Synthesis, Physical 
Characterisation and Solution 
Chemistry of  
N-Acetyl-Cobalt(III)-
 Microperoxidase 8 
 
 
-Desigan Sannasy- 
 
 
 
A dissertation submitted to the Faculty of Science, University of the Witwatersrand, 
Johannesburg, in fulfilment of the requirements for the degree of Master of Science. 
 
 
 
March 2006 
 
 2 
 
 
 
 
 
 
 
 
 
Declaration 
 
 
I declare that this is my own unaided work.  It is being submitted for the Degree of 
Master of Science at the University of the Witwatersrand, Johannesburg.  It has not been 
submitted before for any degree or examination at any other university. 
 
 
 
 
 
 
D. Sannasy 
 
March 2006 
 
 
 3 
Abstract 
 
This dissertation describes the synthesis, physical characterisation and solution chemistry 
of NAc-CoIIIMP8, a biomimetic model compound of vitamin B12a, synthesised from the 
haemoctapeptide derived from horse heart cytochrome c.  Peptic and tryptic digestion of 
horse heart cytochrome c removes much of the globular protein encapsulating the iron 
porphyrin prosthetic group.  The resulting haemoctapeptide fragment retains residues 14 
to 21 of the parent cytochrome (MP8) via thioether linkages to Cys-14 and Cys-17. 
Reductive demetalation of MP8 yielded the metal free MP8.  This was treated with 
cobaltous acetate in an aerated aqueous solution to produce CoIIIMP8.  CoIIIMP8 was 
acetylated by treatment with acetic anhydride and yielded N-acetyl-Co(III)-
 microperoxidase 8 (NAc-CoIIIMP8).  It is well established that acetylation reduces 
aggregation of these haempeptides.  The starting materials and products of each step 
during synthesis were characterised by UV-visible absorption spectroscopy, high 
performance liquid chromatography (HPLC) and fast atom bombardment-mass 
spectroscopy (FAB-MS).  MP8 free base and Co(III)-MP8 were also analysed using 
luminescence spectroscopy.   
 
The molar extinction coefficients of NAcCoIII-MP8 in aqueous and ionic medium were 
determined using inductively coupled plasma-optical emission spectroscopy (ICP-OES) 
and UV-visible absorption spectroscopy.  The extinction coefficient, e, of NAcCoIIIMP8 
(? = 420 nm, pH 7.00, 25 ?C) in distilled water and 1.0 M NaClO 4 was 1.80 + 0.01 x 105 
M-1 cm-1 and 1.66 + 0.01 x 105 M-1 cm-1, respectively.  Beer?s law studies show that 
NAc-CoIIIMP8 remains monomeric in aqueous solution up to concentrations of at least 35 
?M.     
 
The spectroscopic changes observed for NAc-CoIIIMP8 during the course of a 
spectrophotometric titration are very similar to those observed for NAc-FeIIIMP8, with 
both being consistent with six successive ionisations.  By analogy with NAc-FeIIIMP8, 
we attributed the first (pK1 = 2.0 + 0.3) to the coordination of the c-terminal carboxylate 
group (Glu-21) of the appended polypeptide.  The second acid range transition (pK2 = 2.8 
 4 
+ 0.1) for NAcCoIIIMP8 involved the deprotonation of the cationic His-18 and 
concomitant replacement of the c-terminal carboxylate by the neutral heterocyclic base.  
The third and fourth pKa?s are attributed to the ionisation of the haem propanoic acid 
groups (pK3 = 3.9 + 0.03) and (pK4 = 7.5 + 0.03).  Ionisation of the cobalt-bound water 
molecule above neutal pH was assigned to pK5 = 9.2 + 0.04.  Finally, we attributed pK6 
(12.1 + 0.03) to the ionisation of the coordinated histidine trans to the OH- to form the 
histidinate complex (His--CoIII-OH-).   
 
A principal aim of this work was to demonstrate that the kinetics and the thermodynamics 
of the ligand substitution reactions of NAc-CoIIIMP8 can be studied 
spectrophotometrically; a comprehensive investigation of these reactions will be 
undertaken by othe rs.  Towards this end the formation constants between NAc-CoIIIMP8 
and N-methylimidazole and pyridine were determined.  We observed the formation of a 
bis-substituted complex in the reaction of NAc-CoIIIMP8 with the ligands, but only 
mono-substitution with NAc-FeIIIMP8 and B12a.   We attribute this first ligand binding to 
the replacement of the axial water molecule, and the second replacement of the axial 
histidine residue.  The absence of the second reaction with NAc-FeIIIMP8 and B12a 
suggest that the CoIII-N(His) bond in NAc-CoIIIMP8 is significantly weaker than the FeIII-
 N(His) and CoIII-N(dimethylbenzimidazole) bond, respectively.  When comparing the 
formation constants of NAc-FeIIIMP8, NAc-CoIIIMP8 and B12a, we found that the value  
of log K1 for NAc-CoIIIMP8 for these ligands is significantly higher than that reported for 
NAc-FeIIIMP8 and B12a.   
Kinetics studies of NAc-CoIIIMP8 with N-methylimidazole and methylamine were 
investigated.  The data obtained did not follow conventional pseudo-first order kinetics; 
instead there was some evidence for biphasic kinetics.  In the reaction of N-
 methylimidazole with NAc-CoIIIMP8, we observed that the rate of reaction is virtually 
independent of the concentration of the incoming ligand.  The results can be explained if 
the mechanism proceeds through a purely dissociative mechanism, i.e., if the rate of the 
reaction is controlled by the rate at which, firstly, the water molecule dissociates from the 
CoIII centre and, secondly, the histidine dissociates from the metal.  The second order rate 
 5 
constant, k2, could not be determined since the rate of reaction is independent of N-
 methylimidazole concentration.  In the reaction of methylamine with NAc-CoIIIMP8, we 
observed that the rate of reaction is dependent on the concentration and participation of 
the incoming ligand.  We propose that the displacement of water and histidine by 
methylamine involves an interchange mechanism (Id), where the bond forming and bond 
breaking occur simultaneously, and thus the rate of reaction becomes dependent on the 
concentration of the incoming ligand.   
The results showed that the rate of reaction for methylamine with NAc-CoIIIMP8 was 
faster than with N-methylimidazole.  We attributed these differences in rate constants to 
the size of the incoming ligands.  N-methylimidazole is a secondary amine and is 
relatively more bulky than methylamine which is a primary amine; therefore it is easier 
for methylamine to attach to the metal centre compared to N-methylimidazole.  
For comparison, the rate of reaction of B12a with N-methylimidazole and methylamine 
was determined.  The results show that the rate of the reaction between NAc-CoIIIMP8 
and B12a with N-methylimidazole and methylamine are significantly different.  
Furthermore, we observe only mono-substitution in B12a and bisubstitution in NAc-
 CoIIIMP8.  Overall, the results presented in this work do give a general indication on how 
thermodynamically stable a CoIII ion is in a porphyrin ring and also to a very limited 
extent show that a porphyrin does not confer the same kinetic lability on the CoIII ion as 
the corrin ring.   
 
 
 6 
 
 
 
 
 
 
 
 
 
 
 This work in dedicated to God and my loving Family: 
 
My Parents: Jason and Rani Sannasy 
My Brother: Gregory 
My Grandparents: Ma and Nine Daddy, Umma and Shinah 
My Uncle: Vis 
My Fianc?:  Ying 
 
 
Thanks to all of you for the love and support throughout the years. 
 
 
 
 7 
Acknowledgements 
 
 
I wish to thank my supervisors and the following people for all their support during the 
duration of this project. 
 
? Prof Helder Marques and Dr Alvaro de Sousa, for all their assistance and 
insight throughout the course of the project. 
 
? Dr Winile Mavuso, for her constant advice and patience, invaluable help at the 
beginning of this project.   
 
? Prof Ewa Cukrowska and Mr. Ruphert Morena for their help with the ICP-
 OES analysis. 
 
? Mr. Tommy van der Merwe for his help with the FAB-MS analysis. 
 
? Mr. Clint Manuel for all his help with the computer. 
 
? The glassblowers, Mr. Steve Gannon and Mr. Barry Fairbrother for always 
willing to help. 
 
? Mr. Basil Chassoulas, for helping me with the HPLC instrument. 
 
? Christopher Perry, Prethi Vashi, Leanne Knapton and Mabel Conyanis for 
their outstanding enthusiasm and willingness to help.   
 
I would like to thank the University of the Witwatersrand, the National Research 
Foundation (NRF), SACTWU and the Ernst Eriksen Ethel Trust for all their financial 
support.  Last but definitely not least to my fianc? the lovely Miss Ying-Hsuan Wu, 
without whose dedication and support this project would not have been possible.
 8 
Table of Contents 
Preliminaries 
Abstract  3 
Acknowledgements 7 
List of Figures  12 
List of Tables  125 
Abbreviations  16 
 
Chapter 1 Introduction 
1.1 Bioinorganic Chemistry  18 
1.1.1 General Overview 18 
1.1.2 Metal Ions in the Chemistry of Life 20 
1.1.2.1 Aspects of the Chemistry of Iron:  Properties and Special Characteristics 22 
1.1.2.2 Aspects of the Chemistry of Cobalt:  Redox and Magnetic Properties 25 
1.2 Metalloproteins and their Significance 25 
1.3 Porphyrins  29 
1.3.1 Basic Structure  29 
1.3.2 Iron Porphyrins 32 
1.4 Haemproteins  32 
1.4.1 The Cytochromes 33 
1.4.2 The Peroxidases and Catalases 36 
1.4.3 The Cytochromes P450 40 
1.4.4 Haemoglobin (Hb) and Myoglobin (Mb) 44 
1.5 The Microperoxidases (MPs) 46 
1.5.1 Background  46 
1.5.2 Preparation of Microperoxidases 47 
1.5.3 Solution Behaviour of the Microperoxidases 49 
1.5.3.1 Solutio n Structure 49 
1.5.3.2 Aggregation in Aqueous Solution 50 
1.5.3.3 Overcoming Aggregation 52 
 9 
1.5.4 Haempeptide Acid Dissociation Constants 53 
1.5.5 Ligand Binding Studies of Microperoxidases  53 
1.5.6 Demetalation of Haempeptides 56 
1.6 Cobalt Corrinoids: Basic Structure  56 
1.7          Vitamin B12 57 
1.7.1 Background  57 
1.7.2 Structure of Vitamin B12a 58 
1.7.3 Chemistry of Vitamin B12 and its derivatives: 60 
               Ligand Substitution and Kinetic Studies  60 
1.8 Comparison of Porphyrin and Corrin 68 
1.8.1 Structural and Functional 68 
1.8.2 Chemical Properties  69 
1.8.2.1 Spin-splitting energies 71 
1.8.2.2      Thermodynamic stability 72 
1.8.2.3 Geometries and the size of the ring cavities 73 
1.8.2.4 Reduction Potential 74 
1.8.2.5 Reorganisation energies 75 
1.8.2.6 Bond dissociation energies 75 
1.9 Scope and Objectives of the Research Project 76 
 
Chapter 2 Materials and Methods 
2.1         Materials  78 
2.2         General Methods  80 
2.2.1 pH Measurements 80 
2.2.2 Buffers 80 
2.2.3 Concentration and Desalting of Products 81 
2.2.4 Gel Permeation Chromatography 81 
2.3 Physical Techniques of Characterisation 82 
2.3.1 High Performance Liquid Chromatography (HPLC ) 82 
2.3.2 Fast Atom Bombardment ? Mass Spectroscopy (FAB-MS) 83 
2.3.3 Luminescence Spectroscopy 84 
 10 
2.3.4 Determination of Extinction Coefficient using ICP-OES 84 
2.4 Synthesis of products leading to the formation of NAc-CoIIIMP8 85 
2.4.1 Preparation of FeIIIMP8 85 
2.4.2 Synthesis of demetalated MP8  86 
2.4.3      Synthesis of CoIIIMP8 87 
2.4.4      Synthesis of NAc-CoIIIMP8 87 
2.5         Investigating the Solution Chemistry of NAc-CoIIIMP8 88 
2.5.1 Beer-Lambert Plot of NAc-CoIIIMP8 88 
2.5.2 Data Analysis 89 
2.5.3 Determination of Spectroscopic pKas of NAc-CoIIIMP8 89 
2.5.4 Ligand Binding Studies of Selected N-Donor Ligands by NAc-CoIIIMP8 90 
2.5.5 Kinetic Studies of Selected N-Donor Ligands by NAc-CoIIIMP8 91 
2.5.6 Kinetic Studies of Selected N-Donor Ligands by B12a  92 
 
Chapter 3 Results and Discussion 
3.1 Synthesis and Characterisation of products leading to the  
               formation of NAc-CoIIIMP8 93 
3.1.1 Preparation of FeIIIMP8 93 
3.1.2 Synthesis of demetalated MP8  97 
3.1.3 Synthesis of CoIIIMP8 102 
3.1.4       Synthesis of NAc-CoIIIMP8             107 
3.2 Solution Chemistry of NAc-CoIIIMP8 109 
3.2.1 Determination of the Extinction Coefficients of NAc-CoIIIMP8 109 
3.2.2       Beer-Lambert Behaviour of NAc-CoIIIMP8 110 
3.2.3 Determination of Spectroscopic pKas of NAc-CoIIIMP8 112 
3.2.4 Ligand Binding Studies of Selected N-Donor Ligands by NAc-CoIIIMP8 121 
3.2.5 Kinetic Studies of the Reaction of Selected N -Donor Ligands 
               with NAc-CoIIIMP8 128 
3.2.6 Kinetic Studies of Selected N-Donor Ligands by B12a 132 
 
 
 11 
 
               Chapter 4 Conclusions and Future Work 
4.1 Conclusions  138 
4.1.1 The synthesis and characterisation of products leading to the  
               formation of  NAc-CoIIIMP8 138 
4.1.2 Solution Chemistry of NAc-CoIIIMP8 139 
4.2 Future Work 141 
 
References 142 
Appendices  154 
Appendix A: HPLC Data 154 
Appendix B:  FAB-MS Data 155 
Appendix C:  Beer's Law Data             164 
Appendix D: Spectroscopic pKa Data            166 
Appendix E: Derivation of Rate Constant Equation for NAc-CoIIIMP8        169 
 12 
List of Figures                                                                                                          Page 
 
1.1:  Tentative view of overlapping fields in bioinorganic chemistry.   19 
1.2:  Periodic Table of biological elements in life.  20 
1.3:  The structure bacteriochlorophyll a.  21 
1.4:  The structure of the deoxy haem group.   23 
1.5:  The porphine moiety.  29 
1.6:  A typical absorption spectrum observed for porphyrins.  31 
1.7:  3D Structure of oxidised horse heart cytochrome c.  36 
1.8:  Reaction mechanism of most haem peroxidases.  38 
1.9:  Schematic of catalytic cycle of catalase.  39 
1.10:  Ribbon of P450 camphor-substrate complex structure.  40 
1.11:  The catalytic cycle of cytochrome P450.  42 
1.12:  Schematic representation of deoxy T state Hb structure.   45 
1.13:  The overall structure of myoglobin.  46 
1.14:  The structure of the FeIII-MP8, the ferric haemoctapeptide from cytochrome c.  49 
1.15:  The basic structure of the cobalt corrinoids.  57 
1.16:  The structure of the vitamin B12a.  59 
1.17:  The porphyrin (left) and corrin (right) ring systems.  68 
2.1:  Ion formation from an electrospray ionisation source transferred to  
        mass spectrophotometer.  83 
3.1:  Proteolysis of cytochrome c to FeIIIMP8.  The % conversion is calculated  
        from cytochrome c to FeIIIMP8 and assumes 100 % purity of cyt c.  93 
3.2:  HPLC analysis of FeIIIMP8 (retention time of 6.45 minutes) with > 95% purity   
after purification on a column of Sephadex G50.  94 
3.3:  UV/visible spectrum of FeIIIMP8 in distilled water.  94 
3.4:  FAB-MS spectrum of FeIIIMP8 obtained from the proteolytic digestion  
 of horse heart cytochrome c in NBA matrix.  96 
3.5:  Reductive demetalation of FeIIIMP8 to MP8 Free Base.   The % conversion is 
calculated from FeIIIMP8 to MP8 Free Base and assumes FeIIIMP8 is 100 % pure. 97 
3.6:  HPLC chromatogram of MP8 free base (retention time of 5.99 minutes)  
 with 81 % purity.  98 
 13 
3.7:  UV/visible spectrum of MP8 free base in 50 % methanol/water solution and  
 FeIIIMP8 in water.  98 
3.8:  Enlarged view of FAB-MS MP8 Free Base in NBA matrix.  100 
3.9:  Luminescence Spectra of MP8 free base.  101 
3.10:  Metalation of MP8 Free Base to CoIIIMP8.  The % conversion calculated from 
MP8 Free Base to CoIIIMP8 assuming that the metal-free MP8 is 100 % pure.  102 
3.11:  HPLC chromatogram of CoIIIMP8 (retention time of 6.82 minutes) 
          with 79 % purity.               102 
3.12:  UV/visible spectrum of MP8 free base CoIIIMP8 in 0.001 M  
          NH4(CO3)2 buffer and MP8 Free Base in 50 % methanol/water.  103 
3.13:  FAB-MS spectrum of CoIIIMP8 in NBA matrix.  104 
3.14:  Excitation and emission spectra of CoIIIMP8 and emission spectrum of  
   MP8 free base.  105 
3.15:  Acetylation of CoIIIMP8 to NAcCoIIIMP8.  The % conversion is calculated  
   from CoIIIMP8 to NAcCoIIIMP8 assuming CoIIIMP8 is 100 % pure.  106 
3.16:  HPLC chromatogram of CoIIIMP8 (6.59 minutes) and   
   NAc-CoIIIMP8 (5.99 minutes).  106 
3.17:  UV/visible spectrum of NAcCoIIIMP8 and CoIIIMP8 in 0.001 M  
   ammonium carbonate buffer.  107 
3.18:  FAB-MS spectrum of NAcCoIIIMP8 in NBA matrix.   108 
3.19:  Beer?s law plot for an aqueous solution of NAcCoIIIMP8 (pH 7.00, 25 ?C  
   in distilled water and 1.0 M NaClO 4.  111 
3.20:  pH dependence of NAc-CoIIIMP8 in acidic region.  112 
3.21:  Variation of the Soret absorbance (415.05 nm) with pH titration of  
   NAcCoIIIMP8 in acidic region (pH 2 ? 6) with NaOH.  114 
3.22:  pH dependence of NAcCoIIIMP8 in basic region.  115 
3.23:  Variation of the Soret absorbance (415.05 nm) with pH titration of  
   NAcCoIIIMP8 in basic region (pH 6-13) with NaOH.  116 
3.24:  Species distribution as a function of pH at 25 ?C for NAcCoIIIMP8 in an  
   aqueous multicomponent buffer system.  117 
 
 14 
3.25:  Variation of the Soret absorbance (415.05 nm) with pH titration of  
   NAcCoIIIMP8 with pyridine in basic region (pH 6-13) with NaOH.  120 
3.26:  Plot of the titration data for the binding of N-methylimidazole to  
   NAc-CoIIIMP8 at 420 nm. (25 ?C, ? = 1.0 M)  123 
3.27:  Proposed ligand substitution for NAc-CoIIIMP8.  124 
3.28:  Titration data for the binding of Pyridine to NAc-CoIIIMP8 at 420 nm.  
   (25 ?C, ? = 1.0 M)  126 
3.29:  Proposed kinetics of ligand substitution for NAc-CoIIIMP8.  128 
3.30:  Plot of [NMI] against the observed rate constants for the reaction with                     
          NAc-CoIIIMP8.  129 
3.31:  Proposed kinetics of ligand substitution of NMI with NAc-CoIIIMP8.  130 
3.32:  Plot of [MeNH2] against the observed rate constants for the reaction with              
          NAc-CoIIIMP8.  131 
3.33:  Shows schematic of ligand substitution in B12a.  133 
3.34:  Plot of [NMI] against the observed rate constants for the reaction with B12a.  134 
3.35:  Plot of [MeNH2] against the observed rate constants for the reaction with B12a.  135 
 15 
List of Tables                                                                                                            Page 
 
1.1:  Some metalloproteins and their functions.  27 
1.2:  Different Types of Naturally Occurring Porphyrins.   30 
1.3:  Different cytochromes classified by their haem type. 33 
1.4:  A comparison of the ligand formation constants for various haemproteins,  
 and for ferric MP8.   55 
1.5:  The energy difference (kJ mol-1) between the low spin and intermediate-spin 
 states of [MCor/PorImMe) complexes.   71 
1.6:  Metal- ligand bond distances [pm] of the optimised [MCor/PorIm/Me]  
 complexes in their low-spin ground states.  73 
1.7:  Metal- ligand bond distances [pm] of the optimised [MCor/PorIm2]  
 complexes in their low-spin ground states.  74 
2.1:  Materials used in this study.  78 
2.2:  Common Buffers Used in Experiments.  80 
2.3:  Linear Gradient Elution Program.  82 
2.4:  Instrument Settings - VG70-SEQ by Micromass.  84 
3.1:  Summary of ICP-OES results for NAc-CoIIIMP8 .  109 
3.2:  Calculation of Extinction Coefficients using the Beer?s law.  110 
3.3:  Ionisation of iron-bound water in various haemproteins.  118 
3.4:  Absorbance measurements for the reaction of NAc-CoIIIMP8 with varying 
 concentration of N-methylimidazole. (pH 5.45, ? = 1.0 M, 25 ?C)  122 
3.5:  Absorbance measurements for the reaction of NAc-CoIIIMP8 with varying  
 concentration of pyridine. (pH 5.44, ? = 1.0 M, 25 ?C)  125 
3.6:  Formation constants for NAc-FeIIIMP8, NAc-CoIIIMP8 and B12a.  126 
3.7:  Kinetics of NAc-CoIIIMP8 with NMI.  129 
3.8: Kinetics of NAc-CoIIIMP8 with MeNH2.  131 
3.9:  Kinetics of B12a with NMI.  133 
3.10: Kinetics of B12a with MeHH2.  135 
3.11:  Second order rate constants of NAc-CoIIIMP8 and B12a.  136 
 16 
Abbreviations  
 
A Associative mechanism 
AR  Analytical reagent                   
BDE Bond dissociation energy  
 BP  Biologically pure 
CAPS 3-[Cylohexylamino]-1-propanesulfonic acid  
CD Circular dichroism 
CHES 3-[N-Cylohexylamino]-ethansulfonic acid 
CMD Classical molecular dynamics 
CN (Im)Cbl  Cyanoimidazolyl-cobamide 
Co Cobalt 
Cor Corrin  
CP  Chemically pure       
CR  Commercial reagent          
Cyt c Cytochrome c  
D Dissociative mechanism 
DAC Diaquacobinamide 
ESR Electron spin resonance 
ETC Electron transport chain 
FAB-MS Fast atom bombardment-mass spectroscopy 
GR  General reagent 
Hb Haemoglobin 
HOMO Highest occupied molecular orbital 
HPLC High performance liquid chromatography 
HS High-spin 
I Interchange mechanism 
Ia Associative interchange mechanism 
ICP-OES  Inductively coupled plasma?optical emission spectroscopy 
Id Dissociative interchange mechanism 
Im  Imidazole 
 17 
LMCT Ligand to metal charge transfer 
LS Low-spin 
LUMO Lowest unoccupied molecular orbital 
Mb Myoglobin  
MD  Molecular dynamic 
Me Methyl  
MRI Magnetic resonance imaging 
MM Molecular mechanical 
MES 
MOPS 
4-morpholineethanesulfonic acid monohydrate 
3-[N-Mopholino]-propanessulfonic acid 
MP Microperoxidase 
MP11 Fe(III)-microperoxidase-11  
MP8 Fe(III)-microperoxidase-8  
N-Ac-CoIIIMP8 N-Ac-Co(III)-microperoxidase-8 
N-AcMP8 Acetylated MP8 
NAD(P)H  Nicotine adenine dinucleic coenzyme  
NBA Nitro benzoic acid  
NMR Nuclear magnetic resonance 
PDB Protein data base 
Por Porphyrin  
QM Quantum mechanical 
RDS Rate determining step 
Tris Tris-(hydroxymethyl)-aminomethane 
UV Ultraviolet 
UV-Vis Ultraviolet-visible 
?H?  Enthalpy of activation 
?S? Entropy of activation 
 18 
Chapter 1 Introduction  
1.1 Bioinorganic Chemistry 
1.1.1 General Overview 
 
Although bioinorganic chemistry is a relatively new field, reports on metals bound to 
proteins and enzymes date way back into the nine teenth century and may be found in 
earlier centuries if we replace the terms ?proteins? and ?enzymes? with ?animal or plant 
tissues?.1  Scientists by nature self-sort into groups and disciplines.  The artificial division 
of the study of chemical compounds into organic and inorganic chemistry has long been 
established.  Bioinorganic chemistry seems to contradict this artificial division.  In the 
early nineteenth century chemistry was still divided into ?organic? chemistry, which 
included only substances isolated from living organisms, and ?inorganic? chemistry of 
?dead matter?.  This distinction became meaningless after W?hlers? synthesis of 
?organic? urea from inorganic ammonium cyanate in 1828.  Nowadays, organic 
chemistry is defined as the chemistry of carbon compounds, especially of hydrocarbons 
and their derivatives, with possible inclusion of certain heteroelements such as N, O or S 
regardless of the origin of the material. 2 
 
Bioinorganic or biological inorganic chemistry is the discipline dealing with the 
interaction between inorganic substances and molecules of biological interest.  It spans 
many facets of other fields, since it involves the role, uptake, and fate of elements 
essential for life, response of living organisms to toxic inorganic substances, the function 
of metal-based drugs, the synthetic production of functional models, the production of 
pharmaceutical contrast agents in medical applications, the development of theoretical 
models, and so on. 3  Bioinorganic chemistry is a relatively new branch of chemistry and 
only became an independent field after 1960.4 It is, however, a rapidly expanding 
interdisciplinary field that has developed through contributions from chemical physics 
(characterisation techniques), molecular biology (supply materials ), pharmacology (drug-
 inorganic substance interactions), medicine (diagnostic aids and tumour therapy), 
environmental and human toxicology (toxicity of inorganic compounds), crystallography 
and spectroscopy (characterisation techniques) and chemistry (specifically bio-
 19 
coordination chemistry).1,4  Figure 1.1 is a diagrammatic representation of the far 
reaching, interdisciplinary nature of bioinorganic chemistry.  The growing interest in 
bioinorganic chemistry research has become apparent with the emergence of journals that 
are solely devoted to this field of chemistry.  
 
 
Figure 1.1:  Tentative view of overlapping fields in bioinorganic chemistry.1 
 
 
 20 
1.1.2 Metal Ions in the Chemistry of Life 
 
There are approximately thirty elements essential for life.5  These may occur in 
macroscopic, microscopic or trace amounts.6  Of these thirty elements, many are metal 
ions involved in life-sustaining biological systems and biochemical processes.  These 
metal ions play a pivotal role in about one third of enzymes.7  Ions can modify electron 
flow in the substrate or enzyme, thus effectively controlling an enzyme-catalysed 
reaction.  They can serve to bind and orient the substrate with respect to the functional 
groups in the active site, and they can provide a site for redox activity if the metal has 
several valence states.  Many molecules that are involved in the above mentioned 
processes rely on metal ions for their acivity.8  Evolution has made efficient use of the 
available metals and their properties to allow a range of inorganic reactions to be 
accessible to biological systems.  The periodic table illustrated in Figure 1.2 highlights 
the elements that are necessary for life.   
 
 
Figure 1.2:  Periodic Table of biological elements in life.7 
       Key:  Circled Elements  = Bulk Biological Elements 
                   Boxed Elements = Essential Trace Elements 
                  Dashed Box Elements  = Possible Trace Elements Showing Weaker   
                                                           Evidence for Involvement in Biology 
 
 21 
The term ?essential? is given to an element that would cause severe, irreversible damage 
if it were completely eliminated from the organism.4  Some elements such as As, Br, and 
Sn have not been unequivocally determined to be essential since they are present in such 
low concentrations that experiments are difficult to carry out.  The normal concentration 
range of essential metals is narrow and deficiencies or excesses of such metals cause 
pathological changes in an organism.7   
 
The alkali metals sodium and potassium are involved in the relay of electric pulses in the 
nervous system of a number of animals.  The alkaline earth metals, magnesium and 
calcium, are responsible for teeth and bone structure in animals.  Magnesium is also 
important element in the capture of solar energy by plants.  The metal ion is bound in a 
highly conjugated ring system, the chlorin ring, which is bound to a macromolecule and 
together they form chlorophyll as shown in Figure 1.3.  Chlorophyll and other pigments 
such as the cartenoids, work together to absorb a wide range of energies in the visible 
spectrum.8 
 
N
 N N
 N
 M g
 C H 3
 C H 3
 C H 3
 OC O
 O C H 3
 C H 3
 OC H 3
 C H 3
 CO
 O
 C H 2
 H C C
 C H 3
 C H
 C H 3
 C H
 C H 3
 C H
 C H 3
  
Figure 1.3:  The structure bacteriochlorophyll a. 
 
As this project revolves around a particular question in the biological chemistry of two 
specific transition metals namely, iron and cobalt, the focus in this section is primarily on 
them. 
 22 
1.1.2.1 Aspects of the Chemistry of Iron:  Properties and Special Characteristics  
 
Iron is the most abundant metal both in and on the Earth.9  As far as biology is concerned, 
it is one of the most important metals in the first transition series as it is the transition 
metal found in greatest abundance in biological systems. It constitutes one third of the 
Earth?s mass and is present from crustal rocks, to ocean and river waters, and to the 
organisms that exist in those environments.10, 11  The atmosphere as we know, abundant 
in oxygen, has developed over billions of years from a reducing atmosphere that was too 
hostile to support most living organisms.  The only organisms that could exist were those 
that could survive in an anaerobic environment.  In this reducing atmosphere iron was 
present in the ferrous state and was gradually oxidised to the ferric state as oxygen 
became more prevalent.  This increase in oxygen concentration was brought about the 
blue-green algae and the photosynthetic bacteria.12 
 
 
? Redox, Geometric and Magnetic Properties 
 
The oxidation states of iron have been noted to range from -2 to +4, the most common 
being FeII and FeIII in an aqueous environment.13  The change in atmospheric composition 
as described above, has had a dramatic effect on the biology and biochemistry of iron due 
to the different properties of the most common oxidation states of iron. 
 
An interesting property of iron is its ability to be oxidised/reduced over a wide range of 
potentials depending on its coordination environment.  Redox potentials are an indication 
of the relative stability of the oxidation state of iron in the presence of particular ligands.  
Iron is found in the first row of transition metals in the periodic table.  Since the valence 
electrons are in the outer lying d orbitals, FeII is a d6 and FeIII is a d5 metal ion.14   
 
There are a wide range of geometries adopted by both FeII and FeIII.  The sizes of the 
ferrous and ferric ions also affect the geometry of their complexes.  The increase of 
electrons in d orbitals as iron is reduced causes greater repulsion between those electrons, 
 23 
resulting in a slight increase in the ionic radius.  A structural consequence of this is seen 
in haemoglobin where the difference in the ferric and ferrous ionic radius dictates 
whether the ion will fit into the porphyrin ring, this plays an important role in 
haemoglobin function.15  A key feature of it is the behaviour of an individual iron 
porphyrin complex as it goes from the deoxy to the oxygenated state.  In the deoxy state 
the haem complex is distinctly domed, with the iron atom ~0.5 ? out of the porphyrin 
mean plane and with the Fe-N bond to histidine some 8? off perpendicular as shown in 
Figure 1.4.6  However, when an oxygen molecule  becomes bound  the iron comes into, 
or nearly into, the porphyrin plane and the Fe-N becomes more near to perpendicular. 
 
 
 
 
 
 
 
 
 
 
 
Figure 1.4:  The structure of the deoxy haem group.6  
 
Distortions from octahedral geometries arise when there is uneven distribution of 
electrons in the d orbitals.  This occurs when the symmetry about the metal centre is 
decreased.  All of the above mentioned redox, geometric and magnetic properties of the 
FeII and the FeIII ions play a pivotal role in the structure, function and operation of the 
active sites in haemproteins.6  
 
 
 
 
HAEM GROUP 
 
 
HAEM NORMAL  
 24 
? Special Characteristics 
 
Commonly, FeII (d6) and FeIII (d5) complexes are high spin as quite strong ligand fields 
are required to induce low spin electron configurations.  High spin FeII has a 5D ground 
term, which splits into the ground states 5T2 and 5E.  Ligand field (d-d) transitions (5T2  
5E) in octahedral complexes of FeII occur in the visible region of the electronic spectra.   
 
High spin octahedral FeIII has an orbitally non-degenerate 6A1 ground term aris ing from 
6S ground term.  Due to the single orbital occupancy of high spin FeIII, there are no 
excited states of the same multiplicity and therefore all d-d transitions are spin- forbidden 
to the T states, 4T1 and 4T2, and to a nearly degenerate pair, 4A1  and 4E.  The intensities of 
the ligand field transitions 6A1  4T1 (4T2) are so low that the d-d band transitions are 
rarely observed in the electronic absorption spectra of high spin FeIII complexes.  Simple 
monomeric compounds of high spin FeIII have little or no colour at all.  FeIII is electron 
deficient and can participate in the ligand-to-metal cha rge transfer (LMCT) processes that 
have much greater intensity than d-d transitions.5   Strong colours arise from LMCT or 
intraligand transitions.   
 
In the presence of strong octahedral fields there is a stabilisation of the low spin (S = ?) 
state.  The intermediate spin state of (S = 3/2) may be produced in fields of low symmetry.  
High spin FeII and all spin states of FeIII are paramagnetic and therefore ESR active.  The 
iron atom has a nuclear spin of I = ?.6  Although high spin FeII is paramagnetic, it has an 
even number of electrons.  There is no guarantee of seeing an ESR signal because the 
ground term is not a Kramer?s doublet.   
 
 
 
 25 
1.1.2.2 Aspects of the Chemistry of Cobalt:  Redox and Magnetic Properties 
 
Cobalt is a first row transition metal and is found in oxidation states from +1 to +3.    
CoIII is unstable in simple compounds but is much more stable in larger low spin 
complexes.5  Typical geometries for cobalt range from tetrahedral to octahedral 
throughout the range of oxidation states.14  The electronic structure of CoIII is low spin d6 
except for very weak ligands such as the fluoride ion. 5  Thus, CoIII complexes are 
diamagnetic in the ground state.  CoIII complexes have characteristic electronic spectra 
with bands in the visible region due to trans itions from the ground state (1A1g) to the 
excited singlet states (1T1g and 1T2g) representing the d-d transitions.5  Due to its 
diamagnetic electron configuration, CoIII is ESR inactive. 
 
 
1.2 Metalloproteins and their Significance  
 
Many transition elements have been found to serve as prosthetic groups (which are active 
sites of proteins) or as co-factors in metal activated proteins.15  Numerous essential 
biological functions require metal ions and most of these occur in metalloproteins.  Metal 
ions are critical to the function of many proteins, including stabilising their structure.  In 
fact, one third of all proteins are metalloproteins, using metal ions such as iron, copper, 
zinc, molybdenum and vanadium.16  Metalloproteins are an important class of 
biomolecules that play a central role in many biochemical processes, including 
respiration, metabolism, nitrogen fixation, photosynthesis, nerve transduction, muscle 
contraction, signal transduction, and protection against mutagenic agents.  They also 
serve as catalysts in a wide variety of reactions ranging from electron transfer to the 
insertion of oxygen into carbon-hydrogen bonds.  Thus, the role of the metal ions in 
metalloproteins is two fold, as it serves both functional and structural roles.17, 18  In 
performing these actions the metal can function alone, in a cluster or associated with 
complex ring systems.18  
 
 26 
Without the appropriate metal ion, a reaction catalysed by a particular metalloenzyme 
would proceed very slowly if at all.19  The enzymes provide an arrangement of amino acid 
side-chain functional groups which orient themselves to form an appropriate sized hole 
(depending on the metal) to bind the metal.  In addition, preferred groups on the enzyme 
side chains further bind the required metal ion to increase stability.  The optimal number 
of such binding groups is chosen for the particular metal ion, together with an 
environment of appropriate polarity in the binding site.  Metal ions may be bound by 
main-chain amino and carbonyl groups, but specific binding is achieved by the amino 
acid side chains, particularly the carboxylate groups of aspartic and glutamic acid, and 
the ring nitrogen atom of histidine.  Other side chains that bind metal ions include 
tryptophan (ring nitrogen), cysteine (thiol), methionine (thioether), serine, threonine, 
tyrosine (hydroxyl groups), and asparagine and glutamine (carbonyl groups, less often 
amino groups).  There are no general set of rules that describe how a metal will behave in 
an enzyme.20  In essence the function of the protein is to anchor, control and modify the 
fundamental chemistry of the metal. 
 
In recent years, the analysis and classification of metalloenzymes and metalloproteins at 
the interface between chemistry and biology has accelerated.  Although a variety of 
spectroscopic techniques have been used for many years to probe the metal centres in 
metal containing proteins, progress is particularly due to the recent developments in x-ray 
diffraction methods, mass spectroscopy and NMR spectroscopy.  Table 1.1 shows some 
of the metalloproteins that have already been elucidated. 
 
 
 
 
 
 
 
 
 
 27 
Table 1.1:  Some metalloproteins and their functions.18 
 
 Metal  
 
Protein 
 
Function 
Azurin Electron transport protein in 
Alcaligenes xylosoxidans  
Rusticyanin Electron transport protein in 
Thiobacillus ferrooxidans  
Nitrite Reductase Reduces nitrite in the 
denitrification pathway of 
Alcaligenes xylosoxidans  
Superoxide 
Dismutase 
Scavenges superoxide free 
radicals 
Ceruloplasmin  A ferrooxidase found in the blood 
 
 
 
 
 
Cu 
 
 
 
 
 
Haemocyanin 
 
Oxygen transport in the blood of 
molluscs and arthropods 
Haemoglobin Oxygen transport in the blood 
Myoglobin Oxygen storage in muscle tissue  
Transferrins  Iron transport protein found in the 
blood  
FixL  Haem oxygen sensor protein 
symbiotic bacterium Rhizobium 
meliloti  
 
 
 
 
 
 
 
 
Fe 
 
Cytochromes 
 
Participate in important electron 
transfer reactions in ETC 
 28 
 
Peroxidase 
 
Protect the body against 
uncontrolled oxidation by 
hydrogen peroxide 
 
 
 
 
Fe Cytochrome P450 Activation of molecular oxygen 
with insertion of one of its atoms 
into the substrate and reduction of 
the other to form water 
Nitrogenase Catalyses the ATP dependent 
reduction of dinitrogen to 
ammonia  
ModA Periplasmic molybdate binding 
proteins Azotobacter vinelandii 
 
 
Mo  
 
 
 
ModE Molybdate dependent 
transcriptional regulator from 
Escherichia coli 
Mn Catalase Reduces hydrogen peroxide to 
molecular oxygen and water 
Ni 
 
Urease 
 
Hydrolysis of urea to ammonia 
and carbon dioxide 
 
V Nitrogenase 
Catalyses the ATP dependent 
reduction of dinitrogen to 
ammonia 
Zn 
Superoxide 
Dismutase 
Scavenges superoxide free 
radicals  
 29 
1.3 Porphyrins  
1.3.1 Basic Structure  
 
The basic structures of porphyrin consist of four pyrrole units joined by methine bridges. 
In a haemprotein, the iron binds to the porphyrin ring and is also covalently linked by 
axial binding to at least one endogenous ligand. These macromolecules are relatively 
rigid, but flexible, and the porphyrin cavity is able to bind nearly all metal ions.  They are 
ideal building blocks for self-assembled structures due to their pre-organised geometry, 
ease of substitution and selective axial ligand coordination chemistry. Porphyrins are 
susceptible to several chemical reactions typical of aromatic compounds.  The porphyrin 
ring is a planar moiety belonging to the D4h space group with an extensive delocalised p-
 electron system.21  The porphyrin macrocycle is an aromatic system containing 22 p-
 electrons, but only 18 of them are involved in any one delocalization pathway and thus 
conform to the Huckel?s (4n + 2) rule for aromaticity.  Many haemproteins contain 
protoporphyrin IX.  Table 1.2 shows the different porphyrins that arise from the 
substitution that occurs at the peripheral positions in the stable porphine shown in figure 
1.5.21  
 
N
 NH N
 HN 1
 2
 3
 46
 20
 7
 11
 10
 9
 8
 14
 13
 12
 16
 15 17
 18
 19
 5  
Figure 1.5:  The porphine  moiety. 
 
 
 
 
Cmeso 
Ca  
C? 
 30 
Table 1.2:  Different Types of Naturally Occurring Porphyrins.21  
Porphyrin Name  Substituents at C? Position No. 
2      3      7      8      12      13      17     18 
Porphyrin a M     P      P     F      M      Hfar   M      V 
Porphyrin b (porphyrin IX) M     P      P    M      M      V        M     V 
Porphyrin c M     P      P     F      M      C        M      C 
Deuteroporphyrin IX M     P      P     F      M      H        M      H 
Mesoporphyrin M     P      P     F      M      E        M      E 
Coproporphyrin IX M     P      P     F      M      P        M      P 
Hematoporphyrin M     P      P     M     M     SA     M       SA 
Uroporphyrin II A      P      P     A      A      P       A       P 
Aetioporphyrin III M     E      M     E      M      E     M      E 
Chlorocruporphyrin IX M     P      P     M      M      F      M      V 
  
Key:  A = -CH2COOH    C = Cysteinyl S 
E = -C2H5     F = -CHO 
H = -H  Hfar = C17H29O 
M = -CH3  P = -CH2CH2 COOH 
SA = -CH(OH)CH3   V = -CHCH2 
 
 
In addition to the several electrophilic substitutions that can be performed on porphyrins, 
porphyrins can be metalated and demetalated.  Different metals (for example, Zn, Mn, 
Cu, and Ni) can be inserted into the porphyrin cavity using various metal salts.  These 
metals can then be removed with various acids.21 
 
 
 
 
 31 
 
Figure 1.6:  A typical absorption spectrum observed for porphyrins.22 
 
One of the most distinctive properties of haemproteins is their absorption of visible and 
ultraviolet (UV) light.23  The three spectral bands that appear in the visible region are due 
to transitions between p* (vacant orbitals) and p (occupied orbitals).  The electronic 
spectra are dominated by the absorption of the haem moiety, which allows two porphyrin 
p  p* transitions that are mixed together by interelectronic repulsion to give an intense 
signal at 410 nm (called the ? or Soret band) and two less intense signals between 500 nm 
and 600 nm (called a-and ?-bands).  These bands can be explained using Simpson?s 
?perimeter free-electron? model where the porphyrin is idealised as an 18-membered 
circular ring.  Figure 1.6 shows a typical absorption spectrum observed for porphyrins .    
However, more recently, there have been a number of reviews that have described the 
success of the four-orbital model which is based on the earlier free-electron model and 
the simple Huckel?s molecular-orbital model.23  The variations of the peripheral 
substituents on the porphyrin ring often cause minor changes to the intensity and 
wavelength of these absorptions. 22, 23 
 
 
 
 32 
1.3.2 Iron Porphyrins  
 
Iron porphyrins form the prosthetic group (active site) of the biologically important 
haemproteins.  These haemproteins are macromolecules that play very important roles in 
various biological processes, which include oxygen transportation and storage in living 
tissues (as in haemoglobin and myoglobin), the reduction of hydrogen peroxide (as in the 
catalase and peroxidase enzymes), the reduction of oxygen (as in oxidases), the transport 
of electrons (as in the cytochromes b and c in the electron transport chain), and the 
oxygenation of organic substrates (as in cytochrome P-450).  More specifically, 
haemoglobin and myoglobin serve as reversible oxygen transfer proteins, oxidases and 
peroxidases are involved in irreversible, covalent transformation of substrates, and the 
cytochromes b and c function as reversible one -electron transfer agents.  Even though 
these haemproteins have diverse functions, all haemproteins possess a common prosthetic 
group composed of an iron porphyrin complex.  Although the remainder of the protein 
plays a key role in modulation of their biological activity, there is evidence that the 
biological function of the haemproteins is determined by the conformational and  
electronic properties of the haem group , modulated by the protein.23   
 
 
1.4 Haemproteins  
 
Haemproteins (which contain iron porphyrins) are ubiquitous in nature and serve many 
roles as mentioned above.  They can be classified according to their biological function, 
by haem iron coordination, by haem type, and further by sequence similarity.  Studies 
have showed that the hydrophobic interior, hydrogen bonding interactions, and polar and 
non-polar interactions play an integral role in the correct folding around the haem.  
Consequently, there has been much experimental and theoretical effort directed towards 
the elucidation of the detailed nature of the haem group.  Haemproteins are the only 
proteins to have an intimate relationship between structure, spectra and function.  Thus, 
there exists a well-defined relationship between properties of the active site and the 
function of the protein.24 
 33 
1.4.1 The Cytochromes 
 
Cytochromes are haemproteins whose specific bio logical function is electron and/or 
proton transfer by virtue of a reversible valency change of their haem iron.25  They are 
known to play a major role as electron carriers in the electron transport chain (ETC). 
Although this definition takes into account both biological and chemical features and is 
wide enough to encompass both non-autoxidizable and autoxidizable cytochromes, it is 
however too wide to include the catalase and peroxidase group of haemproteins. There 
are however seven main groups of cytochromes that can be distinguished by their haem 
type (see Table 1.3).26    
 
 
 
Table 1.3:  Different cytochromes classified by their haem type.26 
Haem type  Example Haem Structure 
Haem a Cytochrome c oxidase 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
N
 N N
 N
 Fe
 O OHO OH
 HO
 O
 34 
N
 N N
 N
 Fe
 O OHO OH
 S S
 Cys Cys
 Haem b 
 
 
 
 
 
 
 
 
 
 
Cytochromes b 
 
 
 
Haem c 
 
 
 
 
 
 
 
 
 
 
Cytochrome c  
 
 
 
 
 
 
 
 
 
 
Haem d Cytochromes d 
 
 
 
 
 
 
 
 
 
N
 N N
 N
 Fe
 O OHO OH
 N
 N N
 N
 Fe
 O OHO OH
 OH
 OH
 35 
Haem d1 Cytochrome cd1 nitrite 
reductase (d1 domain) 
 
 
 
 
 
 
Haem o Cytochrome  o oxidase 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Haem P460 Hydroxylamine 
oxidoreductase 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
N
 N N
 N
 Fe
 O OHO OH
 O
 O
 HO
 O
 O OH
 N
 N N
 N
 Fe
 O OHO OH
 HO
 N
 N N
 N
 Fe
 O OH
 O
 OH
 S
 S
 Cys
 Cys
 OH
 36 
 
 
 
 
 
 
 
Figure 1.7:  3D Structure of oxidised horse heart cytochrome c.27 
                                        PDB code: 1AKK 
 
In this project cytochrome c (cyt c) was used as the starting material for the preparation 
of microperoxidase 8, so it is important to highlight its significance in nature.  Figure 1.7 
illustrates the 3D structure of cyt c. Cyt c mediates single electron transfer between 
integral membrane complexes in the respiratory chain of eukaryotes.18  It is a protein 
ubiquitous to all eukaryotic organisms and is found associated with the inner 
mitochondrial membrane.  There have been numerous structural studies elucidating the 
structure of cyt c using both X-ray crystallography and NMR spectroscopy.18, 28  Although 
there is a vast amount of work that has been published on cytochromes, this project is not 
focussed on the biochemistry nor the bioenergetics of the cytochromes,  but rather on the 
haem component of the cytochromes. 
 
 
1.4.2 The Peroxidases and Catalases  
Peroxidases and catalases are distributed widely in animals and plants.  However, 
peroxidases are found more commonly in plants.  
? The Role and Biochemistry of Peroxidases 
 
Peroxidases are enzymes which play an important role in large and diverse numbers of 
physiological processes in organisms including humans. They are haem-containing 
enzymes that use hydrogen peroxide (H2O2) or other substrates as the electron acceptor to 
 37 
catalyze a number of oxidative reactions. As such, they are classified as oxidoreductases. 
Toxic molecules such as superoxide and hydroxyl radicals can be found in cells due to 
the presence of oxygen. These are by-products of aerobic respiration. They are eliminated 
by a number of enzymes present inside the cell. Superoxide, for example, is destroyed by 
superoxide dismutase. The degradation, however, produces more hydrogen peroxide, 
which is, in turn, destroyed by peroxidase. Peroxidases reduce H2O2 to water while 
oxidizing a variety of substrates. Thus, peroxidases are oxidoreductases that use H2O2 as 
electron acceptor for catalyzing different oxidative reactions. The overall reaction is as 
follows:  
 
ROOR' + electron donor (2e-) + 2H+ ?  ROH + R'OH             (1.1) 
For many of these enzymes the optimal substrate is hydrogen peroxide, but others are 
more active with organic hydroperoxides such as lipid peroxides. The nature of the 
electron donor is very dependent on the structure of the enzyme. For example, 
horseradish peroxidase can use a variety of organic compounds as electron donors and 
acceptors. Horseradish peroxidase has a broad and accessible active site and many 
compounds can reach the site of the reaction. For an enzyme such as cytochrome c 
peroxidase, the compounds that donate electrons are very specific, because there is a very 
closed active site. Cytochrome c peroxidase is used as a soluble, easily purified model for 
cytochrome c oxidase. 4, 29 
On the basis of sequence and structure similarity, haem-dependent peroxidases can be 
separated into two families: the family of animal peroxidases and the family of plant, 
fungal and bacterial peroxidases.30, 31   
 
 
 
 
 
 
 
 
 
 38 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 1.8:  Reaction mechanism of most haem peroxidases.4, 29 
 
 
In this mechanism, the enzyme reacts with one equivalent of H2O2 to give compound I, a 
porphyrin  cation radical containing FeIV. This is a two-electron oxidation/reduction 
reaction where H2O2 is reduced to water and the enzyme is oxidised. One oxidising 
equivalent resides on iron, giving the oxyferryl (FeIV=O) intermediate and the second 
resides on the porphyrin which becomes a p?radical cation.  Compound I then oxidises an 
organic substrate to give a substrate radical (?AH). Compound I undergoes a second one--
 electron oxidation reaction yielding compound II, which contains an oxyferryl centre 
coordinated to a normal (dianionic) porphyrin ligand. Finally, compound II, is reduced 
back to the native ferric state with concomitant one electron substrate reduction. The 
overall charge on the resting state and compound I is +1, while compound II is    
neutral.4, 29 
 
 
 
 
       AH  + H +    AH2 
F e I I I 
H N 
N 
R e s t i n g   E n z y m e 
F e I V 
O 
H N 
N 
I 
+ F e I V 
O 
H N 
N 
I I 
H O O H 
H 2 O 
   
A H   +  O H - 
A H 2 
    
 39 
? The Role and Biochemistry of Catalases 
Catalase is a haem-containing enzyme that catalyses reaction (1.2). The R represents 
hydrogen, an alkyl or acyl group and the HQOH is a two electron donor.  Catalases can 
utilise hydrogen peroxide (H2O2) both as an electron acceptor and an electron donor 
yielding molecular oxygen (O 2) and water in the disproportionation reaction (1.3).  
ROOH + HQOH  QO + ROH + H2O  (1.2) 
 
H2O2 + H2O2  2H2O + O2                      (1.3)  
The reaction cycle of the catalases begins with the high spin ferric (FeIII) state which 
reacts with the molecule of peroxide to form compound I intermediate, a porphyrin p 
cation radical containing FeI V.  This is followed by oxidation of an electron donor which 
returns compound I to the native resting state.  Both the resting state and compound I of 
catalase are neutral. 
 
 
 
 
 
 
 
Figure 1.9:  Schematic of catalytic cycle of Catalase. 
Most catalases have haem b as the prosthetic group; however, a number of fungal and 
bacterial catalases contain chlorin type haem (haem d).32-34  In all catalases the haem iron 
F e I I I 
O 
R e s t i n g   E n z  y m e 
F e I V 
O 
I 
+ 
O 
H O O H H 2 O 
H O O H H2O + O 2 
 40 
fifth (proximal) ligand is a Tyr residue; a His residue essential for catalysis is located on 
the distal side of the haem. It is thought that its function is to deprotonate H2O2, which is 
bound to H+ (on His) and HO2- (on FeIII).  In all catalase structures, a water molecule 
located close to the sixth coordination of the haem has been observed.35-38  
 
1.4.3 The Cytochromes P450 
 
Cytochromes P450 (P450s) are a superfamily of haem-thiolate enzymes that mainly 
catalyse the monooxygenation of a wide array of apolar substrates.18 This is a two 
electron oxidation of substrate with one of the dioxygen derived O atoms inserted into a 
C-H bond of the substrate.39 Cytochromes P450 are isolated from all living organisms, 
including bacteria, fungi, plants, insects and verterbrates.40 P450s are involved in 
metabolising a wide variety of hydrophobic compounds. In mammals, P450 substrates 
range from endogenous substances such as cholesterol, steroids, and lipids, to exogenous 
ones like environmental pollutants and drugs.   
 
 
 
 
 
 
 
 
 
 
 
Figure 1.10:  Ribbon of P450 camphor-substrate complex structure.41 
                                       PDB code: P450BM-3 
 
 
 41 
? Physical Properties 
The P450s are haemproteins that are made up of between 400 and 500 amino acids and 
contain a single haem prosthetic group.42, 43 The distal axial ligand of this haem moiety is 
formed by a cysteine residue. In the haem system, six- fold coordinated FeIII is generally 
found to be LS and five- fold coordinated FeIII is found in the HS state.44  This is due to 
the fact that the ionic radius is larger for HS FeIII than for LS45 and in the HS state the 
FeIII moves out of the plane of the porphyrin ring as the central cavity is too small. 
Furthermore, it was also thought that the five coordinate FeIII moves out of the porphyrin 
cavity to maximise orbital overlap of the donor atom of the proximal ligand.  The nature 
of the axial haem ligands also has an important effect on the LS-HS balance. A strong 
axial field will bring about a relatively large d-orbital splitting, favouring the LS state. 
This can be seen in practice if one compares the UV absorption spectra of CO complexes 
with haemoglobin (histidine axial ligand) and P450 (cysteine ligand). The axial ligand 
field strength due to histidine is much stronger than that due to cysteine46 and 
correspondingly the UV absorption band due to the d  d transition in CO-hemoglobin 
occurs at 420 nm, while that due to CO-P450 occurs at 450 nm.47  
Ligand-free P450s exhibit a Soret absorption maximum in a UV spectrum at 
approximately 420 nm. This is associated with the LS state of the FeIII. Spectral48, NMR49 
and crystallographic50 data indicate that a water molecule forms a sixth axial ligand of the 
FeIII in the substrate-free form, thus stabilising the LS state of the ion.  The HS and LS 
states are not independent, but exist in equilibrium. The differences in the absorption and 
extinction coefficients between the two bands at 390 and 420 nm allow the equilibrium 
constant between the spin states and hence the fraction of HS character to be determined. 
The equilibrium between HS and LS is only one of a number of microequilibria which 
must be considered in a full description of the substrate binding reaction.50-53  The 
equilibrium between these states may be affected by many factors such as the pH of the 
solution or changes in the conformation of the enzyme which alter the nature of the 
binding of a ligand.  
 
 42 
? Catalytic Cycle 
The principal physiological role of the P450 superfamily of enzymes is that of a 
monoxygenase. The catalytic reaction can be summarised as follows 
RH + O2 + 2H+ + 2e -  ROH + H2O     (1.4) 
where RH can be one of a large number of possible substrates.  
 
     
 
 
 
 
 
 
 
 
 
 
Figure 1.11:  The catalytic cycle of cytochrome P450. 
* The intermediate states enclosed in a dashed box have not been directly observed and are hypothetical. 
 
(1)  Substrate binding  
The binding of a substrate to a P450 causes a lowering of the redox potential by 
approximately 100 mV,54, 55 which makes the transfer of an electron favourable from its 
redox partner, NADH or NADPH. This is accompanied by a change in the spin state of 
the haem iron at the active site.  It has also been suggested that the binding of the 
substrate brings about a conformational change in the enzyme which triggers an 
interaction with the redox component.55 
 43 
(2) The first reduction  
The next stage in the cycle is the reduction of the FeIII ion by an electron transferred from 
NAD(P)H via an electron transfer chain. 
(3) Oxygen binding  
An O2 molecule binds rapidly to the FeII ion forming FeII + O2. There is evidence to 
suggest that this complex then undergoes a slow conversion to a more stable complex 
FeIII ? O2-.56  A comparison between the bond energies of O2, O2-, and O22- suggest that 
the FeIII?O22- complex is the most favourable starting point for the next stage of the 
reaction to occur.57   
(4) Second reduction  
A second reduction is required by the stoichiometry of the reaction. This has been 
determined to be the rate-determining step of the reaction.58 However, evidence from 
resonance Raman spectroscopy indicates the presence of a superoxide (O2-) complex.59 
There is evidence to suggest that this complex then undergoes a slow conversion to a 
more stable complex FeIII ? O22-.56 
(5) O2 Cleavage  
The O22- reacts with two protons from the surrounding solvent, breaking the O-O bond, 
forming water and leaving an (Fe-O)3+ complex. 
(6) Product formation  
The Fe-ligated O atom is transferred to the substrate forming a hydroxylated form of the 
substrate. 
(7) Product release  
The product is released from the active site of the enzyme which returns to its initial 
state.  
 44 
The structures of the transitional states following processes (4) and (5) have never been 
directly observed and are hypotheses based by analogy to other haemproteins and often 
conflicting experimental evidence.  
 
1.4.4 Haemoglobin (Hb) and Myoglobin (Mb) 
 
? Biological Roles  
  
It is well known that invertebrates and vertebrates have different dioxygen coordination 
strategies.  Before dioxygen can be metabolised, it has to be taken up (reversibly) from 
the atmosphere and transported to oxygen-depleted tissues where it must then be stored 
until needed.4   
 
The physiological roles of Hbs in vertebrates are the transport of oxygen from the lungs 
(or gills in fish) to the body tissues, enhancement of carbon dioxide transport in the 
opposite direction and the regulation of blood pH.  Since each haem can bind one oxygen 
molecule, four molecules of oxygen bind to the Hb tetramer.  The affinity for the first 
oxygen is low but the affinity rises with the number of bound oxygen molecules.  This is 
known as the haem-haem interaction or homotropic allosteric effect.18   
 
The physiological roles of Mbs are to buffer the oxygen concentration in the respiring 
tissues.  It is a haemprotein that contributes to intracellular oxygen storage and 
transcellular facilitated diffusion of oxygen.   It is normally found in muscle tissue where 
continuous oxygen is supplied.  The reversibility of oxygen binding is preserved only 
when the haem is kept in the ferrous state which is stabilised by the hydrophobic 
environment of the protein.60  The affinity of myoglobin for oxygen lies between that for 
Hb, which releases oxygen during its passage through respiring tissues, and for the 
cytochromes that make use of molecular oxygen in oxidative respiration.18  
 
 
 45 
? Molecular Structures 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 1.12:  Schematic representation of deoxy T state Hb structure.61 
                                      PDB code:  1HGA 
 
In 1959, Perutz obtained the first electron density map of horse oxyHb, which resulted in 
the determination of its crystal structure.62, 63  Hb consists of four protein units, each of 
which has an iron protoporphyrin IX group.  Two subunits have the same amino acid 
sequence and are designated as a-subunits.  The other two are designated as ?-subunits; 
they are similar but not exactly the same as the a-subunit s.  Despite their differences 
between amino acid sequences, both subunits have the same fold, which is completely 
made up of a-helices. Thus, the four-unit composition of Hb serves a pivotal biological 
role as mentioned earlier.  Hb has an overall compact globular structure with a diameter 
of 50-55 ?, while its subunit is approximately 47 x 35 x 22 ?3 in size.  The helices have 
been traditionally named A through to H from the N- to the C-terminus.18 
 
 
 
 
 
 46 
 
 
 
 
 
 
 
 
Figure 1.13:  The overall structure of myoglobin.64 
                                                     PDB code:  2MBW 
 
In 1960, Sir John Kendrew and his co-workers published the first solved crystal structure 
of Mb and consisted of a number of right-handed a-helices.65  It was the first protein 
structure to be revealed at the atomic level.  Mb consists of a single peptide chain 
combined with an iron porphyrin.  It comprises eight helices packed in the pattern tha t 
leaves a pocket for haem. These helices are typically named from A to H.  This helical 
haem-binding domain is known as the globin fold.  The tertiary structure of the protein is 
kept intact by sulphide bridges, hydrogen bonds, van der Waals forces and ionic 
interactions.66  The overall shape of the molecule is best described as an oblate ellipsoid 
with dimensions 45 x 45 x 25 ?3.18  
 
 
1.5 The Microperoxidases (MPs) 
1.5.1 Background   
 
One of the greatest challenges facing bioinorganic chemists and biochemists is  
explaining how the protein controls and modifies the basic properties of the active site to 
give rise to such wide-ranging functions.67  One method of doing this is to compare the 
properties of protein- free porphyrins and the haemproteins.  However, with this approach 
come two main complications.  The first is solubility limitations, where porphyrins like 
haemin have limited solubility in aqueous and neutral solutions68 and tend to aggregate in 
 47 
aqueous alkaline solutions.69  Thus work had to be performed in non-aqueous solvents70-
 73 and detergents74, 75 to ensure a monomeric system.  These non-aqueous solvent 
conditions for studying the physical and chemical properties of haemproteins are 
inadequate and are not physiologically relevant.  The second complication is that many 
haemproteins possess either one or two histidine residues as axial ligands of the metal 
ion; these include Hb, Mb, peroxidases and the different cytochromes c, b, f and 
cytochrome c oxidase.67  The addition of ligands like imidazole and its derivatives to 
solutions of iron porphyrins almost always results in the formation of the bis- imidazole 
complex with only transient existence of the mono - imidazole species because of the 
thermodynamic drive from HS to LS.76, 77   
 
Thus, the need arose to synthesise model haemproteins whose iron porphyrin active site 
incorporated the ligand combination found in the protein itself.  Various model 
haempeptides with a variety of axial ligands were synthesised.  These include (i) 
haempeptides prepared by adding small mono-, di-, and tri-peptide groups to the 
propanoic acid side chains or other suitably functionalised groups of the porphyrin; (ii) 
synthetic haempeptide fragments in which the invariant Cys14-X15-X16-Cys17-His18 
sequence of the cytochrome c is attached to the haem periphery and (iii) the range of 
haempeptide fragments that may be obtained by treating cytochrome c with suitable 
proteolytic enzymes.67  The haempeptides in (ii) and (iii) incorporate a His residue as an 
axial ligand, and therefore afford the possibility of mixed ligation of the haem iron (i.e. 
the formation of the mono- and bis-complexes together).67  However, the flexible ligand-
 bearing side-chains attached to the porphyrin periphery has the tendency to dimerise and 
form higher aggregates in solution.78   
 
 
1.5.2 Preparation of Microperoxidases 
 
Microperoxidases (MPs) are haem-based minenzymes that form a new generation of 
biomimics.  They consist of a haem c cofactor covalently bound to an oligopeptide, 
which contains histidine, the axial ligand to the haem iron. It is obtained by controlled 
 48 
proteolytic digestion of horse heart cytochrome c.  It is an appealing model for 
mimicking the structure and the function of the active site of haem-containing enzymes 
since it retains an axial histidine ligand.79  The haempeptides provide an opportunity for 
exploring aspects of the fundamental chemistry of an iron(III) porphyrin in which one of 
the axial coordination sites of the metal ion is occupied by a His residue which is from 
the parent protein and the other position contains a readily replaceable water molecule.67  
This provides an opportunity for quantitatively studying in aqueous solution the 
coordination chemistry of an iron porphyrin with a single accessible coordination site.80   
Like many iron porphyrins, microperoxidase-8 is monomeric only at very low 
concentrations in aqueous solution.81, 82   
 
Although MP8 is commercially available, it is expensive.  The hydrolysis of cytochrome 
c was historically performed to help determine the structure of the parent protein.83-85  
The first haempeptides that were isolated MP11 (residues 11-21), MP9 (residues 14-22) 
and MP10 (residues 13-22).85-87  Further proteolysis resulted in the formation of MP8 
(residues 14-21) and MP6 (residues 14-19).88, 89  The peptides are usually prepared by 
digestion of ferric cytochrome c with appropriate enzymes (in this project for the 
preparation of MP8, the enzymes trypsin and pepsin were used).  This is then followed by 
size-exclusion or ion exchange chromatographic techniques to separate the different MPs 
(in this project we separated the different MPs using a Sephadex G-50 size exclusion 
column).  Analytical high performance liquid chromatography (HPLC) can be used to 
purify MP preparations effectively.  The products can be characterised or identified using 
a multitude of   characterization techniques including amino acid analysis90, FAB-MS, 
liquid secondary ion mass spectrometry91, UV-vis spectrophotometry67 and NMR.92   
 
Recently, there has been a considerable amount of work and development in the area of 
biomimetic models for haemproteins. Most of the chemical and biochemical models 
show catalytic activities closely related to the activities of the proteins they mimic. Thus, 
metalloporphyrin models can act as oxygen binding and/or oxygen transport molecules; 
they can transfer electrons, and also catalyse peroxidase and/or monoxygenase 
 49 
reactions.80  The biomimetic modelling approach is the key to unlocking many facts and 
theories that scientists dealing with proteins have encountered through the decades. 
 
 
 
 
 
 
 
 
 
Figure 1.14:  The structure of the FeIII-MP8, the ferric haemoctapeptide from 
        cytochrome c. 
 
 
1.5.3 Solution Behaviour of the Microperoxidases 
1.5.3.1 Solution Structure  
 
Although there have been many attempts, none of the MPs have been crystallised; thus 
no solid state data are available.  However, there has been a study of the 1H NMR 
spectroscopic study of the monomeric cyanide complex MP11 which has suggested that 
most of the amino acid side chain from Lys -13  to Glu-21 lies in an a-helix conformation 
beneath the coordinated FeIII atom and the proximal His 18 ligand in the shift cone 
produced by the diamagnetic ring current of the porphyrin ring.93  This suggestion was 
later confirmed by Mondeli et al.94 who studied monomeric MP11 in the paramagnetic, 
FeIII low spin state by NMR-restrained molecular dynamics calculations in water-
 trifluoroethanol and water-methanol mixtures with various iron distal ligands at different 
temperatures.  Ten structures obtained showed the polypeptide side chain preferred the 
right-handed helical secondary structure.  
 
 50 
In another study involving MP11 it was found that at low pH, His-18 becomes protonated 
and does not coordinate FeIII.  As the solution pH is raised His-18 is deprotonated and 
coordinates FeIII with a pKa of 3.4 at 25 ?C.95  In addition to this a second pKa of 5.8 was 
found in the acid region which was attributed to the deprotonation and intermolecular 
coordination of either the a-NH2 group of Val-11 or the e-NH2 group of Lys-13.95  
However, more recently there were three pKas  observed in the acidic region and the 
neutral region for MP11 at 3.4, 5.8 and 7.6 at 20 ?C, and attributed to the deprotonation 
and intermolecular coordination of His-18, to the deprotonation and intermolecular 
coordination of a-NH2 group of Val-11 and to the deprotonation and intermolecular 
coordination of the e-NH2 group of Lys-13, respectively.96   
 
Most studies reported on MP11 reveal a severe problem due to intermolecular 
coordination and aggregation of the haempeptide, and compromised the findings.68  
Similarly, physical studies on aqueous solutions of MP8 show its behaviour is also 
complicated by the aggregation phenomena and the solutions of MP8 have been found to 
deviate from Beer?s Law as a result of dimerisation at concentrations as low as 2 ?M at 
pH 7.00 at 25 ?C91; since the intensity of the Qv band at 494 nm in the visible region of 
the electronic spectra increased at the expense of the charge-transfer band at 622 nm, 
which is indicative of the conversion of the high-spin aqua complex to the low spin 
species, thus is was concluded that dimerisation by intermolecular coordination of a -NH2 
group of the Cys-14 was likely.  Supporting this evidence were the results obtained by 
Urry and Pettegrew97 who used optical rotary dispersion (ORD), circular dichroism (CD), 
and difference absorption spectroscopy with a range of solutions of MP8.  It was found 
that dimerisation of MP8 also involved p-p interactions between the haem groups.  
 
 
1.5.3.2 Aggregation in Aqueous Solution 
 
In 1955, Ehrenberg et al. demonstrated that at pH 8.7 the haempeptides became 
extensively aggregated with an apparent molecular weight of 10 000.88  These workers 
also showed that at pH 2.3, no such aggregation occurred and proposed that this was due 
 51 
to the intermolecular coordination of a-NH2 group of the terminal Val-11 residue.  
Furthermore, the addition of exogenous ligands such as His, ammonia and imidazole, 
which compete for the distal coordination site was shown to disperse the aggregation.88,89  
Further evidence for pH dependant aggregation phenomena of MP11 was confirmed by 
Harbury and Loach89, who demonstrated that the  aggregation was concentration 
dependent.  The authors also discovered that the acetylation of Val-11 and the e-NH2 
group of Lys-13 eliminated aggregation, which suggested that the observed aggregation 
was due to intermolecular coordination of the amino groups of the peptide chains.88, 89   
 
In addition, further evidence that aggregation of MP11 in aqueous solution occurs due to 
intermolecular coordination of the free amino groups of the polypeptide chains was 
provided by magnetic studies90, 98-101, which all show that Fe(III) is low spin above pH 6.  
Aron et al.102 studied the aggregation of MP8 (due to the intermolecular coordination of 
the N-terminal Cys-14 residue ) qualitatively in aqueous and aqueous methanolic solutions 
at pH 7.00 and 12.00 by observing deviations from Beer?s Law that occurs in the 
concentration of MP8 increased.  The workers proposed that the deviations could be 
accounted for by a simple equilib rium between two monomers and a dimer.102  Similar 
results were obtained by Baldwin et al.103 who demonstrated that MP9 is monomeric  in 
50% aqueous methanol solution up to a concentration of 35 ?M.  It was thought that 
aggregation of MP9 occurred due to intermolecular coordination of Lys-22 to the FeIII 
centre of an adjacent molecule.104   
 
It has well been established that the aggregation of MP8 and MP11 aqueous solutions 
occur due to non-covalent p- interactions between the distal faces of two or more 
haempeptides.97, 105  Urry and coworkers97, 105 studied this second form of aggregation by 
means of optical rotation, CD and absorbance spectroscopy, and demonstrated that 
aggregation could be minimised by both dilution of the haempeptide and by elevated 
temperatures.  It was suggested that dimerisation of MP11 proceeds through a head-to-
 tail alignment of the haem planes, whereas dimerisation in MP8 was thought to be 
through the stacking of haems. 
 
 52 
1.5.3.3 Overcoming Aggregation 
 
Monomerisation in aqueous solution may be achieved by using alcohols or detergents to 
disperse aggregates, or simply by using microperoxidase solution concentrations < 2 ?M.  
Although these techniques are commonly used to ensure that MPs remain monomeric in 
solution, these conditions are far from ideal.68  It has long since been established that rate 
constants, equilibrium constants, and other physical parameters measured in mixed 
solvent systems usually depend on the permittivity of the medium; thus it becomes more 
difficult to compare the results of the model studies in mixed solvent systems with data 
obtained on studies of haemproteins in aqueous solution. Furthermore, the low 
concentrations required to ensure monomeric conditions limit physical studies to UV-vis 
spectroscopy using long pathlength cells.68 
 
In 1959, Harbury and Loach106 reported that the acetylation of the a-NH2 group of the 
terminal Val-11 residue and e-NH2 group of Lys-13 minimised aggregation of MP11 in 
aqueous solution.  Further evidence that acetylation of MP8 prevents aggregation was 
provided by the ESR data collected by Yang and Sauer.107  A further suggestion by Wang 
and Wart108 suggested that acetylation prevents intermolecular coordination in their 
studies of MP8 using optical and resonance Raman spectroscopy.  However good the 
above results, they failed to provide conclusive evidence that acetylation prevents 
aggregation.   
 
The most comprehensive study that showed definitive evidence that acetylation of the 
Cys-12 of the a-amino group of MP8 prevents aggregation was provided in a detailed 
solution chemistry study by Marques and Munro.109  Marques and co-workers 
demonstrated that acetylated MP8 (N-AcMP8) remains monomeric in aqueous solution at 
concentrations below 3 ? 10-5 M, with a self-association constant an order of magnitude 
lower than that for MP8,102 showing evidence that the N-acetyl protecting group 
diminishes intermolecular coordination.  These results are in agreement with those of 
Wang et al.110 who reported tha t N-AcMP8 forms aggregates above concentrations of 1 ? 
10-5 M. 
 53 
1.5.4 Haempeptide Acid Dissociation Constants 
 
It has been well established that for all haempeptides for which the solution chemistry has 
been investigated there are corresponding pH dependant electronic spectra. 91  However, 
in most cases the pKas associated with these spectral changes are complicated by 
formation of aggregates.  In the study carried out by Marques and Munro109 who 
investigated the solution chemistry of N-AcMP8 as a monomeric aqueous solution, N-
 AcMP8 was found to undergo six spectroscopically observed pH dependant transitions.  
The authors suggested that the first transition (pKa = 2.1) is due to the deprotonation of 
the c-terminal carboxylic group, which subsequently displaces the proximal H2O ligand 
of the bis-aqua complex.  Furthermore, it is reported that the electronic spectra of these 
two species show that both are predominantly high spin.  The second transition occurs 
when the pH is raised above 3.21, and is attributed to the deprotonation of the His-18 
with subsequent binding to the lower axial coordination site of the ferric centre.  The 
electronic spectra of the third species suggest that it is predominantly high spin, but also 
has some low spin character.  The third and fourth pKas at 4.95 and 6.1, respectively, are 
thought to be due to the deprotonation of the two propionic acid groups, with the first 
also corresponding to the ionisation of the c-terminal carboxylic acid group of the peptide 
chain.  It is however probable that these two processes are closely overlapped and not 
completely resolved.  The pKa at 9.59 is attributed to ionisation of the water molecule 
occupying the distal position, and the final pKa at 12.71 corresponds to the deprotonation 
of the bound His-18 forming the histidinate complex (His-Fe(III)-OH-).  The electronic 
spectra show that Fe(III) in N-AcMP8 is predominantly in the S = 5/2 state at pH 7, while 
the hydroxo complex at pH 10.5 is an equilibrium mixture of the S = 5/2 and S = ? states.  
Above pH 12.71, the ferric ion is predominantly in the low spin state.  
 
 
1.5.5 Ligand Binding Studies of Microperoxidases 
 
There have been many studies aimed at understanding the chemistry that occurs at the 
active site of haemproteins, and to elucidate the effect of the protein on the chemistry of 
the prosthetic group, a wide range of ligand binding studies have been performed on both 
 54 
the native haemproteins and their corresponding model compounds.  By comparing the 
formation constants of a given ligand by a haemprotein and by a MP, it provides vital 
information about the effect of the protein on the ligand binding properties of the iron 
porphyrin.91  The model compound, in this case MP, serves as useful tool for the 
prosthetic  group of haemproteins such as haemoglobin, myoglobin and the peroxidases.  
However, these model compounds have to be monomeric in solution.  In these systems, 
the lower distal His-18 ligand is retained as part of the amino acid side chain and the 
upper axial ligand is a readily replaceable water molecule.  This makes it possible for a 
ligand substitution to be investigated for a wide variety of exogenous ligands.  This 
process is defined by Equation 1.5: 
 
                         His-Fe-H2O + L                   His-Fe-L   + H2O                         (1.5) 
K = [His-Fe-L] / [His-Fe-H2O][L] 
 
The nature of the distal ligand in haemproteins and MPs determines the spin states of the 
complexes.  Ligands that produce predominantly low spin complexes includes cyanide 
100, 107, 111, 112 , imidazole and its derivatives100, 107, 110, 111, 113 -115, and pyridine and its 
derivatives 82, 116.  Weaker field ligands such as azide107, 111, 117, 118, and sulphite114 produce 
complexes that are in spin equilibrium.  Predominantly high spin complexes are formed 
with weak field ligands such as cyanate119 and fluoride100, 107, 118 anions.  Table 1.4 shows 
a comparison of equilibrium constants for various ligands coordinated by MP8 and 
various haemproteins.  Krel, which is defined a Khaemprotein/KMP8, is generally much larger 
for the anionic ligands than for neutral ones.  The distal environments of most 
haemoglobins and myoglobins contain the side chains of Val, Phe, and the distal His. 
 
 
 
 
 
 
 
 55 
Table 1.4:  A comparison of the ligand formation constants for various haemproteins,   
                   and for ferric MP8.120 
Ligand (L) Haempeptide/ Haemprotein Log K Krelb Reference 
F- MP8 
Sperm whale Mb+ c 
Human Hb+ 
Aplysia  Mb+ d 
-0.022 
1.85 
1.77 
1.7 
 
74 
62 
53 
121 
122 
122 
123 
SCN- MP8 
Human Hb+ 
Horse Mb+ 
Aplysia  Mb+ 
-0.52 
2.6 
2.42 
1.05 
 
1300 
870 
37 
120 
122 
124 
123 
NCO- MP8 
Sperm whale Mb+ 
Human Hb+ 
Horse Mb+ 
0.083 
3.3 
3.19 
2.58 
 
1600 
1300 
310 
120 
122 
125 
124 
N3- MP8 
Sperm whale Mb+ 
Human Hb+ 
Horse Mb+ 
Aplysia  Mb+ 
D. dendriticum Hb+ e 
Hb+ MHyde Park f 
Hb+ MIwatef 
1.45 
4.98 
5.41 
4.46 
3.4 
3.1 
5.17 
5.27 
 
3400 
9100 
1000 
89 
45 
5200 
6600 
120 
126 
122 
124 
123 
122 
122 
122 
Imidazole  MP8 
Sperm whale Mb+ 
Human Hb+ 
Horse Mb+ 
Aplysia  Mb+ 
4.45 
2.1 
2.6 
1.6 
2.3 
 
0.0045 
0.014 
0.0015 
0.0071 
102 
127 
128 
129 
127 
2-methyl 
imidazole  
MP8 
Horse Mb+ 
2.61 
1.3 
 
0.049 
120 
129 
Benzimidazole MP8 
Sperm whale Mb+ 
2.71 
1.8 
0.12 120 
130 
CN- MP8 
Sperm whale Mb+ 
Human Hb+ 
7.58 
8.4 
8.74 
 
6.6 
14 
112 
130 
131 
a   The equilibrium constant for the process defined in Equation 4. 
b   Krel = Khaemprotein/ KMP8 
c   Mb+ and Hb+ refer to the ferric (met) myoglobin and haemoglobin, respectively. 
d   In myoglobin for the mollusc  Aplysia limacine the distal His is replaced by Val.  
e    Dicrocoelium dendriticum is a small liver fluke; the distal His is replaced by Gly. 
f   Mutant Hb; in Hb MHyde Park  the distal His of the ? chains is replaced by Tyr; in the Hb M Iwate, 
    the same mutation conditions but in the a chains. 
 56 
1.5.6 Demetalation of Haempeptides 
 
In order to gain new insights into the structure and function of metalloproteins, the iron in 
the iron porphyrins has been substituted with other metals.  Iron has been substituted with 
manganese in MP8.79, 80  In other studies iron has been substituted with copper, zinc, 
manganese, nickel and cobalt in cytochrome c.132  There are three methods of iron 
abstraction that have been reported.  The first method involves removing the entire haem 
moiety from the polypeptide chain and substituting it by another porphyrin containing the 
desired metal.133  This route is both complicated and quite involved, with many steps.  
The second method involves the removal of iron by treatment with hydrogen fluoride at 
temperatures below 0 oC.80, 132  This path is hazardous and not user friendly as it involves 
the use of hydrogen fluoride.  The final and much safer method which was adopted 
involved the removal of iron by reductive demetalation reported by Primus et al.79  In the 
project the insertion of the cobalt ion is accomplished by refluxing the free base (metal-
 free MP8) with cobaltous acetate which is followed by oxidation in air.   
 
 
1.6 Cobalt Corrinoids : Basic Structure  
 
Much of the chemistry discussed in this project revolves around the unique nature of the 
corrin ring and why life has preserved this apparently evolutionary antique.  The 
corrinoids are a group of compounds containing four reduced pyrrole rings joined into a 
macrocyclic ring by links between their a positions; three of these links are formed by a 
one-carbon unit (methylidyne radicals) and the other by a direct (Ca-Ca bond).  They 
include various B12 vitamins, factors and derivatives based upon the corrin skeleton, 
C19H22N4.134 
 
 57 
N
 N N
 N
 Co
  
 
                               Figure 1.15:  The basic structure of the cobalt corrinoids. 
 
 
 
1.7   Vitamin B12 
1.7.1 Background 
 
Vitamin B12 is a vital biological compound as it is involved in several biochemical 
processes, particularly in bacteria.  The nature of the biologically active forms of vitamin 
B12, which involve a cobalt-to-carbon s-bond, is most remarkable. This organometallic 
bond is a source of some very interesting coordination chemistry found in the derivatives 
of vitamin B12.4  It is not fully understood what vitamin B12 does in cells, but it is known 
to be essential in preventing pernicious anaemia, a fatal disease of the red blood cells.135  
The history of the discovery of vitamin B12 and its biological function can be traced back 
to the early 1920s when Minot and Murphy, two American physic ians, started to cure 
pernicious anaemia patients with a raw liver diet, and Castle observed that the stomach 
juice contains a protein factor he called intrinsic factor, which strongly enhances the 
curing effect of the liver or its extracts administered orally.136,137  Following this 
discovery, and for the next twenty years, liver was the main source of this unknown 
curing factor which was prepared, as time passed, in more and more concentrated form.  
Finally, in 1948 the isolation of pure crystalline factor was announced by two 
independent teams from the United States and England.138,  139   
 
It was Folkers and his co-workers at Merck Laboratories and Smith and Parker at Glaxo 
Laboratories who purified and isolated pure crystals of vitamin B12. Small red crystals of 
 58 
vitamin B12 were then grown by Lester Smith and given to Dorothy Hodgkin for crystal 
structure analysis.  All that was known at that stage was the approximate empirical 
formula.  A crystal structure on a molecule of this size and complexity had never been 
attempted before, it was a huge and complex task, since crystal structure determinations 
were not the routine tasks that they are today, and the techniques were still being 
developed, both the X-ray and the computer equipment were tedious and difficult to use. 
Thus the X-ray crystal structure which emerged from this study between 1950 and the 
early 1960s was the first determination of a chemical formula by x-ray diffraction, and 
the first determination of the structure of a metalloenzyme.138, 139   However, it is possible 
to trace several lines of study, independent in origin, which were well underway to the 
discovery of vitamin B12. 
 
 
1.7.2 Structure of Vitamin B12a 
 
The cobalt corrinoids contain a cobalt ion in the +3 oxidation state coordinated in the 
equatorial plane by four nitrogen atoms of the corrin macrocycle.23  Vitamin B12 is the 
only known biomolecule with a stable carbon-metal bond  as shown in Figure 1.16. The 
core of the molecule is a corrin ring with various attached side chains. The ring consists 
of 4 pyrrole subunits, joined on opposite sides by a C-CH3 methylene link, on one side by 
a C-H methylene link, and with the two of the pyrroles joined directly. It is thus like a 
porphyrin, but with one of the bridging methylene groups removed. The nitrogen of each 
pyrrole is coordinated to the central cobalt atom. The sixth ligand below the ring is a 
nitrogen of a 5,6-dimethylbenzimidazole. The other nitrogen is linked to a five-carbon 
sugar, which in turn connects to a phosphate group, and thence back onto the corrin ring 
via one of the seven amide groups attached to the periphery of the corrin ring. The base 
ligand thus forms a 'strap' back onto the corrin ring.  
 
 59 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 1.16:  The structure of the vitamin B12a. 
 
Generally, vitamin B12 and other cobalt corrinoids crystallise in the orthorhombic space 
group P212121 with four molecules per unit cell.  The coordination geometry is 
tetragonally distorted pseudo-octahedral, with four equatorial nitrogen donors from the 
corrin ring.  The corrin ring is quite flexible and folds upward about the C10-Co axis. 
This fold is quantified as the fold angle, defined as the angle between the normal to the 
least squares planes through N21-C4-C5-C6-N22-C9-C10 and through C10-C11-N23-
 C14-C15-C16-N24.140  One important finding from the structural studies of substituted 
cobalamins (R-Cbl?s) is the so called ?inverse trans effect?141 for alkyl-Cbl?s in which 
increasing s -donation from R leads to an increase in the length of both the upper (?) Co-
 L bond and the lower (a) Co-NB3 bond despite expectations to contrary.  There have 
been several studies investigating and describing many different aspects of the chemistry 
of vitamin B12 that are beyond the scope of this project.   These special properties and 
features are described in detail in a review by Brown.142 
N
 N N
 N
 Co III
 CONH2
 CH3
 CH3
 CH3
 H3C
 H3C
 H3C
 H
 CH3
 CONH2
 CONH2
 X
 OHN
 CONH2
 H2NOC
 H3C
 H
 O
 P
 O
 O-
 O
 O
 OH
 HO
 N
 N
 CH3
 CH3
 H3C
 H2NOC
 60 
1.7.3 Chemistry of Vitamin B12 and its derivatives:  
            Ligand Substitution and Kinetic Studies 
 
The inorganic chemistry of the cobalt corrinoids (derivatives of vitamin B12) is vital for 
mankind?s existence and therefore attracts much attention.  The kinetics of the reactions 
of aquacobalamin (vitamin B12a) with imidazole, 4(5)-methyl imidazole, 2-methyl 
imidazole, imidazole 4(5)- lactic acid, histamine and histidine were studied as a function 
of pH at 25 oC, 1.0 M ionic strength (KCl) by UV-Vis spectrophotometry. 143  It was 
found experimentally that the 2-methyl imidazole and the analogous 4-substituted 
tautomers of asymmetrically substituted imidazoles, compared to imidazoles where 
tautomerism was impossible (1-methyl imidazole, 1,5-dimethyl imidazole), 
inconsequential (imidazole), or where the 5-substituted tautomer was the greatly 
predominant species in solution (neutral histamine, anionic histidine), replaced H2O 
insignificantly slowly due to the steric repulsion between the annular substituents and the 
corrin ring.  Thus it was concluded that the rate of substitution of H2O varies linearly 
with the pKa of the incoming ligand.  The positive deviation of imidazole itself from this 
series was attributed to the possibility of its reaction through two annular coordination 
sites.  On the other hand, the negative deviation of 4(5)- lactic acid, cationic histamine and 
neutral histidine from the linear relationship was attributed to the presence of two 
tautomeric forms in solution and this provided the means of finding the equilibrium 
constant, KT , for interconversion between these species.143 
 
The thermodynamics of the coordination of imidazole and its derivatives by 
aquacobalamin were also investigated.144  From the fits of the relevant equations to the 
experimental data and from direct spectrophotometric titrations, it was shown that 
coordination decreases the pKa of the imino group of the imidazoles from >14 to 10 due 
to the polarization by the metal ion.  Also, the pKas of the pendant amino acids of the 
histamine and histidine decreased from 9.96 and 9.18 to 8.89 and 7.99, respectively, on 
coordination.  This was attributed to the  destabilization of the conjugate acids by 
coulombic repulsion with the residual charge at the metal centre.144 
 
 61 
In another study, the activation parameters for the reaction of aquacobalamin with small 
anionic and neutral ligands were determined.145   From the temperature dependence of the 
rate constants, activation enthalpies and entropies were determined for the reaction of 
aquacobalamin with SCN-, S2O32-, NO2- and HSO3-.  These parameters, as well as those 
determined previously for CN-, HCN-, N3- and HN3, demonstrated that there are very 
small differences that exist between most small anionic and neutral ligands in their 
reaction with B12a.  There were also exceptions, though; CN- and SO32- had small and 
large values for ?H? and ?S?, respectively.  These effects were attributed to a possible 
nucleophilic participation of these ligands in the transition state of the reaction. 145 
 
Ligand substitution reactions of aquacobalamin with primary amines have  also been 
investigated.146  Enthalpies and entropies of activation were determined from the 
temperature variation of the spectrophotometrically determined pH-independent rate 
constants for the reaction of nine neutral primary amines with aquacobalamin in aqueous 
solution.  There were compensating changes in ?H? and ?S? for this series of ligands, 
with ? H? decreasing from approximately 81 to 58 kJ mol-1 as ?S? decreased in parallel 
from approximately 46 to -46 J K-1mol-1.  It was also reported that there was no isokinetic 
relationship for the series.  Plots of ?H? and ?S? against pKa revealed that the ligands fall 
into two distinct classes.  Ligands in the first class have significantly larger ?H? and ?S? 
values than those in the second class.  This was ascribed to ligands in the second class 
interacting by hydrogen bonding to the acetamide side chains of the corrin ring thus 
allowing their amino groups to be favorably orientated for interaction with the metal 
ion.146 
 
The rate determining step (RDS) for a purely dissociative process is the first step. Thus 
the overall rate of the reaction should be sensitive to the nature of the leaving group (X), 
but insensitive to the entering group (Y).  Neither purely associative nor dissociative 
mechanisms are generally found.  The difference between Ia and Id is the degree of bond 
formation to the entering ligand, Y. If bond making is more important in transition state 2 
then the mechanism is Ia.  Conversely, if bond breaking is more important then the 
mechanism is Id.147 
 62 
The kinetics of the reaction of aquacobalamin with azide anion and hydrazoic acid, 
producing azidocobalamin, was investigated as a function of ligand concentration, pH, 
temperature and pressure, by stopped-flow spectrophotometry.19  The pH dependant rate 
constants and activation parameters were determined.  However, after a concerted effort 
to try and explain the data obtained, it was not possible to distinguish between a limited 
dissociative and a dissociative interchange mechanism.148 
 
The kinetics of the reaction of aquacobalamin with N-donor ligands, glycine and methyl 
glycinate were investigated as a function of ligand concentration, pH and temperature at 
constant ionic strength by conventional and stopped-flow spectrophotometry.149  The pH 
profile for the reaction with glycine was bell-shaped; this demonstrated that B12a reacts 
only with deprotonated ligand and that hydroxocobalamin is kinetically inert. At 25 oC, it 
was reported that the ligands react at about the same rate, but the enthalpies and entropies 
of activation determined for glycine and methyl glycinate showed that this was 
coincidental and due to a compensating change in the two parameters.149 
 
The kinetics of substitution of bound H2O in aquacobalamin by six different ligands, 
namely hydroxylamine, methyl glycinate, pyridine, 4-methylpyridine, imidazole and 
histamine were investigated as a function of ligand concentration and temperature by 
stopped-flow spectrophotometry at constant ionic strength and pH.150  In all six cases the 
observed pseudo-first-order rate constants for the ionization of bound H2O in B12a showed 
saturation at high ligand concentrations after correcting for the protonation of the N-
 donor ligand. It was shown that there was a compensating change in the ?H? and ?S? 
values of the saturating rate constant, ksat, but there was no general isokinetic relationship 
for the ligands.  It was concluded that the dependence of the value of the saturating rate 
constant ksat, and specifically, the dependence of ?H? and ?S? for this rate constant, on the 
entering ligand indicated that the rate-limiting step in the reaction was not unimolecular 
release of H2O from B12a.  The results were interpreted in terms of a dissociative 
interchange mechanism with the nucleophilic participation of the ligand in the transition 
state.  The compensating effect was then explained on the basis of the extent of bond 
formation between Co and the ligand in the transition state.  Molecular mechanics 
 63 
techniques and molecular volume calculations  were used to show that the steric bulk of 
ligands played an important role in controlling its rate of reaction at the metal centre.150 
 
The position of hydroxide in the thermodynamic trans influence order of the cobalt 
corrinoids were investigated.151  Diaquacobinamide (DAC) was prepared by reduction of 
aquacyanocobinamide with zinc and oxidation using aqueous HCl.  The macroscopic acid 
dissociation constants pKCo1 and pKCo2 for coordinated H2O in DAC were found 
spectrophotometrically to be 5.91 ? 0.04 and  10.30 ? 0.24 at 25 oC; from their 
temperature dependence it was found that ?H = 45 ? 4 and 27 ? 2 kJ mol-1, and ?S = 40 ? 
14 and ?109 ? 8 J K -1 mol-1, for pKCo1 and pKCo2, respectively.  The equilibrium constants 
for the reaction of cobinamide with cyanide were determined in the pH range 8-12 at 25 
oC, ? = 1.0 M (NaClO4).  It was reported that two CN- ligands bind to 
aquahydroxocobinamide and the pH dependence of the observed equilibrium constants 
could be accounted for on the assumption that both dihydroxocobinamide and 
hydroxocyanoamide were inert, while the equilibrium constants for the reaction of the 
labile species aquahydroxocobamide and aquacyanocobamide with HCN were small 
compared to those for the reaction with CN- anion.  A fit of the data showed that the 
macroscopic equilibrium constant, log ?2, for the reaction of 2 CN- with the two 
diastereomers of aquahydroxocobinamide to produce dicyanocobiamide is 19.0 ? 0.1.  
Thus the macroscopic log K for the binding of CN- to the two diastereomers of 
aquahydroxocobinamide was 11 since log K for the binding of CN- to 
aquacyanocobamide is 8.  The equilibrium constants for the binding of azide, pyridine, N-
 methylimidazole and 3-amino-1-propanol to aquahydroxocobinamide were determined at 
25 oC, ? = 1.0 M (NaClO 4), pH 12.  In these cases only a single ligand was bound; the 
macroscopic log K values were N-methylimidazole, 6.06 ? 0.04; 3-amino-1-propanol, 
4.66?0.04; pyridine, 4.19 ? 0.01; and azide, 3.45 ? 0.04.  A comparison of these results 
with others available showed that hydroxide was above water and 5,6-
 dimethylbenzimidazole, but below cyanide, sulfite and alkyl ligands in the trans 
influence order of the cobalt corrinoids.151 
 
The kinetics of substitution of bound H2O in aquahydroxocobinamide by cyanide, azide, 
pyridine and N-methylimidazole were investigated as a function of ligand concentration 
 64 
and temperature by stopped- flow spectrophotometry at pH 12.0.152  The observed pseudo-
 first-order rate constants showed the onset of saturation with ligand concentration for all 
ligands. The relative insensitivity of rate on the identity of the incoming ligand suggests a 
dissociative intimate mechanism.  Furthermore, the observation that the saturation rate 
constant, ksat (and the activation parameters ? H? and ?S? ), depended on the identity of the 
ligands indicated that the incoming ligand participates in the transition state.  This 
allowed the mechanism of the reaction to be identified as a dissociative interchange.152 
 
Further evidence for a dissociation interchange mechanism and factors controlling 
reaction rates in ligand substitution reactions of aquacobalamin were determined.153  The 
substitution of H2O in aquacobalamin by the anionic ligands I-, S2O32-, NO2-, SCN - and 
N3- were determined as a function of ligand concentration and temperature by stopped-
 flow spectrophotometry at constant ionic strength and pH.  The observed pseudo-first-
 order rate constants saturated to a limiting value, ksat , at high ligand concentration which 
is consistent with reactions proceeding through a dissociative activation pathway.  The 
value of ksat  depended on the entering ligand and the activation parameters for ksat, ?H? 
and ?S?, were directly correlated [? H? varied from 26 ? 1 (I-) to 83 ? 4 (N3-) kJ mol-1 
while ?S? varied from ?92 ? 4 (I-) to 94 ? 13 (N3-) J K -1 mol-1].  This is inconsistent with 
a limiting dissociative (D) mechanism which requires ksat , and hence ?H? and ?S?, to be 
independent of the identity of the entering ligand.  Thus the mechanism was found to be a 
dissociative interchange (Id) mechanism.  Furthermore, there was a direct correlation 
between ? H? (and hence ?S?) and the total Mulliken population of the donor atom of the 
entering ligand, but there was no correlation between these activation parameters and the 
cone angle subtended by the ligand at the metal atom.  These results suggested that for 
anionic ligands, in contrast with neutral N-donor ligands, electronic rather than steric 
effects were primarily responsible for controlling reaction rates.  It was also concluded 
that for anionic ligands, which were capable of bonding through two different donor atom 
types the reaction, occurred primarily through the donor atom with the higher electron 
density.153 
 
 65 
The ligand substitution reactions of coordinated H2O in aquacobalamin by OCN-, SCN-, 
SeCN-, NO2-,  S2O32- and N3- in NaNO3 medium were investigated.25 In the preceding 
report,152 it was reported that the activation parameters for N-donor ligands investigated 
did not correlate to the Mulliken population; the activation parameters for these species 
correlated instead with the steric demands of the ligand at the metal ion.  Thus it suggests 
that the steric demands can override the electronic effects. According to recent studies, it 
has now emerged that the re is a potential problem with these suggestions.  To date the 
ionic adjustor used in all these studies was KCl, because it is UV-transparent it allows 
standard combination glass electrodes to be used for the measurement of pH, and little  
spectroscopic evidence was found for the binding of Cl- to CoIII.  Following this another 
report has shown that at high chloride concentrations there is a significant retardation of 
reaction rates which is attributed to the formation of the more substitution-inert chloro 
complex.  Although these findings do not invalidate the preceding conclusions reached 
concerning the ligand substitution reactions of aquacobalamin, i.e. there is now a general 
agreement that they proceed through an interchange mechanism under dissociative  
activation, however it brings into question any conclusions based on the absolute values 
of the rate constants and their activation parameters.  In this study the previous 
suggestions have been reconsidered concerning the identification of the reacting atom of 
the entering ambidentate ligand.  Attempts have been made to rationalize the nature of 
the outer-sphere or precursor complex formed between aquacobalamin and the incoming 
ligand. The second-order rate constants were corrected for pH effects, for the reactions 
have been determined as a function of temperature, which lead to the calculation of the 
activation parameters. The identity of the donor atom of the ambidentate ligands in the 
rate determining processes, k II, was considered and the correlations found between      
?H?kII, the enthalpy of activation, and ?S? kII, the entropy of activation, respectively. It 
was found that the Mulliken population on the donor atom, and the energy of the highest 
occupied molecular orbital (HOMO) had a  ? symmetry.  The  parameters ?H? kII were 
strongly dependant on the HOMO energy, and more weakly dependant on the electron 
density on the donor atom, while the converse was true for ?S? kII.154 
 
 
 66 
It is surprising to notice that the CoIII in vitamin B12a shows unusual kinet ic lability.  CoIII 
is a classical example of a kinetically inert metal ion, which is largely due to the high 
ligand field stabilization energy associated with its low spin d6 electron configuration. It 
is now generally accepted that the ligand substitution reactions of vitamin B12a proceed 
through a dissociative interchange mechanism (Id), despite earlier suggestions that the 
mechanism is strictly dissociative (D).155  According to this study, the distinction between 
a limiting D mechanism and an interchange Id mechanism for an octahedral complex lies 
in one being to demonstrate directly or indirectly the existence of the five coordinate 
intermediate.  This distinction becomes possible when the pseudo first-order rate constant 
shows saturation with the concentration of the incoming ligand.   The kinetic results 
obtained for vitamin B12a were surprising for two reasons.  Firstly, not only is the CoIII 
encapsulated by the small corrin ring (corrins, unlike porphyrins,  have a direct linkage 
between the A and D ring which results in a smaller macrocyclic cavity) but also the 
substituents on the periphery of the corrin ring provide a formidable steric barrier to the 
incoming ligand.  However, an Id mechanism requires nucleophilic participation by the 
incoming ligand in the transition state.  Secondly, the reactions are surprisingly fast for 
CoIII, which is the classical example of an inert transition metal ion.  A comparison of 
representative second-order rate constants for the substitution of H2O by ligands in CoIII 
complexes with four N-donor equatorial ligands showed that the approximate lability 
ratio of the metal ion towards substitution in corrin, porphyrin, cobaloxime and ammine 
systems was 109 : 106 : 104 : 18.  According to the findings this showed clear evidence for 
a cis labilising effect of the corrin ring, which was a suggested to be a consequence of the 
electronic properties of the cis ligand that have a direct and marked effect on the kinetics 
of the ligand binding properties in the axial position.  This paper aimed to address these 
two questions.  Firstly, investigate the ability of the corrin ring to perturb the properties 
of the metal ion, thus usual inertness of the d6 CoIII had been modified by replacing the H 
atom at C10 with an electron withdrawing group like a nitroso group. This compound 
called 10-nitrosoaquacobalamin was then characterized by NMR and UV-vis 
spectroscopy and by FAB-MS.   Secondly, the steric bulk of the departing ligand was 
increased by making it iodide.  Following this, the kinetics of its substitution by three 
representative ligands, N3-, S2O32- and imidazole were investigated, to determine whether 
 67 
an increase in the bulk of the departing ligand was sufficient to switch from the Id to the 
D mechanism. The pKa and activation parameters were determined and compared to B12a.  
It was found that the electron withdrawing NO group deactivated the metal ion towards 
ligand substitution, and that neither pyridine nor N3- caused a displacement of the 
coordinated H2O, and it was concluded that either the reactions were very slow or the log 
K had been decreased by at least 1 and 4 orders of magnitude for the coordination of 
these two ligands, respectively.  Hence, it was reported that the electronic structure of the 
corrin ring could directly influence the axial ligand binding properties of the metal.155   
 
Ligand substitution reactions of the vitamin B12 analog cyanoimidazolyl-cobamide, CN 
(Im)Cbl, with cyanide were studied.156  Cyanide substitutes imidazole (Im) in the a-
 position more slowly than it substitutes dimethylbenzimidazole in cyanocobalamin 
(vitamin B12).  The kinetics of displacement of Im by CN- showed saturation behaviour at 
high cyanide concentration; the limiting rate constant was found to be 0.0264 s-1 at 25 ?C 
and is characterised by the activation parameters: ? H? = 111 + 6 JK-1 mol-1, ?V? = +9.3 + 
0.3 cm-3 mol-1.  These parameters are interpreted in terms of an Id mechanism. The 
equilibrium constant for the reaction of CN(Im)Cbl with CN- was found to be 861 + 75 
M-1, which is significantly less than that obtained for the reaction of cyanocobalamin with 
CN- which was reported to be 0.99 + 0.05 M-1, which is about 0.9 pH units higher than 
that obtained previously in the case of cyanocobalamin.156  According to the above 
studies it seems that Co(III)?s lability is conferred on it by the corrin ring.   
 
 
  
 
 
 
  
 
 68 
1.8 Comparison of Porphyrin and Corrin  
1.8.1 Structural and Functional 
 
Despite their different functions, the structures of porphyrin and corrin are quite similar.  
The porphine ring has D4h symmetry, with four pyrrole rings connected by methine 
bridges.  As can be seen in Figure 1.17, the only two things that distinguish the two ring 
systems are the absence of one of the four methine bridges in the corrin, a feature that 
lowers symmetry to C2V, and ten saturated carbon atoms at the periphery of the corrin 
ring, which eliminates the conjugation of the outer part of the ring and lowers the 
symmetry to C1.157  An important aspect of the corrin ring, when compared to the 
porphyrin, is the relative flexibility of the corrin system. The corrin ring is also less flat 
when viewed from the side than is a porphyrin ring. This adds up to some considerable 
differences between the chemistry of a cobalt porphyrin and a cobalt corrin. In addition, 
the corrin only has a conjugated chain around part of the ring system, whereas a 
porphyrin is delocalised around the whole four pyrrole rings.  
 
 
 
 
 
 
 
 
 
Figure 1.17:  The porphyrin (left) and corrin (right) ring systems.157 
 
It is quite clear that nature seems to a have a preference for iron as the central metal ion 
in porphyrin cofactors, whereas cobalt is normally found in corrins.  Among the first row 
transition metals, cobalt has the lowest abundance in sea water together with scandium,7 
but it is still present in the ubiquitous coenzyme B12.  Cobalt resides between iron and 
 69 
nickel in the first row of the d-block.  It has common oxidation states of +2 and +3, as 
does iron, but cobalt may even be reduced to a formal oxidation number of +1 in vivo, a 
property that iron does not possess.158  On the other hand, iron porphyrin complexes are 
well known for their accessible high-valency states (formally Fe+4 and Fe+5); the first 
plays an important role in the function of the haem oxidases such as P450 (see section 
1.4.3).159   
The corrins exist in nature in the form of cobalamins.  The B12 coenzymes contain a 
corrin ring with a d6 low-spin Co+3 ion in their octahedral resting states.  In most 
cobalamin-dependant enzymes, the imidazole side chain of the histidine residue 
coordinates to the cobalt ion.  In another group of enzymes, the cobalt binds to the 
pendant dimethylbenzimidazole  group of the coenzyme, the properties of which are quite 
similar to those of an imidazole ligand.160  The second axial site is occupied by a methyl 
or 5?-deoxyadenosyl group; these form an organometallic Co-C bond.  This bond is 
broken during the catalytic cycle to form either a five-coordinate Co+2 intermediate and 
an adenosyl radical, or a four-coordinate Co+1 ion, where the imidazole ligand has 
dissociated and the methyl group has been transferred to a nucleophilic substrate.161, 162 
Haem enzymes show a larger variation in their axial ligands (His, Cys, Met, Tyr, Glu, 
Asp, amino terminal, or exogenous ligands), depending on their function.162  The haem 
group can be either five-coordinate with an open coordination site where a substrate 
binds, or six-coordinate with one or more ligands from the protein.  However, the most 
common ligand is a histidine/imidazole group, as is present, for example, in myoglobin, 
haemoglobin, peroxidases, haem oxygenase and most types of cytochromes.4 
 
1.8.2 Chemical Properties 
Recently, density functional calculations have been used to compare various geometric, 
electronic and functional properties of iron and cobalt porphyrin and corrin species.157  
The aim of the  study was to investigate how the chemical properties of cobalt and iron, as 
well as porphyrin and corrin, differ.  Specifically, the authors wanted to understand why 
 70 
cobalt is associated with corrins and iron with porphyrins in nature. This project aims to 
further investigate this observation.  The investigation was concentrated on one typical 
reaction for each of the two coenzymes, namely, the breaking of the Co-C bond as it is a 
typical reaction example of coenzyme B12 metabolism, and electron transfer as it is a 
typical reaction of the haem-containing cytochromes.157 
Others have addressed similar questions.7, 23, 163-165  These investigations suggested that 
corrin was selected to fit the smaller CoIII ion.7, 23, 163  Others have emphasised the 
flexibility of the corrin ring as an important factor in the labilisation of the Co-C bond.165, 
166  Williams et al. have proposed that the low spin CoII is unique among the available 
first-row transition metal ions to provide a stable direct one electron radical. 168  On the 
other hand Pratt et al. have attributed the choice of cobalt to the low 3d to 4s/4p  
promotion energy of CoII, which gives strong Co-C bonds, because the 3d orbitals are too 
small to form strong covalent bonds with carbon.23, 163 Specifically, the mechano-
 chemical trigger mechanism has been a major argument in favour of a specialised 
functio n of corrin systems, based on release of strain energy during catalysis.  However, 
recent experimental170-173 and theoretical173, 174 results have indicated that such a 
conformational change is unlikely to drive a catalytic reaction within corrins.  Finally, 
Rovira et al. compared the geometric and electronic structure of four-coordinate cobalt 
complexes with the corrin and porphine by using theoretical calculations.167  The results 
show that the excitation energy associated with dx2-y2 occupation is much higher in corrins 
than in porphyrins. 
Jensen and Ryde have carried out an extensive comparison of the properties of corrins 
and porphyrins using density functional calculations.157  During recent years, such 
methods have successfully been applied to the study of both iron porphyrin species175-187 
and coenzyme B12 models.167, 174, 188-192  Theoretical methods have the advantage of being 
cheap and fast while giving ?pure? results (well defined reactions in vacuum).  On the 
other hand, so lvation effects and free energies are hard to describe in a consistent way, 
and the accuracy is limited.  This investigation was focused on octahedral MII/III 
complexes (where M is the metal) with two axial imidazole ligands (as a model of b and 
c type cytochromes) or with one imidazole and one methyl ligand (as a model of 
 71 
methylcobalamin). In addition, five-coordinate MII complexes with imidazole ligand and 
four-coordinate MI/II complexes without any axial ligands as models of the intermediates 
in the reaction cycle of coenzyme B12 were investigated.157 
As this project is not concerned with density functional theory and how these results are 
arrived at, we will focus merely on the final conclusions reached in their investigation.    
 
1.8.2.1 Spin-splitting energies 
Several authors have suggested that the porphine and corrin ligands were selected to 
make low-spin (LS) states available for iron and cobalt, because the natural amino acid 
ligands provide too weak a field to drive iron or cobalt into the LS state.7, 168, 193  
However, this is probably most important for CoII, for which the vast majority of ligands 
give rise to a high-spin (HS) state.7  Moreover it has been suggested that the porphyrin 
was selected to keep iron centres close to the crossover point between LS, intermediate-
 spin (IS) and HS states.168 
In order to confirm these suggestions and to compare the relative strength of the porphine 
and corrin ligand fields and the intrinsic preferences of iron and cobalt, an investigation 
was performed to study to differences between the LS, IS and HS states of the octahedral 
imidazole (Im) and methyl (Me) complexes (on both the MII and MIII) and the square-
 pyramidal imidazole complexes (only MII), with all four combinations of Fe/Co and 
Por/Cor.157   
Table 1.5:  The energy difference (kJ mol-1) between the low spin and intermediate-spin       
                    states of [MCor/PorImMe) complexes.157  
Macrocycle Fe II CoII FeIII CoIII 
Porphyrin 39 78 46 82 
Corrin 35 *(a) 92 139 
*(a) No stable octahedral minimum for the quartet state. 
 72 
It was found that only one octahedral Im/Me complex was stable in the HS state, 
[Fe+2PorImMe].  All the other complexes lost the imidazole ligand during optimisation.  
The same applies also to the IS [Co+2CorImMe] complex.  It is important to note that CoII 
has seven electrons and therefore does not have a sextet state comparable to the high state 
quintet state of FeII.  Thus, only the energy difference between the LS and IS states is 
presented in table 1.5.157 
For these octahedral Im/Me complexes, all combinations of metals and ring systems gave 
rise to an LS ground state, in accordance with experimental results for the cobalt corrin 
species.194  The IS states were 46-139 kJ mol-1 higher in energy for the MIII complexes.  
The difference was approximately 40 kJ mol-1 larger for Co than for Fe and also about 50 
kJ mol-1 larger for Cor than for Por.  The results in this report confirmed that both Cor 
and Por gave rise to strong fields and that Fe and Por produced complexes quite close to 
the spin-crossover point.157  Although the splitting energies of four - and five-coordinate 
complexes are reported and interesting, we will not mention the four -coordinate 
complexes, as they are not relevant to our project.  However, overall the spin-splitting 
energy results indicated that cobalt and corrin favour the LS state, whereas Fe and 
porphyrin produced a small splitting between the various states. 
 
1.8.2.2  Thermodynamic stability 
Considering the in vivo abundance of iron porphyrins and cobalt corrins, it is natural to 
address the relative thermodynamic stability of these complexes.  Stability considerations 
may explain why iron forms biological complexes with porphyrin rather than corrin and 
why the opposite is true in the case of cobalt.  The results from the investigation 
suggested that the thermodynamic stability favours the two native combinations of ions 
and ring systems by 8-24 kJ mol-1 over the non-native combinations for all complexes, 
except the MII with Im/Me complexes.  Thus  it was reported that there was some intrinsic 
thermodynamic reason to choose the native combinations.157 
 
 73 
1.8.2.3  Geometries and the size of the ring cavities 
It has been suggested that CoIII prefers the corrin cavity because it is smaller.7, 23, 163, 195  
The metal- ligand distances of the optimised Im/Me and Im2 models were collected in 
Table 1.6 and 1.7, respectively. The results confirmed suggestions made in earlier studies 
that the central cavity of the corrin ring is smaller than that of the porphine ring.  
Therefore, the corrin ring favoured the low-spin states of CoIII and CoII, and even the 
four-coordinate low-spin CoI ion which fits into the corrin ring.  Thus, CoCor species are 
always low spin and all reactants and intermediates in the cobalamin reaction cycles 
involve small strain in the ring.  On the other hand, the cavity in the porphine was too 
large for low-spin cobalt and iron ions.  However, it seemed to be more appropriate for 
intermediate-spin states.  Consequently, intermediate- and high- spin states are available 
for the FePor species, and they are important for many biochemical reactions.157 
 
Table 1.6:  Metal-ligand bond distances [pm] of the optimised [MCor/PorIm/Me]                      
                         complexes in their low-spin ground states. 
Structure  M-C M-NIm M-Neq. av 
[FeIIPorImMe]- 200.5 222.8 201.5 
[FeIICorImMe]0 201.0 226.0 192.1 
[CoIIPorImMe]- 195.2 220.4 202.0 
[CoIICorImMe]0 195.9 225.2 192.8 
[FeIIIPorImMe]0 198.9 224.9 201.8 
[FeIIICorImMe]+ 199.6 229.0 193.3 
[CoIIIPorImMe]0 195.7 221.2 200.6 
[CoIIICorImMe]+ 196.6 225.0 192.0 
 
 
 74 
Table 1.7:  Metal- ligand bond distances [pm] of the optimised [MCor/PorIm2]   
                            complexes in their low-spin ground states. 
Structure  M-NIm1 M-NIm2 M-Neq. av 
[FeIIPorIm2]- 202.4 204.5 204.5 
[FeIICorIm2]0 206.0 209.1 193.5 
[CoIIPorIm2]- 240.0 240.1 210.2 
[CoIICorIm2]0 247.3 251.0 192.1 
[FeIIIPorIm2]0 201.2 202.6 202.6 
[FeIIICorIm2]+ 204.7 204.8 193.5 
[CoIIIPorIm2]0 197.9 198.1 200.6 
[CoIIICorIm2]+ 199.5 201.6 192.5 
 
1.8.2.4  Reduction Potential 
Another possib le reason for the selection of certain combinations of ions and rings is the 
differences in reduction potential between two metal ions and the two ring systems.  CoIII 
in aqueous solution is more easily reduced than iron.  However, the results indicated that 
in tetrapyrrole rings, this tendency was reversed, so that cobalt consistently gave 
potentials that are 0.1-0.3 eV lower than those for iron.  Furthermore, there were also 
pronounced differences in the reduction potentials of the corrin and porphine rin gs, in 
that the former produced higher potentials.  It was suggested that this was a result of 
differing charge on the ring system and therefore strongly dependant on solvation effects.  
As an effect of these two opposing tendencies, the native FePor and CoCor combinations 
often produced quite similar reduction potentials at high dielectric constants, but the latter 
had a higher potential in most solvents.157 
 
 75 
1.8.2.5 Reorganisation energies 
The octahedral [FeII/IIIPor/CorIm2] complexes produced much lower inner-sphere 
reorganisation energies (8-9 kJ mol-1) than the corresponding CoII/III complexes (179-197 
kJ mol-1).  It was proposed that the reason for this was that the occupation of the dx2 
orbital in the low-spin CoII (d7), leads to a large difference in the dis tances to the axial 
ligands in the two oxidation states.157  Therefore, it was concluded that the CoII/III couple 
is useless for electron-transfer reactions, whereas the octahedral FeII/III complexes (d5/6) 
formed excellent electron carriers, also in comparison with other metal sites.157 
 
1.8.2.6 Bond dissociation energies 
The homolytic Co-C bond dissociation energy (BDE) for CoCor in coenzyme B12 was 
investigated.  Equation 1.6 shows a typical reaction for CoCor in coenzyme B12 and it is 
the energy released in this equation we are determining. 
[M IIICor/PorImMe]                [M IICor/PorIm] + Me .                                     (1.6) 
Also the bond dissociation energy (BDE) of the hydrolysis of the six-coordinate Im/Me 
complexes was investigated, and is shown in Equation 1.7. 
[M IIICor/PorImMe]    + H2O         [M IIICor/PorImOH] + CH4                  (1.7) 
It was found that [Cor/PorImMe] had the largest homolytic Co-C BDE of the studied 
complexes with a variation of 12 kJ mol-1.  It was suggested that this is compensated by 
the largest resistance towards hydrolysis.  In addition, the cobalt complexes were more 
stable towards hydrolysis than the iron complexes (by approximately 40 kJ mol-1).  This 
is thought to explain why cobalamins are employed in nature for organometallic 
reactions.157 
The work reported in this article answers many questions and gives several good 
explanations for why iron is associated with porphine in nature whereas cobalt is 
  
 76 
associated with corrin, and also why FePor species are used for electron transfer and 
oxygen activation whereas CoCor species are involved in organometallic Co-C reactions 
(methyl transfer and radical formation).157 
 
 
1.9 Scope and Objectives of the Research Project  
To date, the Bioinorganic Research Group at the University of the Witwatersrand has 
done extensive research on the chemistry of haempeptides, particularly FeIII-
 microperoxidase-8 (MP8) and FeIII-microperoxidase-11 (MP11).  The group has also 
done extensive research on the cobalt corrinoid derivative, Vitamin B12a.  One of the 
significant results is the demonstration that CoIII in cobalamins is labile. This project aims 
to explore aspects of the chemistry of  N-Ac-CoIII-microperoxidase-8, a model compound 
in which Co(III) is coordinated in its equatorial plane by a porphyrin ligand and to 
compare the results with the extensive data available for vitamin B12a and summarized 
above. 
The specific objectives of the project are: 
 
? Synthesise MP8 using the standard methodology the Wits Bioinorganic 
Research Group has developed. 
? To remove FeIII from MP8 using the reductive demetalation procedure 
adapted from Primus et al.79 
? To insert Co(III) into the porphyrin by treating with cobaltous acetate; this is 
followed by aerial oxidation.79 
? Developing conditions where N-Ac-CoIIIMP8 is monomeric in solution.  It has 
been anticipated that at the very least protection of the peptide N-terminus by 
acetylation will be required. 
? Characterise the all the products leading to the synthesis of N-Ac-CoIIIMP8 
using a battery of techniques like UV-visible spectroscopy, HPLC (High 
Performance Liquid Chromatography), ICP-OES (Inductively Coupled 
Plasma ? Optical Emission Spectroscopy), Luminescence spectroscopy and 
 77 
FAB-MS (Fast Atom Bombardment ? Mass Spectroscopy). 
? To describe the solution chemistry of N-Ac-CoIIIMP8 by determining the 
behaviour of CoIII-MP8 as a function of pH and ionic strength.  
? Determine the feasibility of studying the kinetics and thermodynamics of 
ligand substitution reactions on CoIII ion. 
 
This project will set the foundation for future work in understanding the unique chemistry 
of the porphyrin and corrin.  After determining the Beer?s Law behaviour  and developing 
conditions where N-Ac-CoIII-MP8 is monomeric in solution, we would then be able to  
begin exploring,  the hypothesis that the key to kinetic lability of CoIII in the corrin lies in 
the corrin ligand itself, i.e.; the equatorial ligand imparts lability on the axial coordination 
site of the metal ion.  One way to test this hypothesis is to replace the equatorial 
macrocycle whilst maintaining the axial coordination site constant.  Towards this end, the 
N-Ac-CoIII-MP8 fulfils these requirements: 
 
? The metal is CoIII. 
? The equatorial system features porphyrin rather than corrin. 
? The axial ligand (His) is a close match, in so far as electronic effects go, to 
the axial ligand in B12a (dimethylbenzimidazole). 
 
Although it will take a long time until we fully understand these fundamental 
evolutionary antiques, therein lies opportunity for endeavours where thermodynamics of 
ligand coordination, kinetics of substitution processes, molecular modelling, reduction 
and self-exchange reactions, catalytic studies and various other studies on the solution 
chemistry of N-Ac-CoIIIMP8 can be unravelled and the results compared to the extensive 
data available for the cobalt corrinoids.  In the long term we should be in a better position 
to explain the origin of the kinetic lability of CoIII in the corrins, and have a better idea 
about the unique properties this ligand imparts on CoIII. 
 78 
Chapter 2 Materials and Methods 
2.1  Materials 
 
Deionised water was produced by a Millipore RO unit, and further purified by reverse 
osmosis using a Millipore MilliQ Ultra-Pure system (18 M? cm).  All glassware was 
washed thoroughly with liquid soap, rinsed with dist illed water, rinsed with commercial 
acetone and dried in an oven set at 100 ?C.  The reagents used are listed in Table 2.1.   
 
Table 2.1:  Materials used in this study. 
Reagent Supplier Grade  
Acetic Acid BDH AR 
Acetic Anhydride Merck GR 
Acetone Merck CR 
Acetonitrile Agros-Organics AR 
Ammonia Merck GR 
Ammonium hydrogen carbonate BDH AR 
Ammonium sulphate Saarchem AR 
Argon Afrox AR 
Cobaltous Acetate Saarchem AR 
Ctyochrome c Seravac BP 
Helium Gas Afrox AR 
Hydrochloric Acid Saarchem GR 
Iron(II)chloride Merck AR 
Liquid Nitrogen Afrox  
Methanol BDH AR 
Methyl Ammonium Hydrochloride Sigma AR 
 79 
Reagent Supplier Grade  
Nitrogen Gas Afrox AR 
Nitric acid  Saarchem GR 
Pepsin Sigma CP 
Phosphoric acid Aldrich AR 
Potassium chloride Aldrich AR 
Potassium dihydrogen phosphate Merck GR 
Potassium hydrogen phthalate Saarchem GR 
Potassium Permanganate Sigma AR 
Sephadex G50 Sigma Superfine 
Silver chloride Effective lab supplies Ltd AR 
Sodium chloride Merck CP 
Sodium cyanide Merck GR 
Sodium dithionite Hopkin & Williams Ltd GR 
Sodium hydrogen carbonate Merck CP 
Trypsin Sigma CP 
Vitamin B12a  (hydroxocobalamin)  Roussel BP 
 
 
Key:  AR = Analytical Reagent  CR = Commercial Reagent          
CP = Chemically Pure   GR = General Reagent 
BP = Biologically Pure 
 
 
 
 
 80 
2.2  General Methods 
2.2.1 pH Measurements 
  
pH measurements were performed using a Metrohm 6.05 pH meter and a 6.0234 
combination glass electrode.  The instrument was calibrated against the standard buffers 
potassium hydrogen phthalate (pH 4.004), potassium dihydrogen phosphate/dipotassium 
hydrogen phosphate (pH 6.863) and borax (pH 9.183) at 25.0 ?C.  The acid and base that 
were used to adjust the pH were H3PO4 and NH3, respectively.  After calibration, the pH 
was tested with standards of known pH. 
 
 
2.2.2 Buffers 
The buffers used throughout this work are tabulated in Table 2.2.  
Table 2.2:  Common Buffers Used in Experiments. 
Buffers pKa pH Range 
Ammonium hydrogen carbonate 7.8 6.8 ? 7.0 
CHES 9.3 8.6 ? 10.0 
MES 6.1 5.5 ? 6.7 
Tris/HCl 8.1 7.0 ? 9.0 
MOPS 7.2 6.5 ? 7.9 
Multi-component buffer (CHES, MES, Tris, 
MOPS, Potassium hydrogen phthalate) 
- 4.0 ? 10.0 
 
 
 
 
 
 81 
2.2.3 Concentration and Desalting of Products 
 
The concentrating and desalting of the various products formed during the synthesis of 
NAc-CoIIIMP8 was performed using a 50 ml Amicon stirred cell connected to a nitrogen 
cylinder under ca. 70 psi pressure (0.45 ?m, 45 mm).  Each sample was concentrated by 
allowing it to stir until a minimal volume of sample was left in the cell.  When the 
samples were desalted, additional deionised water was added to the concentrated sample 
and the process repeated until all salts were removed.  To test for the presence of salts, 
0.5 ml of 0.1 M NaCl was added to the sample initially and the presence of chloride ions 
in the eluent was monitored using AgNO3.  It was assumed that if all chloride ions were 
washed out then all other salts present were also washed out or were present in very 
negligible concentrations. Desalting continued until the eluent test for Cl- ions was 
negative. 
 
 
2.2.4   Gel Permeation Chromatography 
 
Gel permeation chromatography separates molecules on the basis of their size.  The 
smaller molecules enter pores (with specified size) in the solid phase, while the larger 
molecules travel in the mobile phase and are eluted first.  Sephadex G50 was used in a 3 
x 40 cm column allowing small molecules (between 1 500-30 000 kDa) to pass through 
the pores in the solid phase and the larger molecules (> 30 000 kDa) to travel in the 
mobile phase.  The resin was degassed before packing. After packing, the column was 
equilibrated by passing (NH4)2CO3 (0.1 M) at a flow rate of 10 ml h-1 for at least 48 hrs.  
A Watson-Marlow 101U peristaltic pump was used for elution and the fractions were 
collected using an ISCO 1850 fraction collector. 
 
 
 
 
 
 82 
2.3 Physical Techniques of Characterisation 
2.3.1 High Performance Liquid Chromatography (HPLC) 
 
HPLC was performed using a Phenomenex 5? micron (150 X 4.6 mm) C18 reverse phase 
analytical column, a Spectra-Physics SP8800 ternary gradient pump, a Linear UV/vis 200 
detector (set at 410 nm), and a Varian  4290 integrator.  10 ?l samples were injected onto 
the column using a Rheodyne valve and eluted by a linear gradient elution shown in 
Table 2.3 below. 
 
Table 2.3:  Linear Gradient Elution Program. 
Time % 
50 mM Phosphate 
Buffer/1 mM NaCN 
% 
Deionised Water 
% 
Acetonitrile 
0 98 0 2 
2 98 0 2 
4 75 0 25 
6 65 0 35 
8 65 0 35 
10 75 0 25 
12 98 0 2 
14 98 0 2 
 
 
2.3.1 UV-vis Spectrophotometry 
 
All UV-vis spectra were recorded on either a Cary 1E or a Cary 3E spectrophotometer, 
using quartz curvettes with a path length of 1 cm.  The cell compartment temperature was 
 83 
kept constant with a water-circulating bath kept at 25.0 ?C.  The instruments were 
allowed to warm up for approximately 1 hour prior to ensure optimum performance. 
 
2.3.2 Fast Atom Bombardment ? Mass Spectroscopy ( FAB-M S) 
 
All products leading to the synthesis of NAc-CoIIIMP8 were analysed using Fast Atom 
Bombardment-Mass Spectroscopy (FAB-MS).  Figure 2.1 illustrates the proposed 
mechanism for the formation of molecular ions.  FAB has been particularly useful in the 
analysis of polar biomolecules and natural products like those used in this project.  
Figure 2.1:  Ion formation from an electrospray ionisation source transferred to  
                     mass spectrophotometer.196  
 
The sample preparation involved dissolving the products in a minute amount of a nitro 
benzoic acid (NBA) matrix. This was then inserted into the VG70-SEQ mass 
spectrometer with a MASPEC II data system and bombarded with 8-15 keV Cs+ ions.  
The instrumental parameters are outlined in Table 2.4.   Following ionisation, the selected 
positive or negative ions are extracted, accelerated, and then mass analysed. The low 
resolution FAB mass spectra are characterised by peaks corresponding to matrix cluster 
ions, analyte ions and ions representing impurities.  The observed positive pseudo-
 molecular ion peaks are typically adducts formed with cations (i.e., [M+H]+ and 
[M+Na]+), whereas typical negative pseudo-molecular ions observed are [M-H]-.196  
 
 
 
Mass Analyser 
 84 
Table 2.4:  Instrument Settings - VG70-SEQ by Micromass. 
Parameters  Settings 
Polarity Positive 
Ionisation FAB 
Regulation Field 
Resolution 1000 
Mass Range 3429 amu (7 kV) 
Scan Rate 4 secs/decade (External)  
Start Mass 2000 
End Mass 100 
 
 
2.3.3 Luminescence Spectroscopy 
Solutions of MP8 free base and CoIIIMP8 of arbitary concentration were prepared for 
luminescence spectroscopy analysis.  The respective solutions were put into a 1.0 cm 
pathlength luminescence curvettes and the emission and excitation spectra were recorded 
using a Cary Eclipse luminescence spectrophotometer.   
 
2.3.4 Determination of Extinction Coefficient using ICP-OES 
 
 Inductively Coupled Plasma-Optical Emission Spectroscopy-(ICP-OES) is a fast multi-
 element technique with a dynamic linear range and moderately- low detection limits (~0.2 
- 100 ppb). The instrument uses an ICP source to dissociate the sample into its constituent 
atoms or ions, exciting them to a level where they emit light of a characteristic 
wavelength. Up to 60 elements can be screened per single sample run of less than one 
minute and the samples can be analysed in a variety of aqueous or organic matrices. 
 85 
There is less chemical interference than with Fast Atom Absorption Spectroscopy 
(FAAS), but some spectral interference is possible and there are some element 
limitations.196 
  
In this project a Spectro-Ciros ICP-OES instrument was employed for the determination 
of extinction coefficient of N-AcCoIIIMP8.  It is assumed that the amount of cobalt 
detected is proportional to the concentration of N-AcCoIIIMP8 once a blank correction is 
taken into account. NAc-CoIIIMP8 was accurately weighed out using a six figure 
Sartorius  CP2P balance and transferred quantitatively into a 25 ml A grade volumetric 
flask kept at 25.0 ?C, which was then filled to the mark with distilled water.  A set of 
calibration standards are prepared using a 10 ppm Spectro-multielement stock solution. 
The calibrating solutions are sprayed into the plasma and standard curves obtained. The 
samples are then sprayed into plasma and the average  concentration of cobalt determined 
using data from calibration standards and appropriate dilution factors. 
 
 
2.4 Synthesis of products leading to the formation of NAc-CoIIIMP8 
2.4.1 Preparation of Fe IIIMP8 
 
The synthesis of FeIIIMP8 was carried out according to the procedure developed by 
Munro and Marques109 with some modifications and is based on the preparation of ferric 
MP8 described by Aron and co-workers.102  Ferric horse heart cytochrome c (300 mg) 
was treated with pepsin (15 mg; activity: 3200  units/mg) in 0.1 M phosphate buffer (10 
ml) at pH 2.1 and at 35.0 ?C.  The reaction mixture was incubated for 1 hr, after which a 
further 15 mg of pepsin was added.  The pH was adjusted to 2.1, and the solution was left 
for a further 3 hrs.  The pH was then raised to 4 with concentrated NH4OH, at which 
point the precipitation of MP11 was visible.  Solid (NH4)2SO4 was added to the stirred 
suspension on ice, until no more could be dissolved; saturated (NH4)2SO4 solution was 
then added dropwise.  The reaction mixture was centrifuged for 45 min at 7000-8000 
rpm.  The supernatant was decanted, and the MP11 pellet was dissolved in ca. 10 ml 
deionised water.  Concentrated NH4OH was then added to adjust the pH to 8.5 at 35.0 ?C.  
Trypsin (15 mg; activity: 12700 units/mg) was then added and the solution was incubated 
 86 
overnight under gentle stirring.  MP8 was precipitated by lowering the pH to 3.5 with 
dilute H3PO4.  (NH4)2SO4 was added as before, and the reaction mixture was centrifuged 
for 45 min at 7000-8000 rpm.  The supernatant was decanted and the MP8 pellet was 
dissolved in 5 ml of deionised water, with dilute NH4OH to ensure complete dissolution 
of MP8.  The pH was readjusted to 7.00, desalted, and washed using an Amicon stirred 
cell. 
 
2.4.2 Synthesis of demetalated MP8  
 
The demetalation procedure was adapted from a previously published method developed 
by Primus and co-workers for the synthesis of MnIIIMP8.79  During the process, care was 
taken to avoid direct exposure to light. The reaction was carried out using a vacuum line 
connected to an Edwards RV12 vacuum pump and an argon cylinder. While the stopcock 
between the argon line and the reaction vessel was open the following reactants were 
added to the reaction vessel: FeIIIMP8 (10.00 mg, 6.64 ?mol); 10.00 ml glacial acetic 
acid; 166 ?l concentrated HCl.  The stopcock to the argon line was closed and the 
reaction mixture frozen with liquid nitrogen.  After ensuring that the reactants were 
thoroughly frozen, the stopcock between the vacuum line and the reaction vessel was 
open for ca. 2-3 minutes and then closed.  A Dewar flask containing liquid nitrogen was 
then removed from under the reaction vessel and the contents were allowed to thaw at 
room temperature. After the reactants completely thawed, the stopcock to the argon line 
was opened for about 30 seconds and closed again. The reaction mixture was allowed to 
undergo two more freeze-thaw cycles followed by purging with argon.  After the final 
freeze-thaw cycle and after the reaction mixture had completely thawed, the reaction 
vessel was covered in aluminium foil ensuring that the reaction mixture was not exposed 
to white light.  Using a spatula and an argon saturated powder funnel, iron(II)chloride 
(10.00 mg, 80.16 ?mol, 12.07 equiv.) was added to the reaction mixture and was stirred 
with a magnetic stirrer for about 1 hour at room temperature, resulting in a change of 
color from red to purple.  The foil- covered product was frozen in liquid nitrogen and 
stored in a freezer set at ?60 oC.  The solvent was removed under reduced pressure on a 
vacuum line attached to an Acatel vacuum pump.  The residue was dissolved in deionised 
 87 
water and immediately re-evapourated in order to completely remove residual HCl. This 
washing was repeated twice and the final residue was solubilised in 5 ml of 50 mM 
NH4HCO3 (pH 8.00) and desalted with a Sepak C18 cartridge.  The demetalated MP8 
obtained was then lyophilised using a Virtis-UnitrapII freeze dryer. 
 
 
2.4.3   Synthesis of CoIIIMP8 
 
The metalation procedure was adopted from a previously published method developed by 
Primus and co-workers for the synthesis of MnIIIMP8.79  Demetalated MP8 was dissolved 
in 2 ml of water and allowed to reflux in a paraffin oil bath set at 95.0 ?C.  Cobaltous 
acetate (2.33 mmol, 2.33 M) was dissolved in 1 ml methanol was prepared before hand.  
12 ?l cobaltous acetate solution was then added to the reaction mixture containing 
demetalated MP8 (4.82 ?mol).  The reaction mixture was allowed to reflux for 3 hours.  
After the refluxing process the sample could be exposed to white light.  The product was 
then loaded onto the Sepak cartridge and washed with copious amounts of water to 
remove salts and once all salts were removed the product was eluted with a 1:1 
acetonitrile:water solution.  The volume of the metalated product obtained was decreased 
using an Amicon stirred cell and then lyophilised. 
 
 
2.4.4    Synthesis of NAc-CoIIIMP8  
 
The acetylation procedure was adopted from a previously published method developed 
by Marques and co-workers for the synthesis of NAc-MP8.109  This involved the 
preparation of a 50 ml of 0.4 M carbonate buffer at pH 9.30.  6 mg of lyophilized 
CoIIIMP8 was transferred into a round-bottomed flask. To this 20 ml of the carbonate 
buffer was added to dissolve the CoIIIMP8.  To aid dissolution approximately 500 ?l of a 
1 mM (NH4)2CO3 was added.  A 1000 molar excess of acetic anhydride was used to 
acetylate CoIIIMP8 (it was assumed that there was a 100% conversion from cytochrome c 
to MP8).  The CoIIIMP8/carbonate solution was cooled to approximately 3.0 ?C, and  
acetic anhydride was slowly pippetted into the reaction flask.  Care was taken not to 
 88 
allow the reaction to become too vigorous.  The reaction flask was allowed to stand for 5 
min at 3.0 ?C after completely adding acetic anhydride.  The temperature was set up to 
35.0 ?C and the solution was allowed to heat at the rate of the bath and allowed to remain 
35.0 ?C for 3 hours.  Finally, the solution was purified and desalted an Amicon cell to 
ensure that all carbonate and acetate species were removed.  The final product obtained 
was then lyophilised. 
 
 
2.5 Investigating the Solution Chemistry of NAc-CoIIIMP8 
2.5.1 Beer-Lambert Plot of NAc-CoIIIMP8 
 
A concentrated stock solution of NAc-CoIIIMP8 was prepared in distilled water.  Small 
increments of stock NAc-CoIIIMP8 solution were added to a 1.0 cm pathlength curvette 
that contained distilled water. The absorbance at 415.05 nm at 25.0 ?C for each increment 
was recorded using a Cary 3E spectrophotometer. After the absorbance reached 2.0 
AUFS, 200 ?l of the final 1.0 cm pathlength curvette mixture was transferred to a 0.1 cm 
pathlength curvette.  Similarly, small increments of stock NAc-CoIIIMP8 solution were 
added to the 0.1 cm pathlength curvette. The absorbance at 415.05 nm at 25.0 ?C for each 
increment was again recorded using a Cary 3E spectrophotometer.  A Beer-Lambert plot 
of NAc-CoIIIMP8 in 1.0 M NaOCl4 was determined using the same methods.   
 
It became clear that there was a discrepancy between the absorbance of solutions in the 
1.0 cm and the 0.1 cm pathlength curvettes, as far as the absorbance of the same solution; 
when corrected for different pathlengths, was different.  We  suspect (but have not 
proved) that this discrepancy arises from the scaling of the curvettes in the housing of the 
spectrometer that give pathlengths that are, in effect, different to the actual pathlength of 
the curvette.  To correct for this, a solution of arbitrary concentration of KMnO4 was 
prepared; 100 ?l of the KMnO4 solution was added to a 1.0 cm pathlength curvette 
containing 1900 ?l distilled water.  The solution was mixed carefully using a glass rod.  
The absorbance at 337.9 nm at 25.0 ?C was recorded using a Cary 3E spectrophotometer.  
This process was then repeated a further four times.  200 ?l of each of the five solutions 
 89 
from the 1.0 cm pathlength curvettes were transferred to 0.1 cm pathlength curvette and 
the absorbance for each of the solutions were recorded at 337.9 nm at 25.0 ?C using a 
Cary 3E spectrophotometer.  The mean correction factor for when transferring from the 
1.0 cm to 0.1 cm pathlength curvette was then calculated.  Finally the all data collected 
were corrected accordingly. 
 
 
2.5.2 Data Analysis 
All results were analysed with the computer programme CurveFit197, which employs a 
Newton-Raphson procedure.   
 
2.5.3 Determination of Spectroscopic pKas of NAc-CoIIIMP8 
 
The titration procedure for the determination of spectroscopic pKas of NAc-CoIIIMP8 was 
adopted from a previously published method developed by Munro and co-workers109 for 
the synthesis of NAc-FeIIIMP8.  Spectroscopic pKas were determined by titrating 40 ml of 
a 8.01 ?M stock solution of NAc-CoIIIMP8 in a multi-component buffer system 
containing 1.0 mM each of potassium hydrogen phthalate, MES, MOPS and CHES, and 
10 mM Tris as the non-coordinating buffers at a total ionic strength of 0.1 M (KCl used 
as an ionic adjuster) in a thermostated cell at 25.0 ?C.  The pH was adjusted from pH 5.5 
to 13 for the basic region and from 1.9 to 6.3 for the acid region by diffusion of small 
volumes of concentrated HCl or NaOH into the solution from drawn out glass capillaries.  
At each pH, 5 ml of the solution was pumped (using a Watson-Marlow 101U peristaltic 
pump ) into a thermostated 2.0 cm pathlength curvette in a Cary 3E spectrophotometer 
and the absorbance at 415.05 nm was recorded before returning the sample to the 
continuously stirred reaction mixture.  The absorbance at 415.05 nm was recorded as a 
function of solution pH.  
 
 90 
The experimental data were fitted using non-linear least-squares program employing a 
Newton-Raphson procedure to an ionisation isotherm (Equations 2.1 to 2.4), which is 
relevant when three acid/base equilibria are present. 
 
AT =
 A1*[H+]3 + A2*[H+]2*K1 + A3*[H+]*K1*K2 + A4*K1*K2*K3
 K1*K2*K3 + [H+]*K1*K2  + [H+]2*K1 + [H+]3
 (2.3)
 AT =
 A1*[H+] + A2*[H+]*K1
 K1 + [H+]
 (2.1)
 AT =
 A1*[H+]2 + A2*[H+]*K1 + A3*K1*K2
 K1*K2 + [H+]*K1+ [H+]2
 (2.2)
 AT = A1*[H+]4 + A2*[H+]3*K1 + A3*[H+]2*K1*K2 + A4*[H+]K1*K2*K3 + A5*K1*K2*K3*K4
 K1*K2*K3*K4 + [H+]*K1*K2*K3 + [H+]2*K1*K2 + [H+]3*K1 + [H+]4
 (2.4)
  
In Equations 2.1 to 2.4, AT is the absorbance at the absorbance monitoring wavelength, 
K1 to K3 are the acid dissociation constants and Ai is the absorbance of the ith species. 
 
 
2.5.4 Ligand Binding Studies of Selected N-Donor Ligands by NAc-CoIIIMP8 
 
The equilibrium constants for the coordination of the ligands N-methylimidazole and 
pyridine with NAc-CoIIIMP8 were determined by the addition of aliquots of stock N-
 methylimidazole solution (0.02012 M) and pyridine solution (0.02008 M) to NAc-
 CoIIIMP8, contained in a 1.0 cm pathlength curvette in a Cary 1E spectrophotometer.  The 
solutions were buffered with MES (1.0 M, pH 5.5); the ionic strength was maintained 
using KCl (1.0 M) and deionised water was used as the solvent.  The temperature of the 
cell block was maintained at 25.0 ?C by a water circulating bath and measured with an 
electronic thermostat device.  The absorbance changes at 420.0 nm were monitored for 
both reactions , and each absorbance reading was corrected for dilution.  The experimental 
data were fitted using non-linear least-squares program employing a Newton-Raphson 
procedure to Equations 2.5 and 2.6. 
 91 
 
 A0 + A1*K1*[L] + A2*K1*K2*[L]2AT =
 1 + K1*[L] + K1*K2*[L]2
 (2.6)
  A0 + A1*K1*[L]AT =
 1 + K1*[L]
 (2.5)
  
 
In Equations 2.5 and 2.6, AT  is the absorbance at the absorbance monitoring wavelength, 
K1 and K2 are the formation constants, A0 is the initial absorbance and A2 is the final 
absorbance.  After the formation constants were determined they were corrected for pH 
using Equation 2.7.  
 
K = Kobs(1 + 10pKa - pH)                                                          (2.7)  
 
 
2.5.5 Kinetic Studies of Selected N-Donor Ligands by NAc-CoIIIMP8 
 
Preliminary spectrochemical investigations were performed using a Cary 1E 
spectrophotometer.  Using the same equipment, the kinetics of the reactions between the 
ligands N-methylimidazole and methylamine with NAc-CoIIIMP8 were investigated 
under pseudo-first order conditions. NAc-CoIIIMP8 concentrations (< 45 ?M) were 
prepared and buffered with MES (0.1 M) at pH 7.0 and 8.0 for reactions involving N-
 methylimidazole and methylamine, respectively. The ligand concentrations of N-
 methylimidazole and methylamine ranged from 0.05 to 1.01 M and 0.50 to 2.50 M, 
respectively.  Ligand solutions of N-methylimidazole and methylamine were also 
adjusted quantitatively to the same pH as that of the NAc-CoIIIMP8 solution by addition 
of HCl or NaOH, as appropriate.  After mixing specific amounts of ligands with NAc-
 CoIIIMP8 in a 1.0 cm pathlength curvette, and immediately inserting this mixture into the 
spectrophotometer, the absorbance change was monitored at 420.0 nm (the maximum of 
the Soret band for NAc-CoIIIMP8 at pH 5.5).   
 92 
Pseudo-first order rate constants, kobs, were determined by curve- fitting the absorbance 
versus time traces to Equation 2.8 and 2.9 using non-linear least-squares program 
employing a Newton-Raphson procedure. After the first order rate constants were 
determined they were corrected for pH using Equation 2.10. 
k      =    kobs(1 + 10pKa - pH)                                                                (2.10)
 AT        =   -A0*exp(-k1obs*t) - A1*exp(-kobs'*t) + A0 + A1 + A2            (2.8)
 -A1*exp(-kobs*t) + (A0 + A1)AT = (2.9)
  
In the equation A0, A1 and kobs are variable parameters.  Second order rate constants, k2, 
were obtained from a linear least-square plot of k versus [L]. 
 
2.5.6 Kinetic Studies of Selected N-Donor Ligands by B12a 
Similarly ,  preliminary spectrochemical experiments were performed using a Cary 1E 
spectrophotometer.  Using the same equipment, the kinetics between N-methylimidazole 
and methylamine with B12a were investigated under pseudo-first order conditions.  A 
stock solution of B12a (between 50 and 80 ?M) was prepared and buffered with MES  
(0.1 M) at pH 7.0 and 8.0 for the reactions involving N-methylimidazole and 
methylamine, respectively.  The ligand concentrations of N-methylimidazole and 
methylamine ranged from 1.06 ? 10-4 to 2 ? 10-3 M and 2.51 ? 10-1 to  2.51 M, 
respectively.  Ligand solutions of N-methylimidazole and methylamine were also 
adjusted quantitatively to the same pH as that of the B12a solution by addition of HCl or 
NaOH, as appropriate.  After mixing specific amounts of ligands with B12a in a 1.0 cm 
pathlength curvette, and immediately inserting this mixture into the spectrophotometer, 
the absorbance change for the reaction between N-methylimidazole and methylamine  
with B12a was monitored at 358.0 and 352.0 nm (the maximum of the ?-band).  
Similarly, the pseudo- first order rate constants, kobs, were determined using Equation 2.9 
as above and second order rate constants, k2, were obtained from a linear least-square plot 
of k1 versus [L].  
 93 
Chapter 3 Results and Discussion 
3.1 Synthesis and Characterisation of products leading to the formation 
of NAc-CoIIIMP8 
3.1.1 Preparation of Fe IIIMP8 
 
 
 
 
 
 
 
 
 
Figure 3.1:  Proteolysis of cytochrome c to FeIIIMP8.  The % conversion is calculated    
          from cytochrome c to FeIIIMP8 and assumes 100 % purity of cyt c. 
 
Although FeIIIMP8 is commercially available, it is very expensive.  Using the procedure 
developed by Munro and Marques109 with some modifications, FeIIIMP8, the haem-
 containing octapeptide was successfully prepared.  FeIIIMP8 was then purified using size 
exclusion chromatography, where the reaction mixture containing the respective 
microperoxidases (MP11, MP9, MP8 and MP6) were columned on Sephadex G50 to 
separate them according to size.  HPLC was used as an analytical tool to identify 
FeIIIMP8 fractions that were > 95 % pure (the assumption being that the response factor 
of all haempeptides at 410 nm is the same) and these fractions were pooled together. 
FeIIIMP8 was produced in 55 % yield as shown in Figure 3.1.  Figure 3.2 shows a 
chromatogram of a fraction of FeIIIMP8 that was analysed using HPLC.  The pooled 
fractions were desalted using an Amicon stirred cell and lyophillised.  The yield is 5 % 
higher than that reported by Adams and coworkers.102  The synthesis of FeIIIMP8 by 
(300 mg, 24.2 ?mol) 
Cytochrome c 
(20.05 mg, 13.3 ?mol) 
FeIIIMP8 
55 %  
  Trypsic Digestion 
 94 
peptic and trypsic digestion of cytochrome c was performed several times with 
reproducible yields. 
 
 
Figure 3.2:  HPLC analysis of FeIIIMP8 (retention time of 6.45 minutes) with  
                    > 95% purity after purification on a column of Sephadex G50.  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.3:  UV/visible spectrum of FeIIIMP8 in distilled water. 
 
    6                      7 
MP8
 0.0
 0.3
 0.6
 0.9
 1.2
 1.5
 300 350 400 450 500 550 600 650 700
 Wavelength / nm
 Ab
 so
 rb
 an
 ce
 0.00
 0.03
 0.06
 0.09
 450 500 550 600 650 700
 Qv 
Qo 
CT1 CT2 
B 
N 
Time (Min) 
 95 
There are six characteristic peaks observed in the electronic spectrum of FeIIIMP8 shown 
in Figure 3.3.  The first band occurs at 350 nm and is called the N-band and is assigned to 
the a?2u, b2u (p)  eg(p?) transition.15  The B and Q bands involve a2u, a1u (p)  eg(p ?) 
transitions.  These bands are not purely due to orbital transitions but arise as a result of 
the mixing of inter-electronic repulsion terms.15  The Soret band at 395 nm is thought to 
arise from a combination of transitions to two vibronic components.  Usually, a B(1,0) 
satellite, arising from a transition to the first vibronic state of B electronic state, can be 
seen to the lower wavelength side of B band (about 20 nm separation between B and 
B(1,0) bands).15, 198  The Qo and Qv bands observed at 491 nm and 530 nm, respectively, 
are due to different vibrational transitions of the same electronic transition. Qo is the (0,0) 
vibronic component of the a2u,a1u(p) eg(p) transition, whilst Qv is the (1,0) vibronic 
component a2u,a1u(p)  eg(p) transition.  Qv is more intense than Qo band as vibronic 
coupling with the B band enhances its intensity.  The Qv band is thought to arise from the 
vibrational electronic coupling between the B- and Q-states.15, 109, 198, 199  CT1 represents a 
change transfer band due to the spin-allowed porphyrin a2u?(p) ?  eg(dp) promotion; and 
the symbol CT2 represents a change transfer band due to the spin allowed porphyrin 
b2u(p) ? iron eg(dp) transition.  These signals usually occur at 580 nm and 630 nm, 
respectively; however in this system these bands are observed at 567 nm and 622 nm. 
The overall trend is characteristic for ferric haem proteins and may be slightly 
compromised by contributions of both high and low spin components.15  Other p ?  p 
??transitions at 215 nm and 325 nm do occur in absorption spectra of iron porphyrins but 
are obscured by the bands for the aromatic peptides, and hence are not observed in the 
haemproteins.200  
 
FAB-MS has proven to be an extremely useful tool in determining structure of unknown 
haempeptides with respect to both the haem and the peptide portions, and in 
characterising the purity of bulk preparations of known peptides such as the 
microperoxidases.  It has also been used to determine proteolytic hydrolysis sites for 
peptides derived from c-type cytochrome that contain covalently attached haem c.201 
 
 96 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.4:  FAB-MS spectrum of FeIIIMP8 obtained from the proteolytic digestion  
                    of horse heart cytochrome c in NBA matrix. 
 
A typical low resolution spectrum of the haem-octapeptide is presented in Figure 3.4.  
The exact mass determinations are in excellent agreement for the haemoctapeptide, 
FeIIIMP8, where the observed m/z is 1505.4 and the calculated value is 1506.  An 
important observation was the presence of a very intense molecular ion at m/z = 617 in 
all FeIIIMP8 samples analysed.  This ion corresponds to the fragmentation of 
haempeptide covalent bonds and further dehydrogenation/rearrangement to produce 
mono-protonated FeIIIprotoporphyrin IX.201 The observed m/z for the mono-protonated 
FeIIIprotoporphyrin IX was 617.1 and the calculated 617.2.  The above results confirm 
that FeIIIMP8 was prepared successfully with > 95 % purity with reproducible results. 
 
 
Low Resolution m/z 
In
 te
 ns
 ity
  (%
  A
 ge
 ) 
617.1 
1505.4 
 97 
3.1.2 Synthesis of demetalated MP8  
 
 
 
 
 
 
 
 
 
 
Figure 3.5:  Reductive demetalation of FeIIIMP8 to MP8 Free Base.   The % conversion   
                      is calculated from FeIIIMP8 to MP8 Free Base and assumes FeIIIMP8 is  
                      100 % pure. 
 
The method of preparation of MP8 free base was adopted from the procedure of Primus 
and co-workers which involves removing the central metal FeIII ion by reductive 
demetalation. 79  The demetalation process was successful as the expected colour change 
from red to purple  was observed79 and MP8 free base was synthesised in 73 % yield  
shown in Figure 3.5.   
 
The mechanism of reductive demetalation has not been clarified. We suggest that in 
acidic conditions as those provided, the two central nitrogen atoms are protonated 
competitively,21 and FeIII ion is reduced to form FeII ion.  It has been suggested that FeII 
ion is less tightly bound to a porphyrin than FeIII ion; this can be attributed to FeIII having 
a higher charge and therefore bonding more tightly to the highly electron dense porphyrin 
system.  Thus, the FeII is thought to be forced out of the porphyrin macrocycle by 
competition from H+.  
 
 
(10 mg, 6.64 ?mol) 
FeIIIMP8 
(7 mg, 4.82 ?mol) 
MP8 Free Base 
 73 %  
i)    HCl, glacial acetic acid  
ii)   3 x freeze-thaw cycles    
       under vacuum 
iii)  Add FeCl2 
 
 98 
After the MP8 free base was desalted using Sepak cartridges, it was analysed us ing 
HPLC.  The HPLC chromatogram is shown in Figure 3.6 and its UV-vis spectrum is 
shown in Figure 3.7. 
 
 
Figure 3.6:  HPLC chromatogram of MP8 free base (retention time of 5.99 minutes)                  
                     with 81 % purity. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.7:  UV/visible spectrum of MP8 free base in 50 % methanol/water solution and                       
                     FeIIIMP8 in water. 
0.0
 0.3
 0.6
 0.9
 1.2
 1.5
 300 400 500 600 700
 Wavelength / nm
 Ab
 so
 rb
 an
 ce
 MP8
 Metal Free Base
 0.00
 0.06
 0.12
 450 550 650
 B 
B 
Qv Qo 
Split Qv 
Bands 
Split Qo 
Bands 
N 
    7          6          8 
Time (Min) 
 99 
In demetalated porphyrins (D2h), there are four bands in the visible region and one in the 
ultraviolet region (Soret).198  These absorption bands for metal- free MP8 are observed at 
380 nm, 503 nm,  538 nm, 568 nm and 618 nm as shown in Figure 3.7.  The intense Soret 
band described earlier occurs at 380 nm.  The Qv band is split and the resulting bands 
occur at 503 nm and 538 nm, respectively.  Also, the Qo band is split and the resulting 
bands occur at 564 nm and 616 nm, respectively.  When comparing the spectra of MP8 
and MP8 free base we do not observe charge transfer bands since for a complex to 
demonstrate charge-transfer behaviour, one of its components must have electron 
donating properties and another component must be able to accept electrons. Since there 
is no metal present no charge transfer band will be observed.21  In addition, the N-band 
disappears completely which is a direct result of metal absence. 
 
 
The UV/visible data obtained could not be compared with data obtained by Primus et 
al.79 since their data were obtained  under acidic conditions.  Primus et. al found the Soret 
band at 403 nm and other bands at 552 nm and 595 nm which is typical for a free base 
porphyrin at low pH.  In our results, the UV/visible analysis data were obtained under 
neutral conditions.  However, our data can be compared to Low et al.202 who found a 
very broad Soret band with a maximum at 383 nm as well as the other bands at 508 nm, 
542 nm, 574 nm and 623 nm which are characteristic of free base porphyrins.  Thus, the 
results obtained for UV/visible analysis are consistent with the characteristic peaks for a 
free base porphyrin, confirming that the demeta lation procedure was successful. 
Furthermore, the retention time of 5.99 minutes observed for the free base  porphyrin was 
significantly different from the observed retention time of 6.45 minutes for FeIIIMP8.  In 
addition,  the 81% purity detected by HPLC showed that the sample is relatively pure. 
 
However, the overlap of multiple peaks in the chromatogram (Figure 3.6) means that the 
sample is relatively pure and that this technique is merely an assessment of purity is very 
much an estimate. 
 
 
 100 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.8:  Enlarged view of FAB-MS MP8 Free Base in NBA Matrix. 
The exact mass determinations for the MP8 free base shown in Figure 3.8 were also 
excellent and in agreement where the observed m/z was 1451.2 and the calculated 1451. 
We propose that the signal observed at m/z = 1492.5 are traces of calcium ions since the 
mass difference between this signal and the free base is ~ 41 m/z units. However, the 
other signal where m/z = 1496.5 cannot be explained. 
Low Resolution m/z 
In
 te
 ns
 ity
  (%
  A
 ge
 ) 
1451.2 
 101 
0
 20
 40
 60
 80
 100
 300 400 500 600 700 800
 Wavelength / nm
 In
 te
 ns
 ity
  
Emission MP8 Free Base
 Excitation MP8 Free Base
  
Figure 3.9:  Luminescence Spectra of MP8 free base. 
 
The excitation spectrum of MP8 free base drawn in blue on Figure 3.9 was recorded from 
300 nm to 670 nm with emission ?  = 677 nm.  There were six bands observed, four in the 
visible region and two bands in the ultraviolet region.  The bands at 343 nm and 410 nm 
are characteristic of the N-band and Soret band, respectively, while, the bands at 505 nm, 
541 nm, 565 nm and 614 nm are the same split Q-bands we mentioned earlier.  The 
emission spectra of MP8 free base was recorded from 418.5 nm to 750 nm with 
excitation ?  = 398.5 nm.  There are two bands observed in the visible  region at 617.5 nm 
and 680 nm are characteristic of the transient triplet-triplet absorption observed in free 
base porphyrins and are assigned Qx(0,0) and Qx(0,1), respectively.21  The appearance of 
the Qx bands in the emission spectrum and disappearance of the  split Q bands from the 
excitation spectrum provide more evidence that demetalation was successful.  Metalated 
porphyrins have no significant luminescence spectra because of rapid quenching of the 
excited state by the central metal. 
 
 
N 
B Split Q bands 
Qx(0,0) 
Qx(0,1) 
 102 
3.1.3 Synthesis of CoIIIMP8 
 
 
 
 
 
 
 
 
 
 
Figure 3.10:  Metalation of MP8 Free Base to CoIIIMP8.  The % conversion calculated   
              from MP8 Free Base to CoIIIMP8 assuming that the metal-free MP8 is  
                       100 % pure. 
The metalation procedure was also adapted from the report by Primus and coworkers.79  
However, in this project cobalt was inserted into the porphyrin instead of manganese, so 
as to mimic the cobalt in the vitamin B12a environment. The metalation process was 
successful and CoIIIMP8 was synthesised in 69 % yield shown in Figure 3.10. 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.11:  HPLC chromatogram of CoIIIMP8 (retention time of 6.82 minutes) 
                              with 79 % purity.  
(7 mg, 4.82 ?mol) 
MP8 Free Base 
(5 mg, 3.32 ?mol) 
CoIIIMP8 
69 %  
i)    Dissolve in water 
ii)   Boil in paraffin oil bath 
iii)  Add cobaltous acetated   
      solution 
 
     6                        8 
  
   Time (Min) 
 103 
The HPLC chromatogram of CoIIIMP8 is shown in Figure 3.11 and its UV-vis spectrum 
is shown in Figure 3.12.  When comparing the HPLC retention times of CoIIIMP8 with 
MP8 free base, we find that the retention time of CoIIIMP8 is 6.82 minutes and is 
significantly different from the retention time of MP8 free base which is 5.99 minutes.  
This shows that when the metal is inserted into the MP8 free base, it moves slightly 
slower on the column due to the increased polarity of the newly inserted metal ion. 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.12:  UV/visible spectrum of MP8 free base CoIIIMP8 in 0.001 M         
                      NH4(CO3)2 buffer and MP8 Free Base in 50 % methanol/water. 
Four distinct absorption bands are observed at 350 nm, 425 nm, 533 nm and 567 nm for 
CoIIIMP8 as shown in Figure 3.12.  The characteristic N-band and Soret band occurs at 
350 nm and 425 nm, respectively.  The Qv band occurs at 533 nm and the Qo band occurs 
at 567 nm.  Upon insertion of the cobalt into the MP8 free base, we observe the band 
position of the Soret for the MP8 free base shifting from 380 nm to  425 nm.  
Furthermore, the four split Q-bands that were formed in the MP8 free base, merged to 
form two new Q-bands. The spectrum shows that there is no charge transfer band 
observed for CoIII.  Although there are restricted data available for CoIII, it can be 
0.0
 0.3
 0.6
 0.9
 1.2
 1.5
 300 350 400 450 500 550 600 650 700
 Wavelength / nm
 Ab
 so
 rb
 an
 ce
 Metal Free Base
 Co(III)MP8
 0.00
 0.10
 0.20
 0.30
 470 520 570 620 670
 B B 
Split Q Bands 
Qv Qo 
N 
 104 
compared to another d6 ion like FeII, although the two species are not comparable given 
that FeII is high spin whereas the CoIII system involves a low spin metal ion.  The ligand 
to metal charge transfer in a low spin d6 at 476 nm is due to the transition, a2u (?)  ?  a1g 
(dz2) and another at 320 nm is due to, a?2u (?) ?  b2u (?).  However, for FeII in the high 
spin state the metal to ligand charge transfer band occurs between 758 nm and 930 nm 
which involves the transition, eg (d?) ? eg(?*).203 
From the above spectra it can be clearly seen that there is a shift in peak positions which 
is due to the presence or absence of the central CoIII ion.  Low et al.202 performed a 
similar study where MnIII was inserted into MP8 free base utilising a reductive 
demetalation procedure involving HF.  There were split Soret bands observed which were 
attributed to inequivalent interactions between the porphyrin px and py LUMOs and dp 
electrons of manganese ion.  However, in our study we do not observe this characteristic 
split Soret bands. 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.13:  FAB-MS spectrum of CoIIIMP8 in NBA matrix. 
Low Resolution m/z 
In
 te
 ns
 ity
  (%
  A
 ge
 ) 
 
 
1508.3 
 105 
In addition, FAB-MS was performed on CoIIIMP8.  The exact mass determinations for 
CoIIIMP8 shown in Figure 3.13 were also excellent and in agreement, where the observed 
m/z was 1508.3 and the calculated 1508.  As this was the first time that CoIIIMP8 was 
synthesised there were no comparative data available. 
0
 4
 8
 12
 16
 450 500 550 600 650 700 750
 Wavelength / nm
 In
 te
 ns
 ity
 Emission MP8 Free Base
 Emission Co(III)MP8
 Excitation Co(III)MP8
  
Figure 3.14:  Excitation and emission spectra of CoIIIMP8 and emission spectrum of   
                       MP8 free base. 
The excitation spectrum of CoIIIMP8 shown in red on Figure 3.14 was recorded from 300 
nm to 607 nm with emission ?  = 617 nm.  There were four bands observed, one in the 
ultraviolet region and three bands visible in the region.  The band at 395 nm is the 
characteristic Soret band.  We expected only two bands in the visible region at 530 nm 
and 566 nm, which are characteristic of the Q0 and Qv bands; however three bands are 
observed at 505 nm, 541 nm and 564 nm. This shows that the reaction did not go to 
completion and that not all the free base porphyrin molecules were not metalated.  The 
emission spectra of CoIIIMP8 was recorded from 436 nm to 750 nm and shows that there 
are no intense bands formed and that the transient triplet-triplet absorption observed for 
free base porphyrins is quenched significantly.21  This provides more evidence that 
metalation was successful.  The fact that the Qx bands do not disappear completely shows 
that there are trace amounts of free base porphyrins in solution that have not been 
metalated. 
B 
Q bands 
Qx(0,0) 
Qx(0,1) 
 106 
3.1.4 Synthesis of NAc-CoIIIMP8 
 
 
 
 
 
 
 
 
 
 
Figure 3.15:  Acetylation of CoIIIMP8 to NAcCoIIIMP8.  The % conversion is calculated   
          from CoIIIMP8 to NAcCoIIIMP8 assuming CoIIIMP8 is 100 % pure. 
 
Acetylation of CoIIIMP8 was carried out according to methods developed by Munro and 
Marques.109  Figure 3.15 shows a schematic of the reaction performed according to the 
above mentioned conditions.  The acetylatio n process was successful and NAcCoIIIMP8 
was synthesised in 58 % yield.  However, the acetylation did not go to completion and 
some of the unreacted CoIIIMP8 remained.  The two chacteristic peaks shown in Figure 
3.16 is indicative that the acetylation did not go to completion.  
 
 
 
 
 
 
 
 
Figure 3.16:  HPLC chromatogram of CoIIIMP8 (6.59 minutes) and 
                                        NAc-CoIIIMP8 (5.99 minutes).  
i)   Dissolve in carbonate buffer 
ii)  Add acetic anhydride  
iii) Cool to 3 ?C for 5 minutes  
      and increase to temp. to 35 ?C. 
 (5 mg, 3.32 ?mol) 
CoIIIMP8 (3 mg, 1.93 ?mol) NAc -CoIIIMP8 
 
58 %  
    7          6          8 
Time (Min) 
 107 
The two major peaks observed at 5.99 minutes and 6.59 minutes have a percentage purity 
of 66 and 27 %, respectively.  When comparing these HPLC results to the retention time 
observed for CoIIIMP8 (6.82 minutes) in Figure 3.11, we find that the peak observed at a 
retention time of 6.59 minutes is the closest match.  Thus if we assume the above to be 
correct, then by elimination the peak that has the retention time of 5.99 minutes is due to 
NAc-CoIIIMP8 and is 66 % pure. 
0.0
 0.3
 0.6
 0.9
 1.2
 1.5
 300 350 400 450 500 550 600 650 700
 Wavelength / nm
 Ab
 so
 rb
 an
 ce
 Co(III)MP8
 N-Ac-Co(III)-MP8
  
Figure 3.17:  UV/visible spectrum of NAcCoIIIMP8 and CoIIIMP8 in 0.001 M   
                                ammonium carbonate buffer. 
 
Four distinct absorption bands are observed at 350 nm, 425 nm, 533 nm and 567 nm for 
NAcCoIIIMP8 as shown in Figure 3.17 which corresponds to those earlier described in 
Figure 3.12.  Although the Soret seems to be slightly shifted by about 4 nm when going 
from CoIIIMP8 to NAcCoIIIMP8, the above spectra show peak positions that were 
expected and confirm that acetylation has no direct effect on the porphyrin electronic 
structure itself. 
 
0.0
 0.2
 0.3
 450 500 550 600 650 700N 
N 
B  
B  
Qv 
Qv Qo 
Qo 
 108 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.18:  FAB-MS spectrum of NAcCoIIIMP8 in NBA matrix. 
 
Furthermore, FAB-MS results gave m/z values expected for the molecular ions formed of 
CoIIIMP8 and NAcCoIIIMP8 as shown in Figure 3.18. The observed m/z for CoIIIMP8 and 
NAcCoIIIMP8 were 1509.5 and 1552.3 and the calculated 1508 and 1552, respectively.  
The molecular ions formed have low intensities, and is due to the small concentration of 
NAcCoIIIMP8 in the matrix.   
 
 
 
 
 
Low Resolution m/z 
In
 te
 ns
 ity
  (%
  A
 ge
 ) 
 
1509.5       1552.3 
 109 
3.2 Solution Chemistry of NAc-CoIIIMP8 
3.2.1 Determination of the Extinction Coefficients of NAc-CoIIIMP8 
 
The extinction coefficients (or molar absorption coefficients), e, were determined using 
ICP-OES and UV-vis spectroscopy.  ICP-OES was used to determine the concentration 
of CoIII ions in 0.210 mg NAc-CoIIIMP8, which was dissolved in 25 ml of 20 % 
acetonitrile/water (v/v) and further diluted 10 times to fall within the range of the ICP-
 OES multi-element calibration standards.  Table 3.1 shows the mean concentration 
obtained for cobalt ions detected during ICP-OES analysis.  Once the blank was 
subtracted, it was assumed that the concentration of CoIII ions detected by ICP-OES was 
equivalent to the concentration of NAc-CoIIIMP8 in solution.  The emission of CoIII ions 
were detected at 237.8 nm.   
 
Table 3.1:  Summary of ICP-OES results for NAc-CoIIIMP8 . 
Statistical Values Co_237.8 / ppm 
Mean Concentration, x 0.0236 
Std. Deviation, s 0.0004 
Relative Std. Deviation, sr 1.5 % 
 
The sample was analysed and the mean concentration observed was 0.0236 ppm with a 
relative standard deviation of 1.5 %.  After taking the dilution factor into account, it was 
calculated that the expected [CoIII] = 0.00319 mg l-1 and the observed [CoIII] = 0.0236 mg 
l-1.  Thus, the % cobalt in solution was 73.98 %.  This suggests that there is 73.98 % 
NAc-CoIIIMP8 in 25 ml solution.  This is expected as the starting cytochrome c was only 
ca. 80 % pure (from manufacturer?s specifications).  The same solution of NAc-CoIIIMP8 
was then put into a 1.0 cm pathlength curvette and using a UV-vis spectrophotometer, the 
extinction coefficients were determined from Beer?s law (e = bc/A, where e represents 
the molar extinction coefficient; b represents the pathlength of the cell; c represents the 
concentration of the solution and A represents the absorbance recorded for that particular 
solution) .  Table 3.2 summarises the results obtained using Beer?s Law.   
 110 
Table 3.2:  Calculation of Extinction Coefficients using Beer?s Law. 
 
 
 
3.2.2   Beer-Lambert Behaviour of NAc-CoIIIMP8 
 
 
The microperoxidases have been used as biomimetic models for haemproteins for many 
years, although many studies have been complicated by the ir tendency to aggregate.  An 
effective method for accessing concentration-dependant aggregation phenomena in 
solutions of haemproteins is to determine the degree of adherence of the solution to 
Beer?s law over the solute concentration range of interest.204  Deviation from Beer?s law 
signals aggregation of the absorbing species, while a single set of isosbestic points reflect 
a simple equilibrium between two species.205  Marques et. al102 carried out a Beer?s law 
study of MP8 in aqueous phosphate buffer solution (pH 7.00, ? = 0.1 M, 25 ?C) and 
revealed the existence of a simple monomer/dimer equilibrium (dimerisation constant KD  
= (1.17 + 0.02) x 105 M-1) below an MP8 concentration of 8.08 x 10-6 M but higher 
aggregates at concentrations above this.  Since low-spin MP8 was obtained at highest 
concentrations, this study suggested the likelihood of an aggregation mechanism 
involving intermolecular coordination by Cys-14 a -amino group of MP8.102  Marques et 
al.109 performed a similar study on NAc-MP8 in aqueous phosphate buffer solution (pH 
7.00, ? = 0.5 M, 25 ?C) and found that the Soret absorbance varies linearly with 
concentration up to ~3 x 10-5 M, hereafter the slope decreases. The limiting slope of 
NAc-MP8 was reported to be 1.55 + 0.01 x 105 M-1 cm-1, and was in agreement with that 
? 
(nm)  
Average 
Absorbance 
Average e  
(? 105) / M-1 cm-1 
Std. 
Deviation 
Relative Std. 
Deviation 
560.8 0.0289  0.139 0.019 1.37 
527.9 0.0286  0.137 0.017 1.27 
415.1 0.3350 1.61 0.052 0.32 
347.7 0.0617 0.296 0.035 1.18 
337.9 0.0560 0.268 0.035                   1.30 
 111 
obtained for monomeric MP8 (1.57 + 0.01 x 105 M-1 cm-1).  It is well established that the  
observed departure from linearity is the consequence of a simple dimerisation 
equilibrium (Equation 3.1, where M and D are monomeric and dimeric forms of NAc-
 MP8, respectively).204, 205 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.19:  Beer?s law plot for an aqueous solution of NAcCoIIIMP8 (pH 7.00, 25 ?C   
                       in distilled water and 1.0 M NaClO 4. 
 
The limiting slope for aqueous solution of NAcCoIIIMP8 (pH 7.00, 25 ?C) in distilled 
water and 1.0 M NaClO4 was 1.80 + 0.01 x 105 M-1 cm-1 and 1.66 + 0.01 x 105 M-1 cm-1, 
respectively.  These values were slightly higher than that obtained for MP8 and NAc-
 MP8 which were mentioned earlier and undoubtedly a consequence of the central metal 
being changed from iron to cobalt.  When the ionic strength of the NAcCoIIIMP8 solution 
was increased to 1.0 M using NaClO 4 there was no deviation from linearity, but there was 
a small decrease in the limiting slope. Figure 3.19 shows that both solutions of 
NAcCoIIIMP8 obey Beer?s law over the concentration range monitored.  Thus, all 
y = 1.66E+05x - 7.84E-02
 R2 = 9.97E-01
 y = 1.80E+05x - 1.20E-01
 R2 = 9.97E-01
 0.00
 2.00
 4.00
 6.00
 8.00
 0.00000 0.00001 0.00002 0.00003 0.00004
 Concentration / M
 Ac
 or
 re
 ct
 io
 n
 N-Ac-Co-MP8 in Distilled Water
 N-Ac-Co-MP8 in 1M NaClO4
 2M D KD =
 [D]
 [M]2
 KD 3.1(3.1) 
N-Ac-Co-MP8 in 1 M NaClO4 
 112 
0
 0.05
 0.1
 500 520 540 560 580 600
 NAcCoIIIMP8 solutions used for future experiments should be carried out with 
[NAcCoIIIMP8] < 40 ?M to avoid any possible aggregation.  This is more than sufficient 
for conventional UV-vis spectroscopic work. 
 
 
3.2.3 Determination of Spectroscopic pKas of NAc-CoIIIMP8 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.20:  pH dependence of NAc-CoIIIMP8 in acidic region. 
B 
Qv       Q0 
0
 0.2
 0.4
 0.6
 0.8
 1
 380 390 400 410 420 430 440 450 460
 Wavlength  / nm 
A
 bs
 or
 ba
 nc
 e
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.7
 0.8
 0.9
 350 450 550 650
 Wavelength / nm
 Ab
 so
 rb
 an
 ce
 pH 2 pH 3
 pH 4 pH 5
 pH 6 pH 7
 0.05
 0.1
 500 520 540 560 580 600
 B 
Qv       Q0 
Qv       Q0 B 
 113 
The spectroscopic pKas of NAc-CoIIIMP8 were determined using UV-vis spectroscopy.  
The pH dependence of the electronic spectrum of NAcCoIIIMP8 in an aqueous 
multicomponent buffer in the acidic region shows a shift of the Soret (B) band  from ?max, 
pH2 = 417.0 nm to ?max, pH6  = 415.1 nm as shown in Figure 3.20.  This change in 
wavelength is also accompanied by increase in the initial intensity by 5.40 %.  Although 
the peak position observed for the Q bands did not change, there were changes in 
intensity.  When changing from pH 2 to 6, the initial intensity Q v and Q0 bands increased 
by 5.79 % and 7.93 %, respectively.  
 
Firstly, the experimental data was fitted to an ionisation isotherm involving one 
spectroscopic pKa (Equation 2.1).  The fit was reasonable with an r2 value of 99.52 % and 
pKa value of 3.68 + 0.02.  The same data were then fitted to an ionisation isotherm 
involving two spectroscopic pKas (Equation 2.2).  The fit was marginally better with an r2 
value of 99.55 %. The values of the spectroscopic pKas determined were 3.4 + 0.5 and 4.2 
+ 0.7.  The experimental data were fitted to an ionisation isotherm involving three 
consecutive equilibria (Equation 2.3).  The r2 value again improved with a fit of 99.61 %.  
The values of the three spectroscopic pKas obtained from the theoretical curve fitting the 
data (Figure 3.21) were pK1 = 2.0 + 0.3, pK2 = 2.8 + 0.1 and pK3 = 3.91 + 0.03.  Finally, 
the experimental data were fitted to an ionisation isotherm with four consecutive 
equilibria (Equation 2.4).  The r2 value was excellent with a fit of 99.86 %.  The values of 
the four spectroscopic pKas were 1.5 + 1.8, 3.3 + 0.1, 4.7 + 0.2 and 6.8 + 1.6.  The r2 
value of the fit involving three pKas is better than the fit involving one and two pKas but 
the r2 value of the fit involving four pKas is better than the fit involving three pKas.  
However, some values obtained for four pKa fit are unrealistic since it produces values 
that are beyond the pH range under which we are working (1.5 + 1.8 and 6.8 + 1.6) with 
significant error.  In addition, glass electrode potentiometry cannot be used below pH of 
about 2 because of the non-Nernstian response.  Therefore, we conclude that the most 
reliable fit was observed for the isotherm with three consecutive equilibria (Figure 3.21). 
 
 114 
0.5
 0.55
 0.6
 0.65
 0.7
 0.75
 1.9 2.9 3.9 4.9
 pH 
Ab
 so
 rb
 an
 ce
  
Figure 3.21:  Variation of the Soret absorbance (415.05 nm) with pH titration of   
                              NAcCoIIIMP8 in acidic region (pH 2 ? 6) with NaOH. 
 
 
The pH dependence of the electronic spectrum of NAcCoIIIMP8 in aqueous 
multicomponent buffer in the basic region shows a shift of the Soret (B) band  from ?max, 
pH7 = 414.1 nm to ?max, pH12 = 419.1 nm.  This change in wavelength is also accompanied 
by increase in the initial intensity by 49.34 %.  The ?max of Qv and Q0 observed at pH 7 
were 529 nm and 561 nm, respectively.  However, ?max of Qv and Q0 at observed at pH 12 
was 533 nm and 566 nm, respectively. Thus there was an overall shift in wavelength.  
Similarly, this change in wavelength is accompanined by an initial increase in intensity of 
Qv and Q0 bands by 50.3 % and 30.6 %, respectively.  
 
 
 
 
 
2.0 
2.8 
3.9 
 115 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.22:  pH dependence of NAcCoIIIMP8 in basic region. 
 
The data were fitted for one, two, three and four spectroscopic pKas.  Firstly, the 
experimental data were fitted to an ionisation isotherm with one spectroscopic pKa 
(Equation 2.1).  The fit was poor with an r2 value of 90.67 % and pKa value of 8.02 + 
0.09.  The same data were fitted to an ionisation isotherm with two spectroscopic pKas 
(Equation 2.2).  The fit was slightly better with an r2 value of 91.23 %. The values of the 
spectroscopic pKas determined were 7.27 + 0.09 and 8.51 + 0.01.  Also, the experimental 
0
 0.2
 0.4
 0.6
 0.8
 1
 1.2
 1.4
 380 390 400 410 420 430 440 450 460
 Wavelength / nm
 Ab
 so
 rb
 an
 ce
 0
 0.05
 0.1
 0.15
 510 520 530 540 550 560 570 580 590 600 610
 0
 0.2
 0.4
 0.6
 0.8
 1
 1.2
 1.4
 310 360 410 460 510 560 610 660
 Wavelength / nm
 Ab
 so
 rb
 an
 ce
 pH 7
 pH 8
 pH 9
 pH 10
 pH 11
 pH 12
 Qv       Q0 
Qv           
             Q0 
B 
B 
 116 
data were fitted to an ionisation isotherm (Equation 2.3) consisting of three consecutive 
equilibria.  The values of the three spectroscopic pKas obtained from the theoretical curve 
fitting the data were pK4 = 7.47 + 0.03, pK5 = 9.19 + 0.04 and pK6 = 12.07 + 0.03.  The r2 
squared value was excellent with a fit of 99.88%.  Finally, the experimental data were 
fitted to an ionisation isotherm containing four consecutive equilibria (Equation 2.4).  
The r2 value was excellent with a fit of 99.92 %.  The values of the four spectroscopic 
pKas were 7.49 + 0.03, 9.24 + 0.01, 11.8 + 0.8 and 13.5 + 0.8.  The r2 value of the fit 
involving three pKas is better than the fits involving one and two pKas but the r2 value of 
the fit involving four pKas is better than the fit involving three pKas.  However, the fourth 
pKa (13.5 + 0.8) is unreliable as glass electrodes typically display Nernstian behaviour up 
to pH values of between 12 and 13.  Moreover this pKa is not observed in other MP8 
studies.102, 109   Therefore, the most reliable fit was observed for the isotherm containing  
three consecut ive equilibria (Figure 3.23). 
 
0.94
 1.04
 1.14
 1.24
 5.3 6.3 7.3 8.3 9.3 10.3 11.3 12.3 13.3
 pH
 Ab
 so
 rb
 an
 ce
  
Figure 3.23:  Variation of the Soret absorbance (415.05 nm) with pH titration of 
                              NAcCoIIIMP8 in basic region (pH 6-13) with NaOH. 
7.5 
9.2 
 12.1 
 117 
 
Figure 3.24:  Species distribution as a function of pH at 25 ?C for NAcCoIIIMP8 in a 
                       aqueous multicomponent buffer system.  
The interpretation of all the spectroscopic pKas is summarised in Figure 3.24.  At pH 
<1.5, we propose, largely by analogy with the FeIII system, that NAcCoIIIMP8 exists as a 
six-coordinate diaqua complex and that the first transition (pK1 = 2.0 + 0.3) involves 
coordination of the c-terminal carboxylate group (Glu-21) of the appended polypeptide.  
It is known that the pKa of the c-terminal carboxylic acid group in proteins is 2.13.206  In 
the present case, the interaction with CoIII-haem centre (with the a residual charge of +1) 
does not appear to  depress this pKa significantly.109  The second acid range transition 
(pK2 = 2.8 + 0.1) for NAcCoIIIMP8 probably involves the deprotonation of the cationic 
His-18 and concomitant replacement of the c-terminal carboxylate by the neutral 
heterocyclic base.  The suggestion that the pKa at 2.8 reflects the coordination of His-18 
is consistent with 1H NMR and other studies on several of the cytochromes c,207-213 as 
well as studies on model haempeptides.95  In Figure 3.24 we have assigned pK3  (3.91 + 
0.03) to the ionisation of one of the haem propanoic acid groups. This is an assumption 
 118 
based on the  pKa (4.31) for this group in the free amino acid,214 which is very close to the 
value assigned to pK3.  The second haem propanoic acid group is assigned to the 
ionization with pK4 = 7.47 + 0.03.  It is assumed that the pKa for this is raised relative to 
that of the free propanoic acid (4.69 + 0.03, 25 ?C, 0.1 M)214 because of the electronic 
repulsion between its conjugate base and the first propanoate group.  It is important to 
note that c-terminal n-propylamide NAcMP8 also shows two haem propanoic pKas in this 
region at 4.95 + 0.07, and 6.1 + 0.3, repectively.109   Indeed, many NMR and redox 
studies on a variety of haemproteins have shown that the ionisation of one haem 
propanoic acid group occurs in the pH range 4.5-7.5, while that of the second, also falling 
within this range, is frequently observed.28, 215-221   
 
pK5 (9.19 + 0.04) was assigned to the ionisation of the cobalt-bound water molecule.  
Table 3.3 shows the pKa values of other metal-bound water molecules; however the 
central metal in these cases is iron.  Comparing pK5 with the tabulated values shown 
below we notice that there is no significant difference between their values although the 
central metal has changed. 
 
Table 3.3:  Ionisation of iron-bound water (pKH2O) in various haemproteins. 
 
 
 
Ferrihaemprotein/Model pK(H2O) 
horse Hb222 8.86 + 0.02 
horse Mb222 9.04 + 0.03 
sperm whale Mb component IV223 8.93 
sperm whale Mb component IIIA223 9.03 
sperm whale Mb component II223 9.07 
NAcMP8109 9.59 + 0.01 
 119 
To confirm that pK5 (9.19 + 0.04) was assigned correctly to the ionisation of the cobalt-
 bound water molecule an investigation was carried out.  Using simple ligand substitution, 
pyridine was used to mask the ionisation of the cobalt-bound water molecule attached to 
the axial position of the cobalt porphyr in.    Figure 3.25 shows the change in absorbance 
at 415.1 nm as a function of pH for an aqueous solution of NAcCoIIIMP8 with pyridine at 
25 ?C (? = 0.1 M KCl).   
 
As before, the data were fitted for one, two, three and four spectroscopic pKas using 
Equations 2.1 to 2.4, respectively.  Firstly, the experimental data were fitted to an 
ionisation isotherm consisting of one spectroscopic pKa.  The fit was poor with an r2 
value of 75.25 % and pKa value of 7.7 + 0.2.  The data were then fitted to an ionisation 
isotherm with two spectroscopic pKas.  The fit was good with an r2 value of 98.47 %.   
The values of the spectroscpic pKas determined were 8.06 + 0.04 and 11.97 + 0.08.  The 
experimental data were fitted to an ionisation isotherm (Equation 2.1) with three 
consecutive equilibria.  The values of the three spectroscopic pKas obtained from the 
theoretical curve fitting the data were pK4 = 7.61 + 0.06, pK5 = 9.16 + 0.04 and pK6 = 
11.78 + 0.03.  The r2 squared value was excellent with a fit of 99.72 %.  Finally, the 
experimental data were fitted to an ionisation isotherm of four consecutive equilibria.  
The r2 value was excellent with a fit of 99.77 %.  The values of the four spectroscopic 
pKas were 7.4 + 0.2, 8.5 + 0.3, 9.9 + 0.4 and 11.73 + 0.05.  The r2 value of the fit 
involving three pKas is better than the fits involving one and two pKas but the r2 value of 
the fit involving four pKas is slightly better than the fit involving three pKas.  Using 
pyridine to bind to the axial position of NAcCoIIIMP8, so as to mask the ionisation the 
water molecule was not successful.  Comparing the fitted data for three spectroscopic 
pKas in the basic region for NAcCoIIIMP8 with pyridine and without pyridine, we found 
that there was no significant change in the value of the pKas.    
 
 120 
0.97
 1.02
 1.07
 1.12
 1.17
 1.22
 6.0 7.0 8.0 9.0 10.0 11.0 12.0 13.0
 pH
 Ab
 so
 rb
 an
 ce
  
Figure 3.25:  Variation of the Soret absorbance (415.05 nm) with pH titration of 
                       NAcCoIIIMP8 with pyridine in basic region (pH 6-13) with NaOH. 
 
Finally, we attribute pKa (12.07 + 0.03) to the ionisation of the coordinated histidine trans  
to the OH- to form the histidinate complex (His--CoIII-OH-).  The pKa values of the free 
histidine and imidazole are 14.4 and 14.2, respectively,224 while coordination of the 
histidine to the transition metals suc h as CoII and CuII reduces the value of these pKas to 
12.5 and 11.7, in each case.225  
 
In a similar study, Marques et. al reported six spectroscopic pKas in the acidic and basic 
regions for NAc-FeIIIMP8; pK1 = 2.1 + 0.3, pK2 = 3.12 + 0.07 and pK3 = 4.95 + 0.07,  pK4 
= 6.1 + 0.3, pK5 = 9.59 + 0.01 and pK6 = 12.71 + 0.05.109  When their data are compared 
to our data, it is clear that the results are quite similar in many instances.  
 
Also of interest is the pKa of water coordinated to CoIII in aquacobalamin.  That pKa is 
reported to be 8.09.155  Rubinson226 and coworkers have reported the pKa of 
aquacobalamin to be 7.8.  However this study was performed in 0.5 M KCl and the Cl- 
7.6 
11.8 
9.2 
 121 
ion may coordinate to the cobalt centre and affect results.  Although the authors have 
suggested that the electrolyte has no effec t on electrochemical potentials used in their 
work. We have decided to use the pKa value of 8.09 for comparative purposes.  The 
ionisation of the metal-bound water molecule for NAcCoIIIMP8 and aquacobalamin 
differs by 1.1 pH units.  This difference is significant and may be attributed to the 
environmental surroundings of the central cobalt ion, namely the porphine and corrin 
macrocyclic environments.   
 
It has been well established that the corrin cavity for cobalamin is much smaller than the 
porphyrin cavity in microperoxidase.7, 23, 163, 167, 195  Although the formal charge of cobalt 
in the porphyrin is +3, the actual charge of the metal depends on the extent the ligands 
donate electron density to the electron deficient cobalt centre.  Now, if we assume to a 
first approximation that the electronic properties of histidine (in NAcCoIIIMP8) and 
dimethylbenzimidazole (in B12a) have the same s donor/p acceptor abilities toward the 
cobalt centre, then the only variable is the difference in the donor ability of corrin and 
porphine macrocycles.  In cobalt corrinoids like B12a, CoIII centre is in close proximity to 
the electron rich corrin macrocycle and becomes more electron rich.  This leads to a 
decrease the electron density on the coordinated water molecule and consequently 
increases the ionisation of the axial water molecule (high Ka) and small pKa values.  In 
NAcCoIIIMP8, the CoIII centre is not in close proximity to the electron rich porphine 
macrocycle (due to the cavity being larger and more rigid) and does not significantly 
change the electron density on the coordinated water molecule.  This results in smaller 
ionisation constants being observed (larger pKa values).  Hence, the results show that the 
corrin moiety is a better electron donor toward CoIII, than the porphine moiety.  
 
 
3.2.4 Ligand Binding Studies of Selected N-Donor Ligands by NAc-CoIIIMP8 
 
Solutions of NAc-CoIIIMP8 in the presence of varying concentrations of ligand were 
allowed to equilibrate overnight in a water bath where the temperature was maintained at 
25.0 + 0.1 ?C.  Since CoIII in NAc-CoIIIMP8 is likely to be inert, the ligand had at least 16 
 122 
hours in which to bind to NAc-CoIIIMP8.  It was our intention to study the 
thermodynamic rather than the kinetic product.  Consequently, the longer equilibrium had 
no effect on the final form.  Absorbance measurements were taken for each solution of 
NAc-CoIIIMP8 with the corresponding ligand and these values were plotted against 
ligand concentration.  Table 3.4 shows the recorded absorbance measurements for the 
reaction of NAc-CoIIIMP8 with varying concentration of N-methylimidazole. 
 
Table 3.4:  Absorbance measurements for the reaction of NAc-CoIIIMP8 with varying    
                   concentration of N-methylimidazole. (pH 5.45, ? = 1.0 M, 25 ?C) 
Absorbance at 420 nm [N-methylimidazole] 
0.3535 0.000 
0.3725 0.000112 
0.3770 0.000335 
0.3824 0.000668 
0.3855 0.00111 
0.3883 0.00167 
0.3903 0.00277 
0.3922 0.00388 
0.3935 0.00498 
0.3959 0.00719 
0.3980 0.00883 
0.4010 0.0127 
0.4028 0.0149 
0.4046 0.0182 
0.4059 0.0214 
0.4075 0.0236 
0.4103 0.0290 
0.4102 0.0344 
 
 
 
 
 123 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 3.26:  Plot of the titration data for the binding of N-methylimidazole to  
                               NAc-CoIIIMP8 at 420 nm. (25 ?C, ? = 1.0 M) 
 
The formation constants between NAc-CoIIIMP8 and N-methylimidazole were 
determined using Equations 2.5 and 2.6.  The plot of the titration data for the binding of 
N-methylimidazole to NAc-CoIIIMP8 is illustrated in Figure 3.26.  The data obtained 
were initially fitted to account for a single formation constant (Equation 2.5).  K1obs was 
determined to be 817 + 264 M-1. The fit was poor, with an r2 value of 91.68 %.   
The pKa of N-methylimidazole is 7.21.  The final pH of the NAc-CoIIIMP8 solution with 
N-methylimidazole was 5.396.  The concentration of ligand in solution is dependent on 
0.35
 0.36
 0.37
 0.38
 0.39
 0.4
 0.41
 0.42
 0 0.005 0.01 0.015 0.02 0.025 0.03
 [N-Me-Im]
 Ab
 so
 rb
 an
 ce
 124 
the pH. Therefore Kobs values had to be converted into pH-independent formation 
constants using Equation 2.7.  After correcting for pH, log K1 was found to be 4.7 + 0.9.   
 
We then proceeded to fit two formation constants using Equation 2.6 and the results were 
excellent with an r2 value of 99.70 %.  We propose that the second formation constant, 
K2, is due to the formation of an outer-sphere complex, which eventually leads to a 
bisubstituted complex shown below in Figure 3.27 (where K2 = KOS*K2?).  Attempts have 
been made to rationalise the nature of the outer-sphere complex formed in aquacobalamin 
and the incoming ligand;227  it was suggested that the complex was formed by p-p 
interactions between the corrin ring (or functional groups on the corrin, presumably the 
amide side chains) and the ligand.  This can be extrapolated to NAc-CoIIIMP8 since it is a 
model for aquacobalamin.  K1obs and K2obs? were determined to be 10539 + 1704 and 50 + 
13.  Similarly, the data were corrected for pH using Equation 2.7.  Log K1 and log K2 
were calculated to be 5.8 + 0.9 and 3.6 + 0.8, respectively. 
 
 
 
 
 
 
 
Figure 3.27:  Proposed ligand substitution for NAc-CoIIIMP8. 
 
As before, absorbance measurements were taken for the reaction of each solution of NAc-
 CoIIIMP8 with varying concentration of pyridine (Table 3.5).  The formation constants 
between NAc-CoIIIMP8 and pyridine were determined using Equations 2.5 and 2.6.  The 
data obtained was initially fitted to account for only one formation constant (Equation 
2.5).  K1obs was determined to be 13 + 2 ? 104 M-1. The fit was reasonable with an r2 value 
of 98.64 %.  The pKa of pyridine is 5.19.  The final pH of the NAc-CoIIIMP8 solution 
with N-methylimidazole was 5.436.  After correcting for pH using Equation 2.7, log K1 
was found to be 4.3 + 0.8. 
Co
 OH2
 N
 L
 Co
 N
 L OH2
 KOS1 K1'
 Co
 N
 L
 K1
 L
 Co
 N
 L
 KOS2
 L
 Co
 NH
 L
 L
 K2'
 K2
 125 
Table 3.5:  Absorbance measurements for the reaction of NAc-CoIIIMP8 with varying    
                   concentration of pyridine. (pH 5.44, ? = 1.0 M, 25 ?C) 
Absorbance at 420 nm [Pyridine] 
0.3606 0.000 
0.3684 0.0000112 
0.3849 0.0000445 
0.3925 0.000111 
0.3956 0.000166 
0.3992 0.000221 
0.4013 0.000331 
0.405 0.000661 
0.4105 0.00186 
0.4112 0.00240 
0.4127 0.00349 
0.4159 0.00676 
0.4187 0.0133 
0.4207 0.0186 
 
We then proceeded to fit two formation constants using Equation 2.6.  The fit was 
excellent with an r2 value of 99.84% as shown in Figure 3.28.  K1obs and K2obs were 
determined to be 2.3 + 0.3 ? 104 M-1 and 2.4 + 0.8 ? 102 M-1.  As before, the data were 
corrected for pH using Equation 2.7.  Log K1 and log K2 were calculated to be 4.6 + 0.8 
and 2.6 + 0.8, respectively.   
 
 
 
 
 126 
          
Figure 3.28:  Titration data for the binding of Pyridine to NAc-CoIIIMP8 at 420 nm. 
                           (25 ?C, ? = 1.0 M)    
Table 3.6:  Formation constants for NAc-FeIIIMP8, NAc-CoIIIMP8 and B12a. 
Formation Constants   NAc-FeIIIMP8  NAc-CoIIIMP8 B12a 
N-methylimidazole 
log K1 
log K2 
 
4.17 + 0.01 
- 
 
5.8 + 0.9 
3.6 + 0.8 
 
4.63 + 0.05 
- 
Pyridine  
log K1 
log K2 
 
2.62 
- 
 
4.6 + 0.8 
2.6 + 0.8 
 
1.23 + 0.07 
 
Munro et al.228 reported the log K1 for the reaction between N-methylimidazole with 
NAc-FeIIIMP8 to be 4.17 + 0.01.  Marques et al.144 reported the log K1 for the reaction 
between N-methylimidazole with aquacobalamin to be 4.63 + 0.05.  These values of log 
K1 for NAc-CoIIIMP8 with N-methylimidazole are significantly higher than those reported 
for NAc-FeIIIMP8 and B12a.    Munro et al.228 reported the log K1 for the reaction be tween 
 127 
pyridine with NAc-FeIIIMP8 to be 2.62.  Knapton et al.229 reported the  log K1 for the 
reaction between pyridine with aquacobalamin to be 1.23 + 0.07.  Similarly, the values of 
log K1 for NAc-CoIIIMP8 with pyridine are significantly higher than those reported for 
NAc-FeIIIMP8 and B12a. 
 
Although log K2 is not reported in the above studies, the formation of a bis-complex is 
well established in MP8 chemistry.  Munro and coworkers found that in their ligand 
binding studies of NAc-FeIIIMP8 with pyridine and other amines, there were multiple 
equilibria observed.228  They also observed that there was a clear relationship between the 
basicity of the incoming ligand (as measured by its pKa) and the log K for the binding of 
that ligand by NAc-FeIIIMP8.  Thus, the affinity of the ligand toward Fe(III) in an iron 
porphyrin parallels its affinity toward the proton.  This was evident for primary amines 
and pyridines.  A similar observation was made by Pratt et al.230  In B12 chemistry 
precedents also exist for substitution of axial base by an exogenous ligand.229  Marques 
and coworkers have made an outstanding contribution and have performed numerous 
studies on aquacobalamin with a variety of ligands which all show that only the axial 
base was replaced and that ligand substitution proceeds through a dissociative 
interchange mechanism.143-146, 148-155        
 
In Table 3.6 we see that the value of formation constants formed between the selected 
ligands and NAc-CoIIIMP8 are higher than those formed for NAc-FeIIIMP8 and B12a.  It 
has already been established that the porphyrin cavity is larger and less flexible than the 
corrin cavity and that the Co(III) ion is smaller than Fe(III) ion. 7, 23  In NAc-CoIIIMP8 we 
have a small Co(III) ion in a large porphyrin cavity.  This results in weaker bonds being 
formed between cobalt and the porphyrin and the axial and proximal residues with the 
metal centre (due to longer bond distances).  This consequently leads to bis-substutution 
in NAc-CoIIIMP8.  However, in NAc-FeIIIMP8 and B12a we observe only mono-
 substitution, which strongly suggests that the Co(III)-His bond in NAc-CoIIIMP8 is 
weaker than the same bond in B12a and the Fe(III)-His bond in NAc-FeIIIMP8.  This result 
is unexpected and clearly merits further, and more detailed investigation.  It was not the 
principal focus of this investigation which was merely to show that investigations are 
 128 
feasible or not. Overall, it is not clear that as to whether the size of the macrocyclic cavity 
and the identity of the central metal play an important role in the chemistry that occurs at 
the metal centre. 
 
 
3.2.5 Kinetic Studies of the Reaction of Selected N -Donor Ligands  
            with NAc-CoIIIMP8 
The pseudo-first order rate constants were determined for the reaction between NAc-
 CoIIIMP8 and N-methylimidazole (NMI) and methylamine (MeNH2) at pH 7.063 and 
8.011, respectively.  As mentioned earlier, the pseudo-first order rate constants, kobs, were 
determined by curve fitting the absorbance versus time traces to Equation 2.8 using non-
 linear least-squares program employing a Newton-Raphson procedure.  It has been well 
established that the concentration of ligand in solution is dependent on the pH.  
Consequently,  kobs, values had to be converted into pH- independent rate constants using 
Equation 3.2. 
kcorrected = kobs(1+10pKa - pH)                   (3.2) 
In the reaction between NAc-CoIIIMP8 with N-methylimidazole we initially noticed that 
the data obtained did not follow convention pseudo-first order kinetics; instead there was 
some evidence for biphasic kinetics.  Table 3.4 below shows the data obtained when 
these data was fitted to Equation 2.8.  The data show that there are two rate constants; we 
proposed the first rate constant, k1,  to be the rate at which the first incoming ligand  
displaces water and the second, k1?, to be the rate constant for another incoming ligand 
which displaces the histidine to form a bis-complex.  
 
 
 
Figure 3.29:  Proposed kinetics of ligand substitution for NAc-CoIIIMP8. 
Co
 OH2
 N
 kos1
 Co
 N
 L OH2
 Co
 N
 L
 k-os1
 kex1
 Co
 N
 L
 L
 Co
 NH
 L
 L
 kos1'
 k-os1'
 kex1'
 129 
In Figure 3.29, kos1  and kos1? represents the rate of formation of the outersphere complexes 
and, kex1  and kex1? represents the exchange rate constants for the first and second ligand 
substitution, respectively.  The derivation of Equations 3.3 and 3.4 is presented in the 
Appendix E. 
k1  = Kos1*kex1                                               (3.3)
 k1' = Kos1'*kex1'                                            (3.4) 
Table 3.7:  Kinetics of NAc-CoIIIMP8 with NMI. 
* Corrected for pH using Equation 3.2 
0.0E+00
 4.0E-04
 8.0E-04
 1.2E-03
 0 0.5 1
 [NMI] / M
 k 
/ s
 -1
 k1
 k1'
  
Figure 3.30:  Plot of [NMI] against the observed rate constants for the 
                                      reaction with NAc-CoIIIMP8. 
[NMI] / M  k1obs  / s -1 k1 / s-1 * k1obs? / s-1 k1? / s-1 * 
0.0503 2.05E-04 4.92E-04 4.42E-06 1.06E-05 
0.1006 3.21E-04 7.70E-04 2.38E-05 5.74E-05 
0.20121 2.87E-04 6.89E-04 4.86E-05 1.17E-04 
0.5030 1.93E-04 4.63E-04 2.74E-05 6.59E-05 
0.7604 2.60E-04 6.25E-04 5.04E-05 1.21E-04 
0.8048 2.07E-04 4.98E-04 2.79E-05 6.71E-05 
1.0060 2.40E-04 5.76E-04 2.32E-05 5.59E-05 
k1 
k1? 
 130 
The plot of k versus [NMI] in Figure 3.30 shows that the rate at which ligand substitution 
occurs is independent of the concentration of the incoming ligand within experimental 
error.   
The results can be explained if the mechanism proceeds through a purely dissociative 
mechanism, i.e., if the rate of the reaction is controlled by the rate at which, firstly, the 
water molecule dissociates from the CoIII centre and, secondly, the histidine dissociates 
from the metal.  Hence, Figure 3.29 should be modified as shown in Figure 3.31.   
 
 
 
 
 
 
 
 
Figure 3.31:  Proposed kinetics of ligand substitution of NMI with NAc-CoIIIMP8. 
This reaction depicts a dissociative mechanism (D) since there is no nucleophilic 
participation of the incoming ligand in the transition state.147, 155   The  second order rate 
constant, k2, cannot be determined since the rate at which the reaction occurs is 
independent of N-methylimidazole concentration. 
Similarly, Table  3.8 below shows the two rate constants obtained when the data for the 
reaction between NAc-CoIIIMP8 with methylamine.  The pKa of methylamine is 10.941.  
Co
 OH2
 N
 k1
 Co + H2O
 N
 NMI
 fast
 Co
 N
 N
 N
 H3C
 Co
 N
 N
 H3C
 NH
 k2
 Co
 N
 N
 H3C
 N
 N
 H3C
 NMI
 fast
 HN
 131 
0
 100
 200
 300
 0 0.5 1 1.5 2 2.5 3
 [Methylamine] / M
 k 
/ s
 -1
     k1
     k1'
 As before the kobs values had to be converted into pH- independent rate constants using 
Equation 3.2. 
Table 3.8: Kinetics of NAc-CoIIIMP8 with MeNH2. 
[MeNH2] / M  k1obs / s -1 k1 / s-1  * k1obs? / s-1 k1? / s-1  * 
0.501 9.79E-03 8.34 4.47E-02 38.09 
1.001 1.56E-02 13.29 6.05E-02 51.55 
1.502 2.26E-02 19.26 4.07E-01 346.82 
1.753 3.27E-02 27.86 1.62E-01 138.05 
2.003 2.44E-02 20.79 4.38E-01 337.24 
2.253 4.78E-02 40.73 3.99E-01 340.00 
2.504 3.58E-02 30.51 4.29E-01 365.57 
* Corrected for pH using Equation 3.2 
  
Figure 3.32:  Plot of [MeNH2] against the observed rate constants for the 
                                    reaction with NAc-CoIIIMP8. 
1 
? 
 132 
The plot of k versus [MeNH2] in Figure 3.32 for the rate constants, k1 and k1?, are 
dependent on the concentration and participation of the incoming ligand and we propose 
that it involves the displacement of water and histidine residue, respectively.  Thus a 
dissociative interchange (Id) mechanism may be operating in both cases.155  We propose 
that the displacement of water and histidine by methylamine involves an interchange 
mechanism (I d), where the bond forming and bond breaking occur simultaneously, and 
thus the rate of reaction becomes dependent on the concentration of the incoming ligand.   
Similarly, the slopes of the above plot gave the second order rate constants, k2 and k2?, to 
be 13 + 3 M-1 s-1 and 177 + 51 M-1 s-1.   In the reaction of N-methylimidazole with NAc-
 CoIIIMP8, we observe that the reaction is independent of ligand concentration and 
therefore must proceed through a dissociative (D) mechanism.  The results show that the 
rate of reaction for methylamine being faster than N-methylimidazole with NAc-
 CoIIIMP8.  These differences in rate constants may be attributed to the size of the 
incoming ligands.  N-methylimidazole is a secondary amine and is relatively more bulky 
than methylamine which is a primary amine; therefore it is easier for methylamine to 
attach to the metal centre compared to N-methylimidazole.  
 
3.2.6 Kinetic Studies of Selected N-Donor Ligands by B12a 
The kinetics involving the ligand substitution of imidazole144 and methylamine146 with 
aquacobalamin has been reported by Marques and coworkers.  However, we have 
decided to repeat these studies under the different conditions so as to have comparative 
data that are relevant to our project.  The pseudo- first order rate constants were 
determined for the reaction between B12a with N-methylimidazole and methylamine at pH 
7.063 and 7.951, respectively.  Similarly, as a result of the reactions being performed at 
different pH, kobs, values had to be converted into pH- independent rate constants using 
Equation 3.2.  The usually inert CoIII-centre is exceptionally labile and the axial water 
ligand easily undergoes substitution with a variety of ligands as shown below  in Figure 
3.33. 
 133 
Co
 OH2
 N
 L
 Co
 N
 L OH2
 Co
 NH
 L
 -H2O
  
Figure 3.33:  Shows schematic of ligand substitution in B12a.155 
Table 3.9 below shows the calculated rate constants obtained when the data for the 
reaction between B12a with N-methylimidazole were fitted to Equation 2.9.   
Table 3.9:  Kinetics of B12a with NMI. 
[NMI] / M  k1obs  / s -1 k1/ s-1 
0.00010055 2.23E-03 5.36E-03 
0.0001508 2.54E-03 6.11E-03 
2.01E-04 2.96E-03 7.11E-03 
2.51E-04 3.34E-03 8.02E-03 
3.02E-04 3.70E-03 8.90E-03 
3.52E-04 2.93E-03 7.03E-03 
4.02E-04 4.72E-03 1.13E-02 
0.0005208 5.53E-03 1.33E-02 
6.03E-04 6.04E-03 1.45E-02 
7.04E-04 7.01E-03 1.69E-02 
0.0008044 7.48E-03 1.80E-02 
0.000905 9.01E-03 2.16E-02 
0.0010055 1.04E-02 2.50E-02 
0.001207 1.14E-02 2.73E-02 
1.41E-03 1.28E-02 3.09E-02 
1.61E-03 1.62E-02 3.90E-02 
0.00181 1.68E-02 4.04E-02 
0.002011 1.98E-02 4.76E-02 
 134 
0
 0.025
 0.05
 0 0.0005 0.001 0.0015 0.002 0.0025
 [NMI]
 k
 1 
/ s
 -1
  
Figure 3.34:  Plot of [NMI] against the observed rate constants for the 
                                      reaction with B12a. 
The slope of the plot of k1 versus [NMI] in Figure 3.34 shows that the rate at which 
ligand substitution occurs is dependant on the concentration of the incoming ligand .  We 
propose that it involves the displacement of water and  that a dissociative interchange (Id) 
mechanism may be operating.155 The second order rate constant, k2, for the reaction 
between N-methylimidazole and B12a is 21.82 + 0.53 M-1 s-1.  However, Marques et al.144 
reported a second order rate constant, k2, for the same reaction to be 16.6 + 0.3 M-1 s-1.  
Marques et al.144 used KCl as an electrolyte and this would have formed a more inert 
chloro-B12a complex which would account for the lower value obtained.    
Sx = 7.6 ? 10-5 
 135 
Table 3.10 below shows the calculated rate constants obtained when the data for the 
reaction between B12a with methylamine were fitted to Equation 2.9.  
Table 3.10: Kinetics of B12a with MeHH2. 
[MeNH2]/ M k1obs / s-1 k1 / s-1 
0.2505 7.24E-01 7.08E+02 
0.5010 5.54E-01 5.42E+02 
0.7515 1.03E+00 1.01E+03 
1.2525 1.70E+00 1.67E+03 
1.5030 1.84E+00 1.80E+03 
1.7535 2.18E+00 2.13E+03 
2.0040 2.21E+00 2.16E+03 
2.2545 2.86E+00 2.79E+03 
2.5050 2.93E+00 2.86E+03 
0
 1000
 2000
 3000
 0 0.5 1 1.5 2 2.5
 [MeNH2] / M
 k1
 / s
 -1
  
Figure 3.35:  Plot of [MeNH2] against the observed rate constants for the 
                                    reaction with B12a. 
Sx = 0.004 
 136 
The slope of the plot of k1 versus [MeNH2] in Figure 3.34 shows that the rate at which 
ligand substitution occurs is dependant on the concentration of the incoming ligand . 
Similarly, we propose that it involves the displacement of water and that a dissociative 
interchange (Id) mechanism may be operating.155   The second order rate constant, k2, for 
the reaction between methylamine and B12a is 1.052 + 0.071 ? 103 M-1 s-1.  In a similar 
study Marques et al.146 reported the second order rate constant for the same reaction to be 
0.927 + 0.0022.   
 
Table 3.11:  Second order rate constants of NAc-CoIIIMP8 and B12a.  
Ligand Second order rate 
constant for 
NAc-CoIIIMP8 / M -1 s-1 
Second order rate 
constant for 
B12a / M-1 s-1 
NMI 
 
 
MeNH2 
- 
- 
 
k2 = 1.32 + 0.3 ? 104 
k2? = 177 + 51 
k2 = 16.6  (a) 
k2 = 21.82 + 0.53 
 
k2 = 9.27 ? 10-1  (b) 
k2 = 1.052 + 0.071 ?103 
a ? data obtained from reference 144 
b ? data obtained from reference 146 
 
Table 3.11 shows that in the reaction of NAc-CoIIIMP8 with N-methylimidazole and 
methylamine, two ligand substitutions  are observed.  In both reactions, we propose that 
the first rate constant is due to the substitution of the axial water molecule and the second 
due to the substitution of the proximal histidine residue giving rise to the formation of a 
bis-complex.  Also when comparing the second order rate constants of both ligands with 
NAc-CoIIIMP8, we observe that the rate of substitution of methylamine is significantly 
faster than N-methylimidazole.  This may be attributed to the relative size of the ligands 
as mentioned earlier.  In addition, the reaction of NAc-FeIIIMP8 with N-methylimidazole 
is independent of ligand concentration which suggests a dissociative mechanism.  While 
 137 
the reaction of NAc-FeIIIMP8 with methylamine the rate of ligand substitution is 
dependant on ligand concentration (suggests Ia mechanism). 
  
Although the observed reaction rates of B12a with N-methylimidazole compare very well 
with literature, the observed reaction rate of B12a with methylamine do not.  The results 
show that the rate of the reaction between NAc-CoIIIMP8 and B12a with N-
 methylimidazole and methylamine are significantly different.  Overall it is clear that the 
solution chemistry of NAc-CoIIIMP8 is complex and does not mimic the chemistry of 
NAc-FeIIIMP8 nor B12a due to bisubstitution of the cobalt centre.   
 
 
 138 
Chapter 4 Conclusions and Future Work 
4.1  Conclusions 
4.1.1 The synthesis and characterisation of products leading to the  
            formation of  NAc-CoIIIMP8 
 
In this project we synthesised NAc-CoIIIMP8 in an attempt to develop a model system to 
probe the chemistry of vitamin B12a.  All products formed during the synthesis of NAc-
 CoIIIMP8 were characterised using HPLC, UV-vis spectroscopy and FAB-MS with good 
results, where appropriate, that compared well with literature values.   
 
We used the procedure developed by Munro and Marques with some modifications to 
synthesise FeIIIMP8.109  The synthesis of FeIIIMP8 was successfully performed in 55 % 
yield.  The removal of the central FeIII ion from FeIIIMP8 was successfully accomplished 
by a process known as reductive demetalation.79  MP8 free base was successfully 
synthesised in 73 % yield.  The MP8 free base was further analysed using luminescence 
spectroscopy and characteristic triplet-triplet absorption band was observed on the 
emission spectrum.   This provided further evidence that the reductive demetalation 
procedure was successful.  CoIII was successfully inserted into the empty porphyrin 
cavity of the MP8 free base using a procedure which was also adopted from the report by 
Primus and coworkers.79  CoIIIMP8 was synthesised in 69 % yield.  CoIIIMP8 was further 
analysed using luminescence spectroscopy, however, the characteristic triplet-triplet 
absorption band was not observed on the emission spectrum.  The quenching of the  
triplet-triplet absorption band provided more evidence that remetalation was successful. 
Finally, acetylation of CoIIIMP8 was carried out according to methods developed by 
Munro and Marques.109  NAc-CoIIIMP8 was synthe sised in 58 % yield.  Even if the 
acetylation step was done very early in the study, better yields would not have been 
obtained, since this would acetylate the nitrogen of the histidine residue instead of Cys-14 
residue.  The overall yield from FeIIIMP8 to was calculated to be 29 %.   
 
 
 139 
4.1.2 Solution Chemistry of NAc-CoIIIMP8 
 
We have used UV-vis spectroscopy to fully characterise the solution chemistry of 
monomeric NAc-CoIIIMP8 over the pH range.  Beer?s law studies shows that NAc-
 CoIIIMP8 remains monomeric in aqueous solution up to concentrations of at least 35 ?M.  
The extinction coefficient, e, of NAcCoIIIMP8 (pH 7.00, 25 ?C) in distilled water and 1.0 
M NaClO 4 was 1.80 + 0.01 x 105 M-1 cm-1 and 1.66 + 0.01 x 105 M-1 cm-1, respectively.   
 
The spectroscopic changes observed for NAc-CoIIIMP8 during the course of 
spectrophotometric titration are very similar to those observed for NAc-FeIIIMP8,109 with 
both being consistent with six successive ionisations.  At pH <1.5, we propose that 
NAcCoIIIMP8 exists as a six-coordinate diaqua complex.  By analogy with NAc-
 FeIIIMP8, we attribute the first (pK1 = 2.0 + 0.3) involves coordination of the c-terminal 
carboxylate group (Glu-21) of the appended polypeptide.  The second acid range 
transition (pK2 = 2.8 + 0.1) for NAcCoIIIMP8 probably involves the deprotonation of the 
cationic His-18 and concomitant replacement of the c-terminal carboxylate by the neutral 
heterocyclic base.  We have assigned pK3  (3.9 + 0.03) to the ionisation of one of the 
haem propanoic acid groups.  The second haem propanoic acid group is assigned to the 
ionization with pK4 = 7.47 + 0.03 and pK5 (9.19 + 0.04) was assigned to the ionisation of 
the cobalt-bound water molecule.  Finally, we attribute pK6 (12.07 + 0.03) to the 
ionisation of the coordinated histidine trans to the OH- to form the histidinate complex 
(His--CoIII-OH-).   
 
To further delineate the solution chemistry of NAc-CoIIIMP8, its reaction with selected 
N-donor ligands was studied by monitoring the dependence of the absorbance at the Soret 
wavelength on the added ligand concentration.  The formation constants between NAc-
 CoIIIMP8 and N-methylimidazole and pyridine were determined successfully.  When 
comparing the formation constants of NAc-FeIIIMP8, NAc-CoIIIMP8 and B12a, we found 
that the value of log K1 for NAc-CoIIIMP8 is significantly higher than that reported for 
NAc-FeIIIMP8 and B12a.  In addition, we observed the formation of a bi-substituted 
complex in the reaction of NAc-CoIIIMP8 with the selected N-donor ligands, and only 
 140 
mono-substituted with NAc-FeIIIMP8 and B12a.  This strongly suggests that the Co(III)-
 His bond in NAc-CoIIIMP8 is weaker than the same bond in B12a and the Fe(III)-His bond 
in NAc-FeIIIMP8.   
 
In the reaction between NAc-CoIIIMP8 with N-methylimidazole and methylamine we 
initially noticed that the data obtained did not follow convention pseudo- first order 
kinetics; instead there was some evidence for biphasic kinetics.  In the reaction of N-
 methylimidazole with NAc-CoIIIMP8, the results can be explained if the mechanism 
proceeds through a purely dissociative mechanism, i.e., if the rate of the reaction is 
controlled by the rate at which, firstly, the water molecule dissociates from the CoIII 
centre and, secondly, the histidine dissociates from the metal.  The second order rate 
constant, k2, cannot be determined since the rate of reaction is independent of N-
 methylimidazole concentration.  In the reaction of methylamine with NAc-CoIIIMP8, the 
rate constants are dependent on the concentration and participation of the incoming 
ligand and we propose that it involves the displacement of water and histidine residue, 
respectively.  The results suggest that a dissociative interchange (Id) mechanism may be 
operating.155  We propose that the displacement of water and histidine by methylamine 
involves an interchange mechanism (Id), where the bond forming and bond breaking 
occur simultaneously, and thus the rate of reaction becomes dependant on the 
concentration of the incoming ligand.   
Overall, the results show that the rate of reaction for methylamine with NAc-CoIIIMP8 
being faster than N-methylimidazole.  As mentioned earlier, these differences in rate 
constants may be attributed to the size of the incoming ligands.  The rate of reaction of 
B12a with N-methylimidazole  and methylamine was also determined.  Although the 
observed reaction rates of B12a with N-methylimidazole compare very well with 
literature, the observed reaction rate of B12a with methylamine do not.  The results show 
that the rate of the reaction between NAc-CoIIIMP8 and B12a with methylamine is  
significantly different.   
 141 
The results obtained from the Beer?s law studies and pKa determinations of NAc-
 CoIIIMP8 are in excellent agreement with similar studies performed with NAc-FeIIIMP8. 
However, it is clear that the ligand substitution and kinetic studies of NAc-CoIIIMP8 is 
complicated due to the formation of a bis-complex.   The results presented in this work 
do give a general indication on how thermodynamically stable a CoIII ion is in a 
porphyrin ring and also to a very limited extent show that a porphyrin does not confer the 
same kinetic lability on the CoIII ion as the corrin ring.   
 
 
4.2 Future Work 
 
This work is sets the foundation for future endevours where much more in depth studies 
of thermodynamics of ligand coordination, kinetics of substitution processes, molecular 
modeling, reduction and self-exchange reactions, catalytic studies and various other 
studies on the solution chemistry of N-Ac-Co(III)MP8 can be unraveled. 
 
 
 
 
 
 142 
References 
 
1. Beinert, H., J. Biol. Chem., 2002, 277, 37967. 
2. Bertini, I., Rosate, A., Proc. Natl. Acad. Sci., U. S. A., 100, 7, 3601. 
3. Williams, R. J. P., Coord. Chem. Rev., 1990, 100, 573. 
4. Kaim, W.; Schwederski, B., Bioinorganic Chemistry: Inorganic Elements in the 
Chemistry of Life, John Wiley and Sons, New York, 1991, Chapters 1 - 4, 7, 8, 10. 
5. Cotton, F. A.; Wilkinson, G., Advanced Inorganic Chemistry, 5th Edition, John 
Wiley and Sons, New York, 1988, Chapters 18, 20 and 30. 
6. Cotton, F. A.; Wilkinson, G., Basic Inorganic Chemistry, 5th Edition, John Wiley 
and Sons, New York, 1987, Chapters 18 and 31. 
7. Frausto da Silva, J. J. R.; Williams, R. J. P., The Biological Chemistry of Elements, 
1991, Claredon Press, Oxford. 
8. Huheey, J. E., Inorganic Chemistry, 3rd Edition, Harper and Row, New York, 1983, 
Chapters 23 and 31. 
9. Thiel, E. C.; Raymond, K. N., Bioinorganic Chemistry, University Science Books, 
California, 1994, Chapter 1. 
10. Martin, H.; Gordon, R. M., Deep Sea Res., 1988, 35, 177. 
11. Fox, L. L., Geochim. Cosmochim. Acta, 1988, 52, 771. 
12. Aisen, P., Iron, Biochemistry and Medicine, 2nd Edition, Academic Press, London, 
1980, Chapter 3. 
13. Spiro, T. G., Proteins of Iron Metabolism, Grune and Stratton, New York, 1977, 
23. 
14. Shriver, D. F.; Atkins, P. W.; Langford, C. H., Inorganic Chemistry, Oxford 
University Press, 1994, p. 246. 
15. Mankinen, M. W; Churg, A. K., in Iron Porphyrins, Part 1; Lever, A. B. P., Gray, 
H. B., Eds.; Addison?Wesley Publishing Company, Inc.: Reading, M. A., 1983, pp. 
141-235. 
16. Silver, J. B. L.; Morrison, I. E. G.; Barnard, P. W. C., Inorg. Chim. Acta, 1983, 78, 
205. 
 143 
17. Wrigglesworth, J. M.; Baum, H., Iron Porphyrins in Biochemistry and Medicine, 
Academic Press London, 1980, pp. 45-67. 
18. Messerschmidt, A.; Huber, R.; Poulos, T.; Wieghardt, K., Handbook of 
Metalloproteins, John Wiley and Sons, Volume 1, 2001. 
19. Glusker, J. R.; Katz, A. K.; Bock, C. W., The Rigaku Journal, 1999, 16, 2. 
20. Glusker, J. R., Adv. Protein Chem., 1991, 42, 1. 
21. Smith, K. M., Porphyrins and Metalloporphyrins, Elserivier Publishing Co., 1975, 
pp. 1-200 and 672. 
22. Ochiai, E., Bioinorganic Chemistry, Allyn and Bacon Inc., 1977, Chapters 1, 5, 6 
and 12. 
23. Pratt, J. M., B12, Volumes 1-2, Dolphin, D., Ed; Wiley, New York, 1982. 
24. Pratt, J. M., Techniques and Topics in Bioinorganic Chemistry, (Ed. McAuliffe, C. 
A.), Macmillan Press Ltd., London, 1971, pp. 109-204. 
25. Lemberg, R.; Barrett, J., Cytochromes, Academic Press, New York, 1973. 
26. Nomenclature Committee of the International Union of Biochemistry (NCIUB)  
Nomenclature of electron transfer proteins. Recommendations 1989. Eur. J. 
Biochem., 1991, 200, 599. 
27. Banci, L.; Bertini, K. L.; Gray, H. B.; Luchinat, C.; Reddig, T.; Rosato, A.; Turano, 
P., Biochemistry, 1997, 36, 9867. 
28. Pettigrew, G. W.; Moore, G. R., Cytochrome c ? Biological Aspects, Springer-
 Verlag, Berlin, 1987. 
29. Levis, D. F. V., Cytochromes P450: Structure, Function and Mechanism, Taylor 
and Francis, London, 1996. 
30. Egawa, T.; Ogura, T.; Makino, R.; Ishimura, Y.; Kitagawa. T., J. Biol. Chem., 
1991, 266, 10246. 
31. Dawson, J.H., Science, 1988, 240, 433.  
32. Jacob, G.S.; Orme  Johnson, W.H.,  Biochemistry, 1979, 18, 2975.  
33. Goldberg, I.; Hochman, A., Arch. Biochem. Biophys., 1989, 268, 124.  
34. Chiu, J.T.; Loewen, P.C.; Switala, J.; Gennis, R.B.; Timkovich, R., J. Am. Chem. 
Soc., 1989, 111, 7046.  
 144 
35. Fita, I.; Silva, A.M.; Murthy, M.R.N., Rossmann, M.G., Acta Crystallogr., 1986, 
B42, 497.  
36. Murshudov, G.N.; Melik  Adamya n, W.R.; Grebenko, A.I.; Barynin, V.V.; Vagin, 
A.A.; Vainshtein, B.K.; Dauter, Z.; Wilson, K.S.,  FEBS Lett., 1992, 312, 127.  
37. Gouet, P.; Jouve, H.M.; Dideberg, O.  J. Mol. Biol., 1995, 249, 933.  
38. Bravo, J.; Verdaguer, N.; Tormo, J.; Betzel, C.; Switala, J.; Loewen, P.C.; Fita, I., 
Structure , 1995, 3, 491.  
39. Ortiz, P. R., Cytochrome P450 Structure ? Mechanism and Biochemistry, 1986, 
Plenum Press, New York and London, pp. 217-71. 
40. Nelson, D. R.; Koymans, L.; Kamataki, T.; Stegeman, J. J.; Feyereisen, R.; 
Waxman, D. J.; Waterman, M. R.; Gotoh, O.; Estabrook, R. W.; Gunsalus, I. C.; 
Nebert, D. W., Pharmacogenetics, 1996, 6, 1.  
41. Haserman, C. A.; Ravichandran, K. G.; Peterson, J. A.; Deseinhofer, J., J.Biol. 
Chem., 1994, 236, 1169. 
42. Omura, T.; Sato, R., J. Biol. Chem.,1964, 239, 2370.  
43. Omura, T.; Sato, R., J. Biol. Chem.,1964, 239, 2378.  
44. Champion, P. M.; Stallard, B. R.; Wagner, G. C; Gunsalus, I. C., J. Am. Chem. 
Soc., 1982, 104, 5469. 
45. Shannon, R. D.; Prewitt, C. T., Acta Cryst., B26, 1970, 1046. 
46. Shimura, Y., Bulletin of the Chemical Society of Japan, 1988, 61, 693. 
47. Hill, H. A. O.; R?der, A.; Williams, R. J. P., Structure and Bonding, 1970, 8, 123. 
48. McMurry, T. J.; Groves, J. T., Cytochrome P-450, Plenum Press, 1986, New York, 
pp. 1-28. 
49. Groves, J. T.; Watanabe, Y., J. Am. Chem. Soc., 1988, 110, 8433.  
50. Poulos, T. L.; Finzel, B. C.; Howard, A. J., Biochemistry, 1986, 25, 5314. 
51. Schenkman, J. B.; Sligar, S. G.; Cinti. D. L., Pharmacology and Therapeutics, 
1981, 12, 43. 
52. Schenkman, J. B.; Kupfer, D., Hepatic Cytochrome P-450 Monoxygenase System, 
Permagon, Oxford, 1982.  
53. Sligar, S. G.; Murray, R. I., Cytochrome P450 and other bacterial P450 enzymes, 
Plenum, New York, 1986, pp. 429-503. 
 145 
54. Ruckpaul, K.; Rein, H.; Blank, J.  Frontiers in Biotransformation, 1989, 1, 1.  
55. Veitch, N. C.; Williams, R. J. P., Frontiers in Biotransformation, 1992, 7, 279.  
56. Archakov, A. I.; Buchmanova, G. I., Cytochrome P-450 and Active Oxygen. Taylor 
and Francis, London, 1990. 
57. Welinder, K.G., Curr. Opin. Struct. Biol., 1992, 2, 388.  
58. Imai, Y.; Sato, R.; Iyanagi, T., Biochemistry, 1977, 82, 1237. 
59. Kimura, S.; IkedaSaito, M.,  Proteins, 1992, 3, 113.  
60. Suzuki, T.; Imai, K., Cell. Mol. Life Sci., 1998, 54, 979. 
61. Kraulis, K., J. Appl. Crystallogr., 1991, 21, 946. 
62. Perutz, M. F.; Rossmann M. G.; Cullis, A. F.; Muirhead, H.; Will, G.; North, A. C. 
J., Nature, 1960, 185, 416. 
63. Perutz, M. F.; J. Mol. Biol., 1965, 13, 646. 
64. Scott, E. E.; Paster, E. V.; Olson, J. S., J. Biol. Chem., 2000, 275, 27129. 
65. Kendrew, J. C.; Dickerson, R. E.; Strandberg, B. E.; Hart, R. G.; Davies, D. R.; 
Philips, D. C.; Shore, V. C., Nature, 1960, 185, 422. 
66. West, E.; Todd, W.; Mason, H. S.; Van Bruggen, T. J., Textbook of Biochemistry, 
1975, 4th Edition.  
67. Hughes, M. N., Comprehensive Coordination chemistry, Vol. 4, 1987, Pergamon 
Press, Oxford, pp. 614-617. 
68. Marques, H. M.; Shongwe, M. S.; Munro, O. Q.; Egan, T. J., S. Afr. J. Chem., 1997, 
50, 166. 
69. Shack, J.; Clark, W. M., J. Biol. Chem., 1947, 171, 143. 
70. Dunford, H. B.; Stillman, J. S., Coord. Chem. Rev., 1976, 19, 187. 
71. Maehly, M. C.; Akeson, A., Acta. Chem. Scand., 1958, 12, 1259. 
72. Davies, T. H., Biochim. Biophys. Acta., 1973, 329, 108. 
73. Hodgkinson, J.; Jordan, R. B., J. Am. Chem. Soc., 1973, 95, 763. 
74. Stein, T. P.; Plane, R. A., J. Am. Chem. Soc., 1969, 91, 607. 
75. Simplicio, J.; Schwenzer, K., Biochemistry, 1973, 12, 1923. 
76. Hambright, P.; Chock, P. B., K., J. Inorg. Nucl. Chem., 1975, 37, 2363. 
77. Walker, F. A.; Lo, M.; Ree, M. T., J. Am. Chem. Soc., 1976, 98, 5552. 
 146 
78. Marques, H. M.; Munro, O. Q.; Crawcour, M. L., Inorg. Chim., Acta., 1992, 196, 
221. 
79. Primus, J.; Marelle, G. B.; Mandon, D.; Boeren, S.; Veeger, C.; Weiss, R.; Rietjens, 
I. M. C. M., J. Biol. Inorg. Chem., 1999, 4, 274. 
80. Momenteau, B.; Rougee, M.; Loock, B., Eur. J. Biochem., 1976, 71, 63. 
81. Primus, J.; Marelle, G. B.; Mandon, D.; Boeren, S.; Veeger, C.; Weiss, R.; Rietjens, 
I. M. C. M., J. Biol. Inorg. Chem., 2002, 7, 7.  
82. Marques, H. M.; Munro, O. Q.; Munro, T.; de Wet, M.; Vashi, P. R., Inorg. Chem., 
1999, 38, 2312. 
83. Tsou, C. L., Nature, 1949, 164, 1134. 
84.  Tsou, C. L., Biochem. J., 1951, 49, 362. 
85. Tsou, C. L., Biochem. J., 1951, 49, 367. 
86. Tuppy, H.; Bodo, G., Montash. Chem., 1954, 85, 1182. 
87. Tuppy, H.; Paleus, S., Acta Chem. Scand., 1955, 9, 353. 
88. Ehrenberg, A.; Theorell, H., Acta Chem. Scand., 1955, 9, 1193. 
89. Harbury, H. A.; Loach, P. A., J. Biol. Chem., 1960, 235, 3640. 
90. Peterson, J.; Silver, J.; Wilson, M. T.; Morrison, I. E. G., J. Inorg. Biochem., 1980, 
13, 75. 
91. Adams, P. A.; Baldwin, D. A.; Marques, H. M., in Scott, R. A.; Mauk, A. G.(Eds.), 
Cytochrome c: A Multidisciplinary Approach, University Science Books, Sausalito, 
California, 1996, pp. 635-692. 
92. Carraway, A. D.; Burkhalter, R. S.; Timkovich, R.; Perterson, J., J. Inorg. 
Biochem., 1993, 52, 201. 
93. Kimura, K.; Peterson, J.; Wilson, M. T.; Cookson, D. J.;  Williams, R. J. P., Inorg. 
Biochem., 1981, 15, 11. 
94. Mondeli, R.; Scaglioni, L.; Mazzini, S.; Bolis, G.; Ranghino, G., Magn. Reson. 
Chem., 2000, 38, 229. 
95. Paleus, S.; Ehrenberg, A.; Tuppy, H., Acta Chem. Scand., 1955, 9, 365. 
96. Wilson, M. T.; Ranson, R. J.; Masiakowski, P.; Czarnecka, E.; Brunori, M.; Eur. J. 
Biochem., 1977, 77, 193. 
97. Urry, D. W.; Pettegrew, J. W., J. Am. Chem. Soc., 1967, 89, 5276. 
 147 
98. Nassif, R.; Baumgartner, C. P.; Sellers, M.; May, L., Naturforsch., 1976, 31C, 232. 
99. Jehanli, A .M. T.; Stotter, D. A.; Wilson, M. T., Eur. J. Biochem ., 1976, 71, 613. 
100. Nassif, R.; Baumgartner, C. P.; Sellers, M.; May, L., Eur. J. Biochem ., 1974, 46, 
625. 
101. DerVartanian, D. V., Biochim. Biophys. Res. Commun., 1970, 41, 932. 
102. Aron, J. A.; Baldwin, D. A.; Marques, H. M.; Pratt, J. M.; Adams, P. A., J. Inorg. 
Biochem., 1986, 27, 227. 
103. Baldwin, D. A.; Marques, H. M.; Mabuya, M. B., S. Afr. J. Chem., 1987, 40, 103. 
104. Arutyunyan, A. M.; Surkov, S. A.; Luchkov, V. A.; Lysko, A. I.; Mironov, A. F.; 
Rumyantseva, V. S.; Evstigneeva, R. P., Mol. Biol. (Moscow), 1986, 20, 502. 
105. Urry, D. W., J. Am. Chem. Soc., 1967, 89, 4190. 
106. Harbury, H. A.; Loach, P. A., Proc. Natl. Acad. Sci., U. S. A.,1959, 45, 1344. 
107. Yang, E. K.; Sauer, K., in Electron Transport and Oxygen Utilisation (Ed.: Chien, 
H.), Elsevier, New York, 1982. 
108. Wang, J.-S.; Van Wart, H. E., J. Phys. Chem., 1989, 93, 7925. 
109. Munro, O. Q.; Marques, H. M., Inorg. Chem., 1996, 35, 3752. 
110. Wang, J.-S.; Van Wart, H. E.; Tsai, A.-L.; Heldt, J.; Palmer, G., J. Biol. Chem., 
1992, 267, 15310. 
111. Owens, J. W.; O?Connor, C. J.; Kassner, R. J., Inorg. Chim. Acta., 1988, 151, 107. 
112. Marques, H. M.; Baldwin, D. A.; Pratt, J. M., J. Inorg. Biochem., 1987, 29, 77. 
113. Baldwin, D. A.; Marques, H. M.; Pratt, J. M., J. Inorg. Biochem., 1986, 27, 245. 
114. Marques, H. M., PhD Thesis, University of the Witwatersrand (Johannesburg), 
1986.  
115. Harbury, H. A.; Cronin, J. R.; Fanger, M. W.; Hettinger, T. P.; Murphy, A. J.; 
Myer, Y. P.; Vinogradov, S. N., Proc. Natl. Acad. Sci., U. S. A., 1965, 54, 1658. 
116. Hamza, M. S. A.; Pratt, J. M., J. Chem. Soc. Dalton. Trans., 1993, 1647. 
117. Blumenthal, D. C.; Kassner, R. J., J. Biol. Chem., 1979, 254, 9617. 
118. Hamilton, G. H.; Owens, J. W.; O?Connor, C. J.; Kassner, R. J., Inorg. Chim. Acta., 
1984, 93, 55. 
119. Jain, A.; Kassner, R. J., J. Biol. Chem., 1984, 259, 10309. 
120. Perry, C. B., PhD Thesis, University of the Witwatersrand (Johannesburg), 2004.  
 148 
121. Huang, Y.-P.; Kassner, R. J., J. Biol. Chem., 1981, 256, 5327. 
122. Antonini, E.; Brunori, M., in Haemoglobin and Myoglobin in Their Reactions with 
Ligands, Elsevier, Amsterdam, 1971. 
123. Giacometti, G. M.; Ascenzi, P.; Bolognesi, M.; Brunori, M., J. Mol. Biol., 1981, 
146, 353. 
124. Sono, M.; Asakura, T., J. Biol. Chem., 1976, 251, 2664. 
125. Scheler, W., Biochem. Z., 1969, 332, 542. 
126. Makino, R.; Yamazaki, I., Arch. Biochem. Biophys., 1973, 157, 356. 
127. Bolognesi, M.; Ascenzi, P.; Giacometti, G. M.; Brunori, M.; Cannillo, E.; Merli, 
A., J. Mol. Biol., 1982, 158, 305. 
128. Uchida, H.; Heystec, J.; Klapper, M. H., J. Biol. Chem., 1971, 246, 2031. 
129. Eaton, D. R.; Wilkins, R. G., J. Biol. Chem., 1978, 253, 908. 
130. Hughes, M. N., in Comprehensive Coordination Chemistry, Vol 4 (Ed.: Wilkinson, 
G.), Pergamon Press, Oxford, 1987, p. 614. 
131. Anusiem, A. C.; Beetlestone, J. G.; Irvine, D. H., J. Chem. Soc. A, 1966, 357. 
132. Erecinska, M.; Vanderkooi, J. M., Methods Enzymol., 1978, 53, 165. 
133. Gray, H. B., Inorg. Chem., 1989, 28, 2074. 
134. Schneider, Z.; Stroinski, A., Comprehensive B12: Chemistry, biochemistry, 
nutrition, ecology and medicine, Walter de Gruyter and Company, 1987, pp. 7-10. 
135. Golding, B. T.; Chem. Brit., 1990, 26, 950. 
136. Minot, C. R.; Murphy, W. P., J. Am. Med. Ass., 1926, 87, 470. 
137. Castle, W. B., Am. J. Med. Sci., 1929, 178, 764. 
138. Rickes, E. L.; Brink, N. G., Koniuszy, F. R.; Wood, T. R.; Folkers, K., Science, 
1948, 107, 396. 
139. Smith, E. L.; Parker, L. F., Biochem. J., 1948, 43, 8. 
140. Brink-Shoemaker, C.; Cruickshank, D. W. J..; Hodgkin, D. C.; Kamper, J.; Pilling, 
D., Proc. R. Soc. London, Ser. A, 1964, 278, 1. 
141. DeRidder, D. J. A.; Zangrando, E.; B?ergi, H. ?B., J. Mol. Struct., 1996, 374, 63. 
142. Brown, K. L., Chem. Rev., 2005, 105, 2075. 
143. Marques, H. M.; Egan, T. J.; Marsh, J. H.; Mellor, J. R.; Munro, O. Q., Inorg. 
Chim. Acta, 1989, 166, 249. 
 149 
144. Marques, H. M.; Egan, T. J.; Marsh, J. H.; Mellor, J. R.; Munro, O. Q., Inorg. 
Chim. Acta, 1990, 170, 259. 
145. Marques, H. M., J. Chem. Soc., Dalton Trans., 1991, 339. 
146. Marques, H. M., J. Chem. Soc., Dalton Trans., 1991, 1437. 
147. http://chemweb.stanford.edu/winter2004/chem151/Handouts/lecture17.pdf. Last 
accessed on 21 Nov 2005 12:32:32 GMT by athena.wits.ac.za 
148. Marques, H. M.; Breet, E. L. J.; Prinsloo, F. F., J. Chem. Soc., Dalton Trans., 1991, 
2941. 
149. Marques, H. M., S. Afr. J. Chem., 1991, 44, 114. 
150. Marques, H. M.; Bradley, J. C.; Campbell L.A., J. Chem. Soc., Dalton Trans., 
1992, 2019. 
151. Marques, H. M.; Bradley, J. C.; Brown, K. L.; Brooks, H., Inorg. Chim. Acta, 1993, 
209, 161. 
152. Marques, H. M.; Bradley, J. C.; Brown, K. L.; Brooks, H., J. Chem. Soc., Dalton 
Trans., 1993, 3475. 
153. Marques, H. M.; Munro, O. Q.; Cumming, B. M.; de Nysschen, C., J. Chem. Soc., 
Dalton Trans., 1994, 297. 
154. Marques, H. M.; Knapton, L., J. Chem. Soc. Dalton Trans., 1997, 3827. 
155. Marques, H. M.; Knapton, L.; Zou, X.; Brown, K. L., J. Chem. Soc., Dalton Trans., 
2002, 3195. 
156. Hamza, M. S. A.; Zou, X.; Banka, R.; Brown, K. L.; van Eldik, R., Dalton Trans., 
2005, 782. 
157. Jensen, K. P.; Ryde, U., Chem. Bio. Chem., 2003, 4, 413. 
158. Wirt, M. D.; Sagi, I.; Chen, E.; Frisbie, S. M.; Lee, R.; Chance, M. R., J. Am. 
Chem. Soc., 1991, 113, 5299. 
159. Scheidt, W. R.; Reed, C. A., Chem. Rev., 1981, 81, 543. 
160. Suko, R. K.; Poppe, L.; R?tey, J.; Finke, R. G., Bioorg. Chem., 1999, 27, 451. 
161. Glusker, J. P., Vitam. Horm., 1995, 50, 1. 
162. Wirt, M. D.; Sagi, I.; Chance, M. R., J. Biophys., 1992, 63, 412. 
163. Pratt, J. M., Pure Appl. Chem., 1993, 65, 1513. 
164. Stolzenberg, A. M.; Stershic, M. T., J. Am. Chem. Soc., 1988, 110, 6391. 
 150 
165. Geno, M. K.; Halpern, J., J. Am. Chem. Soc., 1987, 109, 1238. 
166. Lippard, S. J.; Berg, J. M., Principles of Bioinorganic Chemistry, University 
Science Books, Mill Valley, California, 1994, p. 343. 
167. Rovira, C.; Kunc, K.; Hutter, J.; Parrinello, M., Inorg. Chem., 2001, 40, 11. 
168. Williams, R. J. P., J. Mol. Catal., 1985, 30, 1. 
169. Murakami, Y.; Aoyama, Y.; Tokunaga, K., J. Am. Chem. Soc., 1980, 102, 6736. 
170. Scheuring, E.; Padmakumar, R.; Banerjee, R.; Chance, M. R., J. Am. Chem. Soc., 
1997, 119, 12192. 
171. Dong, S.; Padmakumar, R.; Banerjee, R.; Spiro, T. G., J. Am. Chem. Soc., 1996, 
118, 9182. 
172. Dong, S.; Padmakumar, R.; Banerjee, R.; Spiro, T. G., J. Am. Chem. Soc., 1999, 
121, 7063. 
173. Sirovatka, J. M.; Rapp?, A. K.; Finke, R. G., Inorg. Chim. Acta., 2000, 545, 300. 
174. Jensen, K. P.; Ryde, U., J. Mol. Str., Theochem, 2002, 585, 239. 
175. Rohmer, M.-M., Chem. Phys. Lett., 1985, 116, 44. 
176. Edwards, W. D.; Weiner, B.; Zerner, M. C., J. Am. Chem. Soc., 1986, 116, 2196. 
177. Choe, Y. K-.; Hashimoto, T.; Nakano, H.; Hirao, K., Chem. Phys. Lett., 1998, 295, 
380. 
178. Choe, Y. K-.; Nakajima, T.; Hirao, K.; Lindh, R., J. Chem. Phys., 1999, 111, 3837. 
179. Spiro, T. G.; Kozlowski, P. M.; Zgierski, M. Z., J. Raman Spectrosc., 1998, 29, 
869. 
180. Kozlowski, P. M.; Spiro, T. G.; B?rces, A.; Zgierski, M. Z., J. Chem. Phys. B, 
1998, 102, 2603. 
181. Seno, Y.; Kameda, N.; Otsuka, J., J. Chem. Phys., 1980, 72, 6048. 
182. Obara, S.; Kashiwagi, H., J. Chem. Phys., 1982, 77, 3155. 
183. Yamamoto, S.; Teraoka, J.; Kashiwagi, H., J. Chem. Phys., 1988, 88, 303. 
184. Matsuzawa, N.; Ata, M.; Dixon, D. A.; J. Chem. Phys., 1995, 99, 7698. 
185. Rovira, C.; Kunc, K.; Hutter, J.; Bollone, P.; Parinello, M., J. Chem. Phys., 1997, 
101, 8914. 
186. Sigfridsson, E.; Ryde, U., J. Biol. Inorg. Chem., 1999, 4, 99. 
187. Sigfridsson, E.; Olsson, M. H. M.; Ryde, U., J. Chem. Phys. B, 2001, 105, 5546. 
 151 
188. Andruniow, T.; Zgierski, M. Z.; Kozlowski, P. M., J. Chem. Phys. B, 2000, 104, 
10921. 
189. Andrunio w, T.; Zgierski, M. Z.; Kozlowski, P. M., Chem. Phys. Lett., 2000, 331, 
509. 
190. Jensen, K. P.; Sauer, S. P. A.; Liljefors, T.; Norrby, P. -O., Organometallics, 2001, 
20, 550. 
191. D?lker, N.; Maseras, F.; Lledos, A., J. Chem. Phys. B, 2001, 105, 7564. 
192. Andruniow, T.; Zgierski, M. Z.; Kozlowski, P. M., J. Am. Chem. Soc., 2001, 123, 
2679. 
193. Ludwig, M. L.; Matthews, R. G., Annu. Rev. Biochem., 1997, 66, 269. 
194. Holm, R. H.; Kennepohl, P.; Solomon, E. I., Chem. Rev., 1996, 96, 2239. 
195. Eschenmoser, A.; Angew. Chem., 1988, 100, 5. 
196. Mano, N.; Goto, J., Anal. Sciences, 2003, 19, 3. 
197. Perry, C. B., CurveFit, a computer programme for non- linear least-squares fitting 
using the Marquard algorithm,  University of the Witwatersrand , Johannesburg, 
South Africa, 1998.  
198. Owens, J. W.; O?Connor, C. J., Coord. Chem. Rev., 1988, 84, 1. 
199. Loew, G. H., ., in Iron Porphyrins, Part 1; Lever, A. B. P., Gray, H. B., Eds.; 
Addison?Wesley Publishing Company, Inc.: Reading, M. A., 1983, pp. 1-87. 
200. Moore, G. R.; Pettigrew, G. W., Cytochromes c: Evolutionary, Structural and 
Physiochemical Aspects, Springer-Verlag, Berlin, 1990. 
201. Carraway, A. D.; Burkhalter, R. S.; Timkovich, R.; Peterson, J., J. Inorg. Biochem., 
1993, 52, 201. 
202. Low, D. W.; Abedin, S.; Yang, G.; Winkler; J. R.,Gray, H. B., Inorg. Chem. 1998, 
37, 1841. 
203. Lever, A. B. P., Inorganic Electronic Spectroscopy, Elserivier, Amsterdam, 1968, 
p. 249. 
204. White, W. I., in The Porphyrins, Dolphin, D., Ed.; Academic Press, New York, 
1978, Volume 5, Chapter 7. 
205. Pasternack, R. F.; Huber, P. R.; Boyd, P.; Engasser, G.; Francesconi, L.; Gibbs, E.; 
Fasella, P.; Venturo, G. C.; de Hinds, L. C., J. Am Chem. Soc., 1972, 94, 4511. 
 152 
206. Handbook of Chemistry and Physics, 55th Edition, Weast, R. C., Ed., Chemical 
Rubber Co.: Cleveland, 1974, pp. D126-9. 
207. Cohen, J. S.; Hayes, M. B.; J. Biol. Chem., 1974, 249, 5472. 
208. Cohen, J. S.; Fisher, W. R.; Schechter, A. N., J. Biol. Chem., 1974, 249, 1113. 
209. Moore, G. R.; Williams, R. J. P., Eur. J. Biochem., 1980, 103, 513.  
210. Moore, G. R.; Williams, R. J. P., Eur. J. Biochem., 1980, 103, 523.  
211. Moore, G. R.; Pettigrew, G. W.; Pitt, R. C.; Williams, R. J. P., Biochim. Biophys. 
Acta, 1980, 590, 261. 
212. Boeri, E.; Ehrenberg, A.; Paul, K. G.; Theorell, H., Biochim. Biophys. Acta, 1953, 
12, 273. 
213. Babul, J.; Stellwagen, E., Biochemistry, 1972, 11, 1195. 
214. Smith, R. M. ; Martell, A. E., Critical Stability Constants, Plenum Press, New York, 
1986, Volume 6, Second Supplement, p. 302. 
215. Lommen, A.; Ratsma, A.; Bijlsma, N.; Canters, G. W.; Van Wielink, J. E.; Frank, 
J.; Van Beeumen, J., Eur. J. Biochem., 1990, 192, 653. 
216. Moore, G. R.; Williams, R. J. P.; Peterson, J.; Thomson, A. J.; Mathews, F. S.; 
Biochim. Biophys. Acta, 1985, 829, 83. 
217. La Mar, G. N.; Jackson, J. T.; Dugad, L. B.; Cusanovich, M. A.; Bartsch, R. G., J. 
Biol. Chem., 1990, 265, 16173. 
218. Costa, H. S.; Santos, H.; Tur ner, D. L.; Xavier, A. V., Eur. J. Biochem., 1992, 208, 
457. 
219. Keller, R.; Groundinsky, O.; W?thrich, K., Biochim. Biophys. Acta, 1976, 427, 497. 
220. Goldsack, D. E.; Eberlein, W. S.; Alberty, R. A., J. Biol. Chem., 1966, 241, 2653. 
221. Marques, H. M., Inorg. Chem., 1990, 29, 1597. 
222. George, P.; Hanania, G. I. H., Biochem. J., 1953, 55, 236. 
223. George, P.; Hanania, G. I. H., Biochem. J., 1952, 52, 517. 
224. Yagil, G., Tetrahedron, 1967, 23, 2855. 
225. Martin, R. B., Proc. Natl. Acad. Sci., USA, 1974, 71, 4346. 
226. Rubinson, K. A.; Parekh, H. V.; Itabashi, E.; Mark, H. B., Inorganic Chemistry, 
1983, 22, 458. 
227. Prinsloo, F. F.; Meier, M.; van Eldik, R., Inorg. Chem., 1994, 33, 900. 
 153 
228. Marques, H. M.; Munro, O. Q.; Munro, T.; de Wet, M.; Vashi, P. R., Inorg. Chem., 
1999, 38, 2312. 
229. Knapton, L., PhD Thesis, University of the Witwatersrand (Johannesburg), 2005.  
230. Byfield, M. P.; Hamza, M. S. A.; Pratt, J. M., J. Chem. Soc., Dalton Trans., 1993, 
1647. 
 
 
 154 
Appendices 
Appendix A: HPLC Data 
 
Table A1: HPLC analysis of MP8 in pH 7.00 buffer at ? = 397nm 
Peak No. Area % RT Area BC 
1 95 6.45 217600 01 
2 5 6.15 11453 0.1 
Total 100  229053  
 
Table A2: HPLC analysis of MP8 free base in pH 7.00 buffer at ?  = 403nm 
Peak No. Area % RT Area BC 
1 1.689 4.63 1389 0.1 
2 3.525 5.38 2899 02 
3 9.235 5.64 7595 02 
4 80.592 5.99 66279 02 
5 4.959 6.49 4078 03 
Total 100  82240  
 
Table A3: HPLC analysis of Co(III)MP8 in pH 7.00 buffer at ?  = 403nm 
Peak No. Area % RT Area BC 
1 79.268 6.82 52448 02 
2 9.484 7.43 6275 02 
3 11.248 8.42 7442 03 
Total 100  66165  
 
 
 155 
SCAN GRAPH. Flagging=Low Resolution M/z. 
Scan 5#0:40. Entries=1142. Base M/z=154. 100% Int.=1.52768. FAB. POS. 
Low Resolution M/z
 0 200 400 600 800 1000 1200 1400 1600 1800 2000
 In
 te
 ns
 ity
  (%
 ag
 e)
 0
 10
 20
 30
 40
 50
 60
 70
 80
 90
 100
 136
 154
 307.1
 595.1
 617.1
 1505.4
  (*20) 
Table A4: HPLC analysis of N-Ac-Co(III)MP8 in pH 7.00 buffer at ?  = 397nm 
Peak No. Area % RT Area BC 
1 1.844 5.21 1678 02 
2 4.249 5.47 3867 02 
3 66.103 5.99 60164 03 
4 26.618 6.59 24226 01 
5 1.187 7.05 1080 01 
Total 100  91015  
 
 
Appendix B:  FAB-MS Data 
FAB-MS for MP8 
 
 
Figure B1:  FAB-MS spectra of MP8 sample and matrice. 
 
 156 
 
 
Figure B2: FAB-MS spectra of sample with matrice subtracted (bottom spectra). 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Comparison between "C:\MASPEC\Data\DEN N Ac Co(III) MP8 PREP9(2) LRFAB.ms2" and "C:\MASPEC\Data\NBA 050204 LRFAB.ms2" 
Scan 14#1:50 - 15#1:58. Entries=647. Base M/z=154.1. 100% Int.=1.58368. FAB. POS.  
Scan 8#1:03 - 10#1:19. Entries=530. Base M/z=154.1. 100% Int.=1.408. FAB. POS.  
Difference Scan  
Low Resolution M/z 
0 200 400 600 800 1000 1200 1400 1600 1800 
0 200 400 600 800 1000 1200 1400 1600 1800 
0 200 400 600 800 1000 1200 1400 1600 1800 2000 
Intensity (%age) 
0.0 
0.5 
1.0 
1.5 
2.0 
2.5 
3.0 
3.5 
0.0 
0.5 
1.0 
1.5 
2.0 
2.5 
3.0 
3.5 
0.0 
0.5 
1.0 
1.5 
2.0 
2.5 
3.0 
3.5 
39 107 154.1 
242.1 
307.1 
341.1 
442.1 
465.2 595.1 620.1 
39 107.1 154.1 
249.1 
307.1 
341.1 
424.1 
465.2 595.2 613.2 
55 
69 
149 
205 
279 
345 
425 548 
620 
 157 
SCAN GRAPH. Flagging=Low Resolution M/z. 
Scan 5#0:40. Entries=1142. Base M/z=154. 100% Int.=1.52768. FAB. POS. 
Low Resolution M/z
 1400 1420 1440 1460 1480 1500 1520 1540 1560 1580 1600
 In
 te
 ns
 ity
  (%
 ag
 e)
 0.0
 1.0
 2.0
 3.0
 4.0
 5.0
 6.0
 7.0
 8.0
 9.0
 10.0
 11.0
 1390.3 1439.9
 1462.7
 1488.3
 1505.4
 1508.6
 1528.4
 1547 1565.7 1585.1 1605.8
  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure B3:  Isotopic distribution of sample. 
 
 
Mass
 1502.0 1503.0 1504.0 1505.0 1506.0 1507.0 1508.0 1509.0 1510.0 1511.0 1512.0
 In
 te
 ns
 ity
  (%
 ag
 e)
 0
 10
 20
 30
 40
 50
 60
 70
 80
 90
 100
 1503.52419
 C68.H87.54Fe.N15.O17.S2
 1505.51952
 C68.H87.Fe.N15.O17.S2
 1509.52662
 C65.13C3.H87.Fe.N14.15N.O17.S2
  
 
Figure B4: Theoretical isotopic distribution of sample. 
 
 158 
FAB-MS for MP8 Free Base 
 
 
 159 
 
 160 
Comparison between "C:\MASPEC\Data\DEN COIII MP8 (10-1) LRFAB.ms2" and "C:\MASPEC\Data\NBA 050204 LRFAB.ms2"
 Scan 14#1:50. Entries=1060. Base M/z=154. 100% Int.=1.4912. FAB. POS. 
Scan 8#1:03 - 10#1:19. Entries=530. Base M/z=154.1. 100% Int.=1.408. FAB. POS. 
Difference Scan 
Low Resolution M/z
 0 200 400 600 800 1000 1200 1400 1600 1800
 0 200 400 600 800 1000 1200 1400 1600 1800
 0 200 400 600 800 1000 1200 1400 1600 1800 2000
 In
 te
 ns
 ity
  (%
 ag
 e)
 0.0
 0.5
 1.0
 1.5
 2.0
 2.5
 0.0
 0.5
 1.0
 1.5
 2.0
 2.5
 0.0
 0.5
 1.0
 1.5
 2.0
 2.5
 39 107 154
 249.1
 307.1
 341.1
 424.1
 465.1
 595.2620.1
 684.3 731.1
 39 107.1 154.1
 249.1
 307.1
 341.1
 424.1
 465.2595.2613.2 730.2
 57
 69
 149
 205
 279
 345
 425 548
 620
 684
 FAB-MS for Co(III)MP8  
 
 
 
Figure B8: Fab spectra of sample and matrice. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 161 
SCAN GRAPH. Flagging=Low Resolution M/z. 
Scan 14#1:50. Entries=1060. Base M/z=154. 100% Int.=1.4912. FAB. POS. 
Low Resolution M/z
 1400 1420 1440 1460 1480 1500 1520 1540 1560 1580 1600 1620
 In
 te
 ns
 ity
  (%
 ag
 e)
 0.0
 0.5
 1.0
 1.5
 2.0
 2.5
 3.0
 3.5
 4.0
 4.5
 5.0
 5.5
 1395.5
 1411.4
 1431.2
 1447.1
 1463.9
 1493.3
 1509.4
 1522.2
 1534
 1550.3
 1572.5
 1590.7
 1609.7
 SCAN GRAPH. Flagging=Low Resolution M/z. 
Scan 14#1:50. Entries=1060. Base M/z=154. 100% Int.=1.4912. FAB. POS. 
Low Resolution M/z
 0 200 400 600 800 1000 1200 1400 1600
 In
 te
 ns
 ity
  (%
 ag
 e)
 0
 10
 20
 30
 40
 50
 60
 70
 80
 90
 100
 107
 154
 307.1
 500.1
 620.1
 1508.3
  (*25) 
 
Figure B9: Fab spectra of sample with matrice subtracted(bottom spectra). 
 
 
Figure B10:  Isotopic distribution of sample. 
 
 162 
Mass
 1507.0 1508.0 1509.0 1510.0 1511.0 1512.0 1513.0 1514.0 1515.0
 In
 te
 ns
 ity
  (%
 ag
 e)
 0
 10
 20
 30
 40
 50
 60
 70
 80
 90
 100
 1508.51778
 C68.H87.Co.N15.O17.S2
 1510.52202
 C68.H87.Co.N15.O16.18O.S2
 1512.52488
 C65.13C3.H87.Co.N14.15N.O17.S2
  
Figure B11: Theoretical isotopic distribution of sample. 
 163 
FAB-MS for NAc-CoIII-MP8    
 
SCAN GRAPH. Flagging=Low Resolution M/z. 
Scan 12#1:35. Entries=1173. Base M/z=154.1. 100% Int.=1.52064. FAB. POS. 
Low Resolution M/z
 0 200 400 600 800 1000 1200 1400 1600 1800 2000
 In
 te
 ns
 ity
  (%
 ag
 e)
 0
 10
 20
 30
 40
 50
 60
 70
 80
 90
 100
 136.1 154.1
 307.1
 577.1
 620.1
 1509.5 1552.3
  (*25) 
 
Figure B12:  Fab spectra of sample and matrice. 
 
Comparison between "C:\MASPEC\Data\DEN N Ac Co(III) MP8 PREP2 LRFAB.ms2" and "C:\MASPEC\Data\NBA 050204 LRFAB.ms2"
 Scan 12#1:35. Entries=1173. Base M/z=154.1. 100% Int.=1.52064. FAB. POS. 
Scan 8#1:03 - 10#1:19. Entries=530. Base M/z=154.1. 100% Int.=1.408. FAB. POS. 
Difference Scan 
Low Resolution M/z
 0 200 400 600 800 1000 1200 1400 1600 1800
 0 200 400 600 800 1000 1200 1400 1600 1800
 0 200 400 600 800 1000 1200 1400 1600 1800 2000
 In
 te
 ns
 ity
  (%
 ag
 e)
 0.0
 1.0
 2.0
 3.0
 4.0
 5.0
 6.0
 7.0
 8.0
 0.0
 1.0
 2.0
 3.0
 4.0
 5.0
 6.0
 7.0
 8.0
 0.0
 1.0
 2.0
 3.0
 4.0
 5.0
 6.0
 7.0
 8.0
 39 107 154.1
 205.1
 289.1
 341.1
 424.1
 577.1620.1
 39 107.1 154.1
 249.1
 307.1
 341.1
 424.1
 595.2
 57
 107
 149
 205
 279
 424 620
  
Figure B13: Fab spectra of sample with matrice sub tracted(bottom spectra). 
 164 
SCAN GRAPH. Flagging=Low Resolution M/z. 
Scan 12#1:35. Entries=1173. Base M/z=154.1. 100% Int.=1.52064. FAB. POS. 
Low Resolution M/z
 1400 1420 1440 1460 1480 1500 1520 1540 1560 1580 1600 1620
 In
 te
 ns
 ity
  (%
 ag
 e)
 0.0
 0.5
 1.0
 1.5
 2.0
 2.5
 3.0
 3.5
 4.0
 4.5
 5.0
 5.5
 1403.4
 1417.4 1436.5
 1455.8 1473.7
 1492.6
 1508.3
 1527.7
 1552.3
 1567.3
 1581
 1608.2
 1620.2
  
Figure B14: Isotopic distribution of sample. 
 
Mass
 1534.0 1535.0 1536.0 1537.0 1538.0 1539.0 1540.0 1541.0 1542.0
 In
 te
 ns
 ity
  (%
 ag
 e)
 0
 10
 20
 30
 40
 50
 60
 70
 80
 90
 100
 1535.54125
 C70.H90.Co.N15.O17.S2
 1537.5455
 C70.H90.Co.N15.O16.18O.S2
 1539.54835
 C67.13C3.H90.Co.N14.15N.O17.S2
  
Figure B15:  Theoretical isotopic distribution of sample. 
 
 165 
Appendix C:  Beer?s Law Data 
Table C1: Absorbance measurements for varying concentrations of NAcCoIII-MP8  in            
                 distilled water. 
Absorbance at 420 nm Concentration / M 
0 0 
0.0829 5.49E-07 
0.1724 1.09E-06 
0.2534 1.63E-06 
0.3404 2.15E-06 
0.4226 2.68E-06 
0.5059 3.19E-06 
0.589 3.7E-06 
0.6703 4.21E-06 
0.8379 5.2E-06 
1.0036 6.16E-06 
1.1656 7.11E-06 
1.326 8.03E-06 
1.5058 8.93E-06 
1.6513 9.81E-06 
1.7957 1.07E-05 
1.9357 1.15E-05 
1.733621 1.15E-05 
2.057759 1.34E-05 
2.37931 1.52E-05 
2.707759 1.69E-05 
3.041379 1.85E-05 
3.389655 2.01E-05 
3.681897 2.15E-05 
4.127586 2.43E-05 
4.739655 2.69E-05 
5.156897 2.92E-05 
5.500862 3.14E-05 
5.983621 3.35E-05 
 166 
Table C2: Absorbance measurements for varying concentrations of NAcCoIII-MP8  in            
                 1.0 M NaClO4. 
Absorbance at 420 nm Concentration / M 
0 0 
0.0828 5.49E-07 
0.1684 1.09E-06 
0.2468 1.63E-06 
0.3336 2.15E-06 
0.4152 2.68E-06 
0.4889 3.19E-06 
0.5671 3.7E-06 
0.6413 4.21E-06 
0.81 5.2E-06 
0.9681 6.16E-06 
1.123 7.11E-06 
1.2739 8.03E-06 
1.4251 8.93E-06 
1.5367 9.81E-06 
1.7113 1.07E-05 
1.8456 1.15E-05 
1.606034 1.15E-05 
1.928448 1.34E-05 
2.226724 1.52E-05 
2.567241 1.69E-05 
2.891379 1.85E-05 
3.114655 2.01E-05 
3.469828 2.15E-05 
4.086207 2.43E-05 
4.612931 2.69E-05 
4.768966 2.92E-05 
5.026724 3.14E-05 
5.499138 3.35E-05 
 
 
 167 
Appendix D: Spectroscopic pKa Data 
Table D1: Spectroscopic pKa Data in Acidic Region 
pH Absorbance  pH Absorbance 
2.05 0.698 4.856 0.538 
2.105 0.6998 4.982 0.5325 
2.187 0.7024 5.111 0.5283 
2.22 0.7035 5.243 0.5275 
2.314 0.7045 5.332 0.5245 
2.404 0.7031 5.466 0.5265 
2.51 0.6981 5.614 0.528 
2.576 0.6979 5.751 0.5298 
2.652 0.6949 5.846 0.5355 
2.706 0.6933   
2.805 0.6899   
2.9 0.6824   
3 0.6749   
3.14 0.665   
3.242 0.658   
3.364 0.6472   
3.45 0.6411   
3.532 0.635   
3.587 0.627   
3.644 0.615   
3.755 0.6101   
3.827 0.6046   
3.931 0.5965   
4.024 0.5865   
4.218 0.578   
4.411 0.5672   
4.529 0.5501   
4.646 0.548   
4.762 0.5448   
 
 168 
 
Table D2: Spectroscopic pKa Data in Basic Region 
pH Absorbance  pH Absorbance 
5.49 0.937 8.434 1.0639 
5.573 0.9377 8.504 1.0667 
5.655 0.9384 8.574 1.0695 
5.725 0.9408 8.645 1.0724 
5.795 0.9431 8.742 1.0771 
5.933 0.9444 8.838 1.0825 
6.07 0.9456 8.935 1.0936 
6.142 0.9445 9.03 1.1027 
6.214 0.9433 9.142 1.1128 
6.3 0.9422 9.254 1.1198 
6.395 0.9451 9.326 1.1239 
6.504 0.948 9.398 1.1279 
6.591 0.9504   
6.678 0.9528   
6.766 0.9561   
6.901 0.9641   
7.036 0.973   
7.139 0.9775   
7.322 0.987   
7.402 0.992   
7.481 0.9971   
7.564 1.005   
7.647 1.0129   
7.739 1.0207   
7.914 1.0339   
8.003 1.0441   
8.117 1.0499   
8.231 1.0557   
8.333 1.0598   
 
 
 169 
 
Table D2: Spectroscopic pKa Data in Basic Region with Pyridine 
pH Absorbance  pH Absorbance 
5.958 0.972 9.945 1.1989 
6.101 0.9993 10.135 1.2021 
6.2 0.9929 10.325 1.2052 
6.304 0.9892 10.504 1.2056 
6.41 0.9853 10.683 1.206 
6.515 0.9874 10.855 1.199 
6.61 0.9913 11.027 1.1919 
6.717 0.9957 11.229 1.1805 
6.826 1.0036 11.431 1.169 
6.9 1.0099 11.587 1.1516 
7.01 1.0162 11.743 1.1342 
7.176 1.0226 11.875 1.1219 
7.306 1.0336   
7.435 1.0473   
7.575 1.0548   
7.715 1.0623   
7.856 1.0715   
7.929 1.0827   
8.002 1.0926   
8.289 1.114   
8.576 1.1348   
8.722 1.1436   
8.868 1.1524   
9.014 1.1611   
9.123 1.1714   
9.239 1.1745   
9.356 1.1781   
9.548 1.185   
9.741 1.1919   
 
 
 170 
 
Appendix E: Derivation of Rate Constant Equation 
 
 
A B C
 k1
 k-1
 kex
 dC
 dD
 = kex[B]
 Steady state on [B] :-
 k1[A] = k-1[B] + kex[B]
 = [B](k-1 + kex)
 Therefore, [B] =   k[A]
 k-1 + kex
 It follows that, dC
 dD
 =  kexk1[A]
 k-1 + kex
 Now usually kex << k-1
 Therefore, dC
 dD
 =  kexk1[A]
 k-1
 = kexK1[A]
 Therefore, kobs = kexK1 (or KOS1kex1) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 171