidation state binds  It is this oxygen-bound Fe(II) form (oxyMb or Mb0 ) that is found in vivo or in 2  1  Figure 1.  Structure of heme. Peripheral pyrrole carbons are labelled 1 to 8. Interpyrrole methines are labelled a to 5. In myoglobin, the heme (protoheme IX) has two propionic acid groups at positions 6 and 7 (R=H). In the dimethylester heme (DME) derivative, R=CH . 3  2  The predominant oxidation state in vitro is the Fe(IH) (met- or  newly prepared samples.  ferric-) form as ferroMb readily autoxidizes in solution.  1. Physiology  Mb content is highest in the cytosol of skeletal (including cardiac) muscle cells with aerobic, mitochondrial driven metabolism (Peter et al.,  1972).  There is a correlation  between cytochrome oxidase activity and Mb content in muscle cells (Lawrie, 1953). virtue of its 0  2  By  binding ability, Mb was originally thought to function as a store of 0  in the skeletal muscle cells.  2  While storage may be an important function of Mb in diving  mammals, which have high muscle concentrations of Mb, the storage capacity of most species is limited (Wyman, 1965). From theoretical calculations and experimental models for oxygen diffusion in cells, Wittenberg (1965; 1970) and Wyman (1965) proposed that Mb has a role in facilitating the diffusion of 0  2  from the cell membrane to the mitochondria.  mitochondrial activity  as measured  by steady state oxygen  Mb has no direct effect on consumption and adenosine  triphosphate (ATP) production in isolated mitochondria (Cole et al., 1982).  However, in  isolated cardiac muscle cells (myocytes), oxygen consumption decreases by about a third when Mb  mediated oxygen  transport is completely  monoxide (CO) (Wittenberg and Wittenberg, 1987).  inhibited by low levels of carbon Another function has been proposed  by Galaris et al. (1989) who have suggested that Mb  with ascorbate can act as an  electron sink to protect muscle cells from oxidants such as peroxides. Preparations of Mb0  2  tend to autoxidize slowly to the metMb Fe(III) state in vitro  (Gotoh and Shikama, 1974).  However, in vivo there is no significant amount of metMb  (Ray and Paff, 1930), which indicates the existence of a metMb reduction system in muscle cells.  Erythrocytes have a metHb reductase system that uses nicotinamide adenine  3  dinucleotide (NADH),  a reductase, and cytochrome b ,  metHb (reviewed by Hultquist et al, reductase  which is the direct reductant of  s  Hagler et al. (1979) have isolated a metMb  1984).  with similar characteristics to metHb reductase.  Cytochrome b  5  the reduction of metMb by this reductase in vitro (Livingston et al, Hagler et al. (1979) could not detect any cytochrome b  5  can mediate  1985).  However,  in muscle cells.  2. Genetics  The gene encoding the globin amino acid sequence of Mb cloned from seal (Wood et al, and mouse (Blanchetot et al., copy, which is large (ca.  has been isolated and  1982; Blanchetot et al., 1983), man (Weller et al, 1986).  In each species, the Mb  globin gene is a single  10 kilobases) relative to the required coding  proteins of similar size (153 amino acids). separated by two introns.  sequence for  Each Mb gene is composed of three exons  This pattern is similar to that of the Hb alpha and beta  subunit globin genes (Maniatis et al, human Mb gene in Escherichia  1984),  Varadarajan et al. (1985) have expressed the  1980).  coli using a cDNA clone.  Springer and Sligar (1987) have  expressed a synthetic sperm-whale Mb gene in E. coli with high efficiency. Gilbert (1978) proposed that exons correspond  to the functional domains of proteins.  By specific enzymatic proteolyis of Hb /?-globin subunits, Craik et al. (1980) isolated a peptide the amino acid sequence of which corresponds to that encoded by the central exon of the Hb p subunit gene and that binds heme tightly and specifically. the heme reconstituted Hb fragment was unable to bind oxygen (Craik et al,  However, 1981).  From enzymatic digestion of the globin portion of Mb, De Sanctis et al. (1986, 1988) have isolated a "mini-Mb" peptide corresponding  to the encoded sequence of the central  exon and some (30 amino acids) of the third exon, which encodes the carboxyl terminal sequence.  Mini-Mb not only binds heme, but the heme reconstituted fragment also binds  4  dioxygen and carbon monoxide (1988). Expression  Mb  globin  differentiation (Weller et al.,  1986).  (Longo et al.,  of  the  in  gene  vivo  begins  soon  cell  does not begin to appear until late in fetal  development, mainly in cardiac muscle (review by Kagen, 1973). In infancy, the synthesis of Mb  that is sustained through adulthood (Tipler et al.,  Unlike Hb, there is no  in skeletal muscle increases to a level 1978).  It appears that in adulthood a  is maintained with slow (80-90 days) turnover (Akeson et  al.,  1967), though the activity of some muscles can influence their  Mb  steady state level of Mb 1960; Daly et al.,  muscle  However, in all species studied, including human  1973), a significant level of Mb  fetal form of Mb.  after  content (Hagler et  al.,  regulation of Mb  globin gene expression in muscle cells appears to be through pre-  1980).  Underwood and Williams (1987) demonstrated that the  translational control by correlating Mb  mRNA with muscle activity.  C. Molecular Structure of Myoglobin  The relationship between the structure of Mb characterized as that for any other protein. unique, its heme group is not.  and its functional properties is as well  While the globin (protein) portion of Mb is  The Fe protoporhyrin LX group is a common prosthetic  group found in a diverse number of hemeproteins with different functions and from a variety of animal and bacterial species. as cytochrome  b ), 5  which  Examples include: the 6-type cytochromes (such  are involved  in electron diversity  reactions, catalases and  peroxidases, which  reduce peroxides.  possessing  functional groups indicates that the protein environment (sequence  identical  The  transfer  in function  and structure) dictates the chemical properties of the active site.  5  among hemeproteins  1. Amino Acid Composition and Sequence of Myoglobin  The amino acid compositions and sequences of myoglobins from a variety of species have been reviewed by Kagen (1973) and Romero-Herrera et al. (1978).  Mammalian  myoglobins are composed of 153 amino acids, of which a large number (19 or 20) are the basic amino acid lysine. whale and horse.  The two most studied mammalian myoglobins are those of sperm  The amino acid composition and sequence of sperm whale Mb were  determined by Edmundson (1961; 1965).  The amino acid composition and sequence of  horse heart Mb were originally determined by Dautrevaux et al. (1969) and corrected by Romero-Herrera and Lehmann (1974).  The sequences of these two myoglobins differ by  19/153 amino acid residues (see Figure 2). sequence homology  with  sperm  reported (Goss et al., 1982).  Though horse heart Mb has a high degree of  whale Mb,  differences in ligand  binding  have been  These differences presumably arise from differences in the  amino acid composition and conformation of the heme pocket of the proteins.  2. X-Ray Crystallography of Myoglobin  Sperm  whale metMb was the first protein for which a three-dimensional structure  was determined by X-ray diffraction methods.  This landmark crystallographic work was  performed by Kendrew et al. (1958; 1960) who solved the crystal structure of this protein to a resolution of 2 A. reduced, non-ligated  Nobbs et al. (1966) solved the three-dimensional structure of the  (deoxy-) form of Mb  deoxy- and metMb x-ray diffraction data. structures  were  later refined by Takano  by difference Fourier analysis between the Both the sperm whale metMb and deoxyMb (1977a&b) to 2 A.  The three-dimensional  structure of sperm whale oxyMb was determined and refined by Phillips (1978; 1980) to 1.6 A.  Neutron diffraction studies have been done to determine the position of hydrogen  6  horse whale  1 Gly-Lxu-Ser-Asp-Gly<jlu-Trp^3ta<^ Val Glu Leu  horse whale  21 Ile-Ala-Gly-His-Gly-Gm^31u-Val-Leu-Ile-Arg-Leu^ Val Asp-Ile  horse whale  41 Glu-Lys-Phe-Asp-Lys-Phe-Lys-His-Leu-Lys-Thr-Glu-Ala-Glu-Met-Lys-Ala-Ser-Glu-AspArg  His  Ala  Lys-Ser  61 horse Ixu-Lys-Lys-His<}ly-Thr-Val-Val-lxu-Thr-Ala-Leu-Gly-Gly-Ile-Leu-Lys-Lys-Lys-Glywhale Val-Thr Ala  81 horse His-His-Glu-Ala-Glu-Leu-Lys-Pro-Leu-Ala-Gln-Ser-His-Ala-Thr-Lys-His-Lys-Ile-Prowhale  101 horse Ile-Lys-Tyr-Leu-Glu-Phe-Ile-Ser-Asp-Ala-Ile-Ile-His-Val-Leu-His-Ser-Lys-His-Prowhale Glu Arg  horse whale  121 Gly-Asp-Phe-Gly-Ala-Asp-Ala<Jln<}ly-Ala-Met-Thr-Lys-Ala-I^u-Glu-Leu-Phe-Arg-AsnAsn Lys  horse whale  141 153 Asp-Ile-Ala-Ala-Lys-Tyr-Lys-Glu-Leu-Gly-Phe-Gln-Gly Tyr  Figure 2.  The amino acid sequence of equine (horse) Mb (Dautrevaux et al, 1969) is depicted. For comparision, amino acid differences from the corresponding sperm whale Mb sequence (Edmundson, 1965) are also shown.  7  atoms in sperm whale metMb (Schoenborn, 1971). prepared  by Sherwood  et  al.  From the crystals of horse heart Mb  (1987), Evans and Brayer have determined  the three-  dimensional structure of this protein by x-ray diffraction to a resolution of 2.8 A (1988). They have also recently refined it further to 1.9 A (Evans and Brayer, 1990). In both sperm whale and horse heart metMb, the secondary structure consists of a series of eight alpha-helices, labelled A residues.  to H, which contain most of the amino acid  The tertiary structure has the polypeptide  ellipsoid shape (Figure 3). side-chains  are oriented  chain folded upon itself into an  In general, the folding pattern is such that polar amino acids outward, towards  the solvent  and non-polar  side-chains are  oriented inward. The heme is situated within a cleft or pocket formed by helices B to C and E to G, which have many hydrophobic amino acids in their sequence. has  long  been  considered  compared to the solvent.  hydrophobic,  with  a  presumably  This heme binding region low dielectric  constant  However, MacGregor and Weber (1986) have probed the heme  binding region of apoMb using a polarity-sensitive fluorescent dye.  Their results indicate  that the heme binding region may be less hydrophobic than previously believed owing to contributions from the dipoles of the peptide (amide) bonds lining it. The  heme is held within the heme pocket by van der Waals contacts with the  surrounding  amino acid residues.  On one side of the heme, the Fe(III) is coordinated to  a water molecule, which also forms a hydrogen bond to a "distal" histidine residue (helix position 7E/sequence position 64). to a coordinated  "proximal"  On the other side of the heme plane, the Fe is bound  histidine (8F/93).  In the sperm whale metMb model of  Takano (1977a) the Fe(III) atom is displaced by about 0.4 A out of the mean heme plane towards the proximal histidine residue.  Evans and Brayer (1990) found no such heme Fe  atom displacement in horse heart metMb.  In Mb one of the heme propionates (P7, see  Figure 1) is oriented inward and is hence called the "inner heme propionate" with the  8  Figure  3.  Stereo  drawing  of  the  (aquo)metMb  structure  crystallography  (Evans  histidine His),  residues  and 97,  a-carbon  polypeptide  determined and  around  are depicted  Brayer,  the and  heme,  to 1988). at  labelled.  9  backbone 2.8 For  positions  A  and  reference 93  heme  resolution  of  the  from  the  side  (proximal  His),  equine x-ray  chains 64  of  (distal  other (P6) oriented outward (the "outer heme propionate"). The metMb. atom.  three-dimensional structure of sperm whale deoxyMb is very similar to that of A significant change is the loss of the water molecule coordinated  to the Fe  As a result, the distal histidine residue is shifted away from the heme and the  Fe(II) is extended further from the plane of the pyrrole rings.  In oxyMb, the 0  2  binds  the Fe(II) in a bent geometry with a F e — O — O bond angle of about 115° (Phillips, 1980). The heme prosthetic group in oxyMb is rotated further into the heme pocket with the Fe atom closer to the heme plane.  3. Characteristics of Myoglobin in Solution  Functional comparisons of myoglobin crystals and solutions, which are usually at nonphysiological conditions, have been made.  Observed differences may arise from either  differences in structure or environments of these two states. ligand binding  For example differences in  rates between crystals of Mb/Hb and Mb/Hb in solution (Antonini and  Brunori, 1971) have been observed.  Pain (1983) has stressed that proteins in solution are  less densely packed and have more dynamic structures than indicated by the average crystal structures. through  which  Areas of weak contact within a protein can result in "mobile defects"  small  molecules such  as ligands  can penetrate  variations in conformation have been observed by Frauenfelder  (Pain,  1987).  Such  et al. (1979), who have  collected and analyzed x-ray diffraction data for Mb at a variety of temperatures. One  specific difference between solution and crystal structures of sperm whale Mb is  evident from the NMR  studies of La Mar et al. (1978; 1983; 1984).  'H NMR  spectra of  newly reconstituted metMb exhibited two distinct sets of signals from heme methyl and vinyl groups arising from different orientations of the heme upon its entry into the apoMb pocket.  The two orientations are related by a 180° rotation about the a — T carbon  10  axis of the heme (see Figure 1).  Sperm-whale Mb equilibrates with time (~72 hours at  neutral pH) to a preferred or major form of heme orientation (~ 12:1 found in the crystal structure. mixture of isomers, was initially  Newly reconstituted metMb, which has about a 1:1 reported  to have (after reduction) a higher oxygen  affinity than native Mb (Livingston et al., 1984). studies  by two independent  ratio), which is that  groups  showed  However, later 0  no significant  2  difference  and CO binding between newly  reconstituted (and reduced) Mb and native Mb (Light et al., 1987; Aojula et al., 1987).  D. Functional Studies of Myoglobin  Several approaches have been used to study the manner by which the interactions between the apoprotein and prosthetic group determine the functional properties of Mb. The  kinetics and equilibria with gaseous and anionic ligands have been studied to probe  the steric and electronic environment of the heme pocket. methods have been used to study the reduction processes of Mb.  Electrochemical  and oxidation  and kinetic  (i.e. electron transfer)  Heme prosthetic groups with altered porphyrin structures and/or metal  substitutions have been used to determine the role of the heme within Mb.  Invariably  these studies monitor the spectroscopic properties of the substituted forms of Mb.  1. Heme Replacement  Following the Hb reconstitution work of Hill and Holden (1926), Drabkin (1945) was able to dissociate (met)Mb into its apoMb (globin) and heme components by extraction into acidified acetone.  He then reconstituted Mb by mixing the apoMb and heme and  found no change in the UV-visible spectroscopic characteristics of the product relative to the  native  protein.  Antonini  and Rossi-Fanelli  11  subsequently  showed  (1957) that  reconstituted Mb from Mb  has the same oxygen binding properties as native Mb.  produces a slightly lower density in the resulting apoMb as determined from  ultra-centrifugation  measurements (Breslow,  1963) and a loss of a-helical content as  determined by optical rotational dispersion observations (Breslow et al,  1965).  IX has been inserted into apo-Mb by Breslow et al.  protoporphyrin porphMb.  Heme removal  From circular dichroism  Metal-free  (1967) to form  studies, these authors found no significant difference  in conformation between porphMb and Mb, and they concluded that porphyrin interactions with the globin were a major determinant of Mb tertiary structure. Successful reconstitution of Mb modified  with protoheme has led to the use of "unnatural" or  derivatives of protoheme to study the influence of heme substituent groups on  the structure and function of the protein. work involve chemical modification  Heme modifications employed in this type of  or substitution of the protoheme methyl, vinyl or  propionate groups of the porphyrin ring (see Figure 1). In an extreme example Chang et al. (1984) have reconstituted apoMb with synthetic hemes lacking the methyl and/or vinyl groups.  The most common parameters measured to determine the effects of the heme  modification are spectrophotometric characteristics and ligand binding properties. Yonetani  and co-workers have  (Tamura et al,  1974a&b).  reconstituted  apoMb  esterification of the propionate groups had no effect.  then  determine  crystallography.  modified  hemes  However, substitution of the heme  with methyl groups) resulted in increased  variety of hemes modified  several  The optical absorption spectra of these heme substituted Mbs  showed no major differences from that of native Mb. vinyl groups (e.g.  with  0  2  affinity, while methyl  Miki et al. (1986a&b) have used a  at the vinyl groups to reconstitute sperm whale apoMb and  the three-dimensional In comparing  structures  of these metMb  derivatives by x-ray  the structures of the heme-substituted Myoglobins these  workers observed differences in the heme pocket conformation such as a-helix shifts, which were related to the observed changes in dioxygen binding  12  affinity of the heme  substituted oxyMb derivatives. The coordinated Fe atom of the heme can be substituted with other metals to form various  metal-substituted  forms  of Mb  (reviewed by Hoffman,  1979).  The unique  properties of individual metals can be exploited to study the influence of the apoprotein or porphyrin ring on the metal.  Cobalt (II) substituted Mb  characterized form of metal-substituted Mb. binds dioxygen reversibly  and because  is perhaps the most well  This derivative has been useful because it  the paramagnetic  properties of Co (II) allows  analysis by EPR spectroscopy (Hoffman and Petering, 1970). Electron paramagnetic resonance (EPR) and nuclear magnetic  resonance (NMR) are  sensitive techniques used to probe the electronic and magnetic environment of the heme within the protein. modifications  EPR spectroscopy has been used to characterize the affect of heme  on the paramagnetic  Fe(III) atom  (Tamura  et  al,  1973a).  In  NMR  spectroscopy of metMb the majority of *H resonance signals from the porphyrin ring are shifted downfield (and hence distinct) from the bulk of the *H protein signals. variety  of deuterium ( H) substituted hemes, La Mar et al. 2  identified several porphyrin resonance signals in the *H NMR  Using a  (1980) have specifically spectra for sperm whale  Mb.  2. Ligand Binding  Both ferri- and ferro-Mb are able to bind a variety of exogenous ligands.  The  ligand binding properties of Mb reflect the steric and electronic environment of the heme pocket.  The influence of the protein on ligand binding can be determined by comparison  to model compounds such as the 0 Collman et al. (1974).  2  binding "picket fence" iron porphyrin synthesized by  As described above, the influence of the porphyrin on ligand  binding can be determined at least in part from use of modified hemes.  13  In  iron,  metMb,  though  a  other  visible  spectrum  1971).  This  hydroxyl  anionic  of  has  The  through  of  small  to  allow  must  occur  to  shift  such  to  form  1979),  a  coordinated  a  dimensional  structure  maintained  A  heme  phenyl  bond  remains  iron  reformation  Fourier  studied  in  in  of  a  process  the  transform  absorption  fine  the  can  product  for  a  for  spectroscopy  of  a  binding  and  of  period  carbon  At  of  (Fiamingo  and  (Chance  et  a  time  atom.  by  matrix,  three-dimensional  that  protein  is  (Nobbs,  modelling  have  et  been  al.  bulk  too  conformation  structures  studies  proposed  (1984)  of  have  the  three-  this  group  before  of  1985)  1983).  to  M b .  temperatures  Breakage  variety  Alben,  al,  a  gaseous  protein  monoxide  low  recombination.  by  Fe  defined  determined  The  to  molecule.  photolysis.  followed  14  Ringe  M b  dioxygen  brief  geminate  experiments  of  entry.  chains  crystallography.  flash  been  ligand  in  molecular  side  the  pocket  crystal  M b  residue  the  heme  Antonini,  binding  The  changes  and  in  been  into  the  The  molecule  in  have  the  Mb-ligand  iron  the  a  of  acid  x-ray  to  result,  1985)  allow  pocket  (EXAFS)  entrance  As  amino  heme  infra-red  an  passage  has  binding  to  water.  and  water  involved  metal.  with  bond  ligand  the  transition  the  heme  as  spin  processes  binding  involves  to  low  the  to  broken  referred  to  of  ligand  (Brunori  3)  Mar,  by  dependence  transition  examination  of  displace  ligand  2  La  can  the  0 .  the  process  p H  high  in  azide  coordinating  of  would  to  and  sixth  penetration  be  Fe-CO  structure  as  the  dynamic  suggest  and  enough  ligand  M b C O  within  the  large  1)  matrix,  which  group  of  channel  widely  Fe-CO  C O  a  steps  variety  channel,  a  or  From  a  by  the  such  (Lecomte  cyanide  study  as:  ligands  as  acid-alkaline  metMb)  entry.  spectra  Karplus,  to  protein  (and  of  as  binds  characteristic  an  (1985)  the  entry  to  N M R  and  a  barriers  Wolynes  deoxyMb  such  accompanied  used  energy  2)  structure  that  been  and  (Case  has  is  )  Frauenfelder  1966),  metMb  ( O H  -  normally  ligands  involves  Deoxy M b  diffusion  molecule  dependence  ion  ligands.  water  it  the  free  re-binds  the  and  techniques  and  The  subsequent  such  extended  as  X-ray  3. Oxidation State Changes  Just as the kinetics and equilibria of ligand binding  to Mb  have been studied  intensively, the oxidation-reduction properties of the hemeprotein have also been studied extensively.  The redox equilibrium of Mb was first quantitatively characterized by Taylor  and Morgan (1942) who measured the reduction potential of horse heart Mb.  Brunori et  al. (1971) subsequently reported detailed studies of the reduction potential of sperm whale Mb.  Kinetic studies of Mb usually involve metMb because it is the predominant form in  vitro and it is easier to manipulate. by  use of a variety of reducing  ascorbate  (Tsukahara  The kinetics of metMb reduction has been examined  reagents such as chromium (II) (Huth et al.,  and Yamamoto,  1983),  Fe(EDTA) - and Fe(CyDTA) " (Cassat et al, 2  2  When Mb0 occurs readily.  2  dithionite  (Olivas  al,  1977) and  1975).  is separated from its in vivo reduction system, autoxidation to metMb The rate of autoxidation reflects the stability of the Fe-0  Mb. George and Stratmann (1952) demonstrated that autoxidation kinetics.  et  1976),  Wide variations in the rate of Mb0  2  2  followed  bond within first  order  autoxidation have been measured as a  function of pH (Sugawara and Shikama, 1979), temperature (Sugawara and Shikama, 1980) and  anion  concentration  involves generation  (Satoh  and Sugawara, 1981).  of a superoxide anion  The mechanism of autoxidation  ( 0 ~ ) from the bound oxygen (Gotoh and 2  Shikama, 1976; Tijima and Shikama, 1987). Another possible oxidation state for the heme Fe in Mb is the Fe(IV) or ferryl state. FerrylMb can be formed by use of hydrogen peroxide (H 0 ) as an oxidant of metMb 2  2  (George and Irvine, 1952), deoxyMb (Yusa and Shikama, 1987) or oxyMb (Whitburn, 1987). The ferryl state is proposed to be in the form of F e 0 mechanisms  of ferrylMb  constant stoichiometrics.  formation  are complex  2 +  (George and Irvine, 1952). The  because  the reactions  do not have  Mb cross-linking (Tew and Ortiz de Montellano, 1988) and pH  15  titration  experiments  oxidized  in  studying  heme  catalase.  It  1971;  the  (Uyeda  reaction  as  a  1981)  metMb  enzymes,  serves  L i m and M a u k ,  Peisach,  between  containing  also  and  and  which  starting  have  H  are  0  2  2  shown  .  Ferry 1Mb  involved  material  for  that  in  sulfMb  tyrosine  serves  oxygen  residues  as  a  model  metabolism  preparations  are  for  such  as  et  al,  However,  M b  (Berzofsky  1986).  E. Biological Electron Transfer  The  principal  participates  Fe(II)  in  In  electron  state.  increasing  studied  as  the  by  M b  Metalloprotein  studies  informative  of  transfer  experimental  general,  between  function  they  of  kinetics  ligand  or  during  at  the  and  to  past  by  x-ray  of  and  oxygen.  maintain  it  have  been  reactions  the  element  binding  transport  required  transfer  characterized  reduce  of  bind  as  electron  proteins  greatly  to  reactions  investigation  mechanisms  reaction  is  twenty  years  electron  the  et  have  ambiguity.  transfer  the  (Scott  crystallography  structural  in  both  functional  subject  al,  been  of  1985).  the  most  Similarities  exist  of  which  can  be  equilibrium.  1. Oxidation-Reduction Equilibria  Two  which  in  oxidation  simplest  states  terms  of  is  a  chemical  described  by  the  ox + nc~  The  refers  symbol  to  the  ox  represents  reduced  form  species  can  following  form  oxidized  and  n  form  represents  couple,  (1)  of  the  16  reduction-oxidation  equilibrium:  red  the  a  the  reacting  number  of  species,.  electrons  the  (e~)  symbol  red  transferred  in  the reaction.  Hemeproteins such as Mb can be considered metal centers with complex Then, for example, ox can represent  ligation, which consists of protein and porphyrin. metMbFe(III) and red can represent deoxyMbFe(II).  The free energy of the reaction, AG, can be defined by:  AG = AG° -  *  r  i  n  (  2  )  [ox]  The term AG  0  is the free energy change under standard conditions (one Molar activities),  'R is the gas constant (8.13143 J K" mol") and T is the temperature (in Kelvin). The 1  1  free energy terms can be replaced by electrochemical potential terms by the conversion AG = —nFE to produce the Nernst equation:  B - B o  +  ™ m M  nF  (3)  [ox]  where F is the Faraday constant for the charge/mole electrons (9.64870 x 10 C mol"), 4  1  E is the potential of the system (relative to the standard hydrogen electrode, SHE) and E°  is the standard electrochemical reduction potential for the system.  standard biological conditions, E° is replaced by E , m  the stated conditions. between  Under non-  the midpoint reduction potential at  The direction and magnitude of an electron transfer reaction  different reduction-oxidation  couples,  which is called a cross-reaction, is  determined by the difference between their reduction potentials.  Important examples of  such cross-reactions in biological systems are the electron transport chains of respiration and photosynthesis.  17  The the  structural factors that determine the reduction  focus  of considerable  experimental  potential of metalloproteins are The E  and theoretical consideration.  m  of  hemeproteins vary to some degree depending on the conditions of measurement e.g. pH, ionic strength and temperature. differ widely in reduction  As many hemeproteins have identical heme groups yet  potentials, the protein structure and its interactions with the  heme must be major factors in deteniiining the reduction potential of a given protein. The  influences  on  the reduction  potential  of electron  transfer  proteins  and their  corresponding contributions to the free energy change have been summarized by Moore et al. (1986).  The changes that accompany electron transfer involve:  1) bonding interactions of the reduction-oxidation center (AG  cen  )  2) electrostatic interactions of the center with the protein and solvent (AG ) el  3) oxidation state dependent conformation changes (AG The  latter type of change, represented by AG  conf  conf  )  , is considered negligible for simple  proteins (VM a vis the similarity of metMb and deoxyMb crystal structures) and may be intrinsically linked to the interaction changes represented by A G  c e n  and AG (. e  The electrostatics-linked changes can be further delineated into four components: a) interactions with ions in solution (AG ) ion  b) interactions with the solvent, i.e. water (AG £>) H  c) interactions with charges exposed to the solvent/surface (AG  surf  )  d) interactions with charges within the protein (AG ) int  In terms of free energy,  AG = AG  where  AG  el  = AG  c e n  ion  + AG  el  + AG  + AG ,p + A G H  18  (4)  c o n f  surf  + AG  int  (5).  distinction  A  of  by  water  (within  dielectric  (1978)  has  solvent.  reduction  just  heme  is  studied  (e.g.  E  The  of  iron  manifested  a  oxidation  and/or  of  state  of  extent  groups  of  of  as  (each  of  die  (at  from  and  solvent,  and  polarizability  protein.  pH  of  7)  absence  (measured  Stellwagen  of  heme  each  solution  strength  of  pH>0,  a  function  with  histidine),  water.  the  by  changes  in  the  E  of  m  which  proton(s)  +  a  variety  exposure  odier  conditions  can  the  or  protein  of  to  the  structurally  under  temperature)  an  which  influence  the  can  be  in  a  the  oxidation-  Dutton,  et  al.  and  (1986),  then  be  some  ligands  the  of  1978).  include  some  of  effect  the  such  of  p H  hemeprotein.  characterized  by  the  pK  &  influence.  effect  with  and  potentials  hemeproteins  interactions  hemeproteins  concomitant  1960  nitrogens  Moore  electrostatic  This  in  a  pyrrole  reduction  (Clark,  pAT )  observable  of  midpoint  p H  unique  heme  the  groups  )  of  approach of  hemeprotein.  ( H  die  hemeprotein  have  titratable  the  a  a  the  By  groups,  values  uptake  ionic  in  degree  varied  The  perturbation  (e.g.  bound  titratable  pAT  the  that  presence  hydrophobic  potentials  with  the  difference  more  reduction  exposure.  conditions  some  acids,  dependence  The  the  to  ionizable  amino  is  p H  values  or  the  proteins  pH,  large  in  itself.  vary  heme  m  through  and  the  interactions  the  inversely  compared  non-standard  Titratable  that  vary  than  centre  hemeproteins  of  Mb,  from  water  show  he  potential  Under  on  to  electrostatic  arising  between  including  metalloprotein  reduction  between  protein)  attempted  ways  the  the  However,  other  as  made  constant)  hemeproteins,  types  is  of  may  p H  electron  in  m m  be  determined  involves  the  specific  transfer;  the  former  by  release  can  be  r e p r e s e n t e d as:  oxK +  The  effect  of  degree  p H  is  of  e  protonation  part  of  the  " t i ;  -  in  more  a  red +  protein  general  in  H+  part  effect  19  of  (6)  determines  protein  its  charge  net  on  charge.  E. m  Hence,  Moore  et  the  al.  (1986) consider contribution  electrostatic interactions, represented  to the  reduction  potential of electron  by  AG , el  to  have a  transfer proteins.  substantial  Ionizable  amino  acid residues within hemeproteins give rise to various ionic interactions with the heme, solvent and other residues within the protein (Matthew, 1985). The  AH ,  standard free energy change is composed of a standard enthalpy change,  and entropy change, AS , 0  0  which are defined by:  A G ° = AH  0  — TAS°  where T is the temperature (in Kelvin).'  The  (7)  enthalpic term represents the net energy  change from the chemical changes of the system e.g.  electrostatic bonds.  The  entropic  term represents the change in energy distribution (degree of order) of the system  e.g.  degree of protein  the  reduction This  solvation by  H 0.  Study  2  of the  temperature dependence of  potential allows the components of the free energy change to be  breakdown  allows  better  understanding  As reviewed by Taniguichi et al.  changes.  of  the  mechanisms  (1980), both AH°  reduction of many metalloproteins, including Mb.  The  of  and AS"  quantified.  oxidation  state  are negative for the  increased stability of the reduced  state has been attributed to the lowering of the metal charge within the protein interior.  2. Spectroelectrochemical Determination of Reduction Potentials  To  determine the midpoint potential of a hemeprotein using the Nernst equation, the  ratio of reduced:oxidized forms must be measured. reduction  An  indirect method of determining  potentials is to measure the changes in current  potential. proportional  The  current  represents  the  magnitude  to the change in oxidation states.  20  An  of  that accompanies changes in electron  transfer,  which  is  example of such a "voltammetric"  technique is cyclic voltammetry (Evans et al,  1983).  There are two direct approaches to  measuring the reduction potential.  One  is to control the ratio of oxidation states and  measure the change in potential.  The  other method is to control the potential and  measure the change in ratio of oxidation states. changes can be quantified by monitoring  With both methods, the oxidation state  the accompanying spectrophotometric  changes.  To obtain various ratios of oxidation states, potentiometric titration (Wilson, 1978) or the methods of mixtures (Antonini et al., aliquots  of  oxidation  an  oxidant  state and  hemeprotein. measured.  reductant,  concentration  After  equilibration,  are the  e.g.  ferricyanide, in a  used  to  titrate  electrochemical  In the former method, controlled and  (oxidize/reduce) potential  of  oxidized and  a  solution of  die  solution is  Mixing various volumes of the two solutions allow various ratios of  reduced forms to be obtained.  potential is measured.  After equilibration with an electrode, the  Equilibration times can be reduced if a mediator is added to the  Mediators are small inorganic metal complexes or organic dyes, which react  rapidly with  the titrant, electrode and  hemeprotein, thereby facilitating coupling of the  protein equilibrium to the electrode surface (reviewed by Armstrong et al, The  defined  In the method of mixtures, separate solutions of hemeprotein are either fully  oxidized or reduced.  solution.  or  1964) is commonly used.  1988).  application of a controlled potential across a hemeprotein solution results in a  defined change in the position of the oxidation-reduction equilibrium, which is dependent on the applied potential.  Such changes in oxidation state can be readily monitored  changes in the electronic spectra.  by  However, hemeproteins in general do not react with  electrodes at appreciable rates owing to the insulation of the protein surrounding  the  metal center.  co-  As  a result, electrochemical equilibration times are long.  Hill and  workers have circumvented this problem by chemical modification of electrode surfaces to allow more rapid equilibration (Armstrong et al,  21  1988).  Heineman  et  potentiometric  Electrode  al.  titrations  (OTTLE)  working  interfere  is  reduce  to  molecular  electrode  weight  and  have  spans  with  diffusion  are  the  light  modified  solution  are  and  used  to  an  alternative  of  Optically  cuvettes  (and  transmission.  distances  mediators  reported  development  These  which  significantly  1983)  through  cells.  electrode,  short  (1982,  with  The  hence  equilibration  facilitate  electron  performing  Thin-Layer  semi-transparent  the  pathlength  of  Transparent  a  therefore  method  pathlength)  of  the  but  (~0.2  Redox  active  between  mesh  does  cuvette  time.  transfer  wire  the  not  mm)  small  working  metalloproteins.  3. Electron Transfer Theory  A  bimolecular  electron  acceptor,  electron  A,  can  transfer  be  cross-reaction  represented  as  between  follows  (Scott  an  et  electron  al,  donor,  D,  and  an  1985):  *«  A + D The  initial  reactants.  step  k.  of  the  products.  final  Electron  and  acceptor  ligand  occur  transfer  orbitals,  by  an  the  is  thought  to  by  result  allows  centers  encounter  as  the  A~ + D  an  from  in  an  If  outer-sphere  there  of  is  degree  the  no  of  electron.  complex,  mechanism.  22  precursor  complex  some  activated  or  (8)  +  intramolecular  successor  delocalization  mechanism.  an  an  occurs  transfer,  reactive  occurs  then  of  electron  which  inner-sphere  transfer  formation  transfer  After  et  between  electron  involves  Electron  constant  [AD] —> [A~D+]  such  of  the  with  a  rate  dissociates  to  give  process  then  overlap  If  electron  link  complex  between  there  is  transfer  between  a  is  reactive  donor  bridging  said  to  centers,  The  potential energy states of the nuclei in the complex can be represented as a  function of nuclear configuration in two-dimensions as illustrated in Figure 4. in  nuclear  positions arise from vibrational movements of the nuclei.  Changes  The  energy  difference between the ground states of the precursor and product complex is A£°, is simply the difference in reduction potential between reactants.  The  which  energy difference  between the ground state of the precursor complex and the intersection of coordinates is the activation energy, AE* The  (or AG * ).  energy difference designated AE  their intersection represents  between the unreacted and reacted complexes at  the degree of electron orbital overlap.  reaction, there is sufficient electronic interaction that AE probability  of the  is large.  reaction (electron transfer) proceeding at the  which implies a rapid reaction rate.  In an As  adiabatic  a result, the  intersection is  one,  In a non-adiabatic  reaction, the precursor complex  can exist at high energy levels without electron transfer.  As a result, the probability of  reaction is less than one. The  aforementioned description of electron transfer arose largely from the work of  Taube on the reactions of metal complexes (reviews 1977, considered reactions  to a large extent as elaborate can  metalloproteins  be  analyzed  in  a  Metalloproteins can be  metal complexes, whose  similar manner.  are often non-adiabatic.  1984).  The  Electron  reduction-oxidation  transfer  reactions  reasons for their non-adiabatic  of  behavior  involves the protein environment or medium through which electron transfer occurs.  The  metal center in a metalloprotein is held in a relatively fixed position by the protein (and porphyrin  ring if present).  Proteins effectively act as insulators by  reactive metal centers at finite distances and such cases involves  no  direct contact  fixed orientations.  separating charged  Electron transfer in  or bonding between reactive centers  therefore considered to occur by an outer-sphere mechanism.  23  and  is  A-D  nuclear  Figure 4.  configuration  Potential energy diagram for non-adiabatic electron transfer. D is the electron donor and A is the electron acceptor.  24  According  to transition-state theory (review by Moore and Pearson, 1981), formation  of an activated complex for a bimolecular reaction, such as that described in Equation 8, involves a free energy of activation, A G reaction rate, k ,  This energy term determines the cross-  which is defined by the equation:  n  *  1 2  = *_7-e-<  A G  '^  ( 9 )  Nh The term R is the gas constant, T is temperature, N is Avogadro's number (6.022 x 10 mol" ) and h is Planck's constant (6.625 x 10 .J s). 1  The term RT/Nh in Equation 9  _35  represents the vibrational frequency of the reactants.  23  The free energy of activation  consists of enthalpic and entropic energies of activation symbolized by the terms AH * and A S * respectively.  The activation energies involved in a reaction can be determined from  the temperature dependence of the reaction rate constant defined by Equation 9. Though  the kinetics of electron  transfer reactions  complex mechanisms involved are not obviously defined.  may follow  Equation  9,  the  As reviewed by Newton (1968),  an electron transfer reaction involves three general processes: 1) work to combine reactants and separate products 2) reorganization of the reaction complex to allow electron transfer 3) electron transfer. The  work and reorganization parameters involve energy changes symbolized by the terms  w and A (or AG *) respectively.  The work terms describe the energy involved in moving  R  charged particles together the  geometrical  occur.  l2  or apart  (w i). 2  The reorganization energy  changes in the donor' and acceptor  required  The free energy of the cross-reaction can be defined as: AG  where  (w )  1 2  = AG *+ Aw  (10)  R  AH> = w  12  - w  (11)  21  25  represents  for electron transfer to  4. Marcus Theory  An  alternative  approach  to characterize  reaction  kinetics  has been  developed  specifically for outer sphere electron transfer by Marcus (reviewed by Marcus and Sutin, 1985).  In "relative" Marcus Theory the rate and activation energy of an electron  transfer cross-reaction are viewed as a function of the corresponding parameters for the self-exchange  reactions  of the reactants.  An example  of a self-exchange reaction  involves iron isotopes (Taube, 1977): *Fe(H 0)|  Fe(H 0)|  +  +  2  *Fe(H 0)l  —>  +  2  2  +  +  Fe(H 0)l  +  2  The main feature of a self-exchange reaction is that it has no thermodynamic driving force, i.e. AE°=0.  The cross-reaction and self-exchange rates of electron transfer are  related by the expression:  (12)  *12=V (*11*22^12/) /  where  m  /  (In A T )  (13)  2  12  =  4\n(k k /Z ) 2  n  The  and Jfc^ are the self-exchange rates of the reactants; k  parameters  cross  reaction  involved  22  rate.  They  are determined by the work and reorganization The parameter K  in electron transfer.  n  the electron reaction is bimolecular, it is intramolecular  i.e.  the  energies  In Equation 13, if  Z is the collision frequency between the reactants;  after precursor  frequency defined in Equation 9.  is  is the equilibrium constant of the  cross-reaction, which is determined by A G ° (or AE°) of the reaction.  if  l2  complex  formation, Z  is the vibrational  The term / usually has a value of one if die reactions  involved are adiabatic or are uniformly  non-adiabatic.  26  A  useful feature of relative Marcus theory is that if the characteristics of the self-  exchange reactions are defined, the rate of the cross-reaction can then be predicted with reasonable accuracy.  Alternatively, a self-exchange reaction rate for a reactant can  calculated from the cross-reaction rate and (Wherland &  Gray, 1976).  of its inherent theory  has  reactivity  been used  complexes and  The and  1985, and Scott et al,  the self-exchange rate of the other reactant  self-exchange reaction rate of a reactant is characteristic  hence mechanism of electron transfer.  in this manner in many  metalloproteins  be  Relative Marcus  studies of reactions  (reviews by Wherland and  Gray, 1976,  involving metal  Marcus and  Sutin,  1985).  5. Approaches to Studying Biological Electron Transfer  The  study  of biological hemeprotein electron transfer can  general methods, the study of intermolecular et  al,  1985).  To  must be employed.  study intermolecular A  approached by  intramolecular  two  reactions (Scott  reactions, suitable electron donors or acceptors  variety of electron donors can be generated using flash photolysis molecules e.g.  to electronically excite otherwise unreactive 1982), a variety of flavins (Tollin et al, substituted into hemoglobin, which was 1984).  reactions and  be  Ru(bpy)3  (English et  +  1986). Photoactive zinc protoporphyrin  then used to reduce cytochrome b  5  al,  has been  (Simolo et  Another type of electron donor is the hydrated electron generated by  al, pulse  radiolysis of an aqueous solution (McLendon and Miller, 1985). Another  source of electron donors/acceptors are  stable, substitutionally inert metal  complexes that are well characterized in terms of their structural and  reduction-oxidation  properties.  reaction  corrected  The for the  metalloprotein  observed  kinetic behavior  contributions  to be analyzed by  of  the  protein-complex  of the complex, allowing  the  for example Marcus theory.  27  intrinsic  can  be  features of the  Examples of such metal  complexes  include  the  iron(II/III) emylenediarninetetraacetate  tris(l,10-phenanthroline)cobalt(II/III),  Co(phen)3  Fe(CN)|" ~, which are well characterized and /4  kinetics.  complex,  and  +/3+  Fe(EDTA)  hexacyanoferrate  1  ,  (II/III),  commonly used reactants in metalloprotein  With the use of such reagents, the effects of electrostatic charges and complex  formation on electron transfer rates can be studied. In addition, specific amino acid residues effects of  such substitution on  electrochemical  derivatives  residues.  with  covalently  effects of  modified.  kinetic properties of  dinitrophenyl  The  the resulting  (1983) prepared cytochrome  derivatives  on  specific  lysine  lysine derivative were determined "by  oxidation kinetics with various cobalt complexes.  Another protein modification technique  site-directed mutagenesis,  proteins e.g. (Cutler et al., To  Mb  the  attached  be  location of the  is  The  and  proteins can  For example Butler et al.  derivatives can then be evaluated. c  within  which  allows  substitution of  Val-68 mutants (Varadarajan et al.,  amino  acid  residues  within  1989), cytochrome c Arg-38 mutants  1989).  carry out bimolecular electron transfer reactions, which often occur at high rates,  rapid mixing systems are employed (review by  Chance, 1974).  Such systems mix  the  reactants in a rapid and complete manner to produce a homogeneous reaction whose rate can  be  measured.  To  measure the rates of electron transfer involving hemeproteins,  advantage is taken of their oxidation state dependent spectrophotometric changes. An  alternative approach  reactions  occurring  intervening  within  medium  (e.g.  to a  biological  electron  hemeprotein.  amino  acid  transfer is to  The  residues)  effects on  of  electron  study distance,  transfer  intramolecular orientation, rates  hemeproteins can be studied because the sites of electron transfer are fixed and localized by  x-ray crystallography  proteins with two  models.  There is a paucity  redox active sites that are well characterized.  within can  be  of naturally occurring As  a result, chemical  modification techniques are used to attach a second (detectable) redox-active center at a  28  definable position in a protein.  Examples of such modified hemeproteins are the zinc and  iron porphyrin Hb hybrids (McGourty et al, (Nocera et al,  1983), ruthenium modified cytochrome c  1984) and ruthenium modified Mb (Mayo et al,  1986).  F. Outline and Purpose of Thesis  Previous studies of bovine liver cytochrome b  s  by Reid et al  (1984) demonstrated  that esterification of the heme propionate groups produces a substantial increase in the reduction potential with a significant reduction in the electrostatic potential surface of the hemeprotein.  Based on the structural comparison of hemeprotein three-dimensional  structures by Rossman and Argos (1975), the similarity in structure between cytochrome b  s  and Mb  suggests that the heme propionate groups of Mb  significance in Mb.  may be of comparable  In the three-dimensional structure of horse heart and sperm whale  Mb, the heme propionate groups are located within the. hemeprotein somewhat similar to the positions they occupy in cytochrome b . 5  Heme propionate-7 is oriented inward to  the heme pocket on the proximal side of the heme and forms a hydrogen bond with His97 and serine-92.  Heme propionate-6 is oriented towards the heme pocket opening and  forms a hydrogen bond with arginine-45 in sperm whale and with lysine-45 in horse heart Mb. Comparison  of the three dimensional structures of sperm  whale  metMb  (Takano,  1977a) and deoxyMb (Takano, 1977b) crystals does not reveal any significant change in the environment of the heme. Phillips  (1980) indicates that  becomes more compact  However, similar comparison of deoxy- and oxyMb by with dioxygen bonding, the environment  with decreases in hydrogen bonding lengths.  of the heme This structural  change with ligand binding should stabilize the oxygenated form of the protein relative to  the oxidized  form.  As oxygenation is formally  29  equivalent  to oxidation,  this  observation  suggests the heme propionate esterification  oxidation-reduction  properties of Mb  should  have an effect on the  similar to that previously reported  for cytochrome  by Tamura et al.  (1973a&b) reported  that heme propionate esterification Mb  significant effect on the ligand binding properties of sperm whale Mb.  had no  These workers  however, did not investigate the effects of this modification on the reduction potential Furthermore, the protein that Tamura et al.  or electron transfer kinetics of this protein.  studied appears to have been contaminated by significant amounts of apoMb. et al. (1986) have reported initial results concerning Mb  (prepared  Tsukahara  the reduction of sperm whale DME-  by the method of Tamura et al.) by ascorbate that reveal a significant  increase in the reaction rate relative to that of native Mb. The  present  hemeprotein Mb  study  provides  a  comprehensive  esterification on the electrochemical  in an attempt to resolve these apparently  investigation  of the effect of  and kinetic properties of horse heart  contradictory observations.  In the course  of this work, an efficient method of preparing DME-Mb has been developed.  In addition,  the effect of heme propionate esterification on the rate of autoxidation, the electrostatic properties of Mb and on the pK The  results of this study  patterns  of the heme  &  of the axial water ligand of metMb has been assessed.  establish that  propionates  within  the disruption of normal hydrogen bonding myoglobin  consequences for the functional properties of this protein.  30  by  esterification  has important  EXPERIMENTAL PROCEDURES  A. General Procedures  Reagent grade chemicals were used in all procedures unless otherwise stated. Glassdistilled  water was further purified  by passage through  a Barnstead  purification system until the resistivity was greater than 15.5 Mohm. pH  Nanopure water Measurements of  were obtained with a Radiometer Model PHM84 pH meter and GK2321c combination  electrode.  Conductivity was measured with a Markson Model 10 portable conductivity  meter (Markson Scientific Inc.). Protein  solutions  were  centrifuged to remove  any precipitated  material.  The  centrifuges used were a Sorvall RC-5B superspeed centrifuge (20-30 minutes x 10,000 rpm, SS-34 rotor), which was refrigerated at 4°C, and a Silencer H-25F1 table-top centrifuge (10 minutes x 10,000 rpm), which was used for volumes less than one mL. solutions were dialyzed with Spectrapor type 1 dialysis membranes.  Protein  Ultrafiltration and  concentration of protein solutions were carried out a 4°C with a Millipore Minitan ultrafiltration system (for volumes >500 mL), an Amicon ultrafiltration cell with a YM-5 membrane (volumes >10 mL) and Centricon 10 microconcentrators (volumes <10 mL). Absorbance readings and electronic spectra were recorded on a Shimadzu Model UV250  spectro-photometer  Lauda  model  spectrophotometer  RM3  equipped water  equipped  with a jacketed cuvette holder connected  bath.  Spectra  were  also  recorded  with a jacketed cuvette holder connected  Lauda model RM3 water bath. personal computer. Software  on  to a a  MGW-  Cary-219  to another  MGW-  The Cary-219 was interfaced to a Zenith Model Z-248  for instrument  control, data collection  developed by On-Line-Instrument-Systems (OLIS; Jefferson, Georgia).  31  and analysis was  B. Purification of Myoglobin  Horse heart  myoglobin (Mb) (Sigma Type III) was purified by the procedure of  Tomoda et al. (1981) as follows.  Lyophilized Mb was dissolved in 10 mM  phosphate (KPi) buffer pH 6.8 (4°C).  A 15 x 5 cm CM-50 Sephadex (Pharmacia) column  was equilibrated with the same buffer. ammonium  bisdipicolinatocobaltate,  metMb (Mauk et ah, 1979). material.  To the Mb solution were added a few grains of  [Co(dipic)]NH , 4  to ensure  complete  The solution was then centrifuged  oxidation to  to remove insoluble  The metMb solution was applied to the column and washed with the 10 mM  KPi buffer.  Reddish coloured bands of material eluted with this buffer while the main  metMb band remained bound. mM  potassium  The metMb was eluted with a pH gradient of 100 mL 10  KPi buffer pH 6.8 and 100 mL  Fractions with A 9 / A 2 4 0  g 0  10 mM  dipotassium hydrogen phosphate (dibasic).  absorbance ratios (at pH 6) of greater than 5.6 were pooled,  concentrated, exchanged into water, frozen in liquid nitrogen, and stored at —70°C.  C. Preparation of ApoMb  ApoMb was prepared by the method of Teale (1959) at 4°C as follows.  In a 30 mL  conical glass centrifuge tube fitted with a ground glass stopper, an aqueous solution of metMb was acidified to pH 1.5 with 1 M hydrochloric acid (HCI) and then an equivalent volume of ice-cold 2-butanone (Burdick gentle mixing.  heme, was removed.  containing  Inc.) was added with  After cooling on ice, the less dense 2-butanone and the aqueous solution  separated to form two distinct layers.  of 2-butanone.  and Jackson Laboratories  The 2-butanone layer, which contained  The extraction was repeated twice with successively smaller volumes  The resulting straw coloured  0.6 mM  extracted  apoMb solution was dialyzed against water  sodium bicarbonate (NaHC0 ) and 1 mM 3  32  ethylenediaminetetraacetic  disodium salt and then against water with 0.6 mM  bicarbonate.  solution was centrifuged to remove any precipitate.  After dialysis, the apoMb  The concentration of apoMb was  determined by measurement of absorbance at 278 nm based on an extinction coefficient of 16 (mM cm)" (Tamura et al., 1973a). 1  ApoMb solutions prepared in this manner had  concentrations of 300 to 500 pM and were used immediately afterwards in reconstitution experiments.  D. Reconstitution of ApoMb  Iron (III) protoporphyrin IX dimethyl ester (DME-heme; Porphyrin Products, Logan, Utah) was dissolved in warm methanol (55-60°C) to a concentration of about 1 mg/150 /zl. The solution was centrifuged to remove any undissolved DME-heme, which could be later taken up with a small volume (/*Ls) of methanol.  A  1.25 molar excess of DME-  heme/methanol solution was slowly added to the apoMb solution (at 4°C) with gentle stirring.  After two hours, another equivalent of DME-heme/methanol was added.  total volume of methanol added was <10% of the total protein solution volume. two  After  to three hours, any newly reconstituted metMb was reduced by adding a few drops  of concentrated (49.3 mM) Fe(EDTA) ' solution. 2  Immediately afterwards, the now reddish  solution was dialyzed against two changes of 0.6 mM The  The  dialysate was then centrifuged.  discarded.  NaHC0  3  to remove the methanol.  Excess DME-heme formed a soft pellet, which was  One molar equivalent of DME-heme was then added to the solution with  another one half molar equivalent added two hours later.  In total, the apoMb was  exposed to four molar equivalents of DME-heme. To maintain reconstituted DME-Mb in the reduced state, a few drops of concentrated (49.3 mM) Fe(EDTA) " were added to the solution before it was dialyzed against 15 mM 2  tri(hydroxylmemyl)aminoethane (Tris) buffer pH 8.4 (at 4°C).  33  The dialysate was then  centrifuged to remove excess DME-heme, which formed a firm pellet. solution  was gently  decanted, and the reconstituted  addition of Fe(EDTA) ".  DME-Mb  was reduced with the  Immediately afterwards, the solution was passed through a 10 x  2  1.5 cm DE-52 cellulose (Whatman) column equilibrated with 15 mM A  few  grains  of  recrystallized  potassium  K [Fe(CN) ], was added to the eluant. 3  slow  change in solution  concentrated and exchanged repeated ultrafiltration.  colour  hexacyanoferrrate  from  into 20 mM  red to brown.  407  /A  exchanged into 50 mM  (III) (ferricyanide),  The solution  was then  sodium phosphate (NaPi) pH 7.2 (at 4°C) by  The solution was centrifuged and applied to a 95 x 2.5 cm G-75  Sephadex (Pharmacia) column equilibrated with 20 mM DME-Mb with A  Tris buffer (pH 8.4).  The generation of metDME-Mb was followed by  6  the  The supernatant  278  NaPi pH 7.2 buffer.  Fractions of  ratios greater than 5 (at pH 6) were pooled, concentrated and  sodium chloride (NaCl).  To remove any associated ferricyanide from metDME-Mb, a method similar to that developed by Linder et al. (1978) for hemoglobin was used.  At room temperature, an 8 x  1 cm column of 20-50 mesh Dowex 1-X8 anion exchange resin was cleaned just before use.  The resin was initially  washed with one column  volume each of the following  solvents in the order indicated: water, ethanol and water.  This treatment was followed  by three cycles of washing in the sequence of 1 M sodium hydroxide (NaOH), water, 1 M HCI and water. the  conductivity  The resin was then washed with 1 M NaCl followed by 50 mM of the eluant  matched  that of the 50 mM  concentrated metDME-Mb solution was passed through this column.  NaCl  NaCl until  solution.  The  The resulting eluant  was collected, concentrated, frozen, and stored in liquid N . 2  The  DME-heme  was checked for homogeneity  by using thin layer  chromatography.  Three types of heme were examined, DME-heme, hemin (Fe protoporphyrin LX) and DMEheme extracted from DME-Mb.  The first two heme samples were used as supplied from  commercial sources and dissolved in acidic 2-butanone (pH 1.5).  34  The third sample was  obtained  by extracting an acidified  and desalted DME-Mb  solution  with 2-butanone.  Samples were taken up by capillary tubes and spotted onto an ITLC (Instant Thin Layer Chromatography) silicic spotted  to allow  acid  sheet  visualization.  (Gelman Instrument Co.).  The sheet  was placed  Sufficient  material was  in a Gelman chromatography  chamber pre-saturated with the running solvent, which was 2,6 Lutidine:H 0 2  ratio (Asakura  and Lamson, 1973).  in a 20:1  After the solvent front had moved a sufficient  distance (~8 cm) for significant migration of the samples, the sheet was removed and dried in air.  The Rf values for the samples were determined as the ratio of dieir  migration distances to that of the solvent.  E. Spectroscopic Characterization of DME-Myoglobin 1. Visible Region  The visible absorption spectra of metDME-Mb and oxyDME-Mb were recorded on the Cary-219  spectrophotometer.  Initially,  a sample  of metDME-Mb  was reduced  with  dithionite and converted to oxyDME-Mb by the procedure described in section G (vide infra). The visible absorption spectrum of the oxyDME-Mb (80 pM) was recorded using OLIS software.  Then the sample was oxidized to metMb with the addition of solid  potassium ferricyanide and its visible absorption spectrum recorded. The extinction coefficient of the metDME-Mb (acid form) Soret band was determined using the pyridine hemochrome method described by Fuhrop and Smith (1975) as follows. The Soret absorbance of-a concentrated metDME-Mb solution was measured. portion  (1.35 mL) of this  solution  was transferred to a cuvette  A measured  and the pyridine  hemochrome derivative of DME-heme was formed by mixing in sequence: pyridine (306 /xL), NaOH (144 pL) and solid dithionite.  Immediately after mixing, the absorbance of the  solution was measured at 562 nm, the visible wavelength of maximal absorbance for the  35  pyridine coefficient  hemochrome reported  derivative (Tamura  (Tamura  et  al,  et  al,  1973a).  Based  on the extinction  1973a) for this pyridine hemochrome derivative  (e =34.3 mM^cm" ), the concentration of the metDME-Mb 1  562  solution was determined.  The Soret extinction coefficient was then calculated for metDME-Mb.  For comparison,  this procedure was repeated for native metMb.  2. H NMR Spectroscopy 1  Samples of protein were exchanged into D 0 buffer (KPi in D 0 ( H 0) at pD=7.0 2  2  2  2  and 7=0.1 M) through repeated (>8) cycles of centrifugation in the micro-concentrators. 'H NMR  spectra of horse heart metMb and metDME-Mb were collected by Dr. Colin  Tilcock in the laboratory of Dr. Pieter Cullis of the Biochemistry Dept., U.B.C. Bruker Model WP-200 NMR  spectrometer.  using a  An external sample of 3-2,2-dimethyl-2-2-  silapentane-5-sulfonate (DSS) was used to calibrate the instrument.  Greater than 10,000  transients were collected per sample before Fourier transformation.  3. EPR Spectroscopy  MetMb and metDME-Mb samples were prepared  in sodium phosphate (pH 6.0) and  sodium borate (pH 9.0) buffers (7=0.1 M), placed in EPR tubes (Ace Glass Inc.) and then frozen in liquid N . 2  by  Dr. Linda  EPR spectra of horse heart metMb and metDME-Mb were collected  Pearce  using  a Varian  Model  E109 spectrometer  temperature controller) in the laboratory of Dr. Wu U.B.C. The standard  used  (equipped  of the Faculty of Food Sciences,  was 2,2-Di(4-fm-octylphenyl)-l-picrylhydrazyl  radical (Aldrich Chemical Co., Dziobak and Mendenhall, 1982).  36  with a  (DPPH) free  F. Determination of the Heme pK  a  To determine the pAT of the water molecule bound to the heme Fe in metDME-Mb, a a  spectroscopic pH  titration was  dialyzed overnight against 0.1  performed as follows. M  A  sample of metDME-Mb  was  NaCl and then placed into a custom-modified quartz  cuvette, 4 cm x 1 cm x 1 cm (pathlength).  This cuvette had a 12 cm vertical extension  of glass that ended in a round 4.5 cm diameter opening.  The opening was threaded to  allow placement of a plastic cover (Ace Glass), which had two channels drilled through it to accommodate a MI-412 micro-combination pH glass tip of a Digipet 1.0 mL  ultramicro-pipet (Manostat).  to a Radiometer Model PHM-84 pH meter. M  electrode (Microelectrodes Inc.) and die The electrode was connected  The ultramicro-pipet was used to deliver 0.1  NaOH (Dilut-it, Baker) in small volumes with an accuracy of ± 0.2 ttL.  Mixing was  achieved after each addition of base with a magnetic micro-stirbar placed in the cuvette and driven by the stirrer assembly of the Cary-219 spectrophotometer. The pH  of the initially acidic (pH 5.8 - 6.0) metDME-Mb solution was gradually  raised by addition of ~0.5  /*L volumes of 0.1 M  each addition of base until the pH turned off.  After the pH  using the OLIS software. were observed.  A  NaOH.  The solution was stirred after  reading stabilized, at which  time the stirrer  was  had stabilized again, the spectrum was recorded and digitized The pH  was raised until no further changes in the spectrum  similar titration was performed with native horse heart metMb as a  control.  G. Measurement of Autoxidation Rates  The in vitro autoxidation rate of oxyDME-Mb was measured as follows. was  reduced by excess sodium  MetDME-Mb  dithionite (Na S 0 ) in the presence of air to produce 2  37  2  4  oxyDME-Mb.  As  chromatography  recommended  by  Brown  and Mebine  (1969),  an  ion-exchange  column was used to remove excess dithionite and dithionite by-products,  which accelerate autoxidation of oxyMb.  DEAE-Cellulose (DE-52, Whatman) was chosen  because it also permitted separation of metMb from oxyMb (Gotoh & Shikama, 1974).  To  remove trace metals from solution, the buffer used (NaPi, pH 7.0, 7=0.1 M, containing 0.1 mM  Na EDTA) was passed through a 12 x 1.75 cm column of Chelex 100 chelating resin 2  (Bio-Rad), which had previously been cleaned by 1 M HCI. Solid dithionite (Baker, stored under vacuum) was added to a small volume (30-50 nL) of concentrated metDME-Mb solution.  Immediately  afterwards, the solution was applied  to a 6 x 1 cm column of DEAE-cellulose that had been equilibrated with NaPi buffer (pH 7.0, 7=0.1 M) containing 0.1 mM  Na EDTA. 2  The oxyDME-Mb passed through the column  as a reddish band behind a leading edge of brown colour.  The oxyDME-Mb was collected  into a clean cuvette and immediately placed into the thermostatted (25 °C) sample holder of the Shimadzu spectrophotometer.  After allowing 30 minutes for the solution to reach  thermal equilibrium, the visible spectrum  (700-450 nm) of oxyDME-Mb  repetitively over one hour intervals for 24 hours.  was scanned  The autoxidation rate for native Mb  was measured by identical procedures.  H. Synthesis of Pentaammineimidazoleruthenium (III) trichloride  The synthesis of [Ru(NH ) Im]Cl 3  (1974) as follows.  5  3  was based on the procedure of Sundberg  A 30 mL glass reaction vessel with a glass frit bottom was charged  with 1.2 g freshly prepared zinc amalgam and 20 mL 0.05 M dissolved  72.5  [Ru(NH ) Cl]Cl 3  5  2  et al.  mg  (0.25 mmol)  of  HCI. In this solution was  cUoropentaammineruthenium  (III) dichloride,  (Strem Chemicals Inc.) and 90 mg (1.32 mmol) of imidazole (Im), which  had been recrystallized from benzene.  A stream of deoxygenated (vide infra) argon was  38  bubbled through the frit into the solution. and  diluted with 50  material.  mL  of water.  After combining  2.5  cm  The  the product  through the solution for one hour.  After four hours, the solution was reaction was  from  the two  then repeated to obtain more preparations, air was  HC1  followed by 4 M  The column was  HC1 to elute the [Ru(NH ) Im]Cl . 3  was dried by rotary evaporation to an orange coloured paste. redissolved in a minimal  bubbled  The solution (pH 6.8) was then applied to an 8.5 x  BioRex-70 column, which had been washed with water.  washed with 1 M  filtered  5  The eluant  3  The [Ru(NH ) Im]Cl 3  first  5  3  was  volume of water, crystallized with ethanol, filtered, and dried  under vacuum.  /. Spectroelectrochemical Experiments  Potentiometric  titrations  were  performed  with  electrode (OTTLE) cell as described by Reid et  an  al.  optically  (1984).  transparent  The  thin-layer  electrochemical cell  consists of a lucite frame onto which were mounted two quartz plates (Wilmad) separated by two teflon spacers (Dilectrix Corp.) and a gold mesh (500 lines/inch, Buckbee Mears Co.,  Minneapolis, Minnesota) working electrode.  soldered to the bottom of the gold mesh.  A  5 cm  length of copper wire  was  All components were sealed with epoxy cement  (Epoxi-Patch, The Dexter Corp.). Two filled  receptacles were machined into the top of the OTTLE cell body.  glass adapter  with  Radiometer Model 4112  glass frit  bottom  was  inserted  into  saturated calomel reference electrode (SCE)  water-jacketed salt bridge fitted with a platinum wire junction. filled  with saturated potassium  chloride (KC1),  clamped to a ring stand for support. Model RC3  circulating  each  inserted into one  A  buffer-  receptacle. was  A  placed in a  This salt bridge  was  of the adapters and  The water jacket was connected to a MGW-Lauda  water bath to maintain the reference electrode at 25 °C.  39  A  platinum counter electrode was inserted into the other adapter.  The potential across the  working electrode was controlled by a Princeton Applied Research Model 173 potentiostat and measured with a Keithley Model 177 digital microvoltmeter to an accuracy of ± 0.1 mV. Near UV-visible absorption spectra (700-340 nm) were recorded using the Cary-219 spectrophotometer in the reduced height beam mode. a custom-made jacketed aluminum cell holder.  The OTTLE cell was mounted onto  The cell holder was connected to a  MGW-  Lauda Model RM3 circulating water bath to allow temperature control of the OTTLE cell. The temperature of the OTTLE cell solution was measured to ± 0.2°C using a Fluke Model  2175A  digital  thermometer and a subminiature copper-constantan thermocouple,  which was inserted into a side aperture of the cell. Solutions of metDME-Mb  and native metMb were exchanged into the desired buffer  by use of a Centricon microconcentrator (Amicon Corp.). pH range 5.5 to 8.0. titrated with 0.1 M  From pH 8.0 to 9.5, the buffer consisted of 0.025 M  400  fiL.  boric acid  NaOH to the desired pH; solid NaCl was added to raise the ionic  strength to the desired level. was  NaPi buffer was used in the  Typical  The volume of protein solution used in the OTTLE cell  protein  concentrations were  100-150  pM  with  the higher  concentrations needed at basic pH values to compensate for lower extinction coefficients. Mediators used were [Ru(NH ) Im]Cl 3  native Mb  5  3  for DME-Mb  titrations at concentrations of 0.5 /*M  titrations and [Ru(NH ) ]Cl for 3  and 4  6  respectively.  3  Higher  concentrations of mediator were used at lower temperature or basic pH to compensate for slower equilibration times. Protein solutions were made anaerobic by the action of Rhus vernicifera which  in the reduced  Malmstrom, 1981). the  form  is able  to reduce oxygen  to water  laccase,  (Reinhammar and  This enzyme was a gift from the laboratory of Prof. Harry B. Gray of  California Institute of Technology.  Trace amounts of laccase (0.1 /*M in the protein  40  solution) were used with slightly higher concentrations needed under conditions of low temperature  or acidic  pH  to compensate  measurements of native Mb formation  of a third  equilibrium.  A  required  for lower  activity.  The electrochemical  the presence of the enzyme catalase to prevent  species of Mb  that interfered with  the metMb  and deoxyMb  10:1 dilution of a catalase semi-crystalline solution (Sigma type C-100)  was  prepared using the appropriate buffer. In general 1 /<L of diluted catalase solution  was  added to the protein solution prepared for the OTTLE cell (ca. 200 Sigma units of  activity).  Greater amounts were used under conditions of extreme pH or low temperature  to compensate for lower activity. The procedure used to measure the reduction potentials of DME-Mb and native Mb were identical.  A typical electrochemical measurement of DME-Mb would begin with the  application of a potential of —500 mV vs. SCE across the cell.  This potential reduces  metDME-Mb to deoxyDME-Mb and activates, by reduction, the laccase. The reduction process was monitored by observing the absorbance changes at the Soret wavelengths of the met-and deoxy- forms of DME-Mb. stages by changes of 20 mV  The reduced DME-Mb was then re-oxidized in  starting from a potential approximately 50 mV higher than  the measured midpoint potential to a potential 50 mV  lower.  Complete oxidation of  DME-Mb was achieved by applying a potential of +200 mV vs. SCE to the OTTLE cell. complete spectroelectrochemical  experiment consisted of the fully  reduced and oxidized  spectra and six intermediate spectra representing various ratios of reduced:oxidized Mb.  41  A  DME-  /. Fe(EDTA) ' Reduction Kinetics 2  Kinetic experiments involving  the reduction of DME-Mb  by Fe(EDTA) " were  performed on a Dionex Model D-110 stopped flow apparatus.  Fittings, transfer lines  2  and seals in the apparatus were modified to improve anaerobicity (Reid and Mauk, 1982; Reid, 1984). tungsten-halogen  light  photomultiplier tube.  The optical system was designed by OLIS and consisted of a source,  its power  supply,  a  monochromator,  and a  This system was interfaced with a Data General Nova 2/10  minicomputer (with a Tektronix Model 4006 graphics monitor).  The instrument was  controlled and data were collected and analyzed by software developed by OLIS (Reid and Mauk, 1982; Reid, 1984). Solutions were made anaerobic by bubbling them with nitrogen gas (N ) that had 2  been scrubbed of oxygen and humidified by passage through two gas washing bottles each  containing  a solution  of vanadate(IV)/amalgamated  zinc  (Meites and Meites,  1948), a third gas washing bottle containing a solution of methyl viologen, proflavin and EDTA (Sweetser, 1967) and a condensation trap containing buffer.  All solutions  were transferred using Hamilton gas-tight syringes and stainless steel needles. A  Fe(EDTA) " concentrated stock solution was prepared for each experiment by 2  mixing deoxygenated solutions of NaPi buffer-EDTA and ferrous ammonium sulfate, [Fe(NH ) ]S04, as described by Wherland et al. (1975). The desired concentrations of 4  2  Fe(EDTA) " were made by dilution of the stock solution with deoxygenated NaPi 2  buffer.  Fe(EDTA) " dilutions and protein solutions were prepared under a continuous 2  stream of scrubbed and humidified N stoppers.  2  in serum bottles capped with clean rubber  Protein solutions of 2.5 to 3.0 /*M were initially bubbled gently with  deoxygenated N  2  for 30 minutes, after which the stream of N  2  was passed over the  solution surface. Deoxygenated solutions of protein and Fe(EDTA) ' were drawn into 2  42  separate drive syringes, which were immersed in a water bath. maintained  at constant temperature by  a  Haake  Type  3  The water bath was  circulating  water bath.  Solutions were equilibrated for 20 minutes at 25 °C before use and for 30 minutes at other temperatures. Reactions were performed under pseudo-first order conditions with Fe(EDTA) " in 2  at least 50-fold excess over protein. Soret  wavelength  observed. first  (407  nm)  and  Reduction of metDME-Mb was monitored at the an  absorbance  decrease of 0.12  was  routinely  The average of at least seven traces was used to calculate an observed  order  rate  constant, Jfc . obs  For  each  experiment,  [Fe(EDTA) "] was  used to obtain six values of Jfc .  [Fe(EDTA) "] was  subjected  2  2  obs  a  10  fold  range  Each plot of A:  obs  of  versus  to weighted linear least-squares analysis to determine  the slope, which represented the second order rate constant, £ . 12  v  43  RESULTS  A. Preparation and Isolation of DME-Myoglobin  The apoMb prepared as described in the experimental procedures section showed no significant residual heme as determined by examining the Soret (near UV) region of the spectrum.  Dimethylsulfoxide (DMSO) was initially used as a solvent for reconstitution of  the apoMb with the water insoluble DME-heme because it had been successfully used in the reconstitution of apo-cytochrome b  5  to form DME-cytochrome b  5  (Reid et al.,  1985).  However, the use of DMSO resulted in poor reconstitution of apoMb, with low yield of metDME-Mb.  With DMSO as a solvent, there was an apparent aggregation of the DME-  heme in the presence of apoMb.  This aggregate resulted in a greenish film that stained  glassware, ultrafiltration membranes and Sephadex gels.  The use of warm methanol as a  solvent for the DME-heme reduced the amount of staining.  However, the solubility of  the DME-heme was lower in methanol than in DMSO and this necessitated the procedures described to remove excess DME-heme. Reduction of newly  reconstituted  metDME-Mb to oxyDME-Mb improved die yield  considerably from about 10% to 70% of apoMb used.  This increased yield probably  reflected a more stable interaction between apoMb and the reduced form of DME-heme in the presence of methanol. under  the  conditions  The oxyDME-Mb did not bind to the DEAE-cellulose column  used  to  remove  metDME-Mb, which bound irreversibly. ultrafiltration  membranes,  which  excess  DME-heme  and  Fe(EDTA) ", 2_/1  unlike  However, oxyDME-Mb tended to precipitate onto  necessitated  metDME-Mb prior to concentration.  44  its  oxidation  (by  ferricyanide)  back  to  Ferricyanide was used as an oxidant in spite of its apparent binding to metDME-Mb because the oxidant normally used in this laboratory to oxidize native Mb, [Co(dipic) ]", 2  failed  to  through but  oxidize the  consistent  oxyDME-Mb  Dowex increase  X-8 in  to  any  ion-exchange the  A4 /A 07  significant column  2g0  degree. to  ratio  Passage  of  remove  ferricyanide  of about  2-4%.  metDME-Mb resulted  Failure  in  solutions a  slight  to remove  ferricyanide in this manner resulted in significantly altered electrochemical and kinetic properties of the reconstituted protein.  Dialysis was not effective in removing residual  ferricyanide from metDME-Mb. Thin layer chromatography of the heme samples showed single spots for the DMEheme and DME-Mb extracted heme samples, but the native heme sample was streaked. The Rf values determined for DME-heme, native heme, and DME-Mb extracted heme were 0. 93, 0.085. and 0.93 respectively.  These values agree closely with those reported by  Asakura and Lamson (1973) for the same experimental  conditions and indicate that the  DME-heme within DME-Mb was not modified (i.e. not de-esterified) during reconstitution.  B. Spectroscopic Characterization of DME-Myoglobin 1. Near UV-Visible Region  The  visible electronic spectrum (700-450 nm) of metDME-Mb and oxyDME-Mb are  shown in Figure 5. The  The metDME-Mb spectrum is shown as a function of pH in Figure 8.  deoxyDME-Mb and metDME-Mb spectra can be seen in the spectroelectrochemical  titration in Figure 12.  Comparison of these spectra with the corresponding  Mb spectra  shows only slight shifts in the wavelengths of absorbance maxima, e.g. Soret band is at 407  nm for metDME-Mb and at 409 nm for metMb.  Soret band of metDME-Mb was 168 mM"  1  The extinction coefficient for the  cm" and 190 mM" 1  heart metMb.  45  1  cm" for native horse 1  Figure 5. The absorption spectra of oxyDME-Mb (a and p peaks labelled) and metDMEMb in the visible region. Spectra were recorded under conditions of pH 7.0, 7=0.1 M and 25 °C in sodium phosphate buffer. The concentration of DME-Mb was 80 i*M.  46  2. HNMR l  The  *H NMR  spectra of both native metMb and metDME-Mb were partially obscured  by the strong signal of residual water (around 7 ppm). However, the downfield region of the metMb and metDME-Mb  (pD 7.0, 7=0.1 M) R l  NMR  spectra, which contain the  majority of heme signals, were distinct and are shown in Figure 6 (A & B). NMR al,  The H !  spectrum of horse heart metMb is similar to that of sperm whale metMb (La Mar et 1980) in terms of chemical  between peaks).  shifts (ppm), relative intensities and positions (Appm  Solely on this basis, proposed identification of some of the resonances  in the horse heart metMb spectra are given in Table I. The  corresponding  *H NMR  resonances from that of metMb.  spectra of metDME-Mb, shows a different pattern of However, the methyl resonances are still recognizable  though specific assignments cannot be made (Table I).  The signals for the protons of  the two methyl esters could be represented by the resonance peak labelled 4 (Figure 6A) centered at 51 ppm due to its uniqueness (compared to the native metMb H l  spectrum) and its relatively high intensity (about double that of the methyl peaks).  47  NMR  2  I  3  4  5  A  80  60  40  20  ppm  Figure  6.  The  downfield  labelled parts  A)  per  conditions  and  portions native  million were  of  metMb  (ppm)  7=0.1  the  M  from  H  N M R  spectra  (B).  The  chemical  J  an  potassium  48  external phosphate  DSS  of  metDME-Mb  shifts sample  buffer,  pD  are  (6)  7.0  set  at  and  (spectrum measured 5=0. 25°C.  in The  Table I. Proposed assignments for U NMR resonances labelled in Figure 6. l  protein  peak label  chemical shift (ppm)  met-Mb  a b c d e f g h 1  92 85 71 51 74 58 45 (broad) 30 (broad) 18  methyl H's methyl H's methyl H's methyl H's propionate H a propionate H a vinyl H a and propionate H a vinyl H a and propionate H a propionate H/?s  metDME-Mb  1 2 3 4 5 6 7 8  81 77 69 51 49 61 43 (broad) 32 (broad)  methyl H's methyl H's methyl H's 2 methyl H's methyl H's vinyl or propionate H's vinyl and/or propionate H's vinyl and/or propionate H's  49  proposed heme *H assignment  3. EPR  is  The  first  shown  in  the g - f a c t o r ,  derivative  Figure  of  7.  w h i c h is  the  E P R absorption spectrum o f  The  defined  magnetic  moment  of  a  paramagnetic  h  is  represents  i.e.  the  the  term  Planck's  constant  energy  difference  the  resonance  H  is  t  condition.  the  magnetic  derivative spectrum).  while  ( p H 6.0,  species  is  7=0.1  M)  represented  by  by:  hv = g M B ^ r  where  metDME-Mb  (W)  v  and  is  between  the  the  The  term  fi  field  strength  frequency  two  is  B  at  of  absorption.  orientations  the  Bohr  the  magneton  resonance  (the  T h e E P R spectrum o f m e t D M E - M b  of  term  magnetic  (9.274  inflexion  ( p H 6.0,  The  x  point  10"  hv  moment,  J  2 1  of  T  the  7=0.1  M ) yielded  shown  in  1  ) ;  first-  g=5.9,  g=5.7.  at p H 9 . 0 ,  C. Spectrophotometric pH Titration  The  effect  changes  in  original  2.0  NaOH  each.  volume  and  The  of  the  p H  spectrum  m L  of  This  as  on  a  the  visible  spectrum  of  native  metMb  metDME-Mb  volume  result  of  spectra  were  -  2  0  by  not  the  ± ~  with  metMb  added  e q u i l i b r i u m represented  metDME-Mb H  and  of  metDME-Mb  p H  solutions  NaOH  are  for  less  The  than  titrations  than  a  F i g u r e 8.  20  1%  /iL o f  change  dilution.  spectrophotometric  m e t D M E - M b O H  50  similar.  r e q u i r e d less  represents  corrected  is  -  changes  + H  +  is  as  follows:  (15)  The  of  the  0.1  in  N  total  1000  2000  3000  Magnetic Field (Gauss)  Figure 7. The EPR spectrum of metDME-Mb (0.6 mM). Conditions were pH 6.0, 7=0.1 M sodium phosphate buffer frozen in liquid nitrogen (-180°C). The EPR instrument settings were microwave power of 2 mWatt, magnetic field scanning range of 1000 to 3000 Gauss, and temperature of -180°C (liquid N cooling). 2  51  0.60  "T  '  1  1  1  1  1  r-  0.45 O C  D _Q O CO _Q  0.30  <  0.15  0  — i — i  440  500  560 Wavelength  Figure 8.  620  680  740  (nm)  The visible region absorption spectra of metDME-Mb (43 pM) as a function of pH (NaOH titration). At 25°C and in 0.1 M NaCl, spectra were recorded in sequence at pH 5.87 (spectrum labelled A), 7.47, 8.17, 8.55, 8 82 9 16 10 12 and 10.47 (B). '  52  The concentrations of acid and base forms of metDME-Mb or metMb were derived from absorbance values.  These values were measured at 580 nm, which  for metMb and  metDME-Mb demonstrated the greatest change in absorbance with pH.  The absorbance  values at the extremes of pH were assumed to represent complete (100%) acid or base forms of metDME-Mb or metMb. The relationship used to determine the heme pK is: a  pH = pK + a  n log ~ ~ A a  (16)  A  A-A  b  terms A  The  b  and A  represent  a  the absorbance values of the base and acid  respectively; A represents the absorbance value at an intermediate pH.  forms  The parameter n  represents the number of protons transferred in the reaction and is expected to have a value of one. A  linear  regression plot  titration is shown in Figure 9. this plot and the pK  a  calculated pK  a  of pH  versus  log[(A -A) / (A-A )] for the metDME-Mb b  a  The term n in equation (16) is derived from die slope of  is derived from the y-intercept.  is 8.49 ± 0.05 with n = 0.98(6).  For the metDME-Mb titration the  From a similar analysis of the metMb  titration, the calculated pK is 8.88 ± 0.02 with n = 1.00(2). a  53  Figure 9.  The data from the pH titration of metDME-Mb (Figure 8) plotted according to Equation 16 by linear regression analysis. A and A are the absorption values of metDME-Mb'H 0 (pH 5.87) and metDME-MbOH (pH 10.47) at 580 nm respectively. A is the corresponding absorbance value at intermediate pH values. The spectroscopic pK of metDME-Mb is determined from the yintercept of the linear regression analysis. a  2  a  54  h  D. Autoxidation Rates  The autoxidation of oxyDME-Mb was monitored by recording the visible spectrum of a solution of oxyDME-Mb at one hour intervals, as illustrated in Figure 10. To calculate the oxyDME-Mb  autoxidation rate, the amount of oxyDME-Mb  (relative to the total  amount [DME-Mb] = [oxyDME-Mb] + [metDME-Mb]) has to be determined at each time interval.  The visible spectrum of oxyDME-Mb is characterized by its prominent alpha (a)  and beta (/?) absorption bands (Figures 5 and 10).  The a/p absorbance ratio has been the  traditional standard for determining the proportion of oxyMb in a sample of Mb. The fraction of oxyDME-Mb present in a mixture with metDME-Mb can be correlated to the corresponding o /p absorbance ratio, A[a /p)  C -A(al  [oxyDME-Mb]  x  =  [DME-Mb]  , by the relationship:  p)C RT  C, - 1 +  x  2  (l-C )A(a/p)R-  1  2  The derivation of Equation 17 is given in Appendix D.  The parameter Cj represents the  ratio of the extinction coefficients of metDME-Mb and oxyDME-Mb at the o band; C is 2  the corresponding ratio at the P band.  The parameter R is the ratio of the extinction  coefficients of oxyDME-Mb at its o and p bands. Using data reduction routines of the OLIS software, the absorbance spectra of oxyand metDME-Mb (Figure 5) were used to generate spectra of various oxy- and metDMEMb mixtures.  A total of nine such spectra were generated ranging from 1 part metDME-  Mb^ parts oxyDME-Mb to 9:1 in 1/10 intervals. an A ( o / p) (Micromath  value was determined.  Scientific  Software,  Salt  For each simulated spectrum generated,  The nonlinear least squares fitting program MINSQ Lake City,  Utah) was used  to fit the data to  Equation 17. Calculated values for C i , C and R are listed in Appendix D. 2  55  450  500  550  600  650  700  WAVELENGTH (nm)  Figure  1 0 . The autoxidation of oxyDME-Mb ( 4 5 / J M ) to metDME-Mb depicted by sequential recordings of the absorbance spectrum in die visible region. Spectra were recorded at one hour intervals. Conditions were pH 7 . 0 , 7 = 0 . 1 M and 2 5 °C in sodium phosphate buffer.  56  The A ( o / p)  value of each spectrum in Figure 10 was determined.  The fraction of  oxyDME-Mb present at each time interval was calculated using Equation previously calculated parameter', values.  17 with the  These data were fitted by a linear regression  program to the first order kinetics expression:  -in  [°*y™E-Mb] [DME-Mb]  The  t  =  k  t  ( l g )  c  parameter t refers to time (in hours) and k represents the first order rate constant  of autoxidation.  The ratio [oxyDME-Mb] /[DME-Mb] t  0  represents the amount of oxyDME-  Mb at time t in relation to the starting amount at t fraction of oxy DME-Mb calculated from Equation 17.  = 0 and is equivalent to the  The time t = 0 refers to the time  the first spectrum was recorded and not the time of formation of oxyDME-Mb. Figure 11 shows the plot of the oxy DME-Mb autoxidation data. analyzed by linear regression analysis to obtain a value for the slope k. M  The data were At pH 7, 7=0.1  and 25 °C, oxy DME-Mb has an autoxidation rate of * = 2.6(1) x 10" hr' , which 2  corresponds to a half-life for oxyDME-Mb of 27 hours.  1  Under the same conditions and  method of analysis, oxyMb has an autoxidation rate of k = 9.8(1) x 10" h r , which 3  gives a half life for oxy Mb of 71 hours.  57  1  0  Figure 11.  5.0  10.0 15.0 20.0 time (hours)  25.0  The oxy DME-Mb autoxidation data derived from Figure 10, plotted according to Equation 18 and fitted by linear regression analysis. The autoxidation rate, *, is given by the value of the slope.  58  30.0  E. Synthesis of [Ru(NH ) Im]Cl 3  The  5  3  yield of [Ru(NH ) Im]Cl 3  [Ru(NH ) Cl]Cl , 3  5  was  2  5  54%.  based on  3  The near UV-visible spectrum of [Ru(NH ) Im]Cl , which 3  consists of two absorption bands at 299 et al.  (1974).  cm"  and  1  £  4 3 0  the amount of starting compound used,  and 430 nm,  2  values of 1880 mM"  1  mM"  For comparison, Sundberg et al.  cm" .  1  1  cm"  1  and 250 mM"  cm"  1  1  3  matches that reported by Sundberg  The calculated extinction coefficients (in water) were e 99 =271  5  =  1905  reported  mM"' corresponding  respectively.  F. Reduction Potential Measurements  As  described in the experimental  native Mb  procedures section, electrochemical measurements of  required the presence of catalase.  Without catalase present, the reduced  absorption spectra exhibit a shoulder centered around 600 exhibit a shoulder centered at 630 points  between  the  nm  spectra of deoxyMb  ferrylMb.  decomposition  This identification  (Saunders,  ferrylMb (Tamura et al, The  DME-Mb  1973)  and  and  metMb  completion  beginning  The  is based on  comparison  were not  oxidized Mb  spectra  In addition, isosbestic observed  during  interfering species of Mb  the catalase activity  to the published  visible  of  the was  peroxide  spectrum of  1973a). spectra  obtained  experiment are shown in Figure 12. the  and  instead of a broad peak.  potentiometric titration in the absence of catalase. probably  nm  Mb  of the  titration, the  of the titration.  from  a  representative  spectroelectrochemical  Corresponding spectra of native Mb Soret  There was  absorbance, which indicated that both Mb  only  absorbance was slight (max.  compared 5%)  are similar.  At  to that at the  decrease in the Soret  and DME-Mb were relatively stable under the  experimental conditions used.  59  The  Nemst  equation,  which  defines  the reduction-oxidation  equilibrium, can be  written as:  E = E a  + (2.303 x lO' ) _?L nF 3  m  log  ~ A -A  A r e d  (19)  A  o x  The ratio of absorbance values in Equation 19 is equivalent to the ratio of concentrations in Equation 3.  The terms A  red  , A  o x  and A represent the absorbance values of the fully  reduced, fully oxidized and mixed oxidation state respectively for metDME-Mb or native Mb  at a given wavelength.  Calculations are generally based on the change of absorbance  at the Soret band of met(DME-)Mb because it exhibits the greatest change in absorbance upon change of oxidation state.  The terms E  a  and E  m  (mV vs. SCE) and midpoint reduction potential respectively. a conversion  factor used to adjust for the mV  represent the applied potential The number 2.303 x 10" is 3  scale and base 10 logaridim used in  Equation 19. The  values  corresponding titration.  of  log[(A -A) / (A-A )] ox  were  red  calculated  and  plotted  potentials  Figure 13 is the Nernst plot of the data from Figure 12.  represented were  conversion mV  the  applied potentials to obtain a Nernst plot for each spectroelectrochemical Linear regression  analysis was used tofitthe data and calculate the y-intercept and slope. y-intercept  versus  the midpoint  adjusted  reduction  to the standard  potential (mV  hydrogen  electrode  vs. SHE = mV vs. SCE + 244.4 mV at 25°C.  The calculated  vs. SCE). (SHE) scale  Calculated by the  Under standard conditions  (pH 7.0, 7=0.1 M and 25°C) the midpoint reduction potential of DME-Mb is +100.2 (2) mV (vs. SHE) and that of native Mb is +60.9 (2) mV (vs SHE).  60  Figure 12. The absorbance spectrum of DME-Mb (150 nM) at varying ratios of the oxidation states metDME-Mb (spectrum labelled O) and deoxyDME-Mb (R) as generated by an applied potential (Eapp). The DME-Mb sample was in a spectroelectrochemical (OTTLE) cell as described in text. In sequence, the applied potentials (mV vs. SCE) were: -500.0 (R), -200.0, -180.0, -160.0, -140.0, -120.0, -100.0 and +200.0 mV (O). Conditions were pH 7.0, /=0.1 M, 25°C in sodium phosphate buffer with 0.5 fiM Ru(NH ) Im and 0.1 laccase. 3  61  5  -1.20  -0.80  0.0  -0.40  0.40  0.80  1.20  log((A -A)/(A-A )) 0  Figure 13.  r  The data derived from Figure 12 plotted according to the Nernst Equation (19). A and A are the absorbance values of metDME-Mb and deoxyDMEMb respectively at 407 nm (the Soret band of metDME-Mb). A is the absorbance at 407 nm of various ratios of metDME-Mb and deoxyDME-Mb as determined by the applied potential (£ ). The applied potential values are referenced to both the standard calomel and hydrogen electrodes (SCE and SHE respectively). The midpoint potential is determined from the yintercept of the linear regression analysis fit. Q  r  app  62  The  close agreement between the calculated slope and the coefficient value of the  logarithmic  term  in Equation  19 was used as a criterion for the reliability  of  the  electrochemical potential measurement, e.g. at 25°C, the expected slope is 59 mV. Another criterion was the correlation coefficient of the plotted data to the linear (first order) fit.  Repeated measurements of E  (for a given sample under identical conditions),  m  which met the above criteria, agreed within ± 3 mV.  A list of measured reduction  potentials for both DME-Mb and native Mb (Lim and Mauk, 1986) is given in Appendix 1. All  measured  potentials were derived  from  Nernst  plots with  correlation coefficients  >0.995 and slopes with a <3% difference from the expected slope.  J. pH Dependence of Reduction Potential  The pH dependencies of the midpoint reduction potential for DME-Mb and native Mb are shown in Figure  14.  The pH of the protein solutions after spectroelectrochemical  measurements did not change significantly reduction potential of DME-Mb  from  their initial pH.  At pH 8.0,  the  was measured using both phosphate and borate buffer.  There was no significant difference between the midpoint reduction potentials determined in these two buffers.  This absence of specific buffer effect allowed the continuous fit of  the data gathered throughout the pH range measured for both native and DME-Mb. The  pH dependences of the DME-Mb  and native Mb  initially fitted with a least-squares routine to the equation:  E =E m  + I ln nF K R  0  +  [  H  +  ]  + [H ] +  ox  63  (20)  reduction  potentials were  Equation 20 describes the effect of a single titratable group within a hemeprotein on its measured E  (Dutton, 1978).  m  The proton dissociation constant for this group is in turn The terms K  influenced by the oxidation state of the protein.  ted  and K  ox  represent the  dissociation constant of the group when the hemeprotein is in the reduced and oxidized states respectively.  The term E  Q  is the theoretical reduction potential (and hence free  energy change) independent of contributions from the free energy change associated with the change(s) in protonation upon reduction i.e. at pH 0. From a least-squares analysis based on Equation 20, the fits of the pH dependence data for native and DME-Mb are shown in Figures 14. The pH dependence of the native Mb reduction potential is consistent with an E of 64.6 mV with a pK Q  3.2 x 10" M) and a pK 9  ox  of 7.7 (AT  red  ox  =  8  Q  8  red  = 2.0 xlO" M). The pH dependence of the DME-  ox  Mb reduction potential is best fitted with an E of 134 mV with a pK 1.6 x 10" M) and a pK  of 8.5 (AT  of 6.4 (AT  [ed  of 7.8 (K  red  =  = 4 x 10' M). 7  ox  Another model for this type of numerical analysis was proposed by Pettigrew et al. (1975) based on the method of Clark (1967).  This model assumes two ionizations that  affect the reduction potential in the oxidized protein and only reduced form.  one ionization in  the  This model is applicable to Mb because the heme bound water ligand is  only present in metMb and not deoxyMb.  The dependence of E  m  on pH for this model is  written as:  1'  [H ] +  R T  E = E + m  0  nF  ln  [H ] +  2  2  + K  t  (21)  + tf [H ] + K +  ol  64  ol  K  o2  LxJ X CO  120  > >  Ld  6.0  7.0  8.0  9.0  10.0  PH  Figure 14. The pH dependence of the midpoint potential (E ) of DME-Mb (—O—O—) and native Mb (—•—•—) fitted to Equation 20. Conditions were 7=0.1 M and 25 °C in sodium phosphate buffer. m  65  Figure 15.  The same data as in Figure 14 fitted to Equation 21. The pH dependence of the midpoint potential (E ) for DME-Mb is labelled ( - O - O - ) and that of native Mb is labelled ( — • — • — ) . m  66  The  pH dependence data was fitted to Equation 21 using the least squares fitting  program MINSQ (MicroMath Scientific Software, Salt Lake City, Utah). the value of K  o2  was fixed at the K value for the heme coordinated water molecule of a  native or DME-Mb, which was determined visible absorption spectrum (Section C). Mb  and DME-Mb  In this analysis,  from titration of the metMb or metDME-Mb The fits of the pH dependence data for native  derived from Equation 21 are shown in Figure 15.  For die pH  dependence of the native Mb reduction potential, values of 7.4 and 7.7 were calculated for the pK  and pK respectively with J5 =65.4 mV.  ol  r  0  For DME-Mb, the pK  ol  and pK  t  values were 6.3 and 7.5 respectively with f? =135 mV. 0  2. Temperature Dependence of Reduction Potential  The free energy change of the reduction process defined in Equation 2 is composed of enthalpic and entropic terms.  In an electrochemical cell, these parameters are related  by the equation:  AG  In this relationship, AS° cell  including  terms with Molar  C  = AH° - TAS°  (22)  C  is the reaction entropy  that of the reference electrode.  change for the entire electrochemical In biological applications, the energy  the superscript symbol (°) do not represent standard  concentrations) but  Converting  0  free  energy  conventional  biological  to electrochemical  terms  m  7AS° _ AH F  0  F  67  one  conditions, e.g. pH 7, 7=0.1 M. (by AG°=—nFE ) m  electrochemical version of the van't Hoff equation:  _  conditions (i.e.  (23)  yields the  The  temperature  dependences  of DME-Mb  potentials are shown in Figure 16. the  slope  represents  and y-intercept  and native  Mb  midpoint  reduction  These data were analyzed by linear regression, and  were calculated for each  set.  The value  of the slope  AS° / F and that of the y-intercept represents —AH/F. At pH 7.0 and 7=0.1  M,  C  the oxidation-reduction equilibrium of native Mb is characterized by AS° (—94(2) J mol^K" ) and AH 1  0  = —8.1(2) kcal mol"  1  C  = —22.6(5) eu  (—33.9(8) kJ mol" ). From Equation 2, 1  the free energy change is A G =—1.36 kcal mol" at 25°C. 0  1  The midpoint reduction  potential of DME-Mb exhibited a greater dependence on the temperature than native Mb.  AS° = —35.4(8) cal  The oxidation-reduction equilibrium of DME-Mb is characterized by  C  moHK- or eu (-148(3) J mol^K" ) and AH° = —12.9(2) kcal mol" 1  1  1  (-54.0(8) kJ mol ). 1  The free energy change at 25°C is A G ° = —2.34 kcal mol" . 1  The  AS° values can be corrected for the contribution of die reference standard C  hydrogen electrode (SHE) by the equation:  AS  The  0  = A S ° - (5 +-'/ S ) H  2  (24)  H2  entropic change in the SHE half-cell at 25°C is +15.6 eu (Taniguchi et al, 1980;  Giauque, 1930)  and is regarded as a constant.  After correcting for the SHE entropic  contribution, A S for DME-Mb is —51.0(8) eu (-213(3) J mo^K" ) at pH 7.0 7=0.1 M. 0  1  Similarly, A S for native-Mb is —38.2(5) eu (—160(2) J mol^K" ). 0  1  68  Figure  16. The temperature dependence of the midpoint potential (E ) of DME-Mb (-O-O-) and native Mb (—•—•—) fitted to Equation 23 by linear regression analysis. Conditions were pH 7.0 and 7=0.1 M in sodium phosphate buffer. m  69  3. Ionic Strength Dependence of Reduction Potential  The  ionic strength dependences of DME-Mb and native Mb reduction potentials are  shown in Figure 17.  These data were fitted to the equation derived by Goldkorn and  Schejter (1976):  E  In  = E - 2.3 RTjqj-tf) F  m  n  relationship, the term E  this  Q  reduced states respectively.  and q  T  (25)  is the midpoint  a  strength and the terms q  A /(I)  reduction potential at zero  ionic  are the net charges on the protein in the oxidized and  The constant A is the Debye-Hiickel constant of ionization,  which has a value of 0.509 C^M'^ at 25°C in water. (/) function with the general form (Tanford &  The term J\I) is an ionic strength  Kirkwood, 1957; Beetlestone & Irvine,  1963):  =  M  (26)  —  1 + BR Vl a  The parameter B is the Debye-Hiickel coefficient, which has a value of 0.329 M'^A"  1  in water at 25°C; R  a  ion in solution.  is the distance of closest approach between the protein and a small  Goldkorn and Schejter have used two forms of J\I) depending on the  observed degree of dependence of E 1982)  m  on ionic strength.  For cytochrome b  5  (Reid et al.,  and cytochrome c (Margalit and Schejter, 1973), the midpoint reduction potentials  of which exhibit a relatively small ionic strength dependence, R  a  equivalent to the radius of the protein.  is considered  to be  For cytochrome c-552, the reduction potential of  which exhibits a larger ionic strength dependence (Goldkorn term BR is assigned a value of one. a  70  and Schetjer, 1976), the  In fitting the current data, various values for the radius of Mb  were used.  The  hydrodynamic radius of Mb (assuming a globular shape) was calculated from the following equations derived from Tanford (1961):  if = 3  3  M  4JT/V  (v  2  (27)  + 5jt7f)  In this relationship the term N is Avogadro's number (6.023 x 10 M  23  mol" ) and the term 1  is the molecular weight of native Mb, which is 17,641 Daltons based on its sequence.  The terms tr respectively.  17 2  1  The term  3 1 6  m  e  partial specific volumes of proteins and solvent (water)  5j is the effective solvation of protein.  The standard values  from Rosenberg et al. (1976) are, S = 0.2, t> = 0.73 cm /g, Vf = 1 cm /g. As a result, 3  l  3  2  Equation 27 reduces to:  R = 0.717 V M  (28)  From Equation 28 the calculated hydrodynamic radius for Mb is 23 A. The average axial radius of the native horse heart metMb crystal structure (Evans & Brayer, 1988) was calculated by Mr. Gordon Louie.  First the centre of the metMb crystal  structure was determined from the coordinates of the polypeptide chain atoms. solvent accessible amino acid residues were identified (88 in total).  Then  The distances from  the calculated center to the outermost fixed position atom of each solvent accessible amino acid residue were calculated and averaged. axial radius was obtained by this method.  71  A value of  17.8 A for the average  0  0,20  0.40  0.60  Ionic Strength (M)  Figure  17. The ionic strength dependence of the midpoint reduction potential (E ) for DME-Mb (-O-O-) and native Mb (-•-•-) fitted to Equation 25. Conditions were sodium phosphate buffer, pH 7.0 and 25°C. m  72  ionic strength dependence of E  The  was fitted to Equation 25 using a least squares  m  fitting routine with each of these values for the protein radius.  Figure 17, which shows  the fit for R  = 17.8 A, is representative of all the fits calculated.  the E  of DME-Mb and native Mb are listed in Table II.  a  0  and q  Table II.  Q  Estimated values for  Estimated values for midpoint potential at zero ionic strength (E ) and net charge of the oxidized hemeprotein (q ) as a function of various radius (R ) values used in Equation 26. Q  Q  a  hemeprotein  R (A)  E (mV vs. SHE)  native metMb  3.04* 17.8 23  95.9(4) 150 (2) 169 (2)  + 2.8(2) + 14.0 (3) + 19.9 (4)  metDME-Mb  3.04* 17.8 23  158(6) 250 (3) 281 (3)  + 4.3(3) + 22.7 (4) + 32.6(4)  a  0  * denotes a radius value such that the term BR  a  q  0  = 1 in Equation 26  G. Reduction Kinetics Measurements  The was  effect of residual ferricyanide on the observed kinetics of metDME-Mb reduction  significant.  Samples of metDME-Mb that were not stripped of ferricyanide exhibited  biphasic pseudo-first order plots for the change in A  4 0 7  with time.  Samples that were  stripped of ferricyanide by the Dowex X-8 anion-exchange column gave pseudo-first order plots that were linear for about 3 - 4 half-times.  Figure 18 compares pseudo-first order  plots obtained with and without the presence of ferricyanide.  73  -Fe(CN)l-  + Fe(CN)l-  Figure 18.  Reproductions of the stopped-flow tracings (AA vs. time) and kobs plots (logAA vs. time) for the reduction of metDME-Mb by 2 mM Fe(EDTA) " at pH 7.0, 7=0.1 M and 25°C in sodium phosphate buffer. In the set of tracings labelled +Fe(CN)|~, ferricyanide was present in the metDME-Mb solution. In the set labelled -Fe(CN)|~, ferricyanide was removed as described in the procedures section. 2  74  Figure 19.  The second order plot for the reduction of metDME-Mb by Fe(EDTA) ". The observed pseudo-first order reaction rate (Jfc «) was determined for various concentrations of the reductant ([Fe(EDTA) "]). The second order rate constant was determined from the slope of the linear regression analysis fit. 2  ob  2  75  The  reduction  of metDME-Mb  under all experimental conditions  (stripped of ferricyanide) proceeded as demonstrated  by the identical  observed at all Fe(EDTA) " concentrations used in each experiment.  to completion  changes in A  4 0 7  Figure 19 shows die  2  dependence of metDME-Mb reduction rate on Fe(EDTA) " concentration with the fit from 2  linear regression analysis.  Such analysis of similar data obtained under various reaction  conditions yielded linear fits with y-intercept values hear zero, which is expected for a second order reaction.  The second order rate constant for metDME-Mb reduction by  Fe(EDTA) " under standard conditions (NaPi buffer, pH 7, 7=0.1 M and 25°C) is 1.34(2) x 2  10  3  M'V . 1  7. pH Dependence of Reduction Rate  The  pH dependence of the second order reaction rate for the reduction of metDME-  Mb by Fe(EDTA) " is shown in Figure 20. 2  The variation in  with pH is in part due  to variation in the driving force of the reaction with pH.  The driving force is  equivalent  m  Fe(EDTA). and  to the difference between the reduction  potentials ( A £ )  of DME-Mb and  The A £ can be calculated from the pH dependence of the E M  for Fe(EDTA), which has been measured at 7«0.1  M  m  and 25°C by Kolthoff and  Auerbach (1951) and at 7=0.5 M and 25°C by Reid (1984). The observed k  n  corrected for the change in driving force.  n  ^  l  9  M  A  E  can then be  An adjusted reaction rate, k^K is defined by  an equation derived from Marcus theory (Reid, 1984; Reid et al, 1986):  ^ = k  for DME-Mb  ^  (29)  76  Figure  20.  Frame n  30.  A for  (k )  The  Frame between and  illustrates the  conditions  B  the  reduction were  illustrates  native  p H of  metMb  the and  dependence  of  the  metDME-Mb  by  Fe(EDTA) "  7=0.1  M  results  and  Fe(EDTA) " 2  25°C.  77  order fitted  2  25 ° C  observed  second  in  for  ( L i m and  sodium  the  rate  constant  to  Equation  phosphate  corresponding  Mauk,  1986)  at  buffer. reaction  7=0.5  M  Figure 21. Frame A represents the data for the pH dependence of the second order rate constant adjusted (tf^) for the driving force of the reaction between metDME-Mb and Fe(EDTA) " at 7=0.1 M and 25 °C. Frame B represents the results of corresponding experiments and calculations for the reaction between native metMb and Fe(EDTA) " (Lim and Mauk, 1986) at 7=0.5 M and 25°C. 2  2  78  A table of adjusted reaction rates is provided in Appendix C and the results are plotted in Figure 21.  After correcting for the driving force, it can be seen that about 25% of  the change in rate over the observed pH range was due to the pH dependence of the difference in reduction potential between DME-Mb and Fe(EDTA) ~. 2  The pH dependence of the reaction rate for metDME-Mb reduction by Fe(EDTA) " is 2  based  on the assumption  that a single titratable group in DME-Mb  or Fe(EDTA) " 2  Based on such a model, Rosenberg et al. (1976) have  influences the reaction rate.  derived the following relationship between second order rate constant and pH:  _  "•12 ~  * [H ] + k K  (30)  +  h  a  [H ] + K  a  +  The  term  K  represents the proton  a  a  dissociation  The terms k  titratable group in question.  (acid) equilibrium constant  and k  a  represent the second  h  for  the  order rate  constants for the fully protonated and fully deprotonated forms of DME-Mb respectively. Both the adjusted and unadjusted sets of pH dependence kinetics data for DME-Mb were fitted to Equation 30 using a least-squares fitting procedure.  For the unadjusted set of  data, the values calculated from the fit (Figure 20) were pK =lA  (K =4.0 x 10~ M), 8  a  fc =1918 M ' V a  1  and Jfc = l . l M'V . 1  b  a  From the fit of the adjusted data set (Figure 21),  p/sT=7.6 (tf =2.5 x 10" M), * =1562 M" s' and * =2.8 M" s" . 8  a  a  1  1  1  a  1  b  Measurement of the pH dependence of the reduction potential of native Mb allowed re-examination of the Mb reduction kinetics data of Lim and Mauk (1984). pH  dependence of the reaction rate, k , l2  was adjusted for the driving  between native Mb and Fe(EDTA) ") using Equation 29. 2  Appendix C and plotted in Figure 21.  The observed force ( A E  m  The results are tabulated in  Analysis of the adjusted Jt  12  pH dependence for  native Mb using Equation 30 resulted in the following parameter values (unadjusted values in parentheses): ptf =5.9(5.8), fc =163(60) M ' V a  a  79  1  andfc =13(8.7)M ' V . 1  b  2. Temperature Dependence of Reduction Rate  Equation 9, which relates k  to the free energy of activation by transition state  n  theory, can be redefined in terms of enthalpic (A//*)  and entropic (AS*) activation  energies as: ft,,  In *  1 2  AS*  = _  T  The  R  -  Atf*  . ,_ R  (31)  + In  RT  Nh  term /V represents Avogadro's number (6.022 x 10  constant (6.625 x IO  -34  23  mol" ), h represents Planck's 1  J s), R is the gas constant (8.314 J K^mol' ), and T is the 1  temperature in degrees Kelvin.  Figure 22 shows the Eyring  temperature fitted to Equation 31 by a linear regression analysis.  plot of — l n f t  12  versus  For the reduction of  metDME-Mb by Fe(EDTA) ", AH *= +9.2(3) kcal mol" (38(1) kJ/mol) and AS *= —13(1) eu 2  (-54(4) kJ mol"  1  1  K ). 1  3. Ionic Strength Dependence of Reduction Rate  Analysis  of the ionic  strength  dependence  of metalloprotein  initially based on relationships derived from the Debye-Hiickel ions in solution (Moore and Pearson, 1981).  reaction  rates was  theory for the activity of  However, simple Debye-Hiickel  theory has  proven inadequate for the analysis of metalloprotein kinetics (Feinberg and Ryan, 1981). An  expression  for the ionic  strength  metalloprotein and a small molecular  dependence  of a cross-reaction  involving a  reagent has been derived from Marcus theory by  Wherland and Gray (1976; 1979) and is given in Equation 32:  80  0.25 3.2 1  '  «3.4  1  1  3.6  1/T (1/KELVIN) X 1000  Figure 22. The Eyring plot for the temperature dependence of the second order rate constant (k ) for the reduction of metDME-Mb by Fe(EDTA) ". The data were fitted to Equation 31. The conditions were pH 7.0 and 7=0.1 M in sodium phosphate buffer. 2  l2  81  -BR^I ln k  = ln *  n  1  3.567  i n f  1 + BR y/I  -BR^I  +_!  2  The  parameters Zj  R  +  l  R  of the radii of the reactants and  distance between the reactive centers. 1 and derived  the low  By  from the term A((p-le), l  in Equation 26.  The  term Jfc  inf  term the  convention, the protein is designated reagent The  constant 3.567 is  where e is the electron charge and The  The  is assumed to represent  molecular weight oxidant/reductant is reagent 2.  dielectric constant of water at 25 °C. defined  BR^I  and 7^ represent the apparent net charges of the reagents.  is the sum  2  1+  (32)  e  represents  the  constant B  is the same Debye-Hiickel constant  represents  the second order rate constant at  infinite ionic strength. In a similar manner to the analysis of the ionic strength dependence of the midpoint potential, E , m  several values for the radius of Mb  were used.  average axial radii of Mb  were applied  to Equation 32  dependence kinetics data by  non-linear regression analysis.  The  hydrodynamic and  in fitting the ionic strength Use  of these values implies  that the net charge of the protein influences the reaction rate. Derivation  of  Equation  spherical ions with uniformly  32  involves  the  simplifying  assumption  symmetric charge distributions.  that  proteins  are  However, this equation  does not consider the effect(s) of individual and localized charges in proteins.  If any of  these charges influence  acting  binding  sites  or  the  through  proportional influence on  cross-reaction short-range  rate significantly, for example by electrostatics interactions,  the calculated protein charge, Z j .  The  they  have  a  as dis-  radii of proteins used  in such calculations of ionic strength dependence then become less valid because of the varying influence specific regions of protein surfaces have on reaction rates.  82  To  account for a predominant local rather than net charge effect, Segal and Sykes  (1977) used the arbitrary value of 5 A for the radius (Ri) of the active site of the blue copper protein plastocyanin.  A radius of 5 A for the active site of Mb may be too small  because of the size of the porphyrin molecule.  A  value  of 10 A  ring prosthetic group coordinated  to the iron  was chosen for the current analysis because it  can  accommodate the heme radius and because it is also equivalent to the distances from the heme Fe to the Lys residues 45 and 96, which flank the opening of the heme crevice. The  latter reason would also be consistent with a mechanism that has electron transfer  occurring at the heme edge.  The resulting charges (Zj) calculated for metDME-Mb using  the various radius values are listed in Table III. R  x  =  The fit of the data in Figure 23 with  17.8 A is representative of all the fits because the parameter R  is treated as a  x  constant in Equation 32.  Table III.  Estimated values for net oxidized hemeprotein charge (Zj) and second order rate constant at infinite ionic strength (Jfc ) using various values of hemeprotein radius (R{) in Equation 32. inf  1  R (A)  native metMb  10  +2.0(3)  11(1)  17.8 23  +4(1) +6(1)  12(1) 12(1)  10  +  17.8 23  + 14 (2) + 22 (3)  metDME-Mb  Z  (M'V )  hemeprotein  x  83  x  6.1(4)  fc  inf  2.0 (3) x 10 2.2 (6) x 10 2.2 (7) x 10 2  2 2  12  •  1  0  -  —  *  0.30  » 0.60  Ionic Strength (M)  Figure 23. Frame A represents the ionic strength dependence of the second constant (k ) for the reduction of metDME-Mb by Fe(EDTA) " Equation 32 (pH 7.0, 25°C, sodium phosphate buffer). Frame B similar results for the reduction of native metMb by Fe(EDTA) " Mauk, 1986). 2  12  2  84  order rate fitted by represents (Lim and  ff. Marcus Analysis of Kinetics  The  second order  rate constant  for the reaction between a metalloprotein  substitutionally inert inorganic complex can yield useful mechanistic analyzed  in terms of relative Marcus theory.  and a  information if it is  Such analysis permits,  in  principle,  allowances for the contributions to this rate made by the thermodynamic driving force of the  reaction,  reactivity  the electrostatic interactions  of the inorganic  complex.  between  This  the reactants  and the intrinsic  analysis, however, requires  information about the properties of both reactants.  considerable  This information is readily available  in the present case. The  electrochemical  properties  of Fe(EDTA) have  been  characterized  investigators (Schwarzenbach and Heller, 1951; Kolthoff and Auerbach, 1952;  by several Reid, 1982).  The self-exchange rate of Fe(EDTA) " has been estimated by Wilkins and Yelin (1968) on 2  the basis of cross-reaction with Fe(CyDTA) ".  From the Fe(EDTA) " and DME-Mb  1  electrochemical  and kinetic data,  the various  2  activation energies  their reactions can then be calculated (Appendix  E).  From  of the reactants and the calculated reaction  parameters, the free energy change for the hemeprotein's self-exchange reaction, AG* can be corrected for the electrostatic work term (w ) n  The  1(  to determine AG** (or AG*j ). orr  electrostatics corrected self-exchange reaction rate (at 25 °C and 7=0.1 M), # 1 ° " ,  can then be calculated from AG*i . orr  For DME-Mb an average value of 0.41  was determined forfc^f(from AG*f = 18 kcal mol' ). orr  1  85  M" s" 1  1  DISCUSSION  A. Preparation of DME-Myoglobin  The procedures used to prepare and purify DME-Mb are a significant improvement over those reported by Tamura et al. less than 5% purity  (1973a).  These investigators reported a yield of  while the method developed here provides a 70%  yield.  Furthermore, the  of the product obtained with the current method is significantly  reflected in the higher Soret/A  280  absorbance ratio (>5.0  vs. 3.5).  improved  as  A major problem in  the reconstitution of apoMb with DME-heme is the insolubility of this heme in aqueous solutions that results from esterification of the propionate groups.  Another problem  was  aggregation of the DME-heme and apoMb, which necessitated the additional step of using ion-exchange chromatography with apoMb was (Reid et  al.,  cytochrome b  The type of aggregation observed  not similar to that observed in the reconstitution of apocytochrome  1984). 5  to remove excess heme.  Excess DME-heme could be separated from reconstituted  as a well defined band by gel-filtration on a Sephadex G-75  none of the streaking problems seen with apoMb.  b  5  DME-  column with  This difference was probably due to  non-specific aggregation of apoMb with DME-heme, which could be another cause of the low yield Tamura et al. have reported. Removal of aggregated heme by centrifugation and repeated exposure of the apoMb to freshly dissolved DME-heme presumably also helped improve the yield. step  most  responsible  for the  improvement  in yield  was  the  However, the  reduction  of newly  reconstituted metDME-Mb with Fe(EDTA) " (in presence of air) to oxyDME-Mb. 2  presence  of the DME-heme/methanol, apoMb  appears  reduced  (ferro)DME-heme  (ferri)DME-heme.  than  for oxidized  to have a  consistent with the report of Banerjee and Stetzkowski (1970) who binding fragments of sperm whale Mb  In the  greater affinity for This  observation is  found that the heme-  derived from cyanogen bromide cleavage have much  86  higher affinities for ferroheme than ferriheme.  Tamura et al.  (1973a) also  reduced  reconstituted metDME-Mb  (with dithionite) just before loading it onto a cation exchange  column of CM-50 resin.  However, repetition of their procedure resulted in irreversible  binding of a large proportion of the oxyDME-Mb to the resin. Difficulty experience.  in reconstituting  other  apoproteins  with  DME-heme  is a  common  Asakura and Yonetani (1969) reported difficulty in preparing the DME-heme  derivative of cytochrome c peroxidase (CCP).  Ozols and Strittmatter (1964) were unable  to prepare the DME-heme derivative of cytochrome b . s  Yonetani has suggested that the  heme propionates in CCP direct the heme to the correct position within the heme pocket. In Mb, the heme propionates may have the same function.  Control of heme entry by the  propionates may explain how heme rotational isomers may form in Mb. around the a—i  Heme rotation  axis does not alter the positions of the heme propionates relative to the  protein (see Figure 1).  B. Spectroscopic Properties 1. Near UV-Visible  Comparison of the DME-Mb near UV and visible absorption spectra (Figures 5 and 12) to those published for native Mb  by Tamura et al. (1973a) shows that there are  differences in the wavelengths of absorption maxima, which only differ by only a few nanometers, e.g. the Soret band for native metMb (aquo form) is at 409 nm compared to 407 nm for metDME-Mb (aquo form).  The similarity of their spectra indicates that any  changes in the environment of the heme within the protein upon propionate esterification are not observable based on the near UV-visible spectrum  alone.  The near UV-visible  spectra of other heme derivatives of Mb prepared by Tamura et al. (1973a) also showed only slight differences from that of native Mb.  87  The cm" , 1  Soret extinction coefficient detennined  here for native (aquo)metMb, 192 mM"  1  is similar to that determined by Tamura et al. (1973a), 188 mM^cm" . 1  extinction coefficient detennined  for (aquo)metDME-Mb (at pH 6), 162 mM^cm" , was 1  higher than that reported by Tamura et al. (1973a) of 145 mM^cm" . 1  may  The Soret  This difference  be attributable to the reduced level of apoMb contamination in DME-Mb preparations  obtained by the methods described. The pH dependent changes in the visible spectrum of metMb are due to die proton equilibrium of the water molecule coordinated to the heme Fe(III) i.e. Fe—OH  + H .  -  +  Fe—H 0 2  The pK of this water ligand was found to be 8.9 for native Mb in a  agreement with the findings of Antonini et al. (1971) and Tamura et al. (1973b) who used similar titration procedures.  From titration of the metDME-Mb visible spectrum (Figure  8) the pK of the water ligand in metDME-Mb was found to be 8.5. a  et al. (1973b) reported a pKa of 7.5 for metDME-Mb. calculated from  In contrast, Tamura  However, their pK value was &  a set of data (absorbance change, pH) containing only  three points,  Tamura et al. reported difficulty with  which were measured at pH values less than 8.  DME-Mb stability at alkaline pH, while in the current experiment there were only minor shifts of the isosbestic points at extreme alkaline pH values, indicating denaturation of metDME-Mb.  only  slight  The increase in pK of the heme coordinated water ligand a  in metMb over that of free heme (7.6; Falk, 1964) has been attributed to hydrogen bonding with the distal His-64 residue.  The difference in pK between metDME-Mb and &  native Mb corresponds to a 2.5 fold increase in the K of the water ligand. a  acidity of the water ligand in metDME-Mb  The greater  may result from some weakening in the  hydrogen bonding interaction between the distal His residue and the water ligand e.g. an increased distance of separation between them.  88  2. HNMR l  In the H NMR spectra of sperm whale metMb published by La Mar et al. (1983), the !  minor component of heme rotational isomerization gives rise to absorption peaks with similar chemical shifts as those of the major heme rotational isomer.  There is no  evidence of the minor heme rotational isomer in the (aquo)metMb 'H spectra shown in Figure 6B.  The absence of the minor components signals is probably due to the limited  resolution of the 200 MHz NMR  instrument used, which was unable to resolve the  expected propionic acid methylene proton signals (labelled g & h in Figure 6B). The metDME-Mb *H NMR spectrum at 200 MHz (Figure 6A) also exhibits no evidence the minor  component, which  (aquo)metDME-Mb  suggests that  the heme rotational  equilibrium  is no greater than that of native (aquo)metMb.  ratio of  That their heme  rotational equilibria may be similar is to be expected since the heme methyl and vinyl groups have more influence on the equilibrium between heme rotational isomers by apoMb than the propionate groups (La Mar et al., 1986).  Also, rotation of the heme is about  the a - 7 axis, which does not affect the relative positioning of the heme propionates or propionate esters. The  proton signals of the heme propionate methyl esters in DME-Mb  may be  represented by the broad peak centered at 51 ppm (labelled 4 in Figure 6A), which does not appear in the metMb *H NMR  spectrum.  The 'H NMR  spectrum of the model  compound DME-protoporphyrin IX Fe(Imidazole) (La Mar and Walker, 1979) exhibits a 2  single narrow peak for the proton signals from both methyl esters. seen with DME-Mb  may indicate that the local environments of the heme propionate  esters within DME-Mb metDME-Mb  The peak broadening  are similar though not identical.  One possibility  is that in  the heme propionate esters are both extended out towards the heme pocket  opening unlike the inner and outer heme propionate arrangement found in native metMb.  89  The downfield H NMR spectra of the observable heme signals in metDME-Mb show that !  there are changes in the chemical shift values of all the observable heme methyl groups from  that in native Mb.  These changes indicate alterations throughout the entire  porphyrin environment after propionate esterification though the specific changes can not be determined from the current results.  3. EPR  The EPR spectra of metDME-Mb at pH 6 (Figure 7) and 9 were similar to the corresponding  spectra of native metMb in terms of shape and calculated g value.  These  results agree with those of Tamura et al. (1973a) who calculated an identical g value (6.0)  for both native and metDME-Mb  (and all other  derivatives of Mb) under liquid helium (10 K) temperature.  heme propionate  and vinyl  The similarity between the g  values of metDME-Mb and native metMb indicates that heme propionate esterification did not grossly perturb the heme Fe environment in contrast to its effects on that of the porphyrin.  C. Autoxidation Rate  The measured rate of autoxidation for native horse heart oxyMb at pH 7 (NaPi buffer) and 25°C (0.0098(1) h r ) is slower than the 0.012 hr" reported for both bovine 1  1  (Gotoh and Shikama, 1974) and sperm whale oxyMb (Suzuki and Shikama, 1983). difference is probably  This  not due to a species difference (e.g. sequence) because of the  similarity of the sperm whale and bovine oxyMb autoxidation rates. measured may be due to the different methods used to prepare oxyMb.  The different rates Shikama and co-  workers isolated and purified oxyMb from fresh muscle thus avoiding the need to reduce 90  metMb  with  dithionite, the  oxidation products of which  can  accelerate the observed  autoxidation rate.  That the present study measured a lower autoxidation rate indicates  that dithionite and  its by-products were effectively removed from the oxyMb solution.  The  difference in autoxidation rates may  also be due  to the use of the Chelex  100  chelating resin in the current study to remove trace metals from the buffer solutions, which can also accelerate autoxidation. Analysis of oxyDME-Mb and native oxyMb autoxidation by first order kinetics gave good results. may  However, slight deviations from the calculated fit of the data (Figure 11)  indicate that autoxidation has a more complex mechanism than simple  kinetics.  Based on  the observed pH  Sugawara  (1978) have  proposed  an  dependence of the autoxidation rate, Shikama and "acid-base  catalyzed  three  state model"  mechanism of autoxidation in which there is catalysis by protons at low pH anions at high pH.  first order  for the  and hydroxyl  Protons were proposed to catalyze this reaction by reaction with the  bound oxygen molecule: Fe—0 + H  +  2  The  hydroxyl  anion  Fe(III) + H0 *  0 ~ +  2  promotes autoxidation  via a  2  S2  H  +  nucleophilic displacement of a  N  superoxide anion from the Fe atom:  F e - 0 + OH" -» [ H O — F e — 0 ] -+ Fe-OH + 2  The  three  state  component  involves  2  two  titratable  or  participate in what the authors call the "spontaneous" reactions: Mb0 (AH, BH) -• metMb + 2  Mb0 (A- BH) -» " 2  Mb0 (A- B~) - " 2  91  0 ~ 2  0 ~ 2  "prototropic"  groups,  which  The  values of the pK  for the two groups has been estimated as 6.75 and 10.4.  a  Based on the thermodynamic analysis of the pH dependence, Sugawara and Shikama (1980) identified the prototropic groups as a histidine and tyrosine residue respectively. authors  The  proposed the (distal) His-67 residue, as the specific His residue involved in  autoxidation and that it participates in hydrogen bonding or direct protonation of the Fe bound oxygen molecule in a reaction similar to the one described above.  The action of  the His residue is accompanied by a S 2 displacement of the oxygen molecule (as 0 ) by N  2  an incoming water molecule, which then becomes the aqueous ligand for metMb. et al. (1988) have prepared  a mutant Mb  Olson  in which the distal His residue has been  replaced by a Gly residue.  The resulting mutant exhibits a greater dioxygen dissociation  constant  protein but the effect of the amino acid substitution on  than the wild-type  autoxidation was not studied. The autoxidation rate for oxyDME-Mb at pH 7 and 25°C (0.026 hr" , ^ = 2 7 hours) 1  was  ~2.5 times greater than the corresponding  rate for oxyMb.  However, the observed  rate was much slower than that reported by Tamura et al. (1973a), who estimated the half-life of oxy DME-Mb as only 20 minutes (pH 7, room temperature). heme  propionate  negative  charge  through  esterification  electrostatic interaction with the distal His residue.  would  Loss of the inner  eliminate  any possible  As a result, there may be an acidic  shift of the His residue ptf , which would then increase protonation of the heme Fe a  oxygen ligand leading to an increased autoxidation rate. of the DME-Mb  However, a complete pH profile  autoxidation rate is required to demonstrate such an effect.  possible explanation could be a change in the geometry of the F e — 0  2  Another  bond from that of  native Mb, which is not favorable for electron transfer reactions such as autoxidation (Shikama, 1985). OxyMb prepared from tuna Mb, which has a ~3:2 ratio of heme rotational isomers (Levy  et al.,  1985), has a greater autoxidation rate than sperm whale or bovine Mb  92  species (Brown and Mebine, 1969).  OxyMb prepared from newly reconstituted sperm  whale metMb, which has a high degree of heme rotational isomerization, was observed to autoxidize faster than equilibrated oxyMb (Aojula et dl., that the minor heme rotational isomer, while having  1987).  These results indicate  no difference in 0  2  affinity from  the major isomer, may have a less stable Fe-0 bond with a resulting higher autoxidation 2  rate.  With metDME-Mb however, the *H NMR  results presented previously, even with  their limited resolution, do not show any degree of isomerization as high as that of tuna Mb. that  Unless the predominant rotation of the heme group in DME-Mb is the opposite of in native  Mb, the heme  rotational equilibrium  is not expected  to affect the  autoxidation rate of oxyDME-Mb.  D. Electrochemical Studies  The  initiation  simultaneously  of metMb  or metDME-Mb  reduction  and laccase activity  upon application of the potential across the OTTLE cell.  oxygen was still present in solution when metMb or metDME-Mb Reduction of metDME-Mb  occurred  As a result,  reduction was begun.  initially resulted in formation of oxyDME-Mb.  OxyDME-Mb  gradually formed deoxyDME-Mb presumably through dissociation of the oxygen ligand and subsequent laccase activity. formation  of oxyMb  However, metMb  as determined  reduction  did not result in significant  by periodic examination  of Mb  spectra  during  reduction. The  conditions used for electrochemical measurements of native Mb  those of DME-Mb  in that catalase was needed to prevent the formation  differed from of ferrylMb.  FerrylMb resulted from oxidation of either the ferrous or ferric states of Mb presumably by reaction with hydrogen peroxide (H 0 ) because of the inhibitory activity of catalase. 2  The  H 0 2  2  could  have been formed  2  from  incomplete reduction  93  of oxygen by either  electrode and/or laccase reactions. The during  described  conditions for the formation  the DME-Mb  ferrylDME-Mb  electrochemical  of ferrylMb  experiments.  in the spectra collected.  must have been present  However, there  was no sign of  Tamura et al. (1973a) observed that in the  presence of excess H 0 , the ferryl form of DME-Mb does not form as readily as native 2  2  Mb or other heme derivatives of Mb including monomethyl ester heme substituted Mb. The latter observation indicates that the resistance of DME-Mb to oxidation to the ferryl state could result from the esterification of just one of the heme propionates. observed formation  The  of oxyDME-Mb but not oxyMb within the OTTLE cell is another  indication of the difference in reactivity of these hemeproteins to external ligands such as H 0 2  2  and 0 .  The heme propionate esters may cause these differences by altering  2  the pathway(s) of 0  2  and H 0 2  2  entry into the heme pocket either by steric bulk alone  or more likely by disruption of hydrogen bonding and electrostatic interactions within the polypeptide chain. The measured midpoint reduction potential ( £ ) for native horse heart Mb was +60.9 m  mV (vs. SHE) at pH 7.0, /=0.1 M, NaPi buffer and 25°C. Under identical conditions, a value of +62 mV (vs. SHE) for the E  m  of sperm whale Mb was reported by Ellis (1986)  who used an anaerobic OTTLE cell flushed with nitrogen gas rather than an enzymatic oxygen reduction system. values  were obtained,  In spite of the difference in anaerobic techniques used, similar  which indicate that the small amounts of laccase and catalase  added to the OTTLE cell solution did not affect the measurement of the E DME-Mb.  The similarity in E  m  m  for Mb and  values also indicates that there is no significant species  (amino acid sequence) difference in electrochemical behavior of sperm whale and equine Mb. Comparison  of the present  results with  previously  published  results for Mb is  difficult because of the various solution conditions, techniques of protein purification and  94  electrochemistry that have been employed.  Using an OTTLE cell under an anaerobic  atmosphere, Heineman et al. (1979) measured an E  m  SHE)  at pH 7.0, 7=0.2 M  mediator.  for horse heart Mb of +46.4 mV (vs.  (temperature unreported) using phenazine  Taylor and Morgan  (1942) measured  the E  m  of horse  methosulfate as a heart Mb using  anthraquinone-/?-sulfonate as titrant with toluylene blue and cresyl blue as mediators to obtain a value of +46 mV (vs. SHE) at pH 7, 7=0.2 M and 30°C. Under similar conditions but with Fe(CN)  as a titrant and thionin as a mediator, Behlke and Scheller (1961)  6  Brunori et al. (1971) measured an E  obtained a value of +51 mV (vs. SHE).  m  for sperm-  whale Mb of +47 mV (vs. SHE) at pH 7, 7=0.1 M and 30°C using the method of mixtures with the same mediators as Morgan and Taylor. titrants (at concentrations higher than used  One problem with the use of dyes as  for mediators) however, is their potential  binding to Mb, an interaction that might influence the oxidation-reduction equilibrium. The midpoint reduction potential of DME-Mb was measured as +100 mV (vs. SHE) at pH 7.0, 7=0.1 M and 25°C. Mb  and DME-Mb  The 40 mV difference in reduction potential between native  under standard  solution  conditions indicates a role for the heme  propionate groups in the determination of the Mb reduction potential.  When Tamura et  al. (1973a) prepared DME-Mb they reported no significant change in its 0 that of native Mb.  However, the 0  2  affinity of DME-Mb  2  affinity from  may bear re-investigation  because of the formal relationship between oxidation and oxygenation In comparison to Mb, Reid et al. (1984) reported a 64 mV difference in reduction potential between native cytochrome b (+69  s  (E =+5 m  mV vs. SHE) and DME-cytochrome b  mV vs. SHE) under standard solution conditions.  cytochrome b  5  5  The low potential of the native  and the increase in reduction potential of the DME derivative has been  accounted for by a stabilizing coulombic interaction (over a distance of ~5.4 A) in the oxidized state of cytochrome b  s  Fe (vide infra).  between the inner heme propionate (P7) and the heme  Based on the crystal structure of horse heart metMb (Evans and Brayer,  95  1988), a carboxylic oxygen of the inner heme propionate is ~6.4 A from the heme Fe. salt bridge similar to that in cytochrome b  5  may also be present in Mb.  would be of lower potential (stabilization) than that of cytochrome b of separation.  An alternative explanation  Such a bond  due to the greater  5  distance  A  for the difference between the  reduction potentials of native and DME-hemeproteins is that the reduction potential is simply  a function of the presence and number of heme carboxylate groups, which exert  their influence through an inductive electronic effect. Using reduction  newly reconstituted cytochrome b , 5  potential of the mixture of rotational isomers and from  estimated a 27 mV difference in the E  m  titration  Walker et al. (1988) have measured the  of a mixture  extrapolation have  (7=0.13 M, pH 7.0 and 24°C).  of hemeproteins with  expected to exhibit non-Nerstian behavior.  different  reduction  An electrochemical  potentials would be  Analysis of such a titration using Equation 19  would give values for the slope different from expected values for integer (n) electron transfer.  Results  electrochemical isomerization.  of this  measurements  type of  have  not been  hemeproteins  reported  known  to  in previous have  heme  or current rotational  This absence may be due to either a very small or large difference in  reduction potential between isomers or insensitivity in the techniques  employed in such  titrations to the contribution of the minor isomer.  7. pH Dependence  The measured midpoint reduction potentials for both native Mb and DME-Mb decrease in E  with increasing pH from pH 6 to about 9 (Figure 15).  m  the E  m  Previous measurements of  by Brunori et al. (1971) for sperm whale Mb and Behlke and Scheler (1961) for  equine Mb have shown a similar pattern with a further decrease in E  m  pH.  at more alkaline  This relationship between the pH and reduction potential has been termed the  96  oxidation Bohr effect (Brunori et al., 1971).  The decrease in E  attributed to the ionization of the water ligand coordinated  m  at alkaline pH is  to the heme iron.  The  hydroxyl anion presumably negates the net positive charge of the heme Fe(III) to produce an overall neutral ferric form and as a result the reduction potential of Mb decreases. The observed decrease in reduction potential from pH 6 to 7.5, which is greater for DME-Mb  than native Mb, is inconsistent with  dependent determinant of the reduction potential.  the water ligand being  the only pH  The pAf values for the reduced and a  oxidized states of both native and DME-Mb derived from fitting the pH dependence of E  m  are all less than the pAf values for the water ligands determined from spectroscopic a  titrations.  This non-correspondence between electrochemical and spectroscopic transitions  is not uncommon as discussed by Wilson (1978). is that visible  spectra only  The reason for this non-correspondence  the electronic transitions of the group(s) (e.g.  represent  heme) that undergo changes in oxidation state.  As a result, the visible spectrum is  insensitive to perturbations in other parts of the protein such as pH linked conformation changes that indirectly affect the oxidation states. Analysis of the change in reduction potential with pH by Equation 20, which is based on  the assumption of a single redox-linked titratable group, appears to have fitted the  observed data very well (Figure 14).  However, the contribution of the heme bound water  to hydroxyl transition in met(DME-)Mb has not been taken into account. not all of the change in E  m  group.  The values  overestimated from  As a result,  within the pH range examined is due to a single titratable  for the pK  tei  and pK  0X  for the titratable group are probably  owing to the additional decrease in reduction potential (with increasing pH)  the heme bound hydroxyl  anion.  The lower pAT  a  of (aquo)metDME-Mb (8.5)  compared to that of native (aquo)metMb (8.9) indicates that a larger proportion of the change in E  m  for DME-Mb in the neutral to alkaline pH range was due to the heme  bound hydroxyl than in native Mb.  97  The  one titratable group model of Equation 20 has been used to analyze the pH  dependencies of the E  m  for a variety of cytochromes such as b  5  (Reid et al.,  551 (Moore et al., 1980) and some species of c (Pettigrew et al., 1978). 2  1982), c-  This model is  suitable for die pH dependence of the reduction potential of these cytochromes because the heme group of these proteins do not normally bind external ligands such as water. The  diree ionization model represented by Equation 21 is more appropriate for Mb and  DME-Mb because it accounts for the presence of the heme boimd water/hydroxyl ligand in the oxidized state of Mb. From  Equation 21,  the  pAT  and pAT  ol  r  respectively for native Mb. The pK  &  measured by  J  values  were  calculated  as  7.4  and  values for some histidine residues in Mb have been  H NMR titrations for various species" of Mb (Carver and Bradbury,  Bradbury and Carver, 1984).  7.7  1984;  Three out of the eleven His residues of Mb are within the  heme pocket as determined from the three-dimensional stmcture of horse heart metMb (Evans and Brayer, 1988).  These residues are His-64 (proximal Fe ligand), which does not  titrate in the pH range studied, His-94 (distal Fe ligand) and His-97, which is hydrogen bonded to the inner heme propionate.  The measured pK  and Carver for His-97 (5.54) is less than the pK Another  candidate  propionates. cytochromes  for  the  titratable  In studies examining c  2>  Moore  et al.  a  values reported by Bradbury  measured here for native Mb.  ol  group  in  native  the pH dependence  of  Mb is the  one  of  the  heme  reduction potential of  (1984) proposed a role for the protonation of  heme  propionates in regulation of the oxidation-reduction equilibrium of these proteins.  From  correlation with *H NMR studies these workers conclude that the variation of E  with  m  pH for some species of cytochromes c that exhibit a large (~60 mV) change in E 2  with  m  pH) is attributable to the titration of a heme propionic acid group.  The proposed pK  R  values for these groups were greater than that of heme propionates in solution (pK &4.5; a  Falk, 1964).  The increase in heme propionate pK when within hemeproteins was a  98  Figure 24.  Stereo drawing of selected amino acid residues in the heme pocket of (aquo)metMb determined from the 2.8 A model of Evans and Brayer (1988). The heme is viewed edge on parallel to its a—7 axis. Shown are water molecules, amino acid residue Serine-92, histidine residues within the heme pocket (at positions 64, 94,and 97), and the lysine residues at positions 45 and 96, which flank the heme pocket opening.  99  attributed to the effect of the hydrophobic heme pocket and hydrogen bonding Within Mb, the inner heme propionic acid (P7) may have a pK  residues.  with His  value within  a  this range because the three-dimensional structure of metMb (Evans and Brayer, 1988) indicates  possible hydrogen  bonding  between  this  propionate  and the side-chains of  residues His-97 and Serine-92. For DME-Mb, the calculated pX^ and pK 21  were 6.3 and 7.5 respectively.  values generated by the fit to Equation  r  If the titratable group for native Mb is a heme  propionate, a different group or groups must be titrated in DME-Mb. cytochrome c , pH dependent changes in E 2  m  In some species of  have been attributed by Moore et al. (1984)  to Wstidine residues outside the heme pocket, which can induce conformational changes sufficient to alter the heme environment. have pK  Some of the surface His residues of metMb  values (Bradbury and Carver, 1984) within the range of the pK  a  ox  calculated for  DME-Mb.  2. Thermodynamics  The  standard entropic energy  change (AS ) measured for the reduction of native 0  horse heart metMb to deoxyMb (—38.2(5) eu at pH 7.0 and 7=0.1 M) is in close agreement with the corresponding A S (1986).  0  value measured for sperm whale Mb (—39(1) eu) by Ellis  With myoglobin, the loss of the heme bound water ligand upon reduction would  be expected to result in a positive change in entropy. indicates that there are other processes  That the measured A S  associated with  5  is negative  reduction that have negative  entropy changes e.g. conformational changes in the protein. the reduction of cytochrome b  0  Interestingly, the A S  0  for  (AS°=—37 eu at pH 7 and 7=0.1 M; Reid et al., 1982),  which does not undergo a change in heme Fe coordination upon reduction, is similar to that of Mb.  100  A negative net change in entropy is commonly measured for the reduction of many hemeproteins because hemeprotein conformations are believed to become more "rigid" and "compact" upon reduction (Taniguchi et al, accompanying  1980).  The increase in entropy from the  release of bound water molecules is presumably  offset by the decreased  entropy from solvent-solvent interactions (i.e. hydrogen bonding).  Cytochrome c is one  of the few hemeproteins whose structure has been determined by x-ray crystallography in both  the oxidized  and reduced  states.  Comparison  of the two structures revealed  apparently small differences in conformation (Takano and Dickerson, 1981a & b). solution  however, reduced  cytochrome  c  is more resistant  In  to denaturation than the  oxidized form (references in Dickerson and Timkovich, 1970), indicating that the former form is more stable.  From comparison  of the structures of sperm whale met- and  deoxyMb, Takano (1977a & b) has also reported differences in the conformations of the polypeptide chains and heme.  In neither case has an estimate been made for the entropy  changes arising from any observed redox-linked conformational changes. The entropy change (—51.0(8) eu at pH 7) calculated for the reduction of metDME-Mb to deoxyDME-Mb is greater than that for the reduction of native metMb (—38.2(5) eu). This finding may indicate a greater difference in relative stability between the oxidized and reduced states of DME-Mb than those of native Mb, which may be a reflection of a greater flexibility in the conformation of DME-Mb  than that of native Mb.  This  possibility is supported by the observed greater affinity of apoMb for reduced DMEheme over metDME-Mb.  Also, in this current study, it was observed that metDME-Mb is  less stable than native metMb at high temperature (>35°C, pH 7), as measured by their respective Soret absorbance values over time. The enthalpic change for DME-Mb (—12.9(2) kcal/mol) is lower than that of native Mb (—8.1(2) kcal/mol).  This difference may arise in part from the difference in enthalpy  between a Fe-H 0 and a Fe-OH bond because the pATa of the heme coordinated water 2  101  ligand of DME-Mb (8.5) is lower than that of native Mb (8.9).  Though at pH 7 the  percentage of DME-Mb bound with the hydroxyl ligand is low (~3%) and is not much greater than that of native Mb (~1%).  3. Electrostatics  Table I shows that the calculated net charges of Mb and DME-Mb derived from the use of full protein radius values are both much higher than the net charge of +5.5 based on the amino acid composition and heme charges.  At neutral pH, horse heart metMb has  an apparent net charge close to 0 based on electrophoretic mobility 1974; McLellan, 1984).  (Romero-Herrera,  The equations used in ionic strength dependence analyses are  based on the original work by Kirkwood  and Tanford (1957) who used the following  assumptions: 1) spherical proteins, 2) solvent impenetrable spheres, 3) random (smeared) charge distribution. Though Kirkwood and Tanford in the same paper called these assumptions "invalid" for proteins, Equation 25 has been used with apparent success to analyze the ionic strength  dependence  of the reduction  potential  for various  cytochromes c . (Margalit and Schejter, 1973) and b  s  characteristics  that  hemeproteins  such as  (Reid et al., 1982). Clearly Mb has  do not agree with the above assumptions.  Based  on its three  dimensional crystal structure, the overall shape of Mb is more elliptical than spherical. Also the location of the heme within Mb is more peripheral than central. bound water ligand demonstrates that Mb has some solvent accessibility. molecular weight  (size) of Mb  relative to cytochromes  effects of individual charged groups.  102  c would  The heme The higher  tend to localize the  From Table I, it can be seen that the calculated net charge decreases to seemingly more reasonable  values with the use of decreasing  protein radius values.  This trend  indicates that ions can approach the heme iron center more closely than the radius of Mb would indicate.  For Mb this is not unreasonable considering that the heme Fe does  bind water and other ligands.  As a result, the function /(7)=v //(l+v /) may be the most /  /  appropriate to describe the ionic strength dependence of the reduction potential of both native and DME-Mb.  The choice of the function/(/)=v //(l+v V) /  /  is supported by the work  of Shire et al. (1974) who studied the effect of ionic strength on the pK bound water ligand of metMb.  a  of the heme  Shire et al. found that use of the function f(I) =s/ll(l +\fl)  fitted their ionic strength dependence better than the function f(I)=\/l/(l + 6\/V), which uses an estimate for the full radius of Mb. The  term BR  /(/)=v //(l +V /). /  7  a  in Equation  26 is assigned a value of one to derive the function  One interpretation of this value is that local charges within a small radius  (R ) around the heme determine the E a  Schejter  m  of Mb.  In an alternative view, Goldkorn and  (1976), have interpreted the use of this function as an indication that the  medium (solvent and ionic strength) around the heme is the main determinant of the As Moore et al. (1986) have stated, all of these aforementioned  reduction potential. factors  influence  the reduction  protein to protein. presumably  potential though  their  relative  influence  In the case of DME-Mb, esterification of the heme propionates would  decrease the negative  (or increase the positive) charge around the heme,  which would tend to raise the reduction potential of the hemeprotein. sensitivity of the E  varies from  m  In addition, the  to the ionic strength of the medium has also been increased as  shown by the larger change in reduction potential (AE =65 mV) for DME-Mb over the m  ionic strength range examined (0.05-0.5 M) as compared to native Mb (AE =33 mV). This m  increase  in sensitivity  to ionic  strength  may indicate greater heme exposure  solvent.  103-  to  the  E. Reduction Kinetics  The  presence of ferricyanide resulted in a biphasic reaction between metDME-Mb and  Fe(EDTA) " with A A 2  4 0 7  vs. time (Figure 18) showing an initial rapid phase of reaction  (decrease in A ^ ) followed by a slower phase. more evident  in the corresponding logarithmic  The biphasic nature of the reaction is plot in Figure  (18).  After passage of  metDME-Mb through the Dowex X-8 column to remove any associated ferricyanide, the observed reaction tracing (Figure  with Fe(EDTA) " became monophasic as shown in the A A 2  18) and the corresponding logarithmic plot.  4 0 7  time  Otherwise, the average rate  of the biphasic reaction was about an order of magnitude lower than that observed after ferricyanide removal. Free ferricyanide in solution could have oxidized Fe(EDTA) " thereby decreasing its 2  effective concentration,  which would  result in the slower reaction  phase.  However,  buffer exchange of the protein solution should have removed any free ferricyanide.  The  protein solution itself (<50 uL) was diluted into 20 mL of buffer for use in the kinetic experiments.  In addition, the concentrations of Fe(EDTA) " used were always at least an 2  order magnitude greater than the concentration  of metDME-Mb.  It is therefore unlikely  that any ferricyanide in solution could have caused the observed biphasic kinetics. These results indicate that ferricyanide could metDME-Mb. Mb.  have been bound (non-covalently) to  This is interesting because no such binding has been reported for native  However, ferricyanide has been known to bind ionically to hemoglobin (e.g. Hoffman  and Bull, 1976; Linder et al,  1978) and cytochrome c (Stellwagen and Cass, 1975). If  metDME-Mb was bound with ferricyanide, the following mechanisms could account for the biphasic kinetics: 1) steric hindrance by Fe(CN)|" of the Fe(EDTA) " reaction with heme Fe 2  2) electrostatic repulsion of Fe(EDTA) " by the negatively charged Fe(CN)^" 2  104  3) increased reduction potential of metDME-Mb bound with Fe(CN)|" 4) intramolecular transfer between Fe(CN)^" " and heme Fe(III / II) /3  In the first metDME-Mb.  two cases, the initial  rapid rate would be due to reaction of unbound  In the third case, ferricyanide binding could have resulted in an activated  state-like complex in which the potentials of the two metal centers have changed to a level  intermediate  between  their  unbound  (free) states.  ferricyanide bound metDME-Mb would increase.  The reduction  potential of  As a result, the driving force of the  reaction would become more favorable, which would lead to the higher initial rate. In  the last  mechanism  listed,  ferricyanide  may  have  sterically  Fe(EDTA) " reduction of the metDME-Mb heme iron but not its own.  The resulting  2  ferrocyanide  could  then reduce metDME-Mb.  Such an intramolecular  blocked the  reaction can be  depicted as:  Fe(EDTA) " + Fe(III)(CN) -DME-Mb Fe(III) -» Fe(II)(CN) —DME-Mb Fe(III) + Fe(EDTA) " 2  1  6  6  (33) Fe(II)(CN) -DME-Mb Fe<III) - Fe(III)(CN) -DME-Mb Fe(II) 6  (34)  6  The initial rapid rate observed could be due to reaction of unbound metDME-Mb while the  slower  rate  could  reflect  Alternatively, if metDME-Mb  the rate  limiting  was saturated  step  with bound  could be due to the forward reactions depicted  in Equations  33  and 34.  ferricyanide, the initial phase  in Equation 33 and the slower phase  could be due to the contribution of a back reaction electron transfer from deoxyDME-Mb to ferricyanide, which would be favored based on the difference in reduction potential of deoxyMb and ferricyanide.  105  The  aforementioned possibilities imply  ferricyanide binds  specifically  to DME-Mb.  The crystal structure of native horse heart metMb (Evans and Brayer, 1988) shows that Lys residues 45, which flanks the opening of the heme pocket, forms a salt bridge widi the outer heme propionate  (P6).  Esterification of the propionate  would eliminate its  negative charge and may allow the positively charged Lys residue to bind a ferricyanide anion.  In cytochrome c, ferricyanide is believed to bind a lysine residue near the heme  edge (Stellwagen and Cass, 1975; Eley et al., 1982). The  rate of Fe(EDTA) " reduction metDME-Mb 2  was always greater than that of  metMb under all conditions of temperature, pH, and ionic strength examined. at  all concentrations  completion  of Fe(EDTA) " used the reduction of metDME-Mb 2  in contrast  to the incomplete  concentrations (Lim and Mauk, 1984).  reduction  of metMb  In addition, proceeded to  at low Fe(EDTA) " 2  There are three general reasons for the increase  in reaction rate with heme propionate esterification, all of which are probably significant. First, the driving force for the reaction is more favorable as indicated by the increase in reduction potential of DME-Mb compared to native Mb. between Fe(EDTA) " and metMb are probably 2  propionates  and elimination of the negatively  Second, electrostatic interactions  enhanced by esterification of the heme charged  carboxy late  groups.  As an  example, esterification of the outer heme propionate would release the positive charge of Lys-45  and result  in increased  electrostatic attraction of an approaching  anion.  Finally, the actual mechanism of electron transfer (i.e. pathway) from Fe(EDTA) "  2  2  to the heme iron of metDME-Mb faster electron transfer. heme propionate  Fe(EDTA) "  may be altered (from that in native Mb) to allow  For example, the loss of H-bonding and ionic interactions by  esterification may allow  greater conformational  than native Mb and result in greater heme access by Fe(EDTA) ". 2  106  flexibility  in DME-Mb  1. pH Dependence  The  rate  constant  (t ) 1 2  of metDME-Mb  decreases with increasing (more alkaline) pH. kinetic and thermodynamic.  in reaction rate constant  origin (Figure 21).  by Fe(EDTA) " 2  (Figure 20)  The causes of this behavior are both  Through correction of the second order rate constants for  driving force (AEm), the thermodynamic change  reduction  contribution can be eliminated.  The residual  (Jfcf^) with pH can then be attributed to a kinetic  To explain this change in rate with pH a single titratable group was  assumed to be responsible.  This assumption does not imply that other titratable groups  cannot influence the reduction rate, only that within the narrow pH range studied (6 to 8) one group has a dominant influence. The change in rate constant  observed with changing pH can be attributed to die  properties of DME-Mb and/or Fe(EDTA).  Within the pH range studied, Fe(EDTA) " does 2  not have a titratable group, while Fe(EDTA) " has a bound water molecule with a pAT 1  a  value  of 7.5 (Schwarzenbach and Heller, 1951; Kolthoff and Auerbach, 1952), which  influences the E  m  force  of the Fe(EDTA) " " couple. 1  of the reaction  /2  was corrected  However, this effect on the driving  for in calculating ifcf^.  Fe(EDTA) " was used in the reaction with metDME-Mb. 2  Furthermore  only  As a result, the cause of die  observed pH dependence of the reaction rate probably resides in properties of metDMEMb and not Fe(EDTA) ". 2  This  conclusion  dependence  is reinforced  of Fe(EDTA) " reduction 2  by  the substantial  difference  of native and DME-Mb  between  the pH  (Figure 20).  The pH  dependence of native horse heart metMb reduction by Fe(EDTA) " has been measured and 2  analyzed  (Lim and Mauk, 1984) by similar methods to those used for metDME-Mb.  Esterification of the heme propionates results in two major effects on the pH dependence of the rate constant.  MetDME-Mb appears to have a titratable group with a p/ir value a  107  of 7.6, while esterification  native metMb has a group with  of the heme  propionates,  a lower pK  a  value  of 5.8.  With  the influence of pH on the reduction  rate  constant increased. In native Mb, the outer heme propionate (96), which is directed towards the solvent, may be the titratable group.  Based on the crystal structure of horse heart Mb, Evans  and Brayer have proposed that the outer heme propionate forms a salt bridge with Lys45.  This interaction would negate the positive charge of the Lys residue.  propionic acid groups in free heme is only 4.5 (Falk, 1964). of cytochrome b  5  have estimated  The pK of a  However, the *H NMR data  a pAT value of 5.9 for the solvent accessible outer a  heme propionate of this protein (McLachlan et al,  1986).  The identity of the titratable  group responsible for the pH dependence of reduction by Fe(EDTA) " of metDME-Mb 2  cannot be determined unambiguously because the pAT values for individual amino acid a  residues in DME-Mb are unknown.  The pAT value of 7.6, produced by fitting the rate a  data to Equation 30, is suggestive of a histidine residue. With heme propionate esterification not only was the rate of reduction increased, the The difference between (adjusted) k and  sensitivity of this rate to pH also increased. k  a  increased from ~150 molds' for native Mb to ~1560 m o l ' V 1 for DME-Mb in going 1  b  from pH 6 to 8.  The heme propionates  may be involved in a rate limiting step of  electron transfer such as binding of Fe(EDTA) ". 2  After esterification, the influence of a  titratable group on the reaction rate appears to have been unmasked, accounting  for the  increase in rate and sensitivity of the reaction to pH.  2. Thermodynamics The enthalpy of activation (AZ/*=+ 9.2(3) kcal/mol) for the reduction of metDME-Mb by Fe(EDTA) " is less than that measured for native metMb (+12(1) kcal/mol; Cassat et 2  al, 1975).  The entropy of activation (AS* =—13(1) eu) for the reduction of DME-Mb is 108  identical to that reported for native Mb Cassat et al., from Equation  1975).  in its reaction with Fe(EDTA) " (—13(5) eu; 2  Interpretation and comparison of the activation parameters derived  31 are difficult because one cannot distinguish between  the individual  contributions of reactants. For  comparison, the enthalpy  cytochrome b respectively  and entropy of activation for the reaction  and Fe(EDTA) " at pH 7.0 and 7=0.1 M 2  s  (Reid  and Mauk,  1982).  Under  between  are 5.4-kcal mol" and -29.2 eu 1  the same conditions,  the reaction of  cytochrome c with Fe(EDTA) " is characterized by corresponding activation energies of 5 2  kcal mol" and -20 eu respectively (Hodges et al., 1974). 1  lower than those of metMb  and metDME-Mb  These activation enthalpies are  presumably  because these cytochromes  undergo no change in heme coordination upon reduction.  3. Electrostatics  Analysis of the ionic strength dependence of metDME-Mb reduction by Fe(EDTA) " 2  with  Equation  32 using  full  protein radius values results in an unrealistically  calculated charge for the DME-Mb molecule.  large  Such results are analogous to the large  charges calculated from analysis of the electrostatics dependence potential when full protein radius values were used.  of the Mb  reduction  These inconsistencies are similar  because the assumptions used by Wherland and Gray (1976) in deriving Equation 32 were also used in deriving Equation 25 for the ionic strength dependence reduction potential e.g.  of the midpoint  spherical proteins with symmetrical charge distribution.  assumptions also apply to Fe(EDTA) ", which is not a symmetrical complex. 2  These  Furthermore  there is some controversy over whether the net charge of a protein (expressed by use of full protein radii) or the local (active site) charges control its electrostatic behavior in reactions of this type.  109  However, despite these assumptions, the ionic strength dependence relationship shown in Equation  32 has been used successfully for the analysis of many  especially cytochromes.  metalloproteins,  Calculated charges derived from use of the protein radius have  generally agreed with those derived from sequence.  Based on these results, Feinberg and  Ryan (1981) have concluded that the net charge rather than local charges influences the electrostatic behavior structural  of electron transfer proteins.  features of the cytochromes analyzed  This success  by Feinberg  may be due to the  and Ryan, which when  compared to Mb are generally smaller (lower molecular weight) and more spherical (better agreement  better  Consequently  between  hydrodynamic  and  crystal  structure  radius  values).  the properties of these cytochromes are closer in agreement  with the  assumptions of Equation 32 than are those of Mb. The  apparent success in estimating metalloprotein charges derived from ionic strength  dependencies of reduction potentials is based on their comparison to charges derived from the amino acid and heme composition charge of the assembled protein. agreement  between  coincidental.  sequence  of the protein, which may not reflect the actual  Marcus and Sutin (1985) have pointed out that the  derived  charges  and  calculated  net charges  may  be  If their signs are identical, the sequence derived charge and the local  charge may appear to coincide because they are usually within an order of magnitude of one another.  Segal and Sykes (1977) have questioned  whether ionic strength dependence  equations such as 32 can be applied to reactions involving proteins. Regardless of the view taken (or values used for the radius), Equation the  observed  ionic  metalloproteins.  strength  dependence  of the reaction  32 does fit  kinetics of many  diverse  Equation 32 provides an estimate of the electrostatic contribution to the  reaction kinetics and is particularly useful for comparison of the kinetic behavior of modified  proteins  with  differences involved.  those of the native  form  because  of the smaller structural  The current results for DME-Mb show that esterification of the  110  heme  propionates  significantly  alters the electrostatic behavior  of Mb.  Given the  discrepancy between the calculated and sequence charges of DME-Mb, the local charges around the heme may exert more influence on the kinetics of Mb  than does the net  charge of the protein. The  electrostatic behavior of a metalloprotein, as reflected by the reaction kinetics  (and i ? ) , may be influenced more by its net dipole moment rather than net charge. If m  all the charged groups of a protein contribute to its electrostatic properties, then the net dipole moment may be a more accurate representation than the net charge, especially for a large molecule such as a protein.  Koppenol and Margoliash (1982) have calculated  the dipole moment of horse cytochrome c as being of surprisingly high magnitude (~300 Debye) with the positive end of the vector electron transfer is theorized to occur.  directed towards the heme edge, where  Such an orientation is favorable for interactions  with reactants that have a negative charge such as Fe(EDTA) ". 2  Van Leeuween et al.  (1981) have used the results of Margoliash and Koppenol to analyze the ionic strength dependence of cytochrome c reaction kinetics based on its dipole moment rather tiian net charge. As dipole  dipole calculations of this type has not previously been reported moment  of native  horse  heart  developed by Northrup et al. (1986).  Mb  was calculated  using  for Mb, the  computer programs  This method calculates partial charges and dipole  moments for all the atoms in a protein and then derives an estimate for the net dipole moment.  The calculated dipole moment magnitude of cytochrome c was 245 or 286 Debye  depending on whether formal charges were assigned to the heme propionyl groups.  The  Evans and Brayer (1989) model of horse heart metMb (1.9 A resolution) was used as a source of coordinates  for the dipole moment calculation.  The charge of each heme  propionyl group was set at +0.5. This Mb model was also used in an attempt to simulate the effect of heme propionate esterification by setting the charge of the heme propionyl  111  Figure 25. The dipole moment for native Mb estimated using the methods of Northrup et al. (1986). The vector calculated for native Mb is labelled Mb and is depicted in relationship to a stereo drawing of a model of Mb (Evans and Brayer, 1988). The vector of the calculated dipole moment of Mb after elimination of the charge contribution of the heme propionates is labelled Mb'.  112  groups as zero. The dipole moment calculated for native Mb was 2.8 x 10 Debye with the positive 2  end of the vector pointed away from the heme (see Figure 25) and exiting the surface of the protein near residue glutamine-91. and  This result was surprising because the magnitude  direction of the estimated dipole moment was similar to that of cytochrome c.  Elimination  of the heme  propionates' charge  contribution  (to simulate esterification)  increased the magnitude of the estimated dipole moment to 3.9 x 10 Debye. 2  In addition,  the orientation of the dipole moment vector becomes slightly more parallel to the heme plane and exits the protein surface near residue Serine-92.  Elimination of the heme  propionate charges by esterification may also result in a similar change in dipole moment for DME-Mb assuming no major structural changes from that of native Mb.  Such a  dipole change could explain the increased reactivity of DME-Mb towards anions such as Fe(EDTA) " and Fe(CN)f\ 2  There may be a role in electron transfer for both the net dipole moment and local charges as described by Northrup et al. (1988). dipole  moment  configuration  attracts or steers  i.e. the precursor  reactants  Reports from this group suggest that the  together  complex.  This  in a nonspecific  initial bmding  attainment of the reactive state via reorganization  and unreactive  allows more efficient  of the precursor  complex  through  involvement of shorter range electrostatics interactions.  F. Marcus Theory Analysis  The  values of A^j" calculated using  different values of the protein  radius and  corresponding calculated charge were all within an order of magnitude of each other. Similar  results were obtained  by Wherland  and Gray  (1976) for the reduction of  cytochrome c by a variety of inorganic complexes including Fe(EDTA) ". 2  113  These results  indicate  that  the Wherland-Gray  adaptation  of Marcus  theory  is at least valid for  comparison purposes. As developed by Gray and co-workers (Wherland and Gray, 1976; Cummins and Gray, 1977),  the  electrostatics-corrected  self-exchange  rate  (*i )  for a  orr  metalloprotein  calculated on the basis of its cross-reaction with a substitutionally inert transition metal complex is a useful kinetic parameter to detect mechanistic basis for the usefulness  features of a reaction.  of k^° relative to the second order rate constant  The  resides in  n  the elimination of contributions to this rate of the thermodynamic driving force of the reaction,  the electrostatic  interactions between  protein  intrinsic reactivity of the small molecular reagent.  and small  The A^  orr  molecules  of 0.41 M ' V  and the  at 7=0.1 M  1  calculated for DME-Mb (using full radius value and corresponding calculated charge) is  k\  OTT  larger by an order of magnitude than the calculated  of native Mb, 0.020 M ' V . 1  This difference is significant in that it exceeds the uncertainties involved in measuring and  calculating the various parameters used in their calculation (Appendix E).  result  indicates  reduction  by  that  esterification  Fe(EDTA) " 2  by  of the heme  altering  propionates  the mechanism  increases  of electron  Fe(EDTA) ' and the heme iron to a more favorable pathway. alteration  may  involve  disruption  of the hydrogen  the rate of  transfer between  As discussed previously  2  this  This  bonds  and  electrostatic  interactions of the heme propionate, which may alter the conformation of the protein. At pH 7, 7=0.1 M native  and DME-Mb  ^ =4.3 M ' V orr  1976). axial  1  and 25°C, the electrostatics corrected self-exchange rates for  are both  significantly  smaller  than  those  (Reid and Mauk, 1982) and c, *f =6.2 M ' V orr  1  for cytochrome  b, 5  (Wherland and Gray,  These differences presumably arise from the activation barrier to the change in ligation of the heme Fe that occurs on reduction  of metMb  to deoxyMb as  cytochromes c and b undergo no such redox-linked change in axial ligation. 5  114  CONCLUSIONS  A. Emerging Role of Heme Propionate Groups  In the course of hemeprotein evolution protoheme DC has been used as a prosthetic group in proteins with a wide range of functional properties.  In principal, the heme  propionic acid groups can form specific H-bonds and as propionates can form specific salt bridges either of which can be intra- or intermolecular.  As discussed above, heme  propionate groups in hemeproteins have been shown to function in structural roles, in regulation of reduction potential and in influencing mechanisms of electron transfer. In addition, heme propionates appear to be essential for formation of the initial heme-apoprotein complex of non-covalent  (fc-type) hemeproteins possibly by directing the  heme to the correct position within the heme pocket via hydrogen bonds with amino acid residues.  Once positioned, the heme is held inside the pocket by hydrophobic and Van  der Waals interactions involving the heme vinyl and methyl groups.  The propionates may  interact with the net positive charge of the ferric iron via coulombic thereby  modify  propionates through  the reduction  may  allow  potential of the hemeprotein.  their participation in complex  electrostatic/dipole  interactions  and  thereby  interactions and  The arrangement of the  formation facilitate  with  other  electron  proteins transfer.  Examples of these mechanistic roles for heme propionates are given in the following brief summary.  B. Myoglobin  In the formation of the apoMb—heme complex only one of the heme propionates appears  essential because of the relative ease in preparing the monomethyl ester heme  derivative of Mb  compared  to the dimethyl ester derivative  115  (Tamura et al,  1973a).  Esterification of the heme propionates decreased the stability of the resulting metDMEMb compared to native Mb. with small molecules. lower  in DME-Mb  oxy DME-Mb  The heme propionates can influence the reactions of Mb  For example, the pAT of the water ligand bound to the heme Fe is a  than native  Mb  and the autoxidation  relative to native oxyMb.  reaction is accelerated in  Finally, esterification of the heme propionates  apparently allows high affinity binding of ferricyanide anion to DME-Mb. The reduction potential of Mb is increased by 40 mV (pH 7, 7=0.1 M, NaPi buffer, 25 °C) with heme propionate esterification.  This increase may be due to the loss of the  propionate negative charge stabilizing influence on the positive charge of the heme Fe. Analyses of the ionic strength dependence of both the reduction potential and second order reaction rate constant  with Fe(EDTA) " showed that the propionates influence the 2  apparent net charge of Mb.  The pH dependence of the reduction potential and second  order reaction rate constants  showed that esterification increases the sensitivity of these  parameters to changes in pH and that the calculated pA^s of titratable groups were altered. met  Temperature dependence studies showed a decreased stability of the oxidized  form relative to the reduced deoxy form of DME-Mb.  From application of Marcus  theory, it is apparent that the corrected self-exchange rate of DME-Mb greater than that of native Mb.  is significantly  This result indicates that the heme propionates also  influence the pathway or mechanism of electron transfer exhibited by Mb.  C. Cytochrome b  5  Cytochrome b  s  et al, metHb  is a hemeprotein that functions in fatty acid reduction (Strittmatter  1974), the cytochrome P-450 catalytic cycle in liver (Bonfils et al, reductase  system  microsomal cytochrome b  5  of the erythrocyte.  The three-dimensional  1981) and the structure of  reported by Mathews and Argos (1975) indicates that the heme  116  propionates in cytochrome b  5  the electron density  are oriented in a manner similar to those of Mb.  difference map between  interaction of 5.4 A heme Fe). £  5  b,  the oxidized and reduced cytochrome  Argos and Mathews (1975) proposed that within propionate has a direct stabilizing  Based on  influence  b,  ferricytochrome  on the heme  5  the inner heme  5  Fe(III) via a coulombic  separation (from an oxygen atom of the propionate group to the  The increase in reduction potential of DME-cytochrome b  over that of native  5  can be attributed to the loss of this stabilization (Reid et al, 1984). The  proteins.  outer  heme propionate influences the binding  The increase in rate of reduction of DME-cytochrome b  that of native b  5  propionate  of small  molecules and other 5  by Fe(EDTA) * over 2  (Reid et al, 1984) has been attributed to both a loss of the heme  electrostatic  repulsion  of the Fe(EDTA) " 2  transfer pathway (as represented by * i ) .  in the electron  Salemme (1975) has proposed a structure for  o r r  the complex formed between cytochrome b  s  heme propionate of the b .  and a change  and cytochrome c, which involves the outer  Analysis of the binding between cytochrome c and DME-6  s  (Mauk et al, 1986) has indicated that the outer heme propionate is integral to formation of the Salemme complex but not other possible complexes. reduction  kinetics (with  b —cytochrome 5  c  various  complexes  flavins) between  has also  been  the cytochrome  attributed  b —c and  DME-  s  to the effect of propionate  5  5  and Mauk, 1983) and *H NMR  the  The differences in  esterification on the structure of the cytochrome b -c complex (Eltis et al, 1988). outer heme propionate of cytochrome b  5  has also been implicated by modelling  The  (Poulos  (Livingstone et al, 1985) studies as being involved in  complex formation with metHb and metMb respectively.  117  D. Cytochrome c  Cytochrome oxidation.  c is a component of the electron transport chain of phosphorylative  The three-dimensional  structures for several cytochromes c from  species have been solved (review by Mathews, 1985).  various  In cytochromes c, the heme group  is joined covalently to the polypeptide chain by thioether linkages formed between the heme vinyl groups and (conserved) cysteine residues. the heme propionates  of cytochrome c probably  Unlike the 6-type hemeproteins,  do not have a role in the initial  heme—apoprotein complex, which is believed to be formed by a heme attachment enzyme (Pettigrew and Moore, 1987). In the heme pocket of eukaryotic cytochrome c, the heme propionate directed inward, away from the solvent.  groups are  They contribute to the final conformation of  the heme protein, which is tightly folded, by forming hydrogen bonds with amino acid residue side chains from different helices of the polypeptide chain. c, one propionate  In tuna cytochrome  (P7) forms a salt bridge with an arginine (position 38) residue and  hydrogen bonds with the sidechains of a tyrosine (48) and a tryptophan  (59) residue.  The other propionate (P6) hydrogen bonds to two threonine residues (48 and 78).  The  role of the propionate P7 may be to moderate the positive charge on the arginine-38 residue on the reduction potential of the heme iron of cytochrome c.  Replacement of  this arginine residue with other amino acid residues by site directed mutagenesis results in as much as a 50 mV  reduction of the reduction potential from that of the wild-type  protein (Cutler et al, 1989).  118  E. Cytochrome c Peroxidase  Cytochrome mitochondrial 1976).  c  peroxidase  (CCP) is a  monomelic,  monoheme  (A-type)  protein that catalyzes oxidation of cytochrome c via peroxides  CCP has a i-type heme, which is non-covalently  bound  yeast  (Yonetani,  to the apoprotein.  Yonetani (1964) prepared the dimethyl heme derivative of CCP, DME-CCP, and found its activity to be less than 1% of native.  The three dimensional structure of CCP (Finzel et  al., 1984) shows that the heme propionates in this protein are directed towards the heme pocket  opening  but buried  by H-bonds so as to exclude contact  with  the solvent.  Disruption of this network of H-bonds by site-directed mutagenesis has been shown to alter the heme Fe spin and ligand binding properties of CCP (Smulevich et al., 1988a&b).  F. Further Studies  Interpretation of the effects of heme propionate esterification is hampered by a lack of a three-dimensional  structure for the DME-heme  protein derivatives.  results have shown that metDME-Mb can be crystallized. data, one can use the three-dimensional mechanistic  interpretation.  Preliminary  In the absence of structural  model of native horse heart Mb as a basis for  However, the assumption that the only  their structures is the addition of methyl esters is simplistic.  difference between  Of particular interest is  whether esterification leaves the inner and outer propionate arrangement found in native Mb intact in DME-Mb. Of interest would also be the possible complex between metDME-Mb and ferricyanide. One  means of studying  this complex would be to soak crystals of metDME-Mb  ferricyanide solution to prepare complexes suitable for x-ray diffraction. NMR  in a  Alternatively,  techniques similar to those applied to characterization of ferricyanide binding to  119  cytochrome c (Eley et al., 1982) could be employed to map on the surface of DME-Mb.  A  DME-Mb-ferricyanide complex  electron transfer as a function of distance electron  transfer  kinetics  studies  the ferricyanide binding sites  of  can be used to analyze  in a manner similar to the  hemeproteins  with  small  inorganic  covalently attached to the protein surface (review by Mayo et al., 1986).  120  intramolecular complexes  BIBLIOGRAPHY Akeson, A., v. Ehrenstein, G., Hevesy, G., and Theorell, H. (1960), Arch. Biochem. Biophys., 91, 310-318. Antonini, E. and Brunori, M. (1971), Hemoglobin And Myoglobin In Their Reactions With Ligands, Chapter 5, North-Holland Publishing Co., Amsterdam. Aojula, H.S., Wilson, M.T., and Morrison, I.E.G. (1987), Biochem. J., 243, 205-210. Argos, P. and Mathews, F.S. (1975), /. Biol. Chem., 250, 747-751. Armstrong, F.A., Hill, H.A.O., and Walton, N.J. (1988), Acc. Chem. Res., 21, 407-413. Asakura, T. and Lamson, D.W.  (1973), Anal. Biochem., 53, 448-451.  Asakura, T. and Yonetani, T. (1979), /. Biol. Chem., 244, 4573-4579. Berzofsky, J.A., Peisach, J., and Blumberg, W.E. (1971), /. Biol. Chem., 246, 3,361-2311. Blanchetot, A., Wilson, V., Wood, D., and Jeffreys, A.J. (1983), Nature, 301, 732-734. Blanchetot, A., Price, M., and Jeffreys, A.J. (1986), Eur. J. Biochem., 159, 469-474. Bonfils, C , Balny, C , and Maurel, P. (1981), /. Biol. Chem., 256, 9457-9465. Bradbury, J.H. and Carver, J.A. (1984), Biochemistry, 23, 4905-4913. Breslow, E. (1963), /. Biol. Chem., 239, 486-486. Breslow, E., Beychok, S., Hardman, K.D., and Gurd, F.R.N. (1965), /. Biol. Chem., 240, 304-309. Breslow, E., Koehler, R., and Girotti, A.W. (1967), /. Biol. Chem., 242, 4149-4156. Brown, W.D.  and Mebine, L.B. (1969), /. Biol. Chem., 244, 6696-6701.  Brunori, M., Saggese, U., Rotilio, G.C, Antonini, E., and Jeffries, W. (1971), Biochemistry, 10, 1604-1609. Butler, J., Chapman, S.K., Davies, D.M., Sykes, A.G., Speck, S.H., Osheroff, N., and Margoliash, E. (1983), /. Biol. Chem., 258, 6400-6404. Carver, J.A. and Bradbury, J.H. (1984), Biochemistry, 23, 4890-4905. Cassat, J.C, Marini, CP., and Bender, J.W. (1975), Biochemistry, 14, 5470-5475.  121  Chance, B. (1974), Techniques of Chemistry, vol. 6, 3rd edition, Part II, edited by Lewis, E.S., John-Wiley & Sons. Chang, C.K., Ward, B., and Ebina, S. (1984), Arch. Biochem. Biophys., 231, 366-371. Clark, W.M.  (1960), Oxidation-Reduction Potentials of Organic Systems, Chapter 4, Robert  E. Krieger Publisliing Company, New York. Cole, R.P., Sukanek, P.C, Wittenberg, J.B., and Wittenberg, B. (1982), /. Appl. Physiol., 53, 1116-1124. Collman, J.P., Gagne, R.R., Reed, C.A., Robinson, W.T., and Rodley, G.A. (1974), Proc. Natl. Acad. Sci. USA, 71, 1326-1329. Craik, C.S., Buchman, S.R., and Beychok, S. (1980), Proc. Natl. Acad. Sci. USA, 77, 1384-1388. Craik, C.S., Buchman, S.R., and Beychok, S. (1981), Nature, 291, 87-90. Cutler, R.L., Davies, A.M., Creighton, S., Warshel, A., Moore, G.R., Smith, M., and Mauk, A.G. (1989), Biochemistry, 28, 3188-3197. Daly, J.S.F., Little, J.M., Troxler, R.F., and Lester, R. (1967), Nature, 216, 1030-1031. Dautrevaux, M., Boulanger, Y., Han, K., and Biserte, G. (1969), Eur. J. Biochem., 11, 267-277. De Sanctis, G., Falcioni, G., Giardina, B., Ascoli, F., and Brunori, M. (1986), /. Mol. Biol., 188, 73-76. De Sanctis, G., Falcioni, G., Giardina, B., Ascoli, F., and Brunori, M. (1988), /. Mol. Biol., 200, 725-733. Dickerson, R.E. and Timkovich, R. (1970), The Enzymes, vol. 11, Chapter 7. Drabkin, D.L. (1945), /. Mol. Biol., 158, 721-722. Dutton, P.L. (1978), Methods in Enzymology, 54, Chapt. 23, Academic Press. Dziobak, M.P. and Medenhall, G.D. (1982), /. Magnetic Resonance, 50, 274-280. Edmundson, A.B. and Hirs, C.H. (1961), Nature, 190, 663-665.  122  Edmundson, A.B. (1965), Nature, 205, 883-887. Eley, C.G., Moore, G.R., Williams, G., and Williams, R.J.P. (1982), Eur. J. Biochem., 124, 295-303. Ellis, W.R.  (1986), Ph.D. Dissertation, California Institute of Technology.  Eltis, L.E., Mauk, A.G., Hazzard, J.T., Cusonavich, M.A., and Tollin, G. (1988), Biochemistry, 27, 5455-5460. English, A.M., Lum, V.R., DeLaive.D.E., and Gray, H.B. (1982), /. Amer. Chem. Soc, 104, 870-871. Evans, D.H., O'Connell, K.M., Petersen, R.A., and Kelly, M.J. (1983), /. Chem. Educ, 60, 290-293. Evans, S.V. and Brayer, G.D. (1988), /. Biol. Chem., 263, 4263-4268. Evans, S.V. and Brayer, G.D. (1990), /. Mol. Biol., in press. Feinberg, B.A. and Ryan, M.D. (1981), /. Inorg. Biochem., 15, 187-199. Frauenfelder, H., Petsko, G.A., and Tsernoglou, D. (1979), Nature, 280, 558-563. Frauenfelder, H. and Wolynes, P.G. (19S5),Science, 229, 337-345. Fuhrop, J. and Smith, K. (1975), Porphyrins and Metalloporphyrins, Chapter 19, edited by K. Smith, Elsevier Scientific Publishing Co.. Galaris, D., Cadenas, E., and Hochstein, P. (1989), Arch. Biochem Biophys., 273, 497-504. George, P. and Irvine D.H. (1952), Biochem. J., 52, 511-517. George, P. and Stratmann, C.J. (1952), Biochem. J., 51, 103-108. Gilbert, W. (1978), Nature, 271, 501. Goldkorn, T. and Schetjer, A. (1976), Arch. Biochem. Biophys., 177, 39-45. Goss, D.J., LaGow, J.B., and Parkhurst, L.J. (1982), Comp. Biochem. Physiol., 71B, 229-233. Gotoh, T. and Shikama, K. (1974), Arch. Biochem. Biophys., 163, 476-481. Gotoh, T. and Shikama, K. (1976), /. Biochem., 80, 397-399.  123  Hagler, L., Coppes, R.L, and Herman, R.H. (1979), /. Biol. Chem., 254, 6505-6514. Hagler, L., Coppes, R.I., Askew, E.W., Heeler, A.L., and Herman, R.H. (1980), J. Lab. Clin. Med., 95, 222-230. Heineman, W.R.,  Anderson, C.W.,  Halsall, H.B., Hurst, M.M., Johnson, J.M., Kreishman,  G.P., Norris, B.J., Simone, M.J., and Su, S. (1982), Electrochemical and Spectrochemical Studies of Biological Redox Components, Chapt. 1, edited by Kadish, K.M., American Chemical Society. Heineman, W.R.  (1983), /. Chem. Educ, 60, 305-308.  Hill, R. (1933), Nature, 132, 897-898. Hill, R. and Holden, H.F. (1926), Biochem. J., 20, 1326-1339. Hodges, L.H., Holwerda, R.A., and Gray, H.B. (1974), /. Amer. Chem. Soc, 96, 3132-3137. Hoffman, B.M. (1974), /. Amer. Chem. Soc, 97, 1688-1694. Hoffman, B.M. (1979), The Porphyrins, Chapter 9, Vol. VII part B, edited by Dolphin, D., Academic Press. Hoffman, B.M. and Bull, C. (1976), Proc Natl. Acad. Sci. USA, 73, 800-803. Hoffman, B.M. and Petering, D.H. (1970), Proc. Natl. Acad. Sci. USA, 67, 637-643. Hulquist, D.E., Sannes, L.J., and Juckett, D.A. (1984), Curr. Top. Cell. Regul., 24, 287-300. Huth, S.W., Kimberly, K.E., Piszkiewicz, D, and Fleischer, E.B. (1976), /. Amer. Chem. Soc, 98, 8467-8471. Kagen, L.J. (1973), Myoglobin; Biochemical, Physiological, and Clinical Aspects., Columbia University Press. Kendrew, J.C, Bodo, G., Dintzis, H.M., Parrish, R.G., and Wyckoff, H. (1958), Nature, 181, 662-666.  124  Kendrew, J.C, Dickerson, R.E., Strandberg, B.E., Hart, R.G., Davies, D.R., Phillips, D.C, and Shore, V.C (1960), Nature, 185, 422-427. Kolthoff, I.M. and Auerbach, C. (1952), /. Amer. Chem. Soc.,  74, 1452-1456.  Koppenol, W. H. and Margoliash, E. (1982), /. Biol. Chem., 257, 4426-4437. La Mar, G.N., Budd, D.L., Viscio, D.B., Smith, K.M., and Langry, K.C. (1978), Proc. Natl. Acad. Sci. USA, 75, 5755-5759. La Mar, G.N. and Walker(Jensen), F.A. (1979), The Porphyrins,  Chapter 2, Vol. 4,  part B, edited by D. Dolphin, Academic Press Inc. La Mar, G.N., Budd, D.L., Smith, K.M., and Langry, K.C (1980), /. Amer. Chem. Soc,  102,  1822-1827. La Mar, G.N., Davis, N.L., Parish, D.N., and Smith, K.M. (1983), J. Mol. Biol,  168,  887-896. La Mar, G.N., Toi, H., and Krishnamoorthi, R. (1984), /. Am. Chem. Soc,  106,  6395-6401. La Mar, G.N., Emerson, S.D., Lecomte, J.T.J., Pande, U., Smith, K.M., Craig, Wayne, G . , and Kehres, L.A. (1986), /. Amer. Chem. Soc,  108, 5568-5573.  Lecomte, J.T.J, and La Mar, G.N. (1985), Biochemistry, 24, 7388-7395. Light, W.R.,  Rohlfs, R.J., Palmer, G., and Olson, J.S. (1987), /. Biol. Chem. 262, 46-52.  Lim, A.R and Mauk, A.G. (1986), Biochem.  J., 229, 765-769.  Linder, R.E., Records, R., Barth, G., Bunnenberg, E., Djerassi, C , Hedlund, B.E., Rosenberg, A., Benson, E.S., Seamans, L., and Moscowitz, A. (1978), Anal.  Biochem.,  90, 474-480. Livingston, D.J., Davis, N.L., La Mar, G.N., and Brown, W.D. (1983), /. Amer. Chem.  Soc,  106, 3025-3026. Livingston, D.J., McLachlan, S.J., La Mar, G.N., and Brown, W.D. (1985), /. Biol. 260, 15699-15707.  125  Chem.,  Longo, L.D., Koos, B.J., and Power, G.G. (1973), Amer. J. Physiol, 224, 1032-1036.  McGourty, J.L., Blough, N.V., and Hoffman, B.M (1983), /. Amer. Chem. Soc,  105,  4470-4472.  McLachlan S.J., La Mar, G.N., and Sletten, E. (1986), /. Amer. Chem. Soc,  108,  1285-1291. McLellan, T. (1984),  Biochemical Genetics, 22, 181-200.  McLendon, G. and Miller, J.R. (1985), /. Amer. Chem. Soc, MacGregor, R.B. and Weber, G. (1986),  107, 7811-7816.  Nature, 319, 70-73.  Maniatis, T., Fritsch, E.F., Lauer, J., and Lawn R.M. (1980),  Ann. Rev. Genet., 14,  145-178.  Marcus, R.A. and Sutin, N. (1985), Biochim. Biophys. Acta, 811, 265-322 Margalit, R. and Schejter, A. (1976), Eur. J. Biochem, 32, 492-499. Mathews, F.S. (1985),  Prog. Biophys. Molec. Biol, 45, 1-56.  Matthew, J.B. (1985), Ann. Rev. Biophys. Chem. 1985, Chapter 14, 387-417. Mauk, A.G., Coyle, C.L., Bordignon, E., and Gray, H.B. (1979), /. Amer. Chem. Soc,  101,  5054-5056. Mauk, M.R., Mauk, A.G., Weber, P.C. and Matthew, J.B. (1986),  Biochemistry, 25,  7085-7091. Mayo, S.L., Ellis, W.R.,  Crutchley, R.J., and Gray, H.B. (1986), Science, 233, 948-952.  Meites, L. and Meites, T. (1948),  Anal. Chem., 20, 984-985.  Miki, K., II, Y., Yukawa, M., Owatari, A., Hato, Y., Harada, S., Kai, Y., Kasai, N., Hata, Y., Tanaka, N., Kakudo, M., Katsube, Y., Kawabe, K., Yoshida, Z., and Ogoshi, H.  (1986a), /. Biochem. (Tokyo), 100, 269-276. Miki, K., Harada, S., Hato, Y., Iba, S., Kai, Y., Kasai, N., Katsube, Y., Kawabe, K.,  Yoshida, Z., and Ogoshi, H. (1986b), /. Biochem. (Tokyo), 100, 277-284.  126  Moore, G.R., Harris, D.E., Leitch, F.A., and Pettigrew, G.W. (1984), Biochim. Biophys. Acta, 764, 331-342. Moore, G.R., Pettigrew, G.W.,  and Rogers, N.K. (1986), Proc. Natl. Acad. Sci. USA, 83,  4998-4999. Moore, J.W. and Pearson, R.G. (1981), Kinetics and Mechanism, 3  rd  edition, John Wiley &  Sons, Inc. Morner, K.A.H. (1897), Nord. Med. Arkiv., 30, 1. Newton, T.W. (1968), /. Chem. Educ., 45, 571-575. Nobbs, C.L. (1966), Hemes and Hemoproteins, p. 143-147, edited by Chance, B., Estabrook, R.W.  and Yonetani, T., Academic Press.  Nobbs, C.L., Watson, H.C., and Kendrew, J.C. (1966), Nature, 209, 339-341. Nocera, D.G., Winkler, J.R., Yocum, K.M., Bordignon, E., and Gray, H.B. (1984), /. Amer. Chem. Soc, 106, 5145-5150. Northrup, S.H., Reynolds, J.C.L., Miller, CM., Forrest, K.J., and Boles, J.O. (1986), /. Amer. Chem. Soc, 108, 8162-8170. Northrup, S.H., Boles, J.O., and Reynolds, J.C.L. (1988), Science, 241, 67-70. 8170. Olivas, E., de Waal, D.J.A., and Wilkins, R.G. (1977), /. Biol. Chem., 252, 4038-4042. Olson, J.S., Mathews, A.J., Rohlfs, R.J., Springer, B.A., Egeberg, K.D., Sligar, S.G., Tame, J., Renaud, J-P., and Nagai, K. (1988), Nature, 336, 265-266. Ozols, J. and Strittmatter, P. (1963), J. Biol. Chem., 239, 1018-1023. Pain, R.H. (1983), Nature, 305, 581-582. Pain, R.H. (mi),Nature, 326, 247. Peters, J.B., Barnard, R.J., Edgerton, V.R., Gillespie, C.A., and Stempel, K.E. (1972), Biochemistry, 11, 2627-2633. Peterson-Kennedy, S.E., McGouryt, J.L., and Hoffman, B.M. (1984), /. Amer. Chem. Soc, 106, 5010-5012.  127  Phillips, S.E.V. (1978), Nature, 273, 247-248. Phillips, S.E.V. (1980), /. Mol. Biol., 142, 531-554. Ray, G.B. and Paff, G.H. (1930), Amer. J. Physiol., 94, 521-528. Reid, L.S. (1984), Ph.D. Dissertation, University of British Columbia. Reid, L.S. and Mauk, A.G. (1982), /. Amer. Chem. Soc, 104, 841-845. Reid, L.S., Taniguchi, V.T., Gray, H.B. and Mauk, A.G. (1982), /. Amer. Chem. Soc, 104, 7516-7519. Reid, L.S., Lim, A.R. and Mauk, A.G. (1986), /. Amer. Chem. Soc, 108, 8197-8201. Reinhammar, B. and Malmstrom, B.G. (1981), Copper Proteins, Chapter 3, edited by T.G. Spiro, John Wiley and Sons. Ringe, D., Petsko, G.A., Kerr, D.E., and Oritz de Montellano, P. (1984), Biochemistry, 23, 2-4. Romero-Herrera, A.E. and Lehmann, H. (1974), Biochim. Biophys. Acta, 336, 318-323. Romero-Herrera, A.E., Lehmann, H., Joysey, K.A., and Friday, A.E. (1978), Phil. Trans. Roy. Soc London, Ser. B, 283, 61-163. Rosenberg, R.C, Wherland, S, Holwerda, R.A., and Gray, H.B. (1976), /. Amer. Chem. Soc, 98, 6364-6369. Rossi-Fanelli, A., Antonini, E. (1957), Arch. Biochem. Biophys., 72, 243-246. Salemme, F.R. (1976), /. Mol. Biol., 102, 563-568. Satoh, Y. and Shikama, K. (1981), /. Biol. Chem., 256, 10272-10275. Saunders, B.C. (1973), Inorganic Chemistry, Vol. 2, Chapter 28, edited by Eichorn, G.L., Elsevier Scientific Publishing Co., New York. Schwarzenbach, G. and Heller, J. (1951), Helvetica Chimica Acta, 34, 576-591. Scott, R.A., Mauk, A.G., and Gray, H.B. (1985), /. Chem. Ed., 62, 932-938. Segal, M.G. and Sykes, A.G. (1977), /. Amer. Chem Soc, 100, 4585-4592. Sherwood, C , Mauk, and A.G., Brayer, G.D. (1987), /. Mol. Biol., 193, 227.  128  Shikama, K. (1985), Experientia, 41, 701-836. Shire, S.J., Hanania, G.I.H., and Gurd, F.R.N. (1974), Biochemistry, 13, 2974-2979. Simolo, K.P., McLendon, G.L., Mauk, M.R., and Mauk, A.G. (1984), /. Amer. Chem. Soc, 106, 5012-5013. Smulevich, G., Mauro, J.M., Fishel, L.A., English, A.M., Kraut, J., and Spiro, T.G. (1988a), Biochemistry, 27, 5477-5485. Smulevich, G., Mauro, J.M., Fishel, L.A., English, A.M., Kraut, J., and Spiro, T.G. (1988b), Biochemistry, 27, 5486-5492. Springer, B.A. and Sligar S.G. (1987), Proc. Natl. Acad. Sci. USA, 84, 8961-8965. Stellwagen, E. (1978), Nature, 275, 73-74. Stellwagen, E. and Cass R.D. (1975), /. Biol. Chem., 250, 2095-2098. Strittmatter, P., Spatz, L., Corcoran, D., Rogers, M.J., Setlow, B., and Redline, R. (1974), Proc. Natl. Acad. Sci. USA, 71, 4565-4569. Sugawara, Y. and Shikama, K. (1979), Sci. Rep. Tohoku Univ. Ser.IV (Biol), 37, 253-262. Sugawara, Y. and Shikama, K. (1980), Eur. J. Biochem., 110, 241-246. Sundberg, R.J., Bryan, R.F., Taylor, I.F., and Taube, H. (1974), /. Amer. Chem. Soc, 96, 381-392. Sweetser, P.B. (1969), Anal. Chem., 39, 979-982. Tajima, G. and Shikama, K. (1987), /. Biol. Chem., 262, 12603-12606. Takano, T. (1977a), /. Mol. Biol, 110, 537-568. Takano, T. (1977b), /. Mol. Biol., 110, 569-584. Tamura, M., Asakura, T. and Yonetani, T. (1973a), Biochim. Biophys. Acta, 295, 467-479. Tamura, M., Woodrow III, G.V. and Yonetani, T. (1973b), Biochim. Biophys. Acta, 317, 34-49. Tanford, C. (1961), Physical Chemistry of Macromolecules, Chapters 20 & 26, John Wiley and Sons, New York.  129  Taniguchi, V.T., Sailasuta-Scott, N., Anson, F.C, and Gray, H.B. (1980), Pure & Appl. Chem., 52, 2275-2281. Taube, H. (1977), Advances in Chemistry Series, No. 162, Bioinorganic Chemistry II, 128-144. Taube, H. (1984), Science, 226, 1028-1036. Taylor, J.F. and Morgan, V.E. (1942), /. Biol. Chem., 144, 15-19. Teale, F.W.J. (1959), Biochim. Biophys. Acta, 35, 543. Tew. D. and Ortiz de Montellano, P.R. (1988), /. Biol. Chem., 263, 17880-17886. Theorell, H. (1932), Biochem. Z., 252, p. 1 Theorell, H. (1934), Biochem. Z., 268, p. 46. Tipler, T.D., Edwards, Y.H. and Hopkinson, D.A. (1978), Ann. Hum. Genet., London, 41, 409-418. Tollin, G., Meyer, T.E., and Cusonavich, M.A. (1986), Biochim. Biophys. Acta, 853, 29-41. Tomoda, A., Takizawa, A., Tsuji, A., and Yoneyama, Y. (1981), Biochem. J., 193, 181-185. Tsukahara, K., Okazawa, T., Takahashi, H., and Yamamoto, Y. (1986), Inorg. Chem., 25, 4756-4760. Underwood, L.E. and Williams, R.S. (1987), Amer. J. Physiol, 252, C450-C453. Uyeda, M. and Peisach, J. (1981), Biochemistry, 20, 2028-2035. Varadarajan, R., Szabo, A., and Boxer, S.G. (1985), Proc. Natl. Acad. Sci. USA, 82, 5681-5684. Varadarajan, R., Zewert, T.F., Gray, H.B., and Boxer, S.G. (1989), Science, 243, 69-72. Weller, P., Jeffreys, A.J., Wilson, V. and Blanchetot, A. (1984), EMBO J., 3, 439-446. Weller, P.A., Price, M., Isenberg, H., Edwards, Y.H., and Jeffreys, A.J. (1986), Mol. Cell. Biol, 6, 4539-4547.  130  Wherland, S., Holwerda, R.A., Rosenberg, R.C, and Gray, H.B. (1975), /. Amer. Chem. Soc., 97, 5260-5262. Wherland, S. and Gray, H.B. (1976), Proc. Natl. Acad. Sci. USA, 73, 2950-2954. Whitburn, K.D. (1987), Arch. Biochem. Biophys., 253, 419-430. Wilson, G.S. (1978), Methods in Enzymology, 54, 396-410. Wittenberg, B.A. and Wittenberg, J.B. (1987), Proc. Natl. Acad. Sci USA, 84, 7503-7507. Wittenberg, J.B. (1965), /. Biol. Chem., 241, 104-114. Wittenberg, J.B. (1970), Physiol. Rev., 50, 559-636. Wood, D., Blanchetot, A. and Jeffreys, A.J. (1982), Nucleic Acids Res., 10, 7133-7144. Wyman, J. (1965), /. Biol. Chem., 241, 115-121. Yusa, K. and Shikama, K. (1987), Biochemistry, 26, 6684-6688.  131  Appendix A Midpoint reduction potentials (vs. NHE) Potentials were measured in sodium phosphate or boric acid: sodium hydroxide buffers as noted (*).  Midpoint reduction potentials were determined from an unweighted linear  least-squares fitting program using the Nernst equation (19).  The parameter / is ionic  strength and [med.] is mediator concentration. The mediators for native Mb and D M E Mb are [Ru(NH ) ]Cl 3  6  3  and [Ru(NH ) Im]Cl 3  5  3  respectively.  Myoglobin  pH  6.0  temp. (°C)  slope  midpoint potential (mV  4  58.9 ± 0.2  64.3 ± 0.1  5  59.2 ± 0.2  63.1 ± 0.1  7  55.3 ± 0.1  75.0 ± 0.1  15  7  57.9 ± 0.2  70.6 ± 0.1  20  4  58.0 ±0.1  65.3 ± 0.1  25  4  59.6 ± 0.2  60.9 ± 0 . 1 .  30  3  59.4 ± 0.1  55.8 ± 0.1  35  0.5  60.6 ± 0.7  52.4 ± 0.4  0.05  5  59.1 ±0.2  79.4 ± 0.1  0.15  3.5  59.7 ± 0.4  53.3 ± 0.2  0.2  3  58.8 ± 0.1  49.9 ± 0.1  0.3  3  59.4 ± 0.1  45.4 ± 0.1  0.4  2.5  59.6 ± 0.3  42.2 ± 0.2  0.5  2  58.6 ± 0.2  39.0 ± 0.1  0.1  7  59.1 ± 0.1  55.9 ± 0.1  9  59.4 ± 0.2  44.5 ± 0.1  25  /  (M)  0.1  6.5 7.0  7.0  7.5 8.0  10  25  25  0.1  [med.] 0*M)  132  pH  temp. (°C)  I (M)  [med.]  slope  midpoint potential (mV)  (MM)  9  59.8 ±  0.4  33.9  ±  0.2  *9.0  16  59.6 ±  0.3  25.8  ±  0.2  *9.5  20  59.2 ±  0.2  22.6  ±  0.1  *8.5  25  0.1  Dimethyl-ester heme Myoglobin  pH  / (M)  0*M)  0.1  1  60.0 ±  0.6  131.0 ±  6.0  0.75  60.8 ±  0.3  126.4  ±  0.2  6.5  0.75  59.6 ±  0.6  115.1  ±  0.3  0.5  58.0 ±  0.4  124.6  ±  0.2  15  0.5  58.0 ±  0.3  114.8  ±  0.3  20  0.5  57.8 ± 0 . 2  109.0  ±  0.1  25  0.5  59.3 ±  100.2  ±  0.2  30  0.4  60.3 ± 0 . 1  93.4  ±  0.1  35  0.25  61.9 ± 0 . 4  86.3  ±  0.2  0.05  0.5  60.1 ±  0.5  130.7  ±  0.3  0.15  0.5  60.0 ±  0.3  90.9 ±  0.2  0.4  60.6 ±  0.4  82.8  0.3  1  60.2 ±  0.2  75.2 ±  0.4  1  59.5 ±  0.3  70.3  0.5  1  58.8 ±  0.3  65.6 ±  2.5  59.3 ±  0.5  76.0  ±  0.3  6  60.8 ± 0 . 3  63.6  ±  0.2  5.5  7.0  7.0  7.5  8.0  temp. (°C)  25  10  25  25  0.1  0.1  [med.]  slope  133  midpoint potential (mV)  0.4  ±  ±  0.3  0.2  0.2  0.1  0.2  0.2  [med.]  (M)  (uM)  25  0.1  6  59.1 ± 0.4  64.4 ± 0.2  *8.5  5  59.5 ± 0.5  56.6 ± 0.3  *9.0  3  59.7 ± 0.4  52.9 ± 0.2  *8.0  temp.  slope  midpoint potential (mV)  I  (°C)  pH  134  Appendix B Second order rate constants  The  second  order rate constants reported below  metMb and metDME-Mb by Fe(EDTA) ". 2  buffer.  are for the reduction of native  Reactions were performed in sodium phosphate  Second order rate constants were calculated from the slopes of first order plots  of reduction rate versus [Fe(EDTA) "] by a weighted linear least squares program. The 2  data for native Mb has been published previously (Lim and Mauk, 1985) and is presented here solely for reference purposes.  Myoglobin  pH  temp.  (°C) 6.0  25  I  (M-V )  0.5  28.7 ± 0.7  6.5 7.0  7.5 8.0  k\2  (M)  1  16.9 ± 0.3 25  25  0.05  23.4 ± 0.8  0.1  22.5 ±0.5  0.2  14.8 ± 0.6  0.3  14.1 ±0.2  0.4  13.2 ± 0.3  0.5  12.6 ± 0.2  0.5  9.7 ±0.1 8.6 ±0.1  135  Dimethyl-ester  pH  6.0  heme  Myoglobin  temp.  I  (°C)  . (M)  25  0.1  7.0  7.5 8.0  2  (M-V ) 1  1891 ± 33 1630 ± 26  6.5 7.0  *i  25  10  0.05  2475 ± 58  0.1  1341 ± 45  0.2  751 ± 19  0.3  501 ± 2 1  0.4  299 ± 11  0.5  258 ± 6  0.1  580 ± 18  15  724 ± 25  20  1011 ± 29  30  1803 ± 82  35  2376 ± 28  25  0.1  861 ± 42 350 ± 19  136  Appendix  C  Second order rate constants adjusted for driving force The  second order rate constants for the reduction of metDME-Mb and metMb  Fe(EDTA) " were adjusted for the driving force.  by  The difference in midpoint reduction  2  potential (AE) between the reactants under the same conditions of pH, ionic strength (J) The adjusted reaction rate, fcj2 » as  and temperature was defined as the driving force.  J  defined from Equation 29.  pH  E  Fe(EDTA) "£ 2  m  (mV)  (mV)  m  AE (V)  DME-Mb  6.0 6.5 7.0 7.5 8.0  126.4 115.1 100.2 76.0 63.6  115.5 108.9 96.6 84.4 66.6  —0.0109 —0.0062 —0.0036 +0.0084 +0.0030  native Mb  6.0 6.5 7.0 7.5 8.0  64.3 63.1 60.9 55.9 44.5  104. l 98.4 95.0 86.8 86.9  +0.0512 +0.0458 +0.0357 +0.0285 +0.0221  at conditions of 7=0.1 M " 7=0.1 M " 7=0.5 M " 7=0.5 M " 7=0.5 M  s  C  b  d  (M'V )  and 25°C and 25°C from Kolthoff and Auerbach (1951) and 25°C and 25°C from Reid (1984) and 25°C from Lim and Mauk (1986)  137  k®  *12  (M"V  1  1891 1630 1341 861 350  1529 1444 1250 1014 371  a  28.7 16.9 12.6 9.7 8.6  e  77.8 41.2 25.2 16.9 13.2  w  0  Appendix D Derivation of Equation 17 Equation  17  and  its derivation applies to the analyses of both native and  DME-Mb  autoxidation experiments.  Equations:  By definition  [Mb] = [oxyMb] + [metMb]  By the Beer-Lambert Law:  Where A  A = E C L  is the absorbance at a particular wavelength, E  that wavelength, C  is the extinction coefficient at  is the concentration of chromophobe, and  L  is the pathlength of  absorption. In a mixture of oxyMb and metMb, the observed absorptions at the o (581 nm) (543 nm) bands, A  and A  a  and  p  respectively, can be defined as follows (assuming a pathlength  b  of 1 cm):  K  xy  et  = £ f ([oxyMb] + (Ef / y  eX  Let the constants C  x  ^  A  = £^ [oxyMb] + ^ [metMb]  a  = ^  x  y  and C  2  ) [metMb])  b  = ^ [ o x y M b ] + ^'[metMb]  = ^  ([oxyMb] + (£g" / JSg"?) [metMb]) et  represent the ratios Ef^/E *?  ( [oxyMb] + Cj [metMb])  0  A  and U  /£  ) x y b  = £ g ( [oxyMb] + C [metMb]) xy  b  m e t  2  138  respectively.  Let the ratio of the of absorbances at the a and P bands ( A / A ) be represented by a  b  A(a/p).  £T ([oxyMb] + Ci[metMb]) y  A { a / p ) =  £g y([oxyMb] + C [metMb]) x  2  Let R represent the ratio £ >'/£g . >x  xy  a  A ( a / / 3 ) =  [oxyMb] + Ci ([Mb] - [oxyMb])  R  [oxyMb] + C ([Mb] - [oxyMb]) 2  (1 _ d ) [oxyMb] + CjfMb] ( 1 - C ) [oxyMb] + C [Mb] 2  2  The term [oxyMb]/[Mb] is isolated.  A(aip)R-  (1 - C ) [oxyMb] +  1  x  A(a/p)R- C [}Ab] 1  2  = (1 - C i ) [oxyMb] - Ata/Z^/T'C^Mb]  [oxy](A(a//?)(l-C2)JR-1-  ( l _ d )) =  [oxyMb] [Mb]  W>\(C -A{aip)C R- ) x  x  2  C -A(alp)C R-  1  x  =  2  C j - 1+ (l-C )A(a//?) 2  Substituting DME-Mb for Mb gives Equation 17. Using  data  from  the reference  spectra and the MINSQ  following values for the parameters were calculated:  native Mb  1.0525  0.19529  0.38058  1.0588  0.24435  0.42068  P  DME-Mb  139  least  squares analysis the  Appendix E Marcus Theory Calculation of  A^i  r r  References: Wherland and Gray (1976); Marcus and Sutin (1985) 1. Calculate the free energy change of the cross-reaction (AG* ) from the electrochemical 2  reduction potential difference (AE) between the reactants.  AG* = - n F(AE) 2  The terms n and F are the stoichiometry of the electron transfer reaction and the Faraday constant respectively.  2. Calculate the Marcus theory activation energies, AG* and AG , from the first 2  22  equation and the conversion between transition state (Eyring) and Marcus formulism.  k 2=  KT (—AG*/RT) e  AG* =AG*+ RT In ^ KT  The term K is Boltzmann's constant; T is temperature (Kelvin); h is Planck's constant; Z is the collision frequency.  140  3. Calculate the electrostatic work terms w , w2i, w , and w from: xl  12  -BR s/l  -BR s/l  1 + BR y/l  1 + BR s/l  h  It  *ab  22  a  a  b  This equation describes the work required to bring together two reactants (a and with charges of Z and  and radii of R and R respectively. The charges are  a  a  b  dependent on the ionic strength /. The term t is the dielectric constant of the solvent, water; R  ab  is the sum of R and R ; a  b  is the electron charge.  4. Calculate the electrostatics-independent free energy of activation terms, A G * and A G  2 2  from:  A G j = A G j — Wji  5. Calculate the electrostatics-independent free energy change, AG°, from:  AG° = A G f — w 2  l2  + w  2l  6. Calculate the electrostatics-independent free energy of activation, AGjJ, by solving the quadratic:  A(AG\\) + B(AG\\) + C = 0 2  141  2  where A = 4  B = 8AG*2 + 4AG ° — SAG** R  and C = 4AG* * (AG22 + AG ° - 2AG*2) + ( A G ° ) 2  2  R  7. Calculate the free energy of activation corrected for the electrostatics-work, AG*j , orr  from:  AG*r  = AG** + H.  U  8. Calculate the corrected self-exchange rate, k\[  n  from Equation 9:  h  Equation 9, which is derived from transition-state theory, is used with a Marcus equation derived parameter  (AG*\ ) OTr  because the transition-state in a self-exchange  reaction is considered similar to a mononuclear reaction, which can be analyzed by Equation 9.  142  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items