USF Libraries
USF Digital Collections

Time-resolved thermodynamics studies of heme signaling proteins and model systems

MISSING IMAGE

Material Information

Title:
Time-resolved thermodynamics studies of heme signaling proteins and model systems
Physical Description:
Book
Language:
English
Creator:
Mokdad, Audrey
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Thermodynamic profiles
Signaling proteins
Photoacoustic calorimetry
Heme domain
Porphyrin
Spin crossover
Debye-Hückel equation
Dissertations, Academic -- Chemistry -- Doctoral -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Heme-based gas sensor proteins have the ability to sense diatomic molecules such as O₂ (FixL, EcDos or HemAT), CO (CooA, a CO-sensing protein of Rhodospirillum rubrum) and NO (guanylate cyclase) molecules and subsequently regulate numerous important biological processes in prokaryotic and eukaryotic organisms. The sensing function of these proteins is initiated by the binding of an effector (i.e., O₂, CO, etc⁵) to the heme iron which then leads to a cascade of conformational events which gives rise to changes in kinase activity, DNA-binding activity, etc...In order to better understand the mechanism heme-based signaling, time resolved photothermal methods as well as transient optical techniques were utilized to obtain thermodynamic profiles for ligand binding/release in heme based signaling proteins including HemAT from Bacillus subtilis (aerotactic transducer), FixL from Sinorhizobium meliloti (regulation of the nitrogen fixation) and CooA from Rhodospirillum rubrum (transcriptional activator). In addition, a number of model systems were examined to understand the underlying thermodynamic processes involved in heme ligation. The variation of volume and enthalpy changes associated with spin state change of the iron from high-spin to low-spin where examined using the spin crossover Fe(III)(salten)(mepepy) complex. In addition, the experimental determination of the volume change due to electrostriction events were using Ru(II)(L)₃ and the Debye-Hückel equation.Finally, different model heme proteins were studied to understand how a signal is conformationaly transmitted within a heme protein matrix. Sandbar shark hemoglobin was examined as an example of a non-signaling an allosteric protein. Two different peroxidases (horseradish and soybean) which have a direct channel between the heme pocket and the solvent involving no barrier energetic for the photodissociated ligand leaving the heme pocket were examined as example of non-signaling, non-allosteric proteins. The results show that each protein has a unique thermodynamic profile to conformationaly transmit signals subsequent to photodissociation of CO, even within the same class of protein (i.e. PAS domains, globins, etc...).
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2009.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Audrey Mokdad.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 235 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002069242
oclc - 608044377
usfldc doi - E14-SFE0003063
usfldc handle - e14.3063
System ID:
SFS0027379:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200421Ka 4500
controlfield tag 001 002069242
005 20100420110325.0
007 cr mnu|||uuuuu
008 100420s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003063
035
(OCoLC)608044377
040
FHM
c FHM
049
FHMM
090
QD31.2 (Online)
1 100
Mokdad, Audrey.
0 245
Time-resolved thermodynamics studies of heme signaling proteins and model systems
h [electronic resource] /
by Audrey Mokdad.
260
[Tampa, Fla] :
b University of South Florida,
2009.
500
Title from PDF of title page.
Document formatted into pages; contains 235 pages.
Includes vita.
502
Dissertation (Ph.D.)--University of South Florida, 2009.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
520
ABSTRACT: Heme-based gas sensor proteins have the ability to sense diatomic molecules such as O (FixL, EcDos or HemAT), CO (CooA, a CO-sensing protein of Rhodospirillum rubrum) and NO (guanylate cyclase) molecules and subsequently regulate numerous important biological processes in prokaryotic and eukaryotic organisms. The sensing function of these proteins is initiated by the binding of an effector (i.e., O, CO, etc) to the heme iron which then leads to a cascade of conformational events which gives rise to changes in kinase activity, DNA-binding activity, etc...In order to better understand the mechanism heme-based signaling, time resolved photothermal methods as well as transient optical techniques were utilized to obtain thermodynamic profiles for ligand binding/release in heme based signaling proteins including HemAT from Bacillus subtilis (aerotactic transducer), FixL from Sinorhizobium meliloti (regulation of the nitrogen fixation) and CooA from Rhodospirillum rubrum (transcriptional activator). In addition, a number of model systems were examined to understand the underlying thermodynamic processes involved in heme ligation. The variation of volume and enthalpy changes associated with spin state change of the iron from high-spin to low-spin where examined using the spin crossover Fe(III)(salten)(mepepy) complex. In addition, the experimental determination of the volume change due to electrostriction events were using Ru(II)(L) and the Debye-Hckel equation.Finally, different model heme proteins were studied to understand how a signal is conformationaly transmitted within a heme protein matrix. Sandbar shark hemoglobin was examined as an example of a non-signaling an allosteric protein. Two different peroxidases (horseradish and soybean) which have a direct channel between the heme pocket and the solvent involving no barrier energetic for the photodissociated ligand leaving the heme pocket were examined as example of non-signaling, non-allosteric proteins. The results show that each protein has a unique thermodynamic profile to conformationaly transmit signals subsequent to photodissociation of CO, even within the same class of protein (i.e. PAS domains, globins, etc...).
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
590
Advisor: Randy W. Larsen, Ph.D.
653
Thermodynamic profiles
Signaling proteins
Photoacoustic calorimetry
Heme domain
Porphyrin
Spin crossover
Debye-Hckel equation
690
Dissertations, Academic
z USF
x Chemistry
Doctoral.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.3063



PAGE 1

Time-Resolved Thermodynamic Studies of Heme Signaling Proteins and Model Systems by Audrey Mokdad A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Chemistry College of Arts of Sciences University of South Florida Major Professor: Ra ndy W. Larsen, Ph.D. Brian Space, Ph.D. David J. Merkler, Ph.D. Gloria C. Ferreira, Ph.D. Martin Muschol, Ph.D. Date of Approval: July 13th 2009 Keywords: thermodynamic profiles, signaling proteins, photoacoustic calorimetry, heme domain, porphyrin, spin cross over, Debye-Hckel equation Copyright 2009, Audrey Mokdad

PAGE 2

Dedication For my parents, brother and sister

PAGE 3

Acknowledgements First and foremost, I would like to express my sincere appreciation to my advisor Dr. Randy Larsen for challenging me in this very interesting area of biophysical chemistry, and for allowing me to complete my doctorate studies in his research laboratory with his wisdom and guidance through these past few y ears. I would like to thank my committee, Dr. Brian Space, Dr. David Merkler, and Dr Gloria Ferreira, for their constructive criticism and insight, and Dr. Martin Muschol for se rving as chairperson. Finally, I would like to thank Carissa Vetromile, Meagan Small and Billy Maza, my fellow labmates for their helpful comments and encouragements. Thank you again for your support! I also would like to thank every person (organic, inorganic, computation’s members) who helps me by their knowledge, or just by listening to me during these past years and especially Farid Nouar. Finally, I would like to dedica te this dissertation to my parents for their support, for always encouraging me to reach high goals to surpass myself in my studies and to seek to be the best that I could be. Without their love encouragement, patience and listening ears, I would never have survived these years. I also dedicate this dissertation to my brother and my sister whom I have negl ected over the past few years in order to achieve this goal. Last but not least, my wo rds can not effectively express my sincere appreciation to Jason for his love, support, patience and understanding during this difficult time that is the doctorate years.

PAGE 4

i Table of Contents List of Tables v List of Figures vii List of Schemes xvii Abstract xviii Chapter I – Introduction 1 1.1 Introduction 1 1.2 References 6 Chapter II – Methods 9 2.1 Photothermal Spectroscopy 9 2.1.1 General background 11 2.2 Transient Absorption Spectroscopy 21 2.3 References 25 Chapter III – Models Systems 27 3.1 References 31 3.2 Photophysical Studies of the Trans to Cis Isomerization of the Push Pull Molecule: 1-(Pyridin-4-yl)-2-( N methylpyrrol-2-yl)ethene (mepepy) 32 3.2.1. Introduction 32 3.2.2. Materials and Methods 35 3.2.2.1. Synthesis of mepepy 35 3.2.2.2.Computational Methods 36 3.2.3. Results and Discussion 37

PAGE 5

ii 3.2.3.1. Optical properties 37 3.2.3.2. Theoretical Structural Analysis 41 3.2.3.3. Photothermal Studies 43 3.2.4. Conclusion 47 3.2.5. References 48 3.3 Photothermal Studies of the Room Temperature Photoinduced Spin State Change in the Fe(III) Salten Mepepy Complex 51 3.3.1. Introduction 51 3.3.2. Materials and Methods 54 3.3.3. Results 57 3.3.4. Discussion 60 3.3.5. Conclusion 67 3.3.6. References 69 3.4 Time-Resolved Thermodynamic Calorimetry and DebyeHckel equation: Determining Electrostriction Associated with Excited State Ru(II)(L)3 Complexes 72 3.4.1. Introduction 72 3.4.2. Materials and Methods 75 3.4.3. Results 75 3.4.4. Discussion 77 3.4.5. Conclusion 88 3.4.6. References 89 Chapter IV – Signaling Proteins 91 4.1 Introduction 91 4.2 References 98

PAGE 6

iii 4.3 Evidence for fast conformational change upon ligand dissociation in the HemAT class of bacterial oxygen sensors 100 4.3.1. Introduction 100 4.3.2. Materials and Methods 101 4.3.2.1. Protein expression, isolation and purification 101 4.3.2.2. Sample preparation 102 4.3.3. Results 103 4.3.4. Discussion 111 4.3.5. Conclusion 117 4.3.6. References 118 4.4 Thermodynamic of conformational changes coupled to CO photodissociation from the CO -sensing transcriptional activator CooA 120 4.4.1. Introduction 120 4.4.2. Materials and Methods 124 4.4.3. Results and Discussion 125 4.4.4 Conclusion 134 4.4.5 References 135 4.5 Photothermal studies of Carbon Monoxide Ligand Photodissociated from FixL Sinorhizobium meliloti 138 4.5.1. Introduction 138 4.5.2. Materials and Methods 143 4.5.3. Results 143 4.5.4. Discussion 154 4.5.5. Conclusion 167 4.5.6. References 168

PAGE 7

iv Chapter V – Model Proteins 170 5.1 Kinetic Properties of Polymorphic Hemoglobin from the Sandbar Shark Hemoglobin ( Carcharhinus plumbeus ) 172 5.1.1. Introduction 172 5.1.2. Materials and Methods 174 5.1.3. Results 175 5.1.4. Discussion 187 5.1.5. Conclusion 194 5.1.6. References 195 5.2 Photothermal Studies of CO photodissociation from Peroxidases from Horseradish and Soybean 197 5.2.1. Introduction 197 5.2.2. Materials and Methods 201 5.2.3. Results 201 5.2.4. Discussion 207 5.2.5. Conclusion 222 5.2.6. References 223 Chapter VI Conclusion 227 6.1 Summary and Conclusion 227 6.2 Futures Directions 234 About the Author End page

PAGE 8

v List of Tables Table 3.1 Configurations of the Global Energy-Minimized Geometric cis and trans -mepepy. 42 Table 3.2 Calculated Energy and Di pole Moment of Global Minima cis and trans -mepepy Isomers associated with the Alternate Local Minimum along the trans Potential Energy Surface. 42 Table 3.3 Variations of the volume and enthalpy for Ru(bpy)3 for in different concentrations of NaCl(aq) (between 10mM to 2M) at pH=7. 80 Table 3.4 Summary of Vi and zi for Ru(bpy)3 after fits to the DebyeHckel equation. 83 Table 3.5 Variations of the volume and enthalpy for Ru(phen)3 for different concentrations of NaCl(aq) (between 10mM to 2M) at pH=7. 86 Table 3.6 Summary of Vi and zi for Ru(phen)3 from fits to the DebyeHckel equation. 88 Table 4.1 Summary of the Soret and Q bands wavelengths for wild type, R200A, R200Q, R200E, R200H, I209M and heme domain of Sm FixL. 146 Table 4.2 Summary of the rate constant associated with CO rebinding to wild type, R200A, R200Q, R200E, R200H, I209M mutants and heme domain of Sm FixLH. 148 Table 4.3 Summary of the CO rebinding rate constant for Bj FixLH, truncated Bj FixLH and Sm FixLH. 148 Table 4.4 Summary of photoacoustic re sults for the wild type, R200A, R200Q, R200E, R200H, I209M and heme domain of Sm FixL. 157

PAGE 9

vi Table 4.5 Summary for photoacoustic results for Bj FixLH, truncated Bj FixLH and Sm FixLH. 160 Table 5.1 Variation of volume and en thalpy of the CO photodissociation at pH= 6 and 7. 176 Table 5.2 Variation of activat ion volume and enthalpy of CO photodissociation associated with their rate constant. 181 Table 5.3 Different time relaxations of human hemoglobin. 188 Table 5.4 Table 5.5 Summary of PAC results for CO photo-release from ferrous HRP. Summary of PAC results for CO photo-release from ferrous SBP. 208 209 Table 5.6 Table 5.7 Summary of Hstr and Vstr associated with CO photo-release from ferrous HRP. Summary of Hstr and Vstr associated with CO photo-release from ferrous SBP. 213 214 Table 6.1 Summarized of the variation of volume and enthalpy for different heme-based sensor proteins. 233

PAGE 10

vii List of Figures Figure 1.1 Diagram of the oxygen sensor domain. 3 Figure 1.2 Thermodynamic profiles (var iation of volume and enthalpy) of the signal transmitted along th e sensor, linker and kinase domain. 4 Figure 2.1 Overlap between the samp le and the reference photoacoustic signals showing no frequency shift. 18 Figure 2.2 Overlap between the samp le and the reference photoacoustic signals showing a frequency shift. 19 Figure 2.3 Transient absorption spec tra after excitati on at 440 nm. 23 Figure 3.1 Structures of the trans and cis conformations of mepepy. 34 Figure 3.2 Optical absorption spectra of the steady-state of mepepy in water (containing <1% DMSO) (solid line) and aqueous 0.01 N NaOH (containing <1% DMSO; dotted line). [mepepy] = 270 M. 38 Figure 3.3 Diagrams of the molecu lar orbital viewing the HOMOs and LUMOs related with the observed to transitions of the trans (left) and cis (right) isomers of mepepy. 39 Figure 3.4 Top panel : UV-Vis spectra of the mepepy equilibrium before and after steady-state photolysis (from 10 to 70 mins). The sample was solubilized in de ionized water containing <1% DMSO. Bottom panel : UV-Vis spectra of the trans -mepepy equilibrium before and after 5, 10, 15, and 20 laser pulses ( excit = 355 nm, 100 J/pulse, and 7 ns fwhm). 40

PAGE 11

viii Figure 3.5 Ground-state energy surfaces of the trans / cis mepepy as a function of the pyridine and pyrro le ring angles as defined by deviation from planar configura tion. See text for details. 43 Figure 3.6 Structure of the global and local trans geometries (left and right, respectively). 43 Figure 3.7 Overlay of the normalized acoustic wave of the reference Fe4SP and the trans -mepepy in water containing <1% DMSO at 35C. Both absorbencies at the excitation wavelength (355 nm) were 0.4. 44 Figure 3.8 Thermodynamic profile for the photoinduced is omerization of the trans to cis mepepy in aqueous solution. 46 Figure 3.9 Structure of the Fe(III)( salten)(mepepy) complex before and after photolysis. Distances: a= 2.035 , b=1.977 , c=1.965 , d=1.888 , e=1.892 and f=2.022 . 52 Figure 3.10 Structure of the Trans and Cis Isomers of Mepepy. 53 Figure 3.11 Steady-state optical absorption spectra of the Fe(III)(salten)(mepepy) complex (orange dotted line) and the mepepy ligand (bleu solid line) in 85% (V/V) CH3CN:H2O Path length= 1cm. [mepepy]=[Fe(III)](salten)(mepepy)] = 9 M. 58 Figure 3.12 Steady-state optic al absorption spectrum of Fe(III)(salten)(mepepy) photolysis over time. 59 Figure 3.13 Overlay of the normalized acoustic wave of the reference Fe4SP and mepepy in 85% (V/V) CH3CN:H2O at 35C. The absorbance of the sample and reference at the excitation wavelength (355 nm) was ~0.4. 61 Figure 3.14 Plot of Eh vs. Cp / for mepepy photoisomerization. The slope of these lines provides that change in molar volume ( V = slope/ in mL mol-1) and the intercept provides the enthalpy ( H = (Eh Q)/ in kcal mol-1). 61

PAGE 12

ix Figure 3.15 Overlay of the normalized acoustic wave of the reference Fe4SP and the Fe(III)(salten)(mep epy) complex in 85% (V/V) CH3CN:H2O at 35C. The absorbance of the sample and reference at the excitation wavelength (355 nm) was ~0.4. 65 Figure 3.16 Plot of Eh vs. Cp / for Fe(III)(salten)(mepepy) complex photoisomerization. The slope of these lines provides that change in molar volume ( V = slope/ in mL mol-1) and the intercept provides the enthalpy ( H = (Eh Q)/ in kcal mol1). 65 Figure 3.17 Thermodynamic profile fo r the photoinduced isomerization of the trans to cis mepepy in acetonitrile:water solution. 68 Figure 3.18 Thermodynamic profile for the photoinduced Fe(III)(salten)(mepepy) complex in acetonitrile:water solution. 68 Figure 3.19 Overlay of the acoustic waves of the Ru(bpy)3 (bleu dashed line), the reference Fe(III)4SP (bla ck solid line), the fit (green dotted line) and the residue (red dashed and fit line) Ru(bpy)3 in 10 mM NaCl(aq) (pH 7.0) at 34C.. 78 Figure 3.20 Overlay of the acoustic waves of the Ru(phen)3 (purple dashed line), the reference Fe(III)4SP (bla ck solid line), the fit (green dotted line) and the residue (red dashed and fit line) Ru(phen)3 in 10 mM NaCl(aq) (pH 7.0) at 34C. 79 Figure 3.21 Plot of V versus ln(1+b I) for the 1s t phase (dark bleu square) and 2nd phase (green dot) from Ru(bpy)3 between 10mM and 2M NaCl(aq) (pH 7.0) which gives an intercept equal to the partial molar volume at infinite dilution Vi (mL mol-1) and the slope is equal to (zi 2 Av / 2b), which will give the ion charge zi. 83 Figure 3.22 Plot of V versus ln(1+b I) for the 1s t phase (purple square) and 2nd phase (orange dot) from Ru(bpy)3 between 10mM and 1.5M NaCl(aq) (pH 7.0). 87

PAGE 13

x Figure 4.1 PAS domain (PDB entry 1lsw). 94 Figure 4.2 GCS domain (PDB entry 1or6). 95 Figure 4.3 CooA domain (PDB entry 2hkx). 96 Figure 4.3 HNOB domain (sGC) (PBD entry 1xbn) 97 Figure 4.5 Equilibrium optical absorption spectra of the Bs HemATHD: as isolated (red dotted line), deoxy Bs HemATHD (bleu dashed line) and COBs HemATHD (black solid line). Bs HemATHD concentration: ~10 M in 50 mM sodium phosphate (pH 8.0) and 100 mM NaCl. 104 Figure 4.6 Equilibrium CO binding titration for Bs HemATHD. Absorbance changes were measured at 420 nm. Sample concentrations are those described in Figure 4.1. 105 Figure 4.7 Single wavelength transient absorption trace for O2 rebinding to Bs HemATHD. 107 Figure 4.8 Top panel : Single wavelength transient absorption trace for CO rebinding to Bs HemATHD. Bottom panel : Overlay of the kinetic (squares, obtained 4 ms subsequent to photolysis) and equilibrium (solid line) difference spectra (deoxy Bs HemATHD minus COBs HemATHD). Solution conditions are as described in Figure 4.5. 108 Figure 4.9 Eyring plot for O2 recombination to deoxy Bs HemATHD. 109 Figure 4.10 Eyring plot for CO recombination to deoxy Bs HemATHD. 109

PAGE 14

xi Figure 4.11 Top panel : Overlay of photoacoustic traces for COBs HemATHD (solid line) and the Fe3+ tetrakis-(4sulphonatophenyl)porphyrin cal orimetric reference (dotted line). The absence of any freque ncy shifts between sample and reference acoustic signals indicates all of the observed volume/enthalpy changes occur in <20 ns. Bottom panel : Plot of Eh vs. (Cp / ) which gives a intercept equal to the heat, Q returned to the solvent giving H = ( Eh Q)/ (kcal mol 1) and V = slope/ (mL mol 1). 110 Figure 4.12 X-ray crystal structure of the CN bound and free forms of Fe2+ Bs HemATHD (PDF entry 1or6). 112 Figure 4.13 Thermodynamic profiles for the conformational change associated with the Fe(2+) Bs HemATHD photodissociation (PDB entry 1or6). 115 Figure 4.14 Ribbon structure of CooA. Heme pocket amino acid residues, Pro 2 and His 77 are displayed (PDB entry 1ft9). 121 Figure 4.15 Structural diagram of th e CO photodissociation from the heme domain of CooA. 124 Figure 4.16 Equilibrium optical absorption spectra of CooA as isolated (black solid line) and Fe(II)C ooA-CO (purple dotted line). 127 Figure 4.17 Overlay of the acoustic waves for photolysis of CO from CooA (purple dotted line) and th e reference Fe(III)4SP (black solid line). 128 Figure 4.18 Plot of (S/R)* Eh versus Cp / for CO photolysis from CooA 25 mM MOPS and 0.1M NaCl (pH 7.4) between 8C and 32C which gives an intercept equal to the heat (Q), returned to the solvent giving H = (Eh – Q)/ (kcal mol-1) and V = slope/ (mL mol-1). 129 Figure 4.19 Thermodynamic profiles for CO photorelease from Fe(2+)CooA (PDB entry 1ft9). 131

PAGE 15

xii Figure 4.20 Detail of the two heme-bi nding domain of Co oA showing the Pro 2, His 77, the two salt-bridges: Asp 72 / Arg 118 and Glu 59 / Arg 138 (PDB entry 1ft9). 133 Figure 4.21 Heme domain of the Sm FixL (PDB entry 1ewo). 140 Figure 4.22 Equilibrium optical absorption spectra of Sm FixLWT (top panel), Sm FixLR200Q (bottom panel) as isolated (red dash line), reduced Sm FixL (bleu dot line) and reduced CO bound Sm FixL (black solid line). Sm FixLWT and Sm FixLR200Q concentration: ~5 M in 20 mM Tris (pH 8). 144 Figure 4.23 Equilibrium optic al absorption spectra of Sm FixLH as isolated (red dash line), reduced Sm FixLH (bleu dot line) and reduced CO bound Sm FixLH (black solid line). Sm FixLH concentration: ~5 M in 20 mM Tris (pH 8). 145 Figure 4.24 Single wavelength tran sient absorption data for CO recombination to Sm FixLWT, Sm FixLR200A, Sm FixLR200Q, Sm FixLR200E, Sm FixLR200H and Sm FixLI209M at 25C. Excitation wavelength was 532 nm (<20 ps, 20 mJ/pulse, 20 Hz). Sample solution conditions are the same as those reported in Figure 4.22. Time scale: 40 ms. 147 Figure 4.25 Overlay of the acoustic wa ves for the photolysis of CO from Sm FixLWT (top panel), Sm FixLR200Q (bottom panel) (cyan dotted line) and the reference Fe (III)4SP (black solid line). 149 Figure 4.26 Plot of (S/R)* Eh versus Cp / for CO photolysis from Sm FixLWT (top panel) and Sm FixLR200Q (bottom panel) in 20 mM Tris (pH 8) between 10C and 34C. 151 Figure 4.27 Fluorescence of Sm FixLWT and Sm FixLR200Q in the oxy and CO bound form. 153 Figure 4.28 Thermodynamic profiles for CO photorelease from Fe(2+) Sm FixLWT. 166 Figure 5.1 Hemoglobin (PDB entry 1bzo) 170

PAGE 16

xiii Figure 5.2 Direct channel in soybea n peroxidase (PDB entry 1fhf). 171 Figure 5.3 Photographs of the sandbar shark. 173 Figure 5.4 Steady-state optical ab sorption spectra of sandbar shark hemoglobin (as isolated (red dot ted line), deoxy sandbar shark hemoglobin (bleu dashed line) and CO sandbar shark hemoglobin (black solid line) in 50mM Tris (pH=7). [Sandbar shark hemoglobin] = ~10 M. 175 Figure 5.5 Overlay of the normalized acoustic wave of the reference Fe4SP and the sandbar shark he moglobin in 50mM Tris at 10C at pH = 7. The absorbance of the sample and reference at the excitation wavelengt h (532 nm) was ~0.4. 177 Figure 5.6 Figure 5.7 Top panel: Single wavelength transient absorption trace for CO rebinding to sandbar shark hemoglobin (440nm) obtained at pH=6 in 50mM Tris. Bottom panel : Overlay of the transient kinetic in the Soret region (black squares, obtained 15 ms subsequent to photolysis) and equilibrium (bleu empty round) difference spectra (deoxy sandbar shark hemoglobin minus CO sandbar shark hemoglobin). Top panel : Single wavelength transient absorption trace for CO rebinding to sandbar shark hemoglobin (440nm) obtained at pH=7 in 50mM Tris. Bottom panel : Overlay of the transient kinetic in the Soret region (black squares, obtained 15 ms subsequent to photolysis) and equilibrium (bleu empty round) difference spectra (deoxy sandbar shark hemoglobin minus CO sandbar shark hemoglobin). 178 179

PAGE 17

xiv Figure 5.8 Top panel : Single wavelength transient absorption trace for CO rebinding to sandbar shark hemoglobin (440nm) obtained at pH=8 in 50mM Tris. Bottom panel : Overlay of the transient kinetic in the Soret region (black squares, obtained 15 ms subsequent to photolysis) and equilibrium (bleu empty round) difference spectra (deoxy sandbar shark hemoglobin minus CO sandbar shark hemoglobin). 180 Figure 5.9 Figure 5.10 Eyring plot for CO recombination to deoxy sandbar shark hemoglobin at pH = 6. Eyring plot for CO recombination to deoxy sandbar shark hemoglobin at pH = 7 (top pa nel) and 8 (bottom panel). 182 183 Figure 5.11 Single wavelength tran sient absorption data for CO recombination to sandbar shar k hemoglobin at 17C (top panel) and 33C (bottom panel) Excitation wavelength was 532 nm (<20 ps, 20 mJ/pulse, 20 Hz). 184 Figure 5.12 Transient difference spect rum overlaid on equilibrium deoxy minus CO-bound difference spectrum at pH = 6 and 7. 186 Figure 5.13 Thermodynamic profile for CO photolysis from (Fe2+)sandbar shark hemoglobin at pH = 6 and 7. 193 Figure 5.14 Ribbon structures of horsera dish peroxidase (PDB entry 1hch) (top) and soybean peroxidase (P DB entry 1fhf) (bottom). Key distal heme pocket amino acid residues, Arg 38, His 42 and Phe 41 are displayed on the left. 198

PAGE 18

xv Figure 5.15 Top panel : Equilibrium optical absorption spectra of horseradish peroxidase as isol ated (red solid line), reduced HRP (bleu dashed line) and reduced CO bound HRP (black dotted line). HRP in 50 mM Tris and 100 mM NaCl (pH 8.0). Bottom panel : Equilibrium optical ab sorption spectra of the soybean peroxidase as isolated (red solid line), reduced SBP (bleu dashed line) and reduced CO bound SBP (black dotted line). SBP in 50 mM Tris and 100 mM NaCl (pH 8.0). 202 Figure 5.16 Overlay of the acoustic wa ves for the photolysis of CO from HRP (top panel) (bleu dotted line) SBP (bottom panel) (green dotted line) and the reference Fe(III)4SP (black solid line). HRP and SBP in 50 mM Tris and 100 mM NaCl (pH 8.0) at 22C. 204 Figure 5.17 Plot of (S/R)* Eh versus Cp / for CO photolysis from HRP (bleu square) and SBP (green dot) in 50 mM Tris (pH 7.0) between 6C and 34C. 205 Figure 5.18 Illustration of the differen ces in heme structure between CO bound (shown in yellow) and deoxy HRP. 210 Figure 5.19 Illustration of the heme ac tive site of ferri c SBP showing the position of the bound Tris molecule. 215 Figure 5.20 Thermodynamic profiles fo r CO photorelease from ferrous HRP and SBP under the vari ous solvent conditions. 217 Figure 5.21 Illustration showing the larg est internal cavitie s in horseradish peroxidase and sperm whale Mb using CASTp. 220

PAGE 19

xvi Figure 5.22 Summary of the pathway of CO after photolysis in HRP and SBP with the view of the heme pocket for Soybean peroxidase down the direct channel. CO is bound to the heme iron (a). Then CO is photodissociated from the heme iron (b). In the first fifty nanoseconds, CO leaves the heme pocket to the surrounding solvent by the di rect channel and two water molecules are input of into the ligand access channel (c). 221

PAGE 20

xvii List of Schemes Scheme 2.1 Jablonski-Perin diagram. 10 Scheme 2.2 Diagram of photoacousti c calorimetry instrumentation. 15 Scheme 2.3 Deconvolution of photo acoustic calorimetry signals. 20 Scheme 2.4 Diagram of transien t absorption instrumentation. 22 Scheme 3.1 Simple model for CO photolysis from heme proteins. 28 Scheme 3.2 Variation of volume and enthalpy for the bond cleavage COFe, Iron spin state change and CO solvation. 28 Scheme 3.3 Spin crossover for the ir on from low-spin to high-spin. 29 Scheme 4.1 Major hemes. 93 Scheme 4.2 Scheme for the signal tran smission in gas sensor proteins. 94 Scheme 4.3 The schemes for the change in the coordination structure of the heme in wild-type CooA with associated lifetimes. 122 Scheme 5.1 R and T allosteric states of the hemoglobin. 188 Scheme 5.2 General reaction for the catal ytic cycle of plant peroxidases. 197

PAGE 21

xviii Time-Resolved Thermodynamics Studies of Heme Signaling Proteins and Model Systems Audrey Mokdad ABSTRACT Heme-based gas sensor proteins have the ability to sense diatomic molecules such as O2 (FixL, EcDos or HemAT), CO (CooA, a CO-sensing protein of Rhodospirillum rubrum ) and NO (guanylate cyclase) molecules and subsequently regulate numerous important biological processes in prokaryotic and eukaryot ic organisms. The sensing function of these proteins is initiated by the binding of an effector (i.e., O2, CO, etc…) to the heme iron which then leads to a cascade of conformational events which gives rise to changes in kinase activity, DNA-binding activity, etc… In order to better understand the mechan ism heme-based signaling, time resolved photothermal methods as well as transient op tical techniques were utilized to obtain thermodynamic profiles for ligand binding/rele ase in heme based signaling proteins including HemAT from Bacillus subtilis (aerotactic transducer), FixL from Sinorhizobium meliloti (regulation of the nitroge n fixation) and CooA from Rhodospirillum rubrum (transcriptional activator). In addition, a number of model systems were examined to understand the underlying thermodynamic processes involved in heme ligation. The variation of volume and enthalpy changes associat ed with spin state change of the iron from high-spin to low-sp in where examined using the spin crossover Fe(III)(salten)(mepepy) complex. In addition, the experimental determination of the volume change due to electrostriction events were using Ru(II)(L)3 and the DebyeHckel equation.

PAGE 22

xix Finally, different model heme proteins were studied to understand how a signal is conformationaly transmitted within a heme protein matrix. Sandbar shark hemoglobin was examined as an example of a non-signa ling an allosteric pr otein. Two different peroxidases (horseradish and soybean) which have a direct channel between the heme pocket and the solvent involving no barrier energetic for the photodissociated ligand leaving the heme pocket were examined as example of non-signaling, non-allosteric proteins. The results show that each prot ein has a unique thermodynamic profile to conformationaly transmit signals subsequent to photodissociation of CO, even within the same class of protein (i.e. PAS domains, globins, etc...).

PAGE 23

1 Chapter I Introduction 1.1. Introduction Cell signaling is a complex process i nvolving a diverse array of proteins complexes as well as various messenger molecules [1-6] The signaling process is critical for intracellular trafficking of metabolic pathways as well as extracellular environmental monitoring. In large organisms such as plants and mammals both intra and extracellular signaling/sensing involves complex signaling cascades [7-9] An example of the complexity in mammalian signaling is the in sulin pathways. Insulin is a hormone that allows the body to store glucose as glycogen from the blood to the liver or muscle. The insulin pathway which starts from the transmission of the signal (glucose) to the release of insulin can be summarized as follows. Glucose enters the cell through a glucose transporter and enters into glycolysis and subsequent respiratory cycle where several ATP molecules are produced. Depending upon th e concentration of ATP, the potassium channels (K+) will be closed which will, in turn, control the activity of calcium channels (Ca2+) which will remain open to allow a flow of calcium into the cell. This increase in intracellular Ca2+, inside the cell will cause the release of insulin [10-13] In contrast, the signaling pathway associated with E. coli engages only a few proteins. The signal is activated by ligand binding to a methyl-ac cepting chemotaxis protein receptor and transmitted to a kinase domain via a linker pe ptide fragment. The activation of the kinase results in phosphorylation of a histidine kinase (CheA), which will stimulate the phosphorylation of a methylestera se kinase (CheB). In this way, the signal is transmitted to the flagella which will allow the bacteria to translationally re spond to signaling event (Figure 1.1) [14, 15]

PAGE 24

2 The signaling pathways in large orga nisms such as plants and mammals eukaryotes is more complex than small orga nisms such as bacteria which makes the study of bacterial systems very attractive. In mammalian cells, the signaling pathway for the transmission of a signal involves many protei ns. On the other hand, the transmission of a signal for small organisms like bacteria i nvolves only a few proteins. A good example to compare the complexity of the transmission of the signal between the cell signaling in eukaryotes and prokaryotes can be demons trated with the insulin pathway of a mammalian and the chemotaxis pathway of Escherichia coli ( E. coli ). In order to better understand the molecu lar basis for signal transmission, our studies focused on the inves tigation of conformational a nd thermodynamics associated with signaling in heme-based bacterial se nsors. As bacterial communication shares common features with eukaryote cells, (int ra/extracellular communication with others cells or organelles and/or adaptation to e nvironmental changes), the study of bacterial proteins will unravel the sali ent features of the molecular mechanisms of signaling, but with a less complicated system. In bacteria, different types of si gnals (amino acids, pH temperatures, gas molecules, etc…) [16] can initiate signal in the sensing domain, resulting in phosphorylation and a transm ission of the signal to the flagella (Figure 1.1)

PAGE 25

3 Figure 1.1 : Diagram of the oxygen sensor domain. The specific class of heme proteins exam ined in this study responds to diatomic gases including O2, CO and NO. This allows for the exploitation of the well know spectroscopic signatures, wate r solubility and pho tolability of bound ligands (e.g. CO). Well-known examples of heme sensors are th e Per-Arnt-Sim (PAS) domain, the globincoupled sensor (GCS), the CooA heme binding domain, and the heme-NO-binding proteins (H-NOX) [17] Thus, the study of soluble heme oxygen sensors provide an opportunity to study, in detail, the mechanism of small molecule binding to a sensing domain and to understand the transmission of the signal to the effector domain. This can be accomplished by photodissociating bound liga nd and monitoring the response using a variety of spectroscopic and thermodynamic probe s. It was anticipated that each step in

PAGE 26

4 the transmission of the signal from the sensor domain to the effector domain, exhibits a conformational change with an a ssociated variation in volume ( V) and enthalpy ( H), which can be studied using time-resolved photothermal methods (Figure 1.2) Figure 1.2 : Thermodynamic profiles (variation of volume and enthalpy) of the signal transmitted along the sensor, linker and kinase domain. Time resolved photothermal methods have proven in the past to be powerful techniques for determining the magnitudes a nd time scales of molar volume and enthalpy changes associated with physiological ev ents in proteins including ligand binding, folding/unfolding and electron transfer [18-25] In our laboratory, these methods have been used to investigate nanosecond/millisec ond thermodynamics associated with ligand photolysis from heme model systems, liga nd binding to myoglobin and heme copper oxidases, and to the folding/unfolding of apo-myoglobin and several photochemically ‘caged’ peptides [26-42] In this work, time resolved photothermal methods as well as transient optical techniques were utilized to obtain thermodynamic profiles for ligand

PAGE 27

5 binding/release in heme based signali ng proteins including HemAT from Bacillus subtilis FixL from Bradyrhizobium japonicum and Sinorhizobium meliloti and CooA from Rhodospirillum rubrum A number of model systems were also examined to understand the underlying thermodynamic processe s associated with in heme ligation in these proteins.

PAGE 28

6 1.2. References [1] Witzany, G.111 (2000). Life: The Communicative Structure. Norderstedt, Libri BoD. [2] Cell communication: understanding how inform ation is stored and used in cells by Michael Friedman Juvenile Nonfiction – (2005) 48 pages. [3] Cell Communication and Signaling by Bi oMed Central, Vol 1, issue 5 (2003). [4] Cell Communication by Rody P. C ox Science – (1974) 272 pages. [5] Handbook of cell signaling by Ralph A. Br adshaw and Edward A. Dennis Science – (2004) 2000 pages. [6] Molecular Biology of the Cell by Bru ce Alberts Cytology – (1989) 1263 pages. [7] Mohamed O.A., Jonnaert M., Labelle-Dumais C., Kuroda K., Clarke H.J., Dufort D., Proc. Natl. Acad. Sci. U.S.A. 102 (2005), 8579. [8] Clarke M.B., Sperandio V., Am. J. Physiol. Gastrointest. Liver Physiol. 288 (2005), G1105–9. [9] Lin J.C., Duell K, Konopka J.B., Mol. Cell. Biol. 24 (2004), 2041. [10] Bevan, P., Insulin signalli ng. J. Cell Sci., 114 (2001), 1429-1430. [11] Orci L., Thorens B., Ravazzola M., and Lodish H. F., Science Wush. DC (1989) 245295. [12] Bendayan M., Am. J. Physiol. (1993), G187-G194. [13] Shulman GI., J Clin Invest ., 106 (2000), 171-176. [14] Khan S., Spudich J. L., McCray J.A. and Trentham D.R., Proc. Natl. Acad. Sci. USA, 92 (1995), 9757-9761. [15] Bourret R. B., Borkovich K. A. and Simon M. I., Annu. Rev. Biochem. 60 (1991) 401-441.

PAGE 29

7 [16] Berg H. C. and Anderson R. A., Nature 245 (1973), 380-382. [17] Gilles-Gonzalez M. A.; Gonzalez G., J. Biol. Inorg. Chem., 99 (2005), 1 22. [18] Peters K.S., Watson T.and Logan T., J. Am. Chem. Soc. 114 (1992), 4276. [19] Losi A., Michler I., Ga rtner W. and Braslavsky S.E., Photochem. Photobiol. 72 (2000), 590. [20] Di Primo C., Hoa G.H.B., Depr ez E., Douzou P. and Sligar S.G., Biochemistry 32 (1993), 3671. [21] Peters K.S., Watson T. and Mar K., Annu. Rev. Biophys. Biophys. Chem. 20 (1991), 343. [22] Braslavsky S.E. and Heibel G.E., Chem. Rev. 92 (1992), 1381. [23] Westrick J.A., Peters K.S ., Ropp J.D. and Sligar S.G., Biochemistry 29 (1990), 6741. [24] Schulenberg P.J., Gartner W.and Braslavsky S.E., J. Phys. Chem. 99 (1995), 9617. [25] Falvey D.E., Photochem. Photobiol. 65 (1997), 4. [26] Hansen K.C., Rock R.S., Larsen R.W. and Chan S.I., J. Am. Chem. Soc. 122 (2000), 11567. [27] Larsen R.W., Osborne J ., Langley T. and Gennis R.B., J. Am. Chem. Soc. 120 (1998), 8887. [28] Miksovska J., Day J. and Larsen R.W., J. Biol. Inorg. Chem. 8 (2003), 621. [29] Chen R.P.-Y., Huang J.J-T., Chen H.-L ., Jan H., Velusamy M., Lee C.-T., Fann W., Larsen R.W. and Chan S.I., Proc. Natl. Acad. Sci. U.S.A. 101/19 (2004), 7305. [30] Larsen R.W. and Langley T., J. Am. Chem. Soc. 121 (1999), 4495. [31] Larsen R.W., Inorg. Chim. Acta 288 (1999), 74. [32] Miksovska J. and Larsen R.W., J. Prot. Chem. 22 (2003), 387.

PAGE 30

8 [33] Miksovska J. and Larsen R.W., J. Inorg. Chem. 43 (2004), 4051. [34] Miksovska J., Suquet C., Satterlee J.D. and Larsen R.W., Biochemistry 44 (2005), 10028. [35] Kuo N.N.-W., Huang J.J.-T., Miksovska J., Chen R.P.-Y., Larsen R.W. and Chan S.I., J. Am. Chem. Soc. 127 (2005), 16945. [36] Rock R.S., Hansen K.C., Larsen R.W. and Chan S.I., Chem. Phys. 307 (2004), 201. [37] Miksovska J. and Larsen R.W., Inorg. Chim. Acta 355C (2003), 116. [38] Miksovska J., Gennis R.B. and Larsen R.W., FEBS Lett. 579 (2005), 3014. [39] Miksovska J., Gennis R.B. and Larsen R.W., Biochim. Biophys. Acta 1757 (2006), 182. [40] Miksovska J., Norstrom J. and Larsen R.W., J. Inorg. Chem. 44 (2005), 1006. [41] Mokdad A., Nissen M., Satterlee J.D. and Larsen R.W, FEBS Letters, 581 (2007), 4512-4518. [42] Mokdad A., Miksovska J. and Larsen R.W., Photothermal Studies of CO Photodissociation from Peroxida ses Horseradish and Soybean, BBA. (2009) (Accepted).

PAGE 31

9 Chapter II Methods 2.1 Photothermal Spectroscopy Photothermal spectroscopy methods can probe both optical absorption and thermal characteristics of a photoactive sample [1] The principle of photothermal spectroscopy is simple. The molecule studied in solution is excited via absorbing of a photon (h ) resulting in a transition from the gr ound state to an excited state following Fermi’s Golden Rule [21,22] The molecule then dissipates the excess energy via nonradiative decay, emission of a photon (fluor escence) or intersystem crossing to an excited triplet state. The molecule in the tr iplet state can also re lax to the ground state by nonradiative decay or emission of a photon ( phosphorescence). The excitation and delay processes of a molecule by a photon are summ arized by the Jablonski-Perin diagram in Scheme 2.1 [20] The molecule can also relax to the ground state by photochemical events which include formation/cleavage of bonds which affect the molecular volume (van der Waals), the formation or elimination of a charges (electrost riction), change in the metal spin states and/or the release or binding of ligand. These processes also give rise to changes in volume and enthalpy, whic h can be analyzed by phot othermal methods. Photothermal methods probe temperature, pressure and density changes associated with the optical absorption changes of the sample [1] Photothermal methods can be divided in two categories dependi ng on the parameter st udying: photoacoustic methods which are based on the detection of pressure waves and grating and lensing methods which are based on the detecti on of changes in th e refractive index [1] The following projects presented used ph otoacoustic calorimetry methods.

PAGE 32

Scheme 2 10 .1 : Jablons k k i-Perin diag r r am [20]

PAGE 33

11 2.1.1. General background As previously described by Larsen and Miksovska [17] when a molecule dissipates excess energy to ground state subsequent to photoexcitation, th e surrounding solvent will receive thermal energy due to all non-radiative processes [13,14] The classical heat equation describe s the response of the solvent to such a heat impulse is: 2 2 00 Equation 1 T(x, t=0) = F(x) correspond to the initial condition T(x=0;t) = T1 T(x=L,t) = T2 where is the thermal diffusivity, k is the th ermal conductivity and g(x,t) is the heat deposited per unit volume [14] The solution of equation 1 can be written as following: T(x,t) = + (T1) | x=0 d (T2) | x’=L d Equation 2 where the integration limit over x’ are from x’=0 to L and over are from =0 to t, G(x,t|x’, ). The temperature change per excitation pulse as a function of the distance from the center of the Gaussian distribution as a function of time: T(r,t) = { Equation 3 For solvents such as water, the rapid volum e change expansion is caused by the heating within the illuminated volume resulting in bot h an acoustic wave and a change in solution of the refractive index. correspond to the boundary conditions

PAGE 34

12 In the case of photoacoustic calorimetry, with in the illuminated volume, the heat impulse causes a rapid change in the temperature of the solution which results in a physical volume change which generates both an acous tic wave and a change in the refractive index. Equation 2 describes the change in solv ent pressure due to the change in volume: P = 2 faa x Equation 4 where fa is the frequency of the sound wave, a is the acoustic velocity, x is the volume displacement and is the solvent density [15,16] The change in volume of a cylinder with radius R and length L and due to an ad iabatic, isobaric expansion can be expressed as following: R2l – (R + R)2l = V T Equation 5 where is the thermal expansion coefficient. Assuming r=0, exponential part of the equation T(r,t) is negligible: R2l – (R + R)2l = V Equation 6 Assuming R<
PAGE 35

13 An acoustic wave in solution is formed due to the pressure change which is described by the following equation: 2 s 2 2 2 Equation 10 where s is the speed of sound in the medium, (r,t) is the wave amplitude at the observation coordinates r and t, and h(r,t) is a heat source function [15] The solution of the equation 10 can be written as following: = ) dt’’ dr’ g(r,r’,t,t’)h(r,t)dr’ Equation 11 where g(r,r’,t,t’) is a Gree n’s function solving the wave equation for the impulse heat function and h(r’,t’) = (r’)f(t’) at r’ = 0. The acoustic wave amplitude observed at the transducer using equation 11 can then be described as: 0 0s 0 Equation 12 where r0/ s is the propagation delay and 1/r0 is the energy conservation associated with spherical emitters. For a point source (r’=0) with a lifetime the wave amplitude observed at the transducer can be rewritten equation 12: 0 = h00 (-t’’/ ) Equation 13 where (t’) is a Heaviside unit step function, f(t’) = (h0e-t’/ ) (t’) at the point source and is the lifetime of the heat generator. The response of the transducer which is m odeled as an under-damped oscillator to the wave amplitude associated to an impulse response is written as: G(t,t’’) = A sin( (t-t’’))e-(t-t’’)/ 0 Equation 14

PAGE 36

14 where A is the amplitude, is the frequency of the oscillation and 0 the relaxation time. By convoluting the impulse response G( t,t’’) with the acoustic amplitude (r0,t’’), the following equation expresses the transducer response: V(t) = (r0,t’’) dt’’ Equation 15 The transducer response (piezoelectric crystal) to the wave amplitude which represents the relationship between the lifetime of the h eat evolving process and the amplitude at the transducer can be expressed using equation 15 as: 0 0 22 -t/ -t/ 0 Equation 16 where A is the amplitude, is a characteristic os cillation frequency, 0 is a relaxation time, is the lifetime of the heat evolving process. Photoacoustic calorimetry method Photoacoustic calorimetry (PAC) has pr oven to be a powerful technique for determining the magnitude and time scale of conformational changes as well as reaction thermodynamics associated w ith photoinitiated processes [2] The physical principle behind this method is that photoexcited mol ecules dissipate excess energy nonradiatively, resulting in thermal heating of the surrounding solvent. In the case of aqueous solutions this causes rapid volume expansion ( Vth)] resulting in an acoustic wave that can be detected with a sensitive piezoelectric crystal based microphone. In addition, volume changes in the system of interest resu lting from conformational/solvation changes associated with a photoinitiated reaction ( Vcon) also contribute to the acoustic wave. The contributions from Vth and Vcon to the total sample signal, S (acoustic amplitude) can be evaluated by examining the temperatur e dependence of the signal and using a calibration compound (Sref). The theory and instrumentation associated with PAC have been described in detail by [1-5] PAC has been previously employed to determine the thermodynamic profiles for conformational changes associated with the nanosecond

PAGE 37

c h p h h q u N d w s w u c o u ( S h arge trans fe h otodissoci a orse heart m PAC m u artz cuvett e N orthwest) h o etector was w ith a 532 n m w itched Nd: ltrasonic pr e o ntrolled b y sing the tw o S cheme 2.2 ) e r of Ru(bp y a tion of CO m yoglobin a n m easureme n e containin g o using a Pa n facilitated w m or 355 n m YAG laser, e amp (Pana m y VirtualBen c o temperatur e Scheme 2 y )3 [6] conf o from a wide n d sper m w h n ts are perfo r g a sample i n n ametrics V w ith a thin la y m laser pulse 6 ns pulse, < m etrics) and c h software e or multipl e 2 : Diagram 15 o rmational c range of he h ale myoglo b r med (as de s n a temperat u 103 transdu c y er of vacu u (Continuu m < 80 J). Th e recorded us (National I n e temperatu r of photoaco u c hanges ass o me proteins b in [7-12] s cribed prev u re controll e c er. Contact u m grease. P m Minilite I fr e acoustic si g ing an NI 5 1 n strument). T r e method d e u stic calori m o ciated with t including Bj iously [2] ) b e d sample h o between th e P hotoexcitati fr equency d o g nal was a m 1 02 Oscillos T he PAC da t e tailed in th e m etry instru m t he Bj FixL, EcD o b y placing a o lder (Quant u e cuvette an d on was achi e o uble/triplet m plified with cope (15 M H t a were anal y e next sectio n m entation. o s, u m d the e ved Qan H z) y zed n

PAGE 38

16 Multiple temperatures method : The amplitude of the sample acoustic wave is expressed as: Ssamp = KEa (( /Cp )Q + Vcon) Equation 17 where S is the acoustic signal, K is the instrument response parameter, Ea is the number of Einstein absorbed, is the quantum yield of the process, is the thermal expansion coefficient of the solvent (K-1), Cp is the solvent heat capacity (cal g-1K-1), is the solvent density (g mL-1), Q is the amount of heat retu rned to the so lvent (kcal mol-1) and the Vcon is the nonthermal volume change and Vcon represents conformational/electrostriction contributions to the solution volume change (mL mol-1). The amplitude of the reference aco ustic wave is expressed as: Sref = KEa (( /Cp ) Eh Equation 18 The reference converts the energy of the absorbed photon (Eh ) into heat with a quantum yield equal of unity (i.e., Vcon = 0). The volume and enthalpy are determined by taking a ratio of the amplitudes of sample to the reference (Ssamp/Sref) as a function of temperature (Cp, and are temperature dependent s) as expressed below: (Ssamp/Sref)Eh = Q + (Cp / ) Vcon Equation 19 A plot of Eh versus Cp / gives a line with a slope equal to Vcon and an intercept equal to the released heat (Q). Subtracting Q from Eh and dividing by the quantum yield gives H for processes occurring faster than the time resolu tion of the instrument (<20 ns): H = (Eh Q)/ Equation 20 The Q/ values for subsequent ki netic processes represent H for that step (i.e., heat released). H = Q/ Equation 21

PAGE 39

17 Two temperature method : A two temperature methods can also be used to analyze sample and reference acoustic waves. The two temperature methods take the advantage that at = 0 (~4.1C), the wave amplitude of the reference, in aqueous solution, is equal to zero. Thus, Vcon of the sample can be extracted easily using the following equation: Vcon = Eh (Cp / )10C (AT=4.1C samp / AT=10C ref) Equation 22 where is the quantum yield, is the coefficient of thermal expansion of the solvent ( K 1), Cp is the heat capacity (cal.g 1 K 1), is the density (g.mL 1), AT=4.1C samp is the amplitude of the sample at 4.1C, AT=10C ref is the amplitude of th e reference at 10C and Vcon represents conformational/electrostrict ion contributions to the solution volume change. The fraction of heat releases in th e solvent is calculated using the following equation: = (AT=10C samp AT=4.1C samp) / AT=10C ref Equation 23 where is the fraction of heat, AT=10C samp and AT=4.1C samp is the amplitude of the sample at 4.1C and 10C, AT=10C is the amplitude of the reference at 10C. The release of heat deposed in solution is expressed as follows: Q = Eh Equation 24 where Q is the heat released to the solvent. Subtracting Q from Eh and dividing by the quantum yield gives H: H = Eh – Q Equation 25

PAGE 40

18 Time resolution in photoacoustic calorimetry After looking at the overlay for both sa mple and the reference acoustic waves, two different cases which demonstrated or no t a frequency shift can be observed between the sample and the reference. First case : The first case is observed when no freque ncy shift is visible between the sample and the reference as shown on Figure 2.1 This indicates no kinetic event between ~50 ns and ~20 s, so everything happen faster th an ~50 ns or slower than ~20 s. In this case, is equal to the ratio between the amplitude of the sample and the amplitude of the reference. Figure 2.1 : Overlap between the sample and the reference photoacoustic signals showing no frequency shift. 2000 4000 6000 -150000 -100000 -50000 0 50000 100000 150000 Normalized AmplitudeTime (ns) Amplitude

PAGE 41

19 Second case : The second case is observed when a freque ncy shift is visible between the sample and the reference as shown on Figure 2.2 which indicates one or more kinetic events occurring between ~50 ns and ~20 s. In this case, the data wi ll be deconvoluted in order to extract the lifetime, the volume and enthal py changes associated with each phase or event. 20004000 -30000 -20000 -10000 0 10000 20000 Normalized AmplitudeTime (ns) Figure 2.2 : Overlap between the sample and the reference photoacoustic signals showing a frequency shift. Deconvolution corresponds to the convolution between the reference acoustic wave (Tr(t)) and the exponential decay of the heat released to the solvent (H(t)) (Scheme 2.3) In order to deconvolute both reference and sample acoustic waves, noncommercial software pacw98v1 also called larsenware will be used to determine the amplitude and the lifetime of each phase. This software uses the fast iterative convolution equation: Dj+1 = K exp(-tj+1 / )Ij+1 Equation 26

PAGE 42

w r e a r r e w here K is t h e ference wa v r ray using t h e ference wa v R h e amplitude v e and nor m h e initial gu e v e, add the m Scheme 2.3 : Sa m R eference t is the tim m alizing the m e ss. The pro g m together a n Deconvolut i mp le =20 e and the l m to the abs o g ram will th e n d start sim p i on of photo a ifetime. Aft e o rbance, the p e n convolut e p lex in order a coustic cal o e r loading t h p rogram wi l e the new ar r to find the s o rimetry sig n h e sample a n l l create a n e r ays with th e s maller Chi2 n als. n d e w e

PAGE 43

21 2.2. Transient Absorption Spectroscopy Transient absorption (TA) sp ectroscopy follows the kinetics of fast events in real time after excitation of the system under investigation. The ab sorption spectrum is characteristic of the system. The changes i nduced in the sample af ter excitation, involve changes in the absorption spectrum that ar e recorded on the monitor. In transient absorption spectroscopy, a pump pulse will perturb the system with a redistribution of the energy of the molecule. Then, as the probe pulse goes trough the sample, the change in absorption of the sample, at a specific wa velength is recorded as a function of time (Figure 2.3) Transient absorption will give us th e opportunity to extract the different kinetics of the heme after excitation of th e system. In addition, th e results will confirm that the lifetime of the different phases extracted using photoac oustic calorimetry and matching the TA results are due to the dynamics of the heme. TA experiments were performed by monito ring the change in intensity of light from a Xe arc lamp (Oriel) emerging from the sample followed by passage through a 1/4 m single monochromator equipped with an Oriel R928 photo-multiplier tube. The signal was amplified using ether a home-built pre-amplifier (1 MHz bandwidth) or a melles griot which is part of the wide-bandwidth amplifier family used for operating photodiodes in both the photovoltaic and photoc onductive modes, followed by a Stanford Instruments SR445A 350 MHz pos t amplifier. The signal was digitized using a Tektronix TDS7404 4 GHz digital oscilloscope. The samp le was excited with the second harmonic of a Continuum Leopard I Q-switched mode -locked Nd:YAG laser (<20 ps, 20 mJ/pulse, 20 Hz) (Scheme 2.4)

PAGE 44

Scheme 2. 4 4 : Diagram o 22 o f transient a a bsorption i n n strumentati o o n.

PAGE 45

23 Excitation at 440 nm -0.010.000010020.030.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12 Delta AbsorbanceTime (s) Figure 2.3 : Transient absorption spectra after excitation at 440 nm. 360380400420440460 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Reduced form Reduced-CO boundNormalized AbsorbanceWavelength (nm) 360380400420440460 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 Delta AbsorbanceWavelength (nm) Black curve minus red curve

PAGE 46

24 By using transient absorption spectroscopy will also be able to extract the entropy and enthalpy of activation of a reaction. The enthalpy of activation for the ligand bound to the heme will give us the opportunity to extract the enthalpy of activation of the dissociation of the ligand bound to the heme. Then, the complete thermodynamic profiles of the reaction can be drawn using the following equation: H = H dissociation H binding Equation 27 The following Eyring’s equation will be used to calculate the entropy ( S) and enthalpy ( H) of activation for a ligand (CO or O2) binding to a protein: Ln(kobs/kbh/T)) = H /RT + S /R Equation 28 where kb is the Boltzmann’s constant, h is the Planck’s constant, kobs is the observed rate constant and T is the absolute temperature [18,19]

PAGE 47

25 2.3. References [1] Bialkowski S. E., Photothermal spectroscopy me thods for chemical analysis 34 (1996), 1-38. [2] Larsen R. W. and Miksovska J., Coord. Chem. Rev. 251 (2007), 1101-1127. [3] Gensch T. and Viappiani C., Photochem. Photobiol. Sci ., 2 (2003), 699-721. [4] Miksovsk, J.; Larsen, R. W. In Methods in Enzymology: Biophotonics, 360: part A ; Marriott, G., Parker, I., Eds.; Academic Press: New York, 2003; p 302. [5] Barker, B. D.; Larsen, R. W. J. Biochem., Mol. Biol. Biophys 5 (2001), 407434. [6] Miksovska J. and Larsen R.W., Inorg. Chem. 43 (2004), 4051-4055. [7] Miksovska J; Gennis R. B. and Larsen R. W., Protein Sciences 13 (2004), 82-83. [8] Mikšovsk J., Yom J., Di amond B. and Larsen R. W. Biomacromolecules, 7 (2006), 476-482. [9] Mikšovsk J., Suquet C., Satterlee J. D. and Larsen R. W., Biochemistry, 44 (2005), 10028-10036. [10] Miksovska J., Norstrom J. and Larsen R.W., J. Inorg. Chem., 44 (2005), 1006-1014. [11] Miksovska J. and Larsen R.W., Inorg. Chem. 43 (2004), 4051-4055. [12] Miksovska J., Day J.and Larsen R. W., J. Biol. Inorg. Chem., 8 (2003), 621-625. [13] Patel C.K.N. and Tam A.C., Rev. Mod. Phys. 53 (1981), 517. [14] Jackson W.B., Amer N.M., Boccara A.C. and Fournier D., Appl. Optics 20 (1981), 1333.

PAGE 48

26 [15] Rudzki J.E., Goodman J.L.and Peters K.S., J. Am. Chem. Soc. 107 (1985), 7849. [16] Rothberg L.J., Simon J.D., Bernstein M. and Peters K.S., J. Am. Chem. Soc. 105 (1983), 3464. [17] Larsen R. W. and Miksovska J, Coordination Chemistry Reviews 251 (2007), 11011127. [18] van Eldik R., Asano T. and le Noble W.J., Chem. Rev., 89 (1989), 549. [19] Hiromi K., Kinetics of Fast Enzyme Reactions: Theory and Practice, Kodanshi Scientific Books, Kodanshi, Ltd. (1979). [20] Elumalai P., Atkins P., de Paula J. Atkins' Physical Chemistry Oxford University Press, 2002. [21] Dirac, P.A.M., Proc. Roy. Soc. (London) A, 114 (1927), 243–265. [22] Fermi E., Nuclear Physics University of Chicago Press (1950).

PAGE 49

27 Chapter III – Model Systems Different model systems were selected in order to understand the underlying thermodynamic processes in heme ligation. Us ing photothermal methods, the volume and enthalpy changes of a system (molecule or pr otein) after photoexitation can be extracted using equation 9 in the previous section. Th e variations of the observed volume or enthalpy observed ( Vobs and Hobs) originate from two general events. 1) variations in volume or enthalpy corresponding to the photol ysis of a ligand from heme proteins ( Vphotolysis and Hphotolysis) and 2) variations of volume and enthalpy corresponding to the conformational changes associated with the protein matrix ( Vconf and Hconf). As follows: Vobs = Vphotolysis + Vconf Equation 29 Hobs = Hphotolysis + Hconf Equation 30 The variations in volume and enthalpy due to the photolysis event corresponding to a ligand photorelease from a heme prot ein has contributions from: the Fe-L bond cleavage, spin state change of the iron from low-spin to high-spin a nd the solvation of the ligand after its releas e to the surrounding solvent. Thes e processes are summarized in Scheme 3.1

PAGE 50

28 Scheme 3.1 : Simple model for CO photolysis from heme proteins. Each of the three contributions (bond cleavage, spin state change and ligand solvation) is associated with a change in volume and enthalpy as summarized in Scheme 3.2 Scheme 3.2 : Variation of volume and enthalpy for the bond cleavage CO-Fe, Iron spin state change and CO salvation [5-7]

PAGE 51

29 Thus, the variation in volume and enth alpy associated with the corresponding conformational changes with in the protein matrix can be extracted using: Vconf = Vobs Vphotolysis Equation 31 Hconf = Hobs Hphotolysis Equation 32 Previous high pressure results of va rious metallocomplexes have indicated a change of volume of ~10 mL mol-1 for low-spin to high-spin transition. The variation was associated to the expansion of the core of the metal complex due to the expansion of the dz2 orbital [1] Previous studies estimated that the enthalpy change of an iron from lowspin to high-spin from iron complexes was negligible at low temperature. The first model system project was the es timation of the variation of volume and enthalpy of a photo-induced spin state change at room temperature using an iron complex with a push-pull ligand (Scheme 3.3) The iron complex chosen was the Fe(III) salten mepepy complex. Scheme 3.3 : Spin crossover for the iron from low-spin to high-spin. Low spin High spin + CO

PAGE 52

30 The variations in volume and enthalpy of protein conformational changes can correspond to different contribu tions such as van der Waals volume, electrostriction or the variation of volume and enthalpy due to the variation of the protein salvation [2] The following equations summarize the different c ontributions that can be observed for the volume and enthalpy, associated to th e protein conformational changes: Vconf = Vvdw + Velectrostriction + Vsalvation Equation 33 Hconf = Hvdw + Helectrostriction + Hsalvation Equation 34 The second model system investigated in th is study involves the variation of volume associated with van der Waals volume change s versus electrostriction of two different ruthenium complexes using Debye-Hckel theory [3,4]

PAGE 53

31 3.1. References [1] Allen F. H., Acta Crystallogr Sect B, 58 (2002), 380-388. [2] Schmidt R., J. Phys. Chem. A 102 (1998), 9082-9086. [3] Debye P. and Hckel E., Physikalische Zeitschrift, 24 (1923) 185–206. [4] Ananthaswamy J. and Atkinson G., J. Chem. Eng. Data 29 (1984), 81-87. [5] Norris C. L. and Peters K. S., Biophysic. J., 65 (1993), 1660-1665. [6] Traylor T.G., Acc. Chem. Res., 14 (1981), 102-109. [7] R. van Eldik, A. Asano, and W. J. le Noble, Chemical Rev., 89 (1989), 549-688.

PAGE 54

32 3.2. Photophysical Studies of the Trans to Cis Isomerization of the Push Pull Molecule: 1-(Pyridin-4-yl)-2-( N -methylpyrrol-2-yl)ethene (mepepy) 3.2.1. Introduction The study of the class of pushpull molecules (Donor (D) – Acceptor (A)) has improved in the past decades. Their intere st comes from their nonlinear optical (NLO) properties used in high-speed optical modulators, op tical storage media, and fast/ultrafast optical switches, etc. which can be exhibite d by a different dipole moment in the groundstate and excited-state of the push-pull molecule [1-5] A push-pull molecule is a molecule with two different sides linked by a conjugated double bond, one side has an electron-withdraw ing substituent and the other one has an electron-donating substituent. Typically, functional groups like amino, dialkylamino, ether, or oxide (O2 ) form the electron-donating substi tuent while nitro, carbonyl, and cyano groups are employed as the corresponding acceptor group [6, 7] This enhances the polarizability of the double bond region allowing for additional polarization to be induced in the presence of an electric field. A good example is chromophore molecules. Push-pull molecules are very polarized and the conf iguration with a donor and receptor side facilitates the molecular switch between the cis and trans molecule, by decreasing the rotational barrier. The configuration push-pull also stab ilizes the double bond by given a less double bond character [1]

PAGE 55

33 In addition to their nonlinear optical pr operties, push-pull molecules which are liganded to different transition metal complexe s can change photolytically the spin state of the complex [8-13] The push-pull molecule must have some properties such as functional group which can coordinate to th e metal complex and imply a shift in the basicity after photoexitation. The complex will have two different configurations, a lowspin electron configuration which will be mo re basic than the high-spin configuration after photoconversion of the ligand. Actually, these complexes are now synthesized and their optical and magnetic are known. For instance, complexes of the FeII(L4)(X2) type in which L is the photoisomerizable push-pull ligand and of the FeIII(L4)(L )(X) type in which L is the push-pull ligand. The first studies were focused on 4-sterylpyridine (Stpy; 1-phenyl-2-(4-pyridyl)ethane) as well as several phenyl derivativ es of this ligand in order to photoinduce the spin crossover [10-12] A good example of the light-induced spinstate changes, is the FeII( trans -Stpy)4(NCS)2 complex which can induce a high-spin to low-spin transition thermally around 190K by a trans to cis isomerization of the Stpy [10] An idea was to synthesize these complexe s which exhibit liganddriven, light-induced spin-state changes at higher temperatures. For instance, [FeIII(salten)(mepepy)]BPh4 (salten = 4-azaheptamethylene-1,7-bis(salicy lideneiminate); mepepy= 1-(pyridin-4-yl)-2( N -methylpyrrol-2-yl)ethene; BPh4 = tetraphenyl borate) has been shown to exhibit a high-spin to low-spin transition at ro om temperature under visible irradiation [11-13] The difference with FeII(L4)(X2) type complexes discussed above is that they have four photoisomerizable ligands, the FeIII(L4)(L )(X) type complexes (e.g., FeIII(salten)(mepepy)) have only a single isomer izable ligand which is used to alter the ligand field strength. In the case of the mepepy complex, the mepepy will be linked to an Fe(II) complex which has an active spin crossover center. The spin-state transition is triggered by a light-induced a trans to cis isomerization of the li gand. A shift in electron density from the N -methylpyrrol moiety to the pyridine unit will be observed (see Figure 3.1 for a structural di agram of mepepy).

PAGE 56

34 Figure 3.1 : Structures of the trans and cis conformations of mepepy. An example of a push-pull molecule is the mepepy complex (Figures 3.1) The cis / trans configuration of the mepepy ligand ha s two sides conjugated by a double bond: a pyridine forms the electron-withdrawing substituent and a N-methylpyrrole forms the electron-donating substituent. The mepepy ligan d was the first molecule observed with a spin change from a high-spin to a low-spin after excitation associated with a change in energy [11] The electron properties associat ed with the energetics of the cis / trans photoisomerization is still not known very well even if the cis / trans photoisomerization has been demonstrated to alter the ligand field of the chromophore. The trans conformation of mepepy shows an optical sp ectrum with absorption maxima at ~353 nm ( = 22800 M 1 cm 1) and 241 nm ( = 7800 M 1 cm 1) in acetonitrile. These absorption spectra have been displayed to be to transitions [11, 12] After photoisomerization, a slight hypsochromic shift is observed in the absorption band at 353 nm associated with a nearly 50% decrease in the extinction. While only modest changes in extinction of the Trans Mepepy Cis Mepepy

PAGE 57

35 241 nm band are showed. Time-resolved photothermal and computational methods were used in order to study the energetics in addition to th e potential energy surfaces associated with the cis / trans photoisomerization of the mepepy complex after excitation. 3.2.2. Materials and Methods 3.2.2.1. Synthesis of mepepy The procedure described in reference 1 was followed in order to prepare the trans -mepepy ligand. A solution under an argon atmosphere was prepared by adding 905 L of 4-picoline in 4 mL of anhydrous dime thylformamide (DMF) to a suspension of 370 mg of sodium hydride (60% dispersion in mi neral oil) in 10 mL of anhydrous DMF. After stirring the solution at 60 C for 2 h, a solution of 1.01 g of N -methylpyrrole-2carboxyaldehyde in 6 mL of anhydrous DMF wa s added to the red anionic solution and then stirred at 60C overnight. A yellow solid was extracted after been poured onto ice and filtered. The filtrate was then concentrat ed and purified using a silica gel column with a mixture of solvent hexane:ethyl alc ohol at different rati o 2:1, 1:1 and 1:2. The purity of the compound was confirmed by NMR spectroscopy. The quantum yield, associated with the isomerization of the mepepy from the trans to cis form was determined by irradiating, fo r a total of 35 mins a cuvette at the wavelength used to perform the PAC meas urement, which is 355 nm. The cuvette contained 0.19 mM mepepy in water with 1% dimethyl sulfoxide (DMSO). The absorption spectra were recorded every 5 mi ns. A control was also used by performing the same experiment with 0.8 mM azobenzene. The following equation was used in order to determine the quantum yield of the mepepy by comparing the value of the quantum yield with the value alrea dy known of the azobenzene: ( A0 35 mep / A0 35 azo) = {( mep mep C0 mep) / azo azo C0 azo)} Equation 36

PAGE 58

36 where A0 35 mep and A0 35 azo are the changes in absorban ce, respectively, for mepepy and azobenzene at 355 nm, mep is the quantum yield for the trans to cis isomerization of the mepepy, azo is the quantum yield for the trans to cis isomerization of the azobenzene (0.26), mep (2.2 M 1 cm 1) and azo (0.4 M 1 cm 1) are the differences in extinction coefficients, respectively, between the trans and cis forms of mepepy and azobenzene at 355 nm, and C0 mep and C0 azo are the initial concentra tions. Using the equation 22, mep was calculated to be equal to 0.28. 3.2.2.2. Computational Methods Computational calculations were pe rformed by John Belof from Dr Space’s group. The quantum chemistry program GAMESS was used in order to calculate the electronic structure of the mepepy ligand [15] The molecular properties of the azobenzene molecule which is a similar mol ecule to mepepy were already studied. The results were accurate in terms of isomeriza tion studies using density functional theory (DFT) [16-20] The calculations were done by B3LYP hybrid exchange-correlation functional with the augmente d correlation-consistent doublebasis set (aug-cc-pVDZ) [19-22] The effects of the solvent were mimicked by using a polarizable continuum model where the aqueous environment was imitated by surrounding the mepepy molecule into isotropic dielectric field [23] The dielectric constant was estimated to be equal to 78.39 and the solvent radius equal to 1.3850 . In order to establish the minimum of the energy geometries of the trans and the cis mepepy the geometry optimizations were done in the solvation field. In addition, the charge distributions establish the molecular electric dipole moments for the optimized struct ures. Moreover, the ground state of the trans and cis mepepy potential energy surfaces were studie d in order to find any other local energy minima. Each angle of the pyridine and pyrro le rings from a planar configuration was scanned from 0 to 180 in 10 increments along the and vectors leaving the conformation adopting any minimum. The sp ectroscopic calculations using ZINDO/CI method from ArgusLab [24] were completed by using the highest 45 occupied and lowest 45 unoccupied orbitals in th e CI (SCF-CI, restricted Hartree Fock, with STO-6G

PAGE 59

37 minimal basis set). A dielectric constant of 78.3 and a cavity radius of 4.616 (SPCE water model) were also used for a self-consistent reaction field. 3.2.3. Results and Discussion 3.2.3.1. Optical Properties Figure 3.2 shows the steady-state absorption sp ectrum of the mepepy isolated in its trans form and solubilized in water. The ab sorption spectrum shows two bands one at ~355 nm and a shoulder at ~425 nm. The ba nd at 355 nm was designed to be the HOMO to LUMO transition or the to transition. Figure 3.3 shows how the electron density migrated from the HOMO orbitals where the electron density is centered on the fivemembered ring to the LUMO orbitals where the electron density migrated between the five and six-membered rings. Figure 3.3 also shows how the electron density of the cis form of the mepepy migrated from the fi ve-membered ring in the HOMO to the sixmembered ring in the LUMO. The band at 425 nm was attributed to the to transition which corresponds to the prot onation of the nitrogen on the pyridine ring of the mepepy. This hypothesis was confirmed by an experime nt where a dilute NaOH solution (<0.1M) was added to the mepepy solution in water. The decrease of the band at 425 nm was followed on the absorption spectrum (Figure 3.2)

PAGE 60

38 Figure 3.2 : Optical absorption spectra of the steady -state of mepepy in water (containing <1% DMSO) (solid line) and aqueous 0.01 N NaOH (containing <1% DMSO; dotted line). [mepepy] = 270 M.

PAGE 61

39 Figure 3.3 : Diagrams of the molecular orbital viewing the HOMOs and LUMOs related with the observed to transitions of the trans (left) and cis (right) isomers of mepepy. Figure 3.4 demonstrates that after il lumination of the mepepy and photoisomerization from the trans to cis form of the mepepy, a decrease in the molar extinction coefficient of the to transition of the protona ted and deprotonated forms of the mepepy is observed associated with a minor hypsochromic shift of the energy.

PAGE 62

40 300 400 500 00 02 04 06 Abs t=0 Abs after 10 mins Abs after 20 mins Abs after 30 mins Abs after 40 mins Abs after 50 mins Abs after 60 mins Abs after 70 minsAbsorbanceWavelenght (nm) Figure 3.4 : Top Panel : UV-Vis spectra of the mepepy equilibrium before and after steadystate photolysis (from 10 to 70 mins). The sample was solubilized in deionized water containing <1% DMSO. Bottom panel : UV-Vis spectra of the trans -mepepy equilibrium before and after 5, 10, 15, and 20 laser pulses ( excit = 355 nm, 100 J/pulse, and 7 ns fwhm). 300 400 500 00 01 02 03 04 05 06 07 Abs t=0 Abs after 1 min Abs after 2 mins Abs after 3 mins Abs after 4 mins Abs after 5 mins Abs after 6 mins Abs after 7 mins Abs after 8 mins Abs after 9 mins Abs after 10 minsAbsorbanceWavelength (nm)

PAGE 63

41 3.2.3.2. Theoretical Structural Analysis Figures 3.1 and 3.6 represent the different trans and cis structures of mepepy. Figure 3.5 also shows the different energy minima located along the potential energy surface between the trans and cis structure of the mepepy. The details of the geometric are summarized in the Tables 3.1 and 3.2. It has been demonstrated that the trans form of the mepepy has two minima, a global and a local minimum, which are separa ted by an energy barrier equal to 9.12 kcal mol-1 and give to the trans potential energy surfa ce unique characteristics (Figure 3.6) Only one minimum was found for the cis form where the potential energy surface was described as a shallow basin which will give the opportunity of the equilibrium structure to fluctuate out of plane into the solvated ground state. Th e percentage of population for each isomer after photoisomeri zation could not be found. An energy difference of 8.24 kcal mol-1 was found between the cis and the global trans minimum form of the mepepy to compare to the 7.17 kcal mol-1 energy difference between the cis and the local trans minimum form. The differe nce in the dipole moment between the cis and the two different trans form of mepepy we re calculated after evaluating the electrostatic moments. A difference of 0.424 D was found between the cis and the global trans minimum form and 0.0398 D between the cis and the local trans minimum. While the relative populations of the isomers after a photoisomerization event are unknown, these determined differential va lues place an upper and lower bound to the true equilibrium differences. The cis potential energy surface is a shallow basin with a single minimum, where it can be expected that the equilibrium structure will fluctuate out of plane in the solvated ground state. However, unique characteristics of the trans potential energy surface were found, most notably the existence of two ground-state minima separated by an energy barrier, here in referred to as the global and local trans states (geometrically shown in Figure 3.5 ). Complete determination of the isomerization mechanism for mepepy (which is not determined in this work) is required to know whether the transitions from the excited st ate to the ground state favor one isomeric trans form over the other. The minimum en ergy barrier between the cogeneric trans states is 9.12 kcal mol 1, whereas the thermal energy kT at 300 K is only 0.596 kcal mol 1.

PAGE 64

42 Table 3.1 : Configurations of the Global Energy-Minimized Geometric cis and trans mepepya Cis isomer Trans isomer Cis isomer Trans isomer (deg) 126.647 126.490 D() 1.34764 1.35337 (deg) 129.507 127.365 D() 1.44983 1.43931 d() 1.48 689 1.46157 (deg) 90.0 0.0 The bond angles and distances are wi th respect to the labels in Figure 3.1 and is the angle between the pyridine and pyrrole planes. Table 3.2 : Calculated Energy and Dipole Moment of Global Minima cis and trans -mepepy Isomers associated with the Alt ernate Local Minimum along the trans Potential Energy Surface. Isomer Energy (kcal mol-1) Molecular dipole moment (D) cis 335330.354.12 global trans 335338.593.69 local trans 335337.513.08

PAGE 65

43 Figure 3.5 : Ground-state energy surfaces of the trans / cis mepepy as a function of the pyridine and pyrrole ring angles as defined by deviation from planar configuration. See text for details. Figure 3.6 : Structure of the global and local trans geometries (left and right, respectively). 3.2.3.3. Photothermal Studies The overlay of the reference Fe4SP and the mepepy PAC signals at 35C are shown in Figure 3.7 No shift is observed for the photoacoustic signals between the reference Fe4SP and the isomerization of the trans to cis mepepy which demonstrate that all the conformational changes appear to be faster than < ~ 50 ns which is the time response of the PAC instrument After plotting, the equation Eh versus ( Cp / ) (equation 9), the volume and the enthalpy can be extracted as it was mentioned before.

PAGE 66

44 Their profiles are summarized Figure 3.8 with a V = 0.9 0.4 mL mol 1 (1.5 3 / molecule) and a H = 18 3 kcal mol 1. -20000200040006000800010000 -80000 -60000 -40000 -20000 0 20000 40000 60000 Normalized Acoustic AmplitudeTime (ns) Fe4SP Mepepy24681012 72 73 74 75 76 77 78 79 80 E hu (kcal mol-1) C p (kcal mol-1) Figure 3.7 : Overlay of the normalized acoustic wave of the reference Fe4SP and the trans mepepy in water containing <1% DMSO at 35C. Both absorbencies at the excitation wavelength (355 nm) were 0.4.

PAGE 67

45 The change in the molar volume is the a ddition of two different contributions. The first contribution is a volume change due to a variation of the van der Waals volume of the trans to cis isomerization of the mepepy. The va n der Waals volume change for the trans to cis isomerization of the mepepy was calcu lated to be equal to ~ 9 mL mol-1 with the cis conformation having the lower molar volume. The second contribution is a variation of the structure of th e solvent around the mepepy molecule which is a result of a cha nge in the ground state dipole moment and/or in a change of the overall charge of th e mepepy. The following equation will give an estimation of the volume change due to a change in electrostriction: Vel = ( 2 / r3) kT {( + 2)( – 1)/(2 + 1)2}(NA/4 0) Equation 37 where is the dipole moment, r is the cavity radius of the molecule (mepepy), kT is the compressibility of the solvent, is the solvent’s dielectric constant, NA is Avogadro’s number, and o is the vacuum permittivity [25, 26] The variation of the volume due to the electrostriction was calcul ated to be ~ -0.2 mL mol-1 using the variation of the dipole moment for the trans to cis isomerization of the mepepy equal to 0.0398 D, a molecular radius equal to ~ 5 by assuming that the di ameter of the mepepy molecule is ~ 10 , kT is equal to 4.22 10 10 Pa 1 (for water at 25C) and equal to 78.4. The total variation adding the two contribu tions was calculated to be equal to ~ 9.2 mL mol-1. The experimental variation of volum e was estimated to be equal to ~ 0.9 mL mol-1, which is considerably different from the theoretical calculated value. The difference between the theoretical and experi mental value shows th at the interactions between the solvent and the trans and cis mepepy forms are significantly different. Figure 3.3 shows more electron density on th e nitrogen of the pyridine in the trans form of the mepepy than the cis form. This observation can explain that a water molecule will have less probability to form an H-bond with the cis form than the trans form of the mepepy pyridine group. Van Eldik et al. [27] have demonstrated that a bond cleavage will show

PAGE 68

46 an expansion in molar volume of ~ 5mL mol-1 which is consistent with the expansion equal to VH-bond 9 mL mol-1, between the isomerization of the trans to cis mepepy molecule upon the H-bond cleavage. The variation of enthalpy was compared to the results from computational studies. The variation of volume between the cis form and the global energy minima from the trans form of the mepepy shows a difference of ~ 8 kcal mol-1. This value is significantly different from the experimental value which was calculated from PAC results equal to 18 kcal mol-1. The difference can also be explaine d by the cleavage of an H-bond between a water molecule and the mepepy pyridine group in terms of energy. Previous studies have shown that the energy of an hydrogen bond betw een a water molecule and a pyridine is equal to ~ 6-8 kcal mol-1 which is very close to the 10 kcal mol-1 (difference between the theoretical and experimental value) and confirms that th e variation of enthalpy is consistent with the cleavage of an H-bond be tween a water molecule and the nitrogen of the mepepy pyridine group upon the trans to cis isomerization [28-32] The thermodynamic profile is summarized in Figure 3.8 Figure 3.8 : Thermodynamic profile for the photoinduced isomerization of the trans to cis mepepy in aqueous solution.

PAGE 69

47 3.2.4. Conclusion To summarize the results presented for the trans to cis isomerization of the mepepy, the cis form exhibits one energy minimum to compare to the trans form which exhibits two energy minima separated by 9 kcal mol-1 and corresponds to two different conformations of the trans form of the mepepy. The different trans conformations might help to vary the photoinduced spin-state trans itions of the metal complexes coordinated to the mepepy ligand. Secondly, it has also be en demonstrated by computational studies that the dipole moments between the trans and the cis form of the mepepy are very similar which show a very restraint change in the electronic structure of mepepy to compare to azobenzene molecules or other push-pull molecules. Finally, the mepepy pyridine group can form an H-bond with the solvent which suggest that the thermodynamics of the trans to cis isomerization of the me pepy can be altered by the nature of the solvent. For instance, a solvent which can form an H-bond with the mepepy pyridine group will favor the trans form by up to 10 kcal mol-1.

PAGE 70

48 3.2.5. References [1] Bradamante S., Facchetti A. and Pagani G. A., J. Phys. Org. Chem. 10 (1997), 514524. [2] Isaksson G. and Sandstrom J., Acta Chim. Scand ., 27 (1973), 1183-1191. [3] Chemla D. S., Zyss J., Eds. Non-Linear Optical Properties of Organic Molecules and Crystals, Vols. 1 and 2; (1986) Academic Press: New York. [4] Cheng L. T., Tam W., Stevenson S. H., Me redith G. R., Rikken G. and Marder S. R., J. Phys. Chem. 95 (1991), 10631-10643. [5] Oudar J. L., J. Chem. Phys. 67 (1977), 446. [6]. Prasad P. N. and Williams D. J., Intr oduction to Nonlinear Optical Effects in Molecules and Polymers; (1994), John Wiley and Sons: New York,. [7] Quenneville J. and Martinez T. J., J. Phys. Chem. 107 (2003), 829-837. [8] Zarembowitch J., Roux C., Boillot M.-L., Claude R., Itie J.-P., Polian A. and Bolte M., Mol. Cryst. Liq. Cryst. 234 (1993), 247-254. [9] Decurtis S., Gutlich P., Kohler C. P., Spiering H. and Hauser A., Chem. Phys. Lett. 105 (1984), 1-4. [10] Boillot M.-L., Roux C ., Audiere J.-P., Dausse A. and Zarembowitch J., Inorg. Chem. 35 (1996), 3975-3980. [11] Sour A., Boillet M.-L., Riviere E. and Lesot P., Eur. J. Inorg. Chem. (1999), 21172119. [12] Boillot M.-L., Chantraine S., Zarembowitch J., Lallemand J.-Y. and Prunet J., New J. Chem. 23 (1999), 179-184.

PAGE 71

49 [13] Faulmann C., Dorbes S., Garreau de Bonneval B., Molnar G., Bousseksou A., Gomez-Garcia C. J., Coronado E. and Valade L., Eur. J. Inorg. Chem. (2005), 32613270. [14] Larsen R. W. and Miksovska J., Coord. Chem. Rev. 251 (2007), 1101-1127. [15] Schmidt M. W., Baldridge K. K., Boatz J. A., Elbert S. T., Gordon M. S., Jensen J. H., Koseki S., Matsunaga N., Nguyen K. A. and Su S., J. Comput. Chem. 14 (1993), 1347-1363. [16] Crecca C. R. and Roitberg A. E., J. Phys. Chem. 110 (2006), 8188-8203. [17] Cembran A., Bernardi F., Garave lli M., Gagliardi L. and Orlandi G., J. Am. Chem. Soc. 126 (2004), 3234-3243. [18] Tiago M. L., Ismail-Be igi S. and Louie S. G., J. Chem. Phys. 122 (2005), 94311. [19] Becke A. D., J. Chem. Phys. 98 (1993), 5648. [20] Stephens P., Devlin F., Chabalowski C. and Frisch M., J. Phys. Chem. 98 (1994), 11623-11627. [21] Hertwig R. and Koch W., Chem. Phys. Lett. 286 (1997), 345-351. [22] Dunning T. H., J. Chem. Phys. 90 (1989), 1007. [23] Semiat R., Leshinski E., Orell A., Cossi M., Barone V., Cammi R. and Tomasi J., Chem. Phys. Lett. 255 (1996), 327-335. [24] ArgusLab 4.0; Mark A. Thompson, Plan aria Software LLC: Seattle, WA, 2008, http://www.ArgusLab.com [25] Padova J., J. Chem. Phys. 39 (1963), 1552.

PAGE 72

50 [26] Wegewijs B., Paddon-Row M. N. and Braslavsky S. E., J. Phys. Chem. A 102 (1998), 8812-8818. [27] van Eldik R., Asano A. and le Noble W., J. Chem. Rev. 89 (1989), 549-688. [28] Papai I. and Jansco G., J. Phys. Chem. A 104 (2000), 2132-2137. [29] Zega A., Srcic S., Mavri J. and Bester-Rogac M., J. Mol. Struct. 875 (2008) 354363. [30] Marczak W., Lejmann J. K. and Heintz A., J. Chem. T 35 (2003), 269-278.

PAGE 73

51 3.3. Photothermal Studies of the Room Temperature Photoinduced Spin State Change in the Fe(III)(Salten)(Mepepy) Complex 3.3.1. Introduction The ability to modulate the sp in-state within a transition metal complex has enormous implications for the design of novel magnetic mate rials. Of key interest is the ability to switch between metal spin-states using either optical or thermal triggering at room temperature. A number of metal complexe s have now been synthesized utilizing isomerizable ‘push-pull’ ligands in which the field strength of the ligand can be modulated either thermally or through photonics. These complexes are of the FeII(L4)(X2) type or FeIII(L’4)(L)(X) type in which L is the pu sh-pull ligand. One such complex utilizes L = 4-sterylpyridine (Stpy) (1-phenyl -2-(4-pyridyl) ethane) as well as several phenyl derivatives to photo -induce the spin crossover [1-3] It has been shown that the FeII( trans Stpy)4(NCS)2 complex undergoes a thermally induced high-spin to low-spin transition centered near 190 K [1] Subsequent photoexcitation of this complex imbedded within a cellulose acetat e substrate results in a trans to cis isomerization of the Stpy which also induces the high-spin to low-sp in transition at 140K. The drive to produce transition metal complexes with photo-induced spin crossovers at high temperatures has led to the synthesis complexes such as [FeIII(salten)(mepepy)]BPh4] (salten = 4azaheptamethylene-1,7-bis(salicylideneiminate; mepepy= 1-(pyridin-4-yl)-2-(Nmethylpyrrol-2-yl)-ethene; BPh4 = tetraphenyl borate) (Figure 3.9) This complex has been shown to exhibit a photo-induced highspin to low-spin transition at room temperature [2-4]

PAGE 74

52 + BPh4 -O N NH Fe N N O N CH3 + BPh4 -N O N NH Fe N N O CH3 h (III) (III) + BPh4 -O N NH Fe N N O N CH3 + BPh4 -N O N NH Fe N N O CH3 h + BPh4 -O N NH Fe N N O N CH3 + BPh4 -N O N NH Fe N N O CH3 h h (III) (III) Figure 3.9: Structure of the Fe(III)(salten)(mepep y) complex before and after photolysis. Distances: a=2.035 b=1.977 c=1.965 d=1.888 e=1. 892 and f=2.022 [16] The mepepy ligand associated with the Fe(III)(salten)(mepepy) complex (Figure 3.10) is a push-pull type molecule that exhibits either a cis or trans conformation and contains an electron-withdrawing py ridyl group and an electron-donating Nb c a d e f + BPh4 -N O N NH Fe(III) N N O CH3

PAGE 75

53 methylpyrrole group, separated by a isomerizable double bond. Although the trans to cis photo-isomerization process is effective in modulating the ligand field strength due to changes in basicity much less is known about the electronic propert ies and energetics of this process. Figure 3.10: Structure of the Trans and Cis isomers of mepepy. Recently, our lab has utilized photoacoustic calorimetry (PAC) and computational methods to probe the energe tics associated with the trans to cis photo-isomerization of mepepy in aqueous solution. These studies revealed a volume change of 0.9 0.4 mL mol 1 and an enthalpy change of 18 3 kcal mol 1 which were attributed to the loss of a hydrogen bond between a solvent water mol ecule and the pyridyl ring of mepepy associated with the trans to cis isomerization [5] A number of studies have been reported in which the enthalpies of the spin crossover have been measured at low temperatures on powder samples or single crystal using di fferential scanning calorimetry (DSC). Collectively these studies indicate very small enthalpies (i.e., < 1 kcal mol-1) for the spin state transition that are coupled to minor st ructural changes. For example, the thermal induced spin-state transition in an Fe(II) tris(2-(2’-pyridyl)benzimidazole) complex which occurs with a transition temperature of 140K exhibits an enthalpy change of only 0.6 kcal mol-1 (low-spin to high-spin ) and a corresponding increas e in unit cell volume. More recently, thermal and photoinduced spin crossover in novel 3D metal organic trans mepepy cis mepepy

PAGE 76

54 materials was examined in both bulk materials an d thin films. These studies also revealed very small enthalpy changes ranging from ~ 1 to ~ 4 kcal mol-1 at temperatures below 300K [6-8] However, to date, no studies have been reported for the volume and enthalpy change of a spin crossover in solution at room temperature. Here photoacoustic calorimetry (PAC) has been used to study the energetics following th e spin crossover of the Fe(III)(salten)(mepepy) complex by resolv ing the magnitudes and timescales of molar volume and enthalpy changes associated wit h, first, the photoisomerization of the trans to cis conformers of the mepepy ligand and secondl y, the spin crossover of the iron in order to characterize the physical prop erties of the metal complex. 3.3.2. Materials and Methods Synthesis of Mepepy: The synthesis of the trans mepepy ligand has been previously reported by Mokdad et al. [5] Synthesis of the Fe(III)(salten)(mepepy)BH4 complex: Synthesis of the salten ligand: Under reflux, 11 mL salicylaldehyde was a dded to 7.5 mL 1,7-diamino-4-azaheptane in 50 mL of methanol. The solution was re fluxed for 10 minutes under constant stirring. A yellow/orange oil was collected and, after evaporation of the solvent, a yellow solid was obtained (Saltenptn). Synthesis of the Fe(III)(salten)Cl: Under reflux, a solution of 1.7 g of Saltenptn in 20 mL of methanol was stirred at 60C and was treated with a solution cont aining 0.63 g of iron chloride in 25 mL

PAGE 77

55 methanol. The solution was stirred for 10 minut es at 50C, and then for 15 minutes at 60C. To the yellow solution, 3 mL of triethylamine was added. The solution was then stirred for two hours at 60C. The solution was cooled to room temperature. The solution was filtrated and the crystals were washed with methanol and diethyl ether. Brown crystals of Fe(III)(sal ten)Cl were collected. Synthesis of the Fe(III)(salten)(mepepy) Complex: A solution of 77.3 mg of mepepy in 0.5 mL methanol was added to a solution of 90 mg Fe(III)(salten)Cl in 4 mL methanol. Under reflux, the solution was stirred at 60C for 4 hours. The solution was then placed in ice a nd filtered to remove insoluble materials. To the resulting solution was added a solution containing 71.9 mg of sodium tetraphenylborate in 1 mL methanol. The solu tion was stirred under reflux over night at 60C. The solution was then placed in ice, filt ered and the solid material was washed with 3-4 mL of cold methanol, ~ 15 mL of water and ~ 25 mL of ether. The solution was then dried with a vacuum for 6 hours. A brow n solid was collected. The purity for the compound was verified by mass spectra and UV where the UV spectrum matched the one published by Sour et al. [2] Quantum yield: The quantum yield, associated with the isomerization of the trans to cis mepepy was determined by irradiating a c uvette containing ~ 0.2 mM mepepy in both 50:1 and 6:1 (V:V) acetonitrile :water at 355 nm excitation (the wavelength used to perform the PAC measurements). The solution was irradiated for a total of 35 minutes, and the absorption spectra were recorded every 5 minutes. The same experiment was performed with ~ 0.8 mM azobenzene as a control. The value of was then determined using the following equation: ( A0 35 mep / A0 35 azo) = {( mep mep C0 mep) / azo azo C0 azo)} Equation 38

PAGE 78

56 where A0-35 mep and A0-35 azo are, respectively, the changes in absorbance at 355 nm for mepepy and azobenzene, mep is the quantum yield for the mepepy trans to cis isomerization, azo is the quantum yield for the trans to cis isomerization in azobenzene (0.26), mep (3,000 mM 1 cm 1) and azo (550 mM 1 cm 1) are the differences in extinction coefficients at 355 nm between, respectively, for the trans and cis isomers of mepepy and azobenzene, and C0 mep and C0 azo are the initial concentrations. Using this procedure mep was determined to be 0.28. The quantum yield associated with the corresponding Fe(III)(salte n)(mepepy) photo-induced spin st ate was determined using the same procedure with Fe(III)(salt)(mepepy) (22 mM 1 cm 1) giving a value of 0.40. Photoacoustic calorimetry method: The principles behind PAC have been revi ewed in detail in Chapter II. A stock solution of Fe(III)(salten)(mepepy) was prepared in acetonitrile (the Fe(III)(salten)(mepepy) complex is not so luble in water). Samples for PAC were prepared by diluting the stoc k solution from 85% to 99% acetonitrile:water with increment of 1 is used to vary the term of thermal expansion coefficient (see below). A calorimetric reference (Fe(3+)tetrakis( 4-sulphonatophenyl)porphyrin) = Fe4SP was prepared under the same so lution conditions as the sample. The absorbance of the Fe(III)(salten)(mepepy) complex and the refere nce Fe4SP at the exc itation wavelength of 355 nm were both ~ 0.4. The sample was stirred for all acquisitions and seven laser pulses were averaged per trace. The experiments were performed using a Continum MiniLite I frequency triplet Nd:YAG laser (355 nm, ~ 7 ns FWHM, ~ 100 J/pulse, 1 Hz) The acoustic waves were detected using a Panametrics V103 detector coupled to a Panametrics preamplifier and recorded using a PicoScope (50 MHz transient digitizer).

PAGE 79

57 Determination of Cp. / for Acetonitrile-Water Mixtures The Cp / term is temperature dependent in aqueous solution but, in organic solvents, the variations of the Cp / term with temperature is negligible. For mixed solvent systems (such as the acetonitrilewater solutions to be used here), Cp / can be determined by comparing acoustic amplitudes of a reference molecule solubilized in water to those obtained in the mixed solvent system. The mixed solvent Cp / values were determined by performing PAC measurements on ~ 10 M Fe4SP in acetonitrilewater mixtures used for the sample measurements. The value of Cp / for each ratio of acetonitrile:water was then determin ed using the following equation: (Cp / Fe4SP acetonitrile:water) = (Cp / Fe4SP water)* (SFe4SP acetonitrile:water / SFe4SP water) Equation 39 where SFe4SP acetonitrile:water and SFe4SP water are, respectively, the acoustic amplitudes for the reference in acetonitrile:water a nd neat water, respectively. 3.3.3. Results The steady-state absorption spectra of the Fe(III)(salten)(mepepy) complex and that of the trans form of mepepy solubilized in 85% ( volume percent) acetonitrile:water are displayed in Figure 3.11

PAGE 80

58 300350400450500550600 -0.05 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 AbsorbanceWavelength (nm) Mepepy Fe(III)(salten)( mepepy) Complex Figure 3.11 : Steady-state optical absorption spectra of the Fe(III)(salten)(mepepy) complex (orange dotted line) and the mepepy ligand (bleu solid line) in 85% (V/V) CH3CN:H2O Path length= 1cm. [mepepy]=[Fe(III)](salten)(mepepy)] = 9 M. The optical spectrum of the Fe(III)(salten )(mepepy) complex displays transitions centered at ~ 350 nm ( ~ 22 mM-1 cm-1) and at ~ 475 nm. The 350 nm transition corresponds to that of the me pepy ligand with the exception of significantly lower molar extinction coefficients (~ 2 mM-1cm-1 for the mepepy) while the 475 nm band likely arises from a charge transfer transiti on between the iron and the mepepy ligand. The ~350 nm absorption band of the mepepy lig and can be assigned primarily from a transition between the HOMO, with electron density centered primarily on the five membered ring, and LUMO orbitals with elec tron density distributed between the five and six membered rings (i.e., to transition) (Figure 3.11) [5] Upon steady state illumination of the Fe(III)(salten)(mepe py) complex, photoisomerization of the trans to cis form takes place within the mepepy ligand which results in a low to high-spin state change of the iron, giving rise to a hypsoc hromic shift (~10 nm) in the energy of the HOMO to LUMO transition as well as a cons iderable decrease in the molar extinction coefficient (Figure 3.12)

PAGE 81

59 300350400450500550600 0.0 0.2 0.4 0.6 0.8 1.0 1.2 AbsorbanceWavelength (nm) 0 min 5 mins 10 mins 15 mins 20 mins 30 mins 35 mins 40 mins 45 mins Figure 3.12 : Steady-state optical absorption spectru m of Fe(III)(salten)(mepepy) photolysis over time An overlay of the PAC signa ls associated with the trans to cis isomerization for the mepepy ligand, Fe(III)(salten)(mepepy) complex and the Fe4SP reference compound at 35C in 85% acetonitrile :water are displayed in Figures 3.13-3.15 The photoacoustic signals for the trans to cis isomerization of both mepepy and Fe(III)(salten)(mepepy) complex do not show any shifts in frequency re lative to the calorimetric reference. This indicates that the H and V take place within the respons e time of the PAC instrument, i.e., < ~50 ns. From a plot of Eh versus (Cp / ) (equation 9), the volume and enthalpy changes are determined to be V = 0.7 0.3 mL mol 1 and H = 33 9 kcal mol 1 for the trans to cis isomerization of the mepepy ligand (Figure 3.14) and V = 0.9 0.3 mL mol 1 and H = 37 9 kcal mol 1 for the mepepy isomerization and corresponding spinstate transition associated with th e Fe(III)(salten)(mepepy) complex (Figure 3.16)

PAGE 82

60 3.3.4. Discussion Mepepy ligand: The observed volume change associated with the trans to cis isomerization of mepepy in acetonitrile:water solutions may have contributions from changes in van der Waals volume of the mepepy molecule after is omerization as well as to electrostriction effects associated with changes in overall charge due to changes in the ground-state dipole moment. The volume change due to el ectrostriction can be estimated using the following equation [9,10] : Vel = -( 2/r3) T{( + 2)( – 1) / (2 + 1)2}(NA / 4 0) Equation 40 where is the change in dipole moment upon isomer ization, r is the cavity radius of the molecule, T is the compressibility of the solvent, is the solvent’s dielectric constant, NA is Avogadro’s number, and o is the vacuum permittivity. Using the previously calculated dipole change for the trans to cis isomerization of 0.0398 D [5] a molecular radius of ~5 (this assuming the mepepy molecule sweep s out a sphere of diameter ~10 ), T of 6.9 10 10 Pa 1 (85% V:V acetoni trile:water) or T of 7.5 10 10 Pa 1 (for 98% V:V acetonitrile:water) and = 35.97, the electrostriction volume change was calculated to be ~-0.29 mL mol 1 for 85% acetonitrile: water and ~-0.32 mL mol 1 for 98% acetonitrile:water. An average of the volume cha nge due to electrostriction is estimated to be ~-0.3 mL mol 1 over the solvent range used here (85% to 99% acetonitrile). The corresponding van der Waals vol umes change between the trans and cis isomers was previously estimated to be ~9 mL mol-1[5] Giving a total V (i.e., van der Waals volume changes and the electrostricti on change) of ~ -9.3 mL mol 1 which is significantly lower than the observed V = 0.7 mL mol-1 indicating additional cont ributions to the volume change.

PAGE 83

61 Figure 3.13: Overlay of the normalized acoustic wave of the reference Fe4SP and mepepy in 85% (V/V) CH3CN:H2O at 35C. The absorbance of the sample and reference at the excitation wavelength (355 nm) was ~0.4. 34363840424420 30 40 50 60 70 80 V = 0.7 0.3 mLmol-1H = 33 10 kcal.mol-1Eh (kcalmol-1)Cp / (kcal.mL-1) Figure 3.14 : Plot of Eh vs. Cp/ for mepepy photoisomerization. The slope of these lines provides that change in molar volume ( V = slope/ in mL mol-1) and the intercept provides the enthalpy ( H = (Eh Q)/ in kcal mol-1). 2000 4000 6000 -150000 -100000 -50000 0 50000 100000 150000 AmplitudeTime (ns) Fe4SP Mepepy Complex

PAGE 84

62 The Vobs for the trans to cis isomerization in acetoni trile:water mixtures is similar to the volume change previously observed for the trans to cis isomerization of mepepy in neat water ( V ~ 0.9 mL mol-1) [5] The change in molar volume in this case was rationalized by the loss of a hydrogen bond between a solvent water molecule and the N atom associated with the pyr idine of the mepepy. In water, the trans form of mepepy H-bonded exhibits an absorbance at 425 nm which is not present in basic solutions. This band arises from a HOMO to LUMO transition for a mepepy molecule containing a hydrogen bonded/ protonated N atom on the pyridyl unit. Upon photoisomerization to the cis isomer the electron density on the N atom is reduced resulting in a loss of the H-bond between a water molecule and the mepepy pyridyl group. Van Eldik et al. have suggested that on average bond cleavage reactions increase the molar volume of ~ 5-10 mL mol 1 which together with the electrostriction volume and van der Waals volume changes acc ounts for the observed volume change [11] Thus, it is apparent that trans conformation of mepepy solubili zed in acetonitrile:water retains the protonation/hydrogen bonding to the pyridyl N at om and that this interaction is lost upon isomerization to the cis conformation. The corresponding changes in enthalpy result from the energy difference between the cis and trans form of mepepy as well as enthalpy changes associated with solvation including H-bond formation/cleavage. Previous computational results have demonstrated an energy difference between the cis and the global trans isomers of ~ 8 kcal mol 1 [5] The H obtained from PAC measurements fo r the transition between the global trans and cis conformations in acetonitr ile:water is 33 kcal mol-1 and includes the enthalpy of the isomerization, H-bond cleavage and any a dditional solvation effects (i.e., HPAC = Hisom+ HH-bond + Hsolv). Taking into account the enthal py difference between the two mepepy conformers, the enthalpy associat ed with solvation changes and H-bond formation/cleavage is ~25 kcal mol-1. Previous studies have de monstrated that the energy of an H-bond between water and pyrid ine is close to 6-8 kcal mol 1 [12-14] In contrast, Gilli et al. studies suggest a positive charge-assisted H-bond ((+)CAHB) including [N…H…N]+ have an average H-bond energy of ~16.4 kcal mol-1 while an [N…H…O]+ interaction has an average H-bond energy of ~15.2 kcal mol-1 [15] Thus, the remaining

PAGE 85

63 25 kcal mol-1 observed for the mepepy isomerization may be attributed (within the 9 kcal mol-1 uncertainty) to the loss of an (+)CAHB between the mepepy pyridyl group and a water molecule which, in turn, is Hbonded at an acetonitrile solvent molecule. Fe(III)Salten Mepepy Complex : The PAC results indicate that the mepepy trans to cis isomerization and associated spin crossover of the Fe(III)(salten )(mepepy) complex also occurs in < ~ 50 ns as evident from the lack of frequency shift between the PAC signals of the complex and the calorimetric reference. From a plot of Eh versus ( Cp / ) (Figure 3.16) the volume and enthalpy changes are determined to be V = 0.9 0.3 mL mol 1 (1.5 3 / molecule) and H = 37 10 kcal mol 1. The observed volume change has two contributions from the trans to cis isomerization of the mepepy as well as the spin crossover of the center iron ion. In order to extract th e volume change associated with the spin crossover, the volume change associated with the mepepy is omerization must be subtracted from the overall volume change. However, as the N-at om of the pyridine group is coordinated to the iron atom (precluding H-bonding interactions with the solvents only the previously estimated van der Waals and electrostriction contributions are included i.e.,~ -9.3 mL mol 1 Thus, the volume change of the iron spin is then estimated to be ~ 10.2 mL mol-1. Interestingly, high pressure studies of several Ni and Fe ligand complexes have suggested that volume change associated with a low-spin to high-spin transition in is on the order of ~ 10 mL mol 1 [11] This value was attributed to expansion of the metalligand core as well as elect ron repulsion between the 3dz2.orbitals and ligand orbitals. Since the Fe(III)(salten)(mepepy) complex exam ined here undergoes a high-spin to lowspin a V of ~ -10 mL mol 1 would be expected The di fference observed between the previously reported metal complexes and th e Fe(III)(salten)(mepepy) may be due to significant distortion of the complex following the central iron spin state change. The average of the Fe…O and Fe…N bond distances found in th e Cambridge structural database for the Fe(III)(salten)(mepepy) complex are indicated in Figure 3.9 [16] The bond distances between the Fe…O and Fe…N demonstrate that the axial Fe…N bond to the

PAGE 86

64 secondary amine has a higher average bond le ngth indicating a lowe r bond energy. Thus, one possibility for the difference in the observed V associated with spin state changes in metal complexes and that of the Fe(III)(sal ten)(mepepy) complex may be the loss of coordination between the central iron ion and th e N atom associate with the salten amine group. Pixton et al. calculated a V ~24 mL mol 1 for the cleavage of a Fe…N bond between a histidine and heme [17] while a V ~20 mL mol 1 was observed by Laverman et al., for the cleavage of the Fe-NO bond in NO-MetMyoglobin [18] The release of the salten N-group would also result in solvent hydrogen bonding in the mixed acetonitrilewater system. The volume change for hydrogen bond formation has been estimated to be on the order of -10 mL mol 1 [20] Considering the Fe-N bond cleavage step the overall Vobs = VFe-N + VN-H + VHS-LS (after subtracting Vmepepy) which, after rearranging, gives VHS-LS = Vobs VFe-N VN-H = 10 mL mol-1 – 20 mL mol-1 + 10 mL mol-1 ~ 0 mL mol-1. Thus, the overall process follo wing the photoexcitation of the Fe(III)(salten)(mepepy) complex can be summarized by the following processes: 1 ) trans to cis isomerization in the mepepy ligand, 2) cleavage of the Fe…N bond, 3) formation of hydrogen bond between a water molecule and the secondary amine and 3) spin-state change of the iron. The lack of any significant volume cha nge due to the spin state transition is consistent with the fact that, ot her than the loss of th e Fe-N coordination, the iron retains all of the other ligands precl uding any electronic repul sion between the Fe 3dz2 orbital and solvent molecules. This hypothe sis is also consiste nt with volumetric studies of the volume change of spin crossove r within the crystal forms of various iron complexes. These results show an av erage of volume change between 2-3 3 which correspond to a volume change of only 1-2 mL mol-1[20,21]

PAGE 87

65 -20000200040006000800010000 -40000 -30000 -20000 -10000 0 10000 20000 30000 40000 AmplitudeTime (ns) Reference Fe4SP Salten Mepepy Complex Figure 3.15: Overlay of the normalized acoustic wave of the reference Fe4SP and the Fe(III)(salten)(mepepy) complex in 85% (V/V) CH3CN:H2O at 35 C. The absorbance of the sample and reference at the excitation wavelength (355 nm) was ~0.4. 3234363840424440 50 60 70 80 90 V = 0.9 0.3 mL.mol-1H = 37 10 kcal.mol-1C p (mL.mol-1)E h (kcal.mol-1) Figure 3.16: Plot of Eh vs. Cp / for Fe(III)(salten)(mepepy) complex photoisomerization. The slope of these lines provides that change in molar volume ( V = slope/ in mL mol-1) and the intercept provides the enthalpy ( H = (Eh Q)/ in kcal mol-1).

PAGE 88

66 The corresponding enthalpy change has contributions from the trans to cis isomerization of the mepepy ligand as well as the spin state change of the iron and any changes in core conformation, Hobs = Hmepepy + HFe-N + HN-H + HHS-LS. The theoretical enthalpy value for the trans to cis isomerization of the mepepy ligand was calculated to be ~8 kcal mol 1 [5] Subtracting this value fr om the observed enthalpy gives ~ 29 kcal mol-1 for the subsequent Fe-N bond cleavage, hydrogen bond formation and corresponding spin state change (i.e., Hobs Hmepepy = HFe-N + HN-H + HHS-LS). Previous studies have further demonstrat ed that hydrogen bonds between water and pyridine have a bond energy on the order of ~6-8 kcal mol 1 [12-14] while those between water and dimethylamine are on the order of 5-6 kcal mol 1 [23] By taking into account the formation of H-bond between a wate r molecule and the secondary amine (exothermic), the remaining difference between the theoretical and experimental values is equal to ~ 35 kcal mol-1 (i.e., 35 kcal mol-1 = Hobs Hmepepy HN-H = HFe-N + HHSLS). Finally, the enthalpy of dissociation of an Fe…N bond from Fe-NH3 type complexes is on the order of ~ 30 kcal mol-1 [27] which results in a HHS-LS of ~ 5 kcal mol-1 which is negligible when consider the experimental uncertainty.

PAGE 89

67 3.3.5. Conclusion In summary the results presented here de monstrate several key features of the trans to cis isomerization of mepepy and photoinitiated spin crossover of the Fe(III)(salten)(mepepy) complex. First, the mepepy ligand exhibits a volume change equal to V = ~ 0.7 0.3 mL mol 1 which is associated with the loss of H-bond between a water and the mepepy pyridyl group. The enthalpy change, H = ~ 33 10 kcal mol 1, is associated to the cleavage of a positiv e charge-assisted H-b ond between the water and/or acetonitrile and the pr otonated mepepy pyridinium group in addition to the loss of H-bond between a water and mepepy pyridyl group (Figure 3.17) Subsequent PAC studies of the Fe(III)(salten)(mepepy) complex suggest that the volume ( V = ~ 0.9 0.3 mL mol 1) and enthalpy ( H = ~ 37 10 kcal mol 1) changes are consistent with a mechanism in which the Fe(III)(salten)(mepepy) complex undergoes a trans to cis isomerization of the bound mepepy liga nd, followed by the cleavage of an Fe…N bond associated with the salten ligand and high-spin to low-spin transition on the central iron (Figure 3.18) The secondary amine of the salt en ligand forms hydrogen bond with a solvent water molecule subsequent to release from the iron coordination.

PAGE 90

68 Figure 3.17 : Thermodynamic profile for the photoinduced isomerization of the trans to cis mepepy in acetonitrile:water solution. Figure 3.18 : Thermodynamic profile for the photoindu ced Fe(III) salten mepepy complex in acetonitrile:water solution. + BPh4 -O N N H Fe(III) N N O N CH3 N N Fe O O N NH N BPh4

PAGE 91

69 3.3.6. References [1] Boillot, M.-L.; Roux, C.; Audiere, J.-P.; Dausse, A.; Zarembowitch, J. Inorg. Chem. 35 (1996), 3975-3980. [2] Sour A., Boillet M.-L., Riviere E. and Lesot P., Eur. J. Inorg. Chem., 12 (1999), 2117-2119. [3] Boillot, M.-L.; Chantraine, S.; Zarembowitch, J.; Lallemand, J.-Y.; Prunet, J. New J. Chem. 23 (1999), 179-184. [4] Faulmann, C.; Dorbes, S.; Garreau de Bonneval, B.; Molnar, G.; Bousseksou, A.; Gomez-Garcia, C. J.; Coronado, E.; Valade, L. Eur. J. Inorg. Chem., 16 (2005), 32613270. [5] Mokdad A., Belof J. L., Wook Yi S., Shul er S. E., Mc Laughlin M. L., Space B. and Larsen R. W., J. Phys Chem. A, 112 (2008), 8310–8315. [6] Boca R., Boca M., Ehrenberg H., Fuess H., Linert W., Renz F. and Svoboda I., Chemical Physics 293 (2003), 375-395. [7] Rodriguez-Velamazan J. A., Castro M., Pa lacios E., Burriel R., Sanchez Costa J. and Letard J. F., Chem. Phys. Lett. 435 (2007), 358-363. [8] Agusti G., Cobo S., Gaspar A. B., Molnar G., Moussa N. O., Szilagyi P. A., Palfi V., Vieu C., Munoz M. C., Real J. A. and Bousseksou A., Chem. Mater., 20 (2008), 67216732. [9] Padova J., J. Chem. Phys 39 (1963), 1552-1557. [10] Wegewijs B., Paddon-Row M. N. and Braslavsky S. E., J. Phys. Chem. A 102 (1998) 8812-8818. [11] van Eldik R., Asano A. and le Noble W. J., Chem. Rev. 89 (1989), 549-688.

PAGE 92

70 [12] Papai I. and Jansco G., J. Phys. Chem. A 104 (2000), 2132-2137. [13] Zega A., Srcic S., Ma vri J. and Bester-Rogac M., J. Mol. Struct 875 (2008) 354363. [14] Marczak W., Lejmann J. K. and Heintz A., J. Chem. Therm. 35 (2003), 269-278. G Miksovska J., Suquet C., Satte rlee J.D., and Larsen R.W., Biochemistry 44 (2005), 10028-10036. [15] Gilli P., Pretto L., Bertolasi V. and Gilli Gastone, Acc. Chem. Res. 42 (2009), 33-44. [16] Allen F. H., Acta Crystallogr Sect B, 58 (2002), 380-388. [17] Pixton D. A., Petersen C. A., Franke A., van Eldik R., Garton E. M. and Andrew C. R., J. Am. Chem. Soc 131 (2009), 4846-4853. [18] Laverman, L. E., Wanat, A., Oszajca, J., Stochel, G., Ford, P. C. and van Eldik, R. J. Am. Chem. Soc. 123 (2001) 285–293. [19] Franke, A., Stochel, G ., Jung, C. and van Eldik, R. J. Am. Chem. Soc. 126 (2004), 4181–4191. [20] Oswal S. L., Thermochimica Acta 425 (2005), 59-68. [21] Guionneau P., Marchivie M., Brav ic G., Letard J-F and Chasseau D, J. Mater. Chem. 12 (2002), 2546-2551. [22] Reger D. L., Gardinier J. R., Elgin J. D., Smith M. D., Hautot D., Long G. J. and Grandjean F., Inorg. Chem. 45 (2006), 8862-8875. [23] Brink G. and Glasser L., Journal of Computational Chemistry 3, (1982), 47-52. [24] Riordan C. G. and Halpern J., Inorganica Chimica Acta 243 (1996), 19-24.

PAGE 93

71 [25] Kuramshin A. I., Kuramshina E. A. and Cherkasov R. A., Russian Journal of Organic Chemistry 40 (2004), 1265-1273. [26] Traylor T.G., Acc. Chem. Res. 14 (1981), 102-109. [27] Yu C., Hartmann M., Frenking G. Z., Anorg. Allg. Chem. 627 (2001), 985-998.

PAGE 94

72 3.3. Time Resolved Photoacoustic Calo rimetry and Debye-Hckel Theory: Determining Electrostriction A ssociated with Excited State Ru(II)(L)3 Complexes 3.3.1. Introduction Thermodynamic and kinetic profiles are esse ntial to understand reaction mechanisms associated chemical and biological processe s. Reactions of molecules in solution are accompanied by changes, in bond length, charge s, solvent coordination, etc… which will be reflected by a variation of the volume of the solvent. One of the challenges is to interpret the origin of the ch ange by isolating the individual contributions to the reaction. The volume change of a molecule associated with a chemical reaction can be expressed as the sum of the change of the van der Waal s radii and the change in solvation as the sum of electrostriction according to: V = Vintr + Vsolv Equation 41 where Vintr represents changes in bond lengths a nd angles between the reactants and the products which can be approximated by ch anges in the van der Waals volume and Vsolv which considers volume changes associated with electrostriction due to changes in overallcharge between the reactants and products [1] To calculate V, quantitative structure activity relationship (QSAR) can be used for the estimation of Vintr while computational methods including molecular dynamics (MD) simulations and quantum mechanical calculations can be used to estimate Vsolv [2,3] Molecular volume changes can be determin ed using MD experiments to establish the thermodynamic volume of solvated molecules ( Vsolv) along with time-dependent volume changes in the condensed phase [4,5] Principally MD simulations calculate

PAGE 95

73 isothermal-isobaric data for the solvent-solute system for a specific length of time to determine the volume of the system. The solu te molecule is then removed from the solvent and the volume is then recalculated after equilibrium is achieved. The difference of these two volumes calculated from the in itial and final states of the molecule corresponds to the thermodynamic volume ( Vsolv) of a solute. Unfortunately, MD simulations become more expensive with larg er molecules or proteins as the time to perform the MD simmulation increases with the size of the molecule. Experimental, Vintra and Vsolv can be obtained by examing reaction volume changes as a function of ionic strength and using the Debye-Hckel equation. The Debye-Hckel equation, originated from Peter Debye and Erich Hckel [6] expresses the activity coefficients of ions in solutions. Different interactions between ions show discrepancies even at very low con centration. These discre pancies between ideal solutions and dilute solution containing electrolytes are often caused by the interactions occurring between different ions. The Debye-Hckel equation calculates the activity coefficients which are proportional to the concen tration of the electrol ytes and include the energy of interaction between ions in solution. The Debye-Hckel equation is summarized below: Vi = Vi + (zi 2 Av / 2b) ln(1 + b I) Equation 42 where Vi is the apparent partial molar volume of the ionic species expressed as a function of the actual ionic streng th I of the solution, Vi is the partial molar volume at infinite dilution, zi is the ion charge, b = 1.2(kg/mol)1/2 and Av = 1.874 cm3 kg1/2 mol-3/2 at 25 C [7] In order to v erify if the solvent exclud ed volume can be experimentally determine using the Debye-Hckel equati on, the total volume change of a complex or molecule determined using PAC will be plotted versus the ionic strength of the solution. The partial molar volume at infinite dilution (Vi) and the ion charge (zi) will be extracted. This hypothesis will be ve rified by using two ruthenium complexes, Tris-(2,2'bipyridine)-ruthenium(II) (Ru(bpy)3) and Tris-(1,10'-phenanthroline)-ruthenium(II)

PAGE 96

74 (Ru(phen)3), as the photochemical and photophysical properties of these two complexes have been extensively investigated [8-12] Ru(bpy)3 and Ru(phen)3 complexes were of intere st since they are known to absorb light in the visible region with a hi gh intersystem crossing yield. They are also soluble in water and can include a functiona l group on the ligand th at can be protonated to induce a proton transfer in the excited state [13] After photoexitation of the ruthenium complexes, a charge is delocalized from the metal to the ligand (bipyridine or phenanthroline) via a metal to ligand charge transfer (MLCT) with a lifetime between a few picoseconds to a few nanoseconds followe d by intersystem crossi ng to the triplet state will follow and a slow return to th e ground state with a lifetime around 600 ns for Ru(bpy)3 and 900 ns for Ru(phen)3 [14-16] As the charge transfer between the metal and the ligand is reversible, the results ( Vi and zi) for the Debye-Hckel equation between the fast and slow phases can be compared. In addition, the two constants Av and b from the Debye-Hckel equation will also be verified as they are dependent of the so lvent and the ionic stre ngth of the solution. This can be accomplished since the ch arge that is transferred is know.

PAGE 97

75 3.3.2. Materials and Methods Sample Preparation: Ru(bpy)3 and Ru(phen)3 were purchased from Sigma-Aldrich and Fe3+tetrakis(4sulfonatophenyl porphine (Fe(III)4SP) was obtained from Frontier-S cientific, Inc. The samples of Ru(bpy)3 and Ru(phen)3 were solubilized in e ither 10mM, 50mM, 100mM, 150mM, 200mM, 250mM, 500mM, 750 mM, 1M, 1.5M and 2M NaCl(aq) at pH 7.0. The samples for PAC studies were placed in a 1cm quartz cuvette and sealed with a septum cap and subsequently purged for se veral minutes with argon. Optical spectra of the two ruthenium complexes were obta ined using a Shimadzu UV-2401PC UV-Vis spectrophotometer. 3.3.3. Results The UV-Vis spectra of Ru(bpy)3 and Ru(phen)3 with different concentrations of NaCl(aq) are similar with peaks maxima at 425 nm and 454 nm for Ru(bpy)3 and at 421 nm and 448 nm for Ru(phen)3. Thus, high concentrations of NaCl(aq) do not have a significant impact on the electron ic structure of the Ru(bpy)3 and Ru(phen)3. An overlay of PAC traces and fits for both Ru(bpy)3 and Ru(phen)3 and the calorimetric reference co mpound are displayed in Figure 3.19 The graphs show a frequently shift between sample and refere nce acoustics signals wh ich indicates other kinetic events occurring between ~50 ns and ~20 s in addition to events occurring in < 50 ns. In order to calculate the solvent excluded volume of both Ru(bpy)3 and Ru(phen)3, different concentrations of NaCl(aq) (10mM, 50mM, 100mM, 150mM, 200mM, 250mM, 500mM, 750mM, 1M, 1.5M and 2M) at pH=7 were compared with the purpose of probing electrostatic interact ions upon excitation. The Vs calculated for each concentration, after deconvolution of the PA C traces, plotted as a function of ionic strength and fit to th e Debye-Hckel equation (Scheme 3.5) to extract Vi (partial molar volume at infinite dilution) and zi (the ion charge).

PAGE 98

76 The variations in volume and enthalpy in th e different buffers after excitation for both Ru(bpy)3 and Ru(phen)3 are very similar. After deconvolution of the frenquently shift between sample and reference acoustic signals a biphasic relaxation is observed for both Ru(bpy)3 and Ru(phen)3. The biphasic relaxation is attributed to a prompt phase <50 ns subsequent to the excitation and a slower phase with ~ 550 ns for Ru(bpy) 3 and ~900 ns for Ru(phen)3. The H and V associated with these proce sses are obtained after plotting (S/R)Eh versus Cp / (as per equation 9) (Figure 3.21) The average of these results between the different co ncentrations of NaCl(aq) are summarized in Table 3.3 for Ru(bpy)3 and Table 3.5 for Ru(phen)3. The processes occurring in <50 ns, between the different concentrations (10mM to 2M) of NaCl(aq), reveal a H between 43 4 kcal mol1 and 59 1 kcal mol-1 and a V between -2.2 0.1 mL mol-1 and -0.5 0.1 mL mol-1 for Ru(bpy)3 (Table 3.3) The prompt phase for Ru(phen)3 gives similar results with a H between 44 1 kcal mol-1 and 53 0.3 kcal mol-1 and a V between -1.8 0.1 mL mol-1 and -0.5 0.03 mL mol-1 (Table 3.5) The slow phases give also similar results between Ru(bpy)3 and Ru(phen)3.with a H between -43 2 kcal mol-1 and -55 3 kcal mol-1 and a V between 1.6 0.2 mL mol-1 and 0.2 0.1 mL mol-1 for Ru(bpy)3 (Table 3.3) and a H between -43 3 kcal mol-1 and -54 4 kcal mol-1 and V between 2.1 0.2 mL mol1 and 0.6 0.05 mL mol-1 for Ru(phen)3 (Table 3.5) These results also demonstrate that the variation of solution ionic st rength has a slight affect the V and H of the charge transfer between the metal and the ligand. After fitting the volume data to the Debye-Hckel equation, Vi and zi can be extracted for Ru(bpy)3 (Table 3.4) and Ru(phen)3 (Table 3.6) The results show similarity between Ru(bpy)3 and Ru(phen)3, with, respectively, Vi = 2.1/ -1.9 mL mol-1 for the prompt phase and 1.7 / 1.9 mL mol-1 for the slow phase. zi also shows similarity between Ru(bpy)3 and Ru(phen)3, with respectively, 1.5 / 1.5 in the prompt phase and 1.4 / 1.3 in the slow phase.

PAGE 99

77 3.3.4. Discussion The PAC results which take into acc ount the quantum yield (0.96) for the nonradiative decay of the metal ligand char ge transfer (MLCT) states for Ru(bpy)3 and Ru(phen)3, show a biphasic relaxation, subsequent to excitation. The summarized results for Ru(bpy)3 gives, for the prompt phase, a H between 43 4 kcal mol-1 and 59 1 kcal mol-1 associated with a V between -2.2 0.1 mL mol-1 and -0.5 0.1 mL mol-1 for the different concentrations (10mM to 2M) of NaCl(aq) at pH = 7. The slow phase (~ 550 ns) gives a H between -43 2 kcal mol-1 and -55 3 kcal mol-1 and a V between 1.6 0.2 mL mol-1 and 0.2 0.1 mL mol-1 (Table 3.3) These results can be compared with the V and H results already known a nd published for the Ru(bpy)3 in water [17] Miksovska and Larsen have shown that the V and H associated with the formation of the MLCT state for Ru(bpy)3 at pH = 6 is equal to -3.5 0.5 mL mol-1 and 44 2 kcal mol-1 and the subsequent decay is equal to 3.2 0.6 mL mol-1 and -50 3 kcal mol-1, respectively. These results are also similar to the results reported by Borsarelli [18] and Goodman [19] as well as Cherry [20] who calculated the variation of volume and enthalpy of Ru(bpy)3 from the emission maxima.

PAGE 100

78 200030004000 -40000 -30000 -20000 -10000 0 10000 20000 30000 40000 AmplitudeTime (ns) Fe4SP Ru(BPY)3 Fit Residue Figure 3.19 : Overlay of the acoustic waves of the Ru(bpy)3 (bleu dashed line), the reference Fe(III)4SP (black solid line), the fit (green do tted line) and the residue (red dashed and fit line) Ru(bpy)3 in 10mM NaCl(aq) (pH 7.0) at 34C.

PAGE 101

79 Figure 3.20 : Overlay of the acoustic waves of the Ru(phen)3 (purple dashed line), the reference Fe(III)4SP (black solid line), the fit (green dotted line) and the residue (red dashed and fit line) Ru(phen)3 in 10mM NaCl(aq) (pH 7.0) at 34C. 200030004000 -200000 -100000 0 100000 200000 300000 AmplitudeTime (s) Fe4SP Ru(phen)3 Fit Residue

PAGE 102

80 Table 3.3 : Variations of the volume and enthalpy for Ru(bpy)3 for different concentrations of NaCl(aq) (between 10mM to 2M) at pH=7. Conc (mM) 1 (ns) V1 (mL mol-1) H1 (kcal mol-1) 2 (ns) V2 (mL mol-1) H2 (kcal mol-1) 10mM 1 -1.9 0.4 43 4 540 1.4 0.2 -43 2 50mM 1 -2.2 0.1 53 1 350 1.1 0.2 -50 2 100mM 1 -1.7 0.1 54 0.1 460 1.3 0.2 -55 2 150mM 1 -1.3 0.1 50 0.9 400 1.6 0.2 -45 3 200mM 1 -1.2 0.2 52 0.2 390 1.2 0.2 -52 3 250mM 1 -1.2 0.2 56 0.1 450 0.8 0.1 -48 1 500mM 1 -0.7 0.1 57 1 470 0.6 0.3 -55 3 750mM 1 -1.1 0.2 55 2 640 0.5 0.4 -54 7 1M 1 -0.6 0.1 59 1 690 0.3 0.2 -55 3 1.5M 1 -0.5 0.1 58 0.2 670 0.2 0.1 -55 2 2M 1 -0.5 0.1 51 0.9 770 0.2 0.1 -49 2

PAGE 103

81 The V of Ru(bpy)3 increase with ioni c strength of NaCl(aq) (Table 3.3) from -1.9 0.4 mL mol-1 to -0.5 0.1 mL mol-1 due impart to solvent electrostriction where the charge delocalizes from the metal to ligands after excitation. The results are in accord with previous studies of the forma tion of the MLCT state of the Ru(bpy)3 [19, 21] The negative charge after excita tion is delocalized over the three bipyridines around the ruthenium ion. Goodman and Herman demonstrated that the small contraction is due to a shortening of ~ 0.01 of the metal-ligand bond [19] Borsarelli et al. show no dependence between the salt composition of th e solution and the volume change of the MLCT formation. They confirm an internal rearrangement of the ruthenium complex after excitation [22] In our studies, the same salt NaCl was chosen and the variation of the volume change of the MLCT formation was studied as a function of the solution ionic strength. It is not clear why increasing the io nic strength will give rise to an expansion. One explanation is that at low ionic streng th, only a few positive charged atoms will have an electrostriction inte raction around the Ru(bpy)3 complex. Increasing the concentration of NaCl(aq), produces more Na+ ions which will have a greate r electrostriction interaction around the ruthenium complex thus increasing th e contraction. The fact that the charge transfer between the ruthenium and the ligand is a partial charge tran sfer and not a total charge transfer might be the cause of the expansion and not the expected contraction observed when the ionic strength increases,. For the slow phase, the reverse phenome non is observed. The slow phase exhibits a V between 1.6 0.2 mL mol-1 and 0.2 0.1 mL mol-1 (Table 3.3) which is equal and opposite to the V observed in the prompt phase. As explained earlier for the prompt phase, the larger the solution ionic strength, the less change in volume is observed. When the Ru(bpy)3 complex relaxes to the ground state th rough a nonradiative deactivation, the metal-ligand bond will increase and the overall negative charge on the ligands recombine with the Ru(III) ion rela xing the electrostriction. Similarly, H which is between 43 4 kcal mol-1 and 59 1 kcal mol-1, is equal and opposite to the prompt phase, and shows a small dependence between the H and the ionic strength of NaCl(aq).

PAGE 104

82 A plot of V versus the ionic strength of the solution can be plotted in the DebyeHckel equation and both Vi and zi can be extracted: Vi = 2.1 mL mol-1 for the prompt phase and 1.7 mL mol-1 for the slow phase and zi = 1.5 and 1.4 respectively for the prompt and slow phase (Figure 3.21) The ion charge (zi) of Ru(bpy)3 which was calculated from the Debye-Hckel equa tion is equal to 1.5 0.2 mL mol-1 for the prompt phase which is explained by a charge deloca lization from the ruthenium atom to the bipyridine ligands (MLCT). An ion charge equal to 1.4 0.2 mL mol-1 was calculated for the slow phase which correspond to the relaxati on to the triplet st ate to the ground state through a nonradiative deact ivation with a lifetime of ~600 ns for Ru(bpy)3 as demonstrated previously [14,15] Within experimental error, the results for zi can be considered close to 1 indicating that there is one charge transfer occurring in the prompt and slow phase. The results obtained for the excluded volume change are Vi = 2.1 mL mol-1 for the prompt phase and 1.7 mL mol-1 for the slow phase. The results are summarized in Table 3.4

PAGE 105

83 0.00.20.40.60.81.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 V (mL mol-1)ln(1+bVI) 1st Phase 2nd PhaseRu(bpy)3V1 = -2.1 0.1 mL mol-1Z1 = 1.5 0.2 V2 = 1.7 0.1 mL mol-1Z2 = 1.4 0.2 Figure 3.21 : Plot of V versus ln(1+b I) for the 1st phase (dark bleu square) and 2nd phase (green dot) from Ru(bpy)3 between 10mM and 2M NaCl(aq) (pH 7.0) which gives an intercept equal to the partial molar volume at infinite dilution Vi (mL mol-1) and the slope is equal to (zi 2 Av / 2b), which will give the ion charge zi. Table 3.4 : Summary of Vi and zi for Ru(bpy)3 from fits to the Debye-Hckel equation. V (mL mol-1) z 1st phase -2.1 0.1 1.5 0.2 2nd phase 1.7 0.1 1.4 0.2 The experimental electrostriction vol ume change can be extracted from Scheme 3.4 as the solvent excluded volume changes were experimentally determined to be VVDW = 2.1 mL mol-1 for the prompt phase and 1.7 mL mol-1 for the slow phase, and the total volume change was calculated between -1.9 mL mol-1 and 1.4 mL mol-1. The volume change due to electrostriction is then Vel Ru(bpy)3 = -1.9 + 2.1 = 0.2 mL mol-1 for

PAGE 106

84 the prompt phase and Vel Ru(bpy)3 = 1.4 – 1.7 = -0.3 mL mol-1 for the slow phase at 10mM NaCl and Vel Ru(bpy)3 = -0.5 + 2.1 = 1.6 mL mol-1 for the prompt phase and Vel Ru(bpy)3 = 0.2 – 1.7= -1.5 mL mol-1 for the slow phase at 2M NaCl. The calculation for the experimental electrostricti on change are opposite in sign and similar in magnitude between the fast and the slow phases for the Ru(bpy)3. The experimental electrostr iction change can be compared to a theoretical value using the Drude-Nernst equation [23] The Drude-Nernst eq uation describes the contraction of elect rolytes around an ion [24] Using the classical Drude-Nernst equation for electrostriction [25] : Vel = -Bz2 / r Equation 43 where B is a constant with a calcula ted value in water of 4.175 mL mol-1, r is the radius expressed in Angstrms and z is the char ge, the electrostrictio n volume change for Ru(bpy)3 and Ru(phen)3 can be calculated [26] The Drude-Nernst equation gives for Ru(bpy)3 where r = 4.8 [27] a Vel = -3.5mL mol-1. The theoretical calculation of the change in electrostriction shows a higher va lue than the experimental value. This discrepancy which is higher at low ionic strength may be due to solvent effects associated with an increase in the concentration of Na+ and Clions that are not account in the theoretical value. At low ionic strength, the change in overall volume is mostly due to a van der Waals volume change as the me tal-ligand bond shortening and only a few Na+ atoms will interact with the overall negative charge of the ruthenium complex as explained earlier.

PAGE 107

85 The results for the Ru(phen)3 are similar to those obtained for Ru(bpy)3 with a prompt phase associated with a H between 44 1 kcal mol-1 and 53 0.3 kcal mol-1 and a V between -1.8 0.1 mL mol-1 and -0.5 0.03 mL mol-1. The slow phase (~ 850 ns) gives a H between -43 3 kcal mol-1 and -54 4 kcal mol-1 associated with a V between 2.1 0.2 mL mol-1 and 0.6 0.05 mL mol-1 (Table 3.5) As explained earlier, Ru(phen)3 behaves like Ru(bpy)3 also showing that increase th e solution ionic strength of creates a shortening of the metal-ligand bond in the prompt phase associated with a small electrostriction of the Na+ atoms to the overall negative charge of the ligand around the ruthenium atom. As for Ru(bpy)3, the variation of enthal pies show a very small dependence of the ionic strength change. This phenomenon is reversed in the slow phase data (Table 3.5)

PAGE 108

86 Table 3.5 : Variations of the volume and enthalpy for Ru(phen)3 for different concentrations of NaCl(aq) (between 10mM to 1.5M) at pH=7. The excluded volume change (Vi) was calculated to be equal to -1.9 mL mol-1 for the prompt phase and 1.9 mL mol-1 for the slow phase and the ion charge (zi) to be equal to 1.5 0.2 mL mol-1 and 1.3 0.4 mL mol-1 for the prompt and slow phase, respectively (Figure 3.22) Similar to Ru(bpy)3, the ion charge can be consid ered to be close to 1 for the fast and slow phases within e xperimental errors. As for Ru(pby)3, the theoretical volume change can be calculated using the classical Drude-Nernst equation for Ru(phen)3 where r = 4.8 [27] therefore Vel is equal to -3.5 mL mol-1. Finally, as it was shown earlier for Ru(bpy)3, the volume change due to electrost riction effects can be extracted Conc (mM) 1 (ns) V1 (mL mol-1) H1 (kcal mol-1) 2 (ns) V2 (mL mol-1) H2 (kcal mol-1) 10mM 1 -1.8 0.1 50 0.8 762 2.1 0.2 -45 1 50mM 1 -1.7 0.2 44 1 930 1.8 0.3 -50 2 100mM 1 -1.6 0.06 49 0.4 959 1.2 0.2 -51 1 150mM 1 -1.3 0.1 51 0.8 840 1.1 0.2 -47 2 200mM 1 -1.1 0.02 51 0.1 650 1.0 0.1 -46 0.8 250mM 1 -0.6 0.04 53 0.3 774 1.3 0.4 -43 3 500mM 1 -1.1 0.05 53 0.5 590 0.9 0.3 -43 3 750mM 1 -0.5 0.03 52 0.3 990 0.8 0.4 -54 4 1M 1 -0.6 0.1 46 1 700 1.3 0.6 -46 6 1.5M 1 -0.5 0.02 52 0.3 815 0.6 0.05 -45 1

PAGE 109

87 from Drude-Nernst equation, which will give respectively, Vel Ru(phen)3 = -1.8 + 1.9 = 0.1 mL mol-1 for the prompt phase and Vel Ru(phen)3 = 2.1 – 1.9 = 0.2 mL mol-1 for the slow phase at 10mM NaCl and Vel Ru(phen)3 = -0.5 + 1.9 = 1.4 mL mol-1 for the prompt phase and Vel Ru(phen)3 = 0.6 – 1.9 = -1.4 mL mol-1 for the slow phase at 1.5M NaCl. As for Ru(bpy)3,the calculation for the experimental el ectrostriction change are opposite in sign and similar in magnitude between the fast and the slow phases for Ru(phen)3. As with the Ru(bpy)3 complex, the theoretical ca lculation of the change in electrostriction shows a higher value relative to the experimental value. This may also be due to solvent effects associated with the increase of the concentr ation of ion in soluti on not accounted in the theoretical calculation. 0.00.20.40.60.81.0 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 V1 = -1.9 0.1 mL mol-1Z1 = 1.5 0.2 V2 = 1.9 0.2 mL mol-1Z2 = 1.3 0.4ln(1+bVI) 1st Phase 2nd PhaseRu(phen)3V (mL mol-1) Figure 3.22 : Plot of V versus ln(1+b I) for the 1st phase (purple square) and 2nd phase (orange dot) from Ru(bpy)3 between 10mM and 1.5M NaCl(aq) (pH 7.0).

PAGE 110

88 Table 3.6 : Summary of Vi and zi for Ru(phen)3 from fits to the Debye-Hckel equation. V (mL mol-1) z 1st phase -1.9 0.1 1.5 0.2 2nd phase 1.9 0.2 1.3 0.4 3.3.5. Conclusion The confirmation that the Debye -Hckel equation can be used to calculate the solvent excluded volume of molecules, complexes and/or proteins was validated using two ruthenium complexes. These results show a charge transfer between the ligand and the metal for Ru(bpy)3 and Ru(phen)3 close to one as demonstrat ed in previous studies. In addition, the excluded volume change was calcu lated to be equal to ~ 2.1 and -1.9 mL mol-1 for the prompt phase and ~1.7 and 1.9 mL mol-1 for the slow phase, for Ru(bpy)3 and Ru(phen) 3 respectively.

PAGE 111

89 3.3.6. References [1] Schmidt R., J. Phys. Chem. A 102 (1998), 9082-9086. [2] DeVane R., Ridley C., Larsen R. W ., Space B., Moore P. B. and Chan S. I., Biophys. J. 85 (2003), 2801-2807. [3] Ridley C., Stern A. C., Green T., DeVane R., Space B., Miksovska J. and Larsen R. W., Chem. Phys. Lett. 418 (2006), 137-141. [4] Martyna G. J., Tuckerman D. J ., Tobias D. J. and Klein M. L., Mol. Phys. 87 (1996), 1117. [5] Tuckerman M. E. and Nartyna G. J., J. Phys. Chem. B 104 (2000), 159-178. [6] Debye P. and Hckel E., Physikalische Zeitschrift, 24 (1923) 185–206. [7] Ananthaswamy J. and Atkinson G., J. Chem. Eng. Data 29 (1984), 81-87. [8] Houten J. V., Watts R. J., J. Am. Chem. Soc. 98 (1976), 4853-4858. [9] Nagai K. and Takamiya N., Macromol. Chem. Phys. 197 (1996), 2983-2999. [10] Caspar J. V. and Meyer T. J., J. Am. Chem. Soc. 105 (1983), 5583-5590. [11] Baggott J. E. and Pilling M. J., J. Phys. Chem. 84 (1980) 3012-3019. [12] Mazzetto S. E., de Carvalho I. M. M. and Gehlen M. H., J. Lumin. 79 (1998) 47-53. [13] Giordano P. J., Bock C. R., Wrighton M. S., Interrante L. V. and William R. F. X., J. Am. Chem,. Soc. 99 (1977), 3187-3189. [14] Demas J. N. and Grosby G. A., J. Am. Chem. Soc. 93 (1971), 2841-2847. [15] Van Houten J. and Watts R. J., J. Am. Chem. Soc. 97 (1975), 3843-3844.[16] Bradley P. L., Kress N., Hornberger B. A., Dallinger R. F. and Woodruff W. H., J. Am. Chem. Soc. 103 (1981), 7441-7446.

PAGE 112

90 [17] Miksovska J. and Larsen R.W., Inorg. Chem. 43 (2004), 4051-4055. [18] Borsarelli, C. D. and Braslavsky, S. E. J. Phys. Chem. B 102 (1998), 6231-6238. [19] Goodman, J. L. and Herman, M. S., Chem. Phys. Lett. 163 (1989), 417-420. [20] Cherry, W. R. and Henderson, L. J., Inorg. Chem. 23 (1984), 983-986. [21] Habib Jiwan J.-L., Wegew ijs B., Indelli M. T., Scandola F. and Braslavsky S. E., Recl. Tra V Chim. Pays-Bas 114 ( 1995), 542-548. [22] Borsarelli C. D., Corti H., Goldfarb D. and Braslavsky S. E., J. Phys. Chem. A 101(1997), 7718-7724. [23] Drude P and Nernst W. Z., Phys. Chem. 15 (1884), 79-85. [24] Christian Reichardt, Solvents and solvents effects in Organic chemistry Third Updated and Enlarged EditionWiley-VCH [25] Gensch T and Viappiani C, Photochem. Photobiol. Sci. 2 (2003), 699-721. [26] Millero F. J., Chemical Reviews 71 (1971), 147-176. [27] Borsarelli C. D. and Braslavsky S. E., J. Phys. Chem. A 103 (1999), 1719-1727.

PAGE 113

91 Chapter IVSignaling Proteins 4.1. Introduction Heme proteins are a diverse class of metallopr oteins that participate in a wide array of physiological processes including electron tr ansfer (cytochromes), oxygenation (monooxygenases), hydrogen peroxide degradation (p eroxidases, catalases), small molecule sensing (FixL, PAS domain sensors, and HemAT sensors), transcription regulation (CooA type proteins), energy transduc tion (heme/copper oxidases, cytochrome bc1, etc.), oxygen transport and storage (hemogl obins and myoglobins) and polymer synthesis/degradation (lignan peroxidase, etc.) [1-4] The versatility in heme protein functionality is derived largely from the structure of the surrounding protein matrix. The nature of the proximal iron ligand modulates the degree of ligand reac tivity/binding affinity through electron back donation into accepting ligands. Distal heme pocket residues influence the orientation as well as the lability of ligands through H-bond interactio ns, hydrophobicity, etc. Th e protein tertiary structure also provides conduits for ligand acces s to and from the heme active site. These ligand channels provide ready access from the solvent to the distal heme pocket and can be modulated by conformational changes resul ting in ‘gated’ ligand access. Examination of the crystal structures of a wi de range of heme proteins with structurally variable distal heme pockets reveals complex networks of open pockets and access channels through which the gaseous ligands must traverse in order to bind to the heme iron [5-13] Metalloproteins are a generic class of protei n that contains a meta l ion cofactor where iron and/or copper are the most common metal ions. Heme prot eins are a subclass of iron metalloprotein that contain a prosthetic gr oup which consists of an iron centered in a large heterocyclic organic ring or protoporphyrin IX. Heme proteins are one of the most widely distributed metalloproteins in nature The protoporphyrin ring is formed with four

PAGE 114

92 pyrrole rings linked by methene bridges and the iron can be in the ferrous (Fe2+) or the ferric (Fe3+) oxidation state. The protoporphyrin IX is linked to the protein via different amino acids. The most common heme is the he me b, although other hemes also important including heme a, c and o (Scheme 4.1) Heme groups are linked to the protein w ith an axial ligand(s) binding derived from amino acids. The fifth axial position is ty pically a histidine. The proximal side or sixth position can be occupied by an external ligand (gas molecules). Heme protein sensors can reversibly bind sma ll diatomic ligands including O2, CO and NO [14-20] Studies of the kinetics of ligand binding are important in order to understand the biological function of the heme protein. Liga nd binding/release to the heme, the exchange of one internal/external ligand from the heme can result in activation of the protein (Scheme 4.2) The dissociation of the Fe-ligand bond can activate the process and the transmission of the signal. The signal will be transmitted along the protein by conformational change of the protein to the effector domain which will have an appropriate response to the external signal. The bond between the iron and the ligand can be photodissociated with high quantum yield. This photodissociation is not a physiological event but provides an initiation po int for the process of the transmission of the signal along the protein.

PAGE 115

93 Scheme 4.1 : Major hemes Heme o N N N N CH3CH2CH3HOOC O H HOOC Fe N N N N CH3CH3HOOC HOOC SH S H Fe Heme c Heme b Heme a N N N N CH3CH2CH3HOOC O H O H HOOC Fe N N N N CH3CH2CH3HOOC HOOC Fe

PAGE 116

T T p r T S c T here are fo u P T hese protei n r oteins Fix L T he first exa m c heme 4.2 : S c u r classes of P AS domai n Fi g n s contain a L and EcDo s m ple is the p c heme for th e hemeb ase d n g ure 4.1 : P A a heme insi d s are two ex a p rotein Fix L 94 e signal tran d sensors: A S domain d e a four a mples of t h L from R hiz o smission in g (PBD entr y over one h h e PAS dom a o bium meli l g as sensor p r y 1lsw) h elical fold a in which a r l oti which w r oteins (Figure 4.1 r e oxygen s e w as the first h ) .The e nsor. h eme

PAGE 117

95 oxygen sensor discovered and has a 55kDa PAS domain which can bind an oxygen molecule. The FixL protein is an oxyge n-binding hemoprotein and regulates the expression of nitrogen fixation in Rhizobia. In the presence of O2, the kinase activity of FixL is inhibited. In the absence of O2, FixL will transmit a signal to FixJ via a phosphorylation-dephosphylation reac tion with FixJ controlling the expression of other regulatory genes, including nifA, which regula te the transcription of genes required for symbiotic nitrogen fixation [25,26] Secondly; the EcDos from E. coli has a 17kDa heme binding PAS domain and a 90kDa phosphodieste rase domain. The protein exhibits phosphodiesterase activity and is likely to modulate the cyclic-di-GMP second messenger. Globin-Coupled sensor Figure 4.2 : GCS domain (PBD entry 1or6) The myoglobin-like domain of globin c oupled sensors is a recently discovered non-PAS domain protein. The heme is in side a “three over tw o” helical fold (Figure 4.2) HemAT which is an aerotactic transducer, is the best example for the myoglobin-like domain as it was the first myoglobin-like domai n discovered in Archaea. This is also the first heme from the globin coupled sensor family in bacteria [21-24]

PAGE 118

96 HemAT also responds to O2 molecule like FixL or EcDos. The two class examples of aerotaxis transducers: HemATHs from the salt-loving archaeon Halobacterium salinarum with 489 residues and involved in the aerophobic response and HemATBs from the soil bacterium Bacillus subtilis with 432 residues and involved in the aerophylic response. HemAT is a dimer wher e its N-terminal is close to sperm-whale myoglobin and its C-terminal is close in stru cture to the cytoplasmi c signaling domain of Tsr, a methyl accepting chemotaxis protein from Escherichia Coli. CooA protein Figure 4.3 : CooA domain (PBD entry 2hkx) CooA from Rhodospirillum rubrum is a CO-sensing transcriptional activator (Figure 4.3) CooA which is close to the cAMP-receptor prot eins (CRP) is a CO sensor process which activates expression of genes that govern the oxidation of CO to CO2 [27]

PAGE 119

97 Heme nitric oxide binding (HNOB) Figure 4.4 : HNOB domain (sGC ) (PBD entry 1xbn) The soluble guanylyl cyclase (sGC) is a NO-sensing and catalyzes the formation of cyclic guanosine monophosphate (cGMP) (intracellular second messenger) from GTP (Figure 4.4) sGC is a heterodimeric enzyme composed of a mainly alpha-helical domain and has a proximal histidine binding to the heme.

PAGE 120

98 4.2. References [1] Antonini, E, Rossi-Bernardi, L., Chiancone, E. Eds. (1981) Methods Enzymo.,. 76. Hemoglobins (Part A) Academic Press, San Diego. [2] Pettigrew, G. W. and Moore, G. R. (1987) Cytochromes c Biological Aspects. SpringerVerlag, Berlin Heidelberg New York. [3] Smith, A. T., Veitch, N. C Curr. Opin. Chem. Biol., 2 (1998), 269-278. [4] Taylor, B. and Zhulin, I. B Microbiol. Mol. Biol. Rev., 63 (1999), 479-506. [5] Bossa, C., Amadei, A., Daidone, I., Anse lmi, M., Vallone, B., Brunori, M. and Di Nola, A., Biophys., J. 89 (2005), 465-474. [6] Nutt, David R. and Meuwly, M., Proc. Natl. Acad. Sci. USA, 101 (2004), 5998-6002. [7] Cohen, J. and Schulten, K., Biophys. J. 93 (2007), 3591-3600. [8] Bossa, C., Anselmi, M., Roccatano, D., Am adei, A., Vallone, B., Brunori, M., and Di Nola, A., Biophys. J. 86 (2004), 3855-3862. [9] Plattner, N. and Meuwly, M., Biophys. J. 94 (2008), 2505-2515. [10] Cohen, J., Arkhipov, A., Braun, R. and Schulten, K., Biophys. J. 91 (2006), 18441857. [11] Ruscio, J.Z., Kumar, D., Shukla, M., Pr isant, M.G., Murali, T.M., and Onufriev, A. V., Proc. Natl. Acad. Sci. USA, 105 (2008), 9204-9209. [12] Srajer, V., Teng, T-y., Ursby, T., Praderva nd, C., Ren, Z., Adachi, S-i., Schildkamp, W., Bourgeois, D., Wulff, M. and Moffat, K., Science 274 (1996), 1726-1729. [13] Della Longa, S., Arcovito, A., Benfa tto, M., Congiu-Castella no, A., Girasole, M., Hazemann, J.L., and Lo Bosco, A., Proc. Natl. Acad. Sci. USA 100 (2003), 8704-8709. [14] Gilles-Gonzalez M.A. and Gonzalez G, J. Inorg. Biochem. 99 (2005), 1-22. [15] Aono S., Acc. Chem. Res. 36 (2003), 825–831.

PAGE 121

99 [16] Vos M. H., Biochimica et biophysica acta. 1777 (2008), 15-31. [17] Cao W., Ye X., Georgiev G. Y., Ber ezhna S., Sjodin T., De midov A. A., Wang W., Sage T. J. and Champion P. M., Biochemistry 43 (2004), 7017–7027. [18] Chan M. K., Curr Opin Chem Biol. 5 (2001), 216-222. [19] Reedy C. J. and Gibney B. R., Chem. Rev. 104 (2004), 617–650. [20] Spiro T. G. and Jarzecki A. A., Chemical Biology 5 (2001), 715-723. [21] S. Hou, T. Freitas, R.W. Larsen, M. Piatibratov, V. Sivozhelezov, A. Yamamoto, E.A. Meleshkevitch, M. Zimmer, G.W. Ordal and M. Alam, Proc. Natl. Acad. Sci. USA, 98 (2001), 9353–9358. [22] S. Hou, R.W. Larsen, D. Boudko, W. Riley, E. Kara tan, M. Zimmer, G.W. Ordal and M. Alam, Nature 403 (2000), 540–544. [23] M.K. Chan, Curr. Opin. Chem. Biol,. 5 (2001), 216–222. [24] K.R. Rodgers, Curr. Opin. Chem. Biol,. 3 (1999), 158–167. [25] Rodgers K. R., Tang L., LukasRodgers G. S. and Wengenack N. L., Biochemistry 40 (2001), 12932-12942. [26] Gilles-Gonzalez M. A., Ditta G. and Helinski D. R., Nature 350 (1991), 170-172. [27] Aono S., Nakajima H., Saito K and Okada M., Biochem. Biophys. Res. Commun. 228 (1996), 752-756.

PAGE 122

100 4.3. Evidence for fast conformational chan ge upon ligand dissociation in the HemAT class of bacterial oxygen sensors 4.3.1. Introduction Globin coupled sensors (GCS) represent a unique class of oxygen sensing heme proteins found in both archaea and bacteria [1] In comparison, unlike the more omnipresent PAS domain sensors found in bacteria, archaea and eukaryotes, which include an / type fold with an anti-parallel -barrel structure that encapsulates the sensor element, GCS proteins contain an N-terminal heme sensor domain similar to a myoglobin (Mb) type fold [1-4] These proteins also cont ain a C-terminal domain homologous to the cytoplasmic signaling domain of various eukaryotic type chemoreceptors. The first two members reveal ed in the GCS class of sensor were Hs HemAT from the archaea Halobacterium salinarum and Bs HemAT from the gram positive prokaryote Bacillus subtilis [2] Both Hs HemAT and Bs HemAT share structural homology with the globin family. Specifically, both proteins have conserved His (F8), Pro (C2), and Phe (CD1) corresponding to amino acids Pro 56, His 123, Phe 69 ( Bs HemAT numbering). In additi on, both proteins participate in aerotaxis responses in their respective organisms. Hs HemAT has been implicated in an aerophobic response while Bs HemAT is involved in an aerophylic response [2,5] Hs HemAT also undergoes methylation-dependent adaptation via CheR methyltransferase. A recent crystal structure of the Bs HemAT heme domain ( Bs HemATHD) demonstrated a myoglobin-like heme sensing domain with a more traditional ‘three over two’ helical fold typical of gl obins that is structurally di stinct from PAS-domain heme sensors [6] The crystal structure also revealed a dimer in the asymmetric unit similar to the EcDosH [7] Similar to the FixL heme domain s, structural differences between liganded and unliganded forms was modest. The most significant changes were the rotation of the proximal His (His 123) and disp lacement of a distal Tyr residue (Tyr 70).

PAGE 123

101 These results suggest that stru ctural changes associated with the signaling event are quite small or that larger transient conformational changes are not detected by X-ray structures. Like other heme oxygen sensors, conforma tional changes associated with ligand binding/dissociation to heme sensing moietie s are thought to initi ate signaling in the GCS. Thus, characterizing those conformati onal changes is critically important to understand the signal transduction process. We have utilized transient absorption spectroscopy and time-resolved photoacoustic calorimetry (PAC) to examine the thermodynamics and conformational dynamics following CO photodissociation from Fe(2+) Bs HemATHD. 4.3.2. Materials and methods 4.3.2.1. Protein expression, isolation and purification The open reading frame corresponding to the N-terminal 180 amino acids heme binding domain of HemAT ( Bs HemATHD) was amplified from B. subtilis genomic DNA (ATCC 23857D-5). The primers employed for PCR were 5 ATGTTATTTAAAAAAGAC3 (forward) and 5 AAACGCTTCAAGGACAAGCAG3 (reverse). The resulting PCR product was s ubsequently amplified and cloned into pDEST17 (Invitrogen) using Gateway t echnology (Invitrogen) according to the manufacturer’s instruct ions. The primers used for amplification were 5 GGGGACAAGTTTGTACAAAA AAGCAGGCTCCATCGAGGGACGAATGTTATT TAAAAAAGAC3 (forward) and 5 CCCCACCACTTTGTACA AGAAAGCTGGGTTTAAA ACGCTTCAAGGACAAGC AG3 (reverse). The resulting plasmid was named pHemAT180 and contains an Nterminal six residue histidine tag. The inte grity of the HemAT open reading frame was verified by sequencing in the Wa shington State University LBB1. pHemAT180 was established in Escherichia coli strain BL21(DE3) and was used to express the HemAT histidine-tagged N-te rminal domain protein. This strain also

PAGE 124

102 contained pHPEX3 which expresses the hemi n transporter ChuA (a kind gift from Professor D.C. Goodwin (2004), Auburn Univ ersity. Protein Expr. Purif. 35, 76–83). The expression strain with two plasmids was maintained under dualampicillin (pHemAT180) and chloramphenicol (pHPEX) selection. Fo r protein expression, “Terrific Broth” (Current Protocols in Molecular Biology, Au subel et al., Eds.) was inoculated to A560 = 0.05 and incubated with shaking at 37C until A560 increased to 0.6–0.8. Protein expression was induced by addi tion of 0.5 mM IPTG and hemin was added to a final concentration of 50 M. The temperature was reduced to 25C and incubation was continued for 15 h. Following induction and expression the ce lls were collected by centrifugation and resuspended in lysis/wash buffer (50 mM Na phosphate pH 8, 300 mM NaCl, 20 mM imidazole), lysed by two passages through a Fr ench pressure cell op erated at 12 000 psi and cleared by centrifugation. The dark red supernatant was then applied to a Ni2+-NTA column (Qiagen) and the column was washed extensively with buffer. Protein was eluted from the column with a 20–300 mM imidazole gr adient in buffer. Pr otein fractions were pooled then concentrated and exchanged into other buffers using a centrifugal concentrator (Amicon Ultra). Protein pur ity was monitored by SDS–PAGE, MALDITOF Mass Spectrometry (WSU LBB2) and UV–Vis spectroscopy. 4.3.2.2. Sample preparation Samples for PAC were prepared by diluting Bs HemATHD from an ~150 M stock into a buffer containing 50 mM sodium phosphate (pH 8.0) and 100 mM NaCl (the protein is predominantly in the homodimeri c state under these c onditions). The deoxy form of the protein was formed by placing the oxy form of Bs HemATHD in a quartz optical cuvette that was then sealed with a septum cap and purged with Argon. Dithionite was added from a buffered stock solution to give a final concentration of ~100 M. The CO-bound form was obtained by saturating solutions of the deoxy Bs HemATHD with CO resulting in a final solution CO concentrat ion of 1 mM (1 atm pressure). The protein

PAGE 125

103 concentration for PAC samples was ~75 M while those for transient absorption were ~25 M. The equilibrium binding of CO to deoxy Bs HemATHD was performed by titrating aliquots of CO saturated buffer (1 mM) into a solution containing ~10 M deoxy Bs HemATHD. The change in absorption at 420 nm was monitored and plotted versus [CO]. The Ka value for CO binding was obtained using: A420nm = Ka420[ Bs HemATHD]0[CO] / (1 + Ka[CO]) equation 30 where 420 nm is the change in molar extinction at 420 nm (obtained from the fit) and [ Bs HemATHD]0 is the initial concentration of deoxy Bs HemATHD. Equilibrium UV– Vis spectra were obtained using a Shimadzu UV2401 spectrophotometer. 4.3.3. Results The recombinant heme domain of Bs HemAT displays an optical spectrum with a Soret maximum at ~410 nm and visible bands at 578 nm ( -band) and 538 nm ( -band) (Figure 4.5) Upon deoxygenation the Soret band shifts to 425 nm with a broad visible band centered at 555 nm. Incubation of the deoxy protein in the presence of CO results in a Soret band at ~418 nm and visible bands at 573 nm and 535 nm. These spectral changes were utilized to determine the association constant for the recombinant Bs HemATHD (Figure 4.6) Fitting a plot of A420 nm versus CO concentration to a single binding equilibrium yielded a Ka value of (1.5 0.5) 106 M 1 consistent with previous studies of this construct (Alam and Larsen, unpublishe d results) but is slight ly lower than that determined previously by Zhang et al. who reported a value of 6 106 M 1 [9]

PAGE 126

104 300 400 500 600 700 0.0 0.5 1.0 1.5 HemAT oxy form HemAT deoxy form HemAT CO bound formAbsorbanceWavelength (nm)500 600 0.06 0.08 0.10 0.12 0.1 0.16 0.18 0.20 AbsorbanceWavelength (nm) Figure 4.5: Equilibrium optical absorption spectra of the Bs HemATHD: as isolated (red dash line), deoxy Bs HemATHD (blue dot line) and CO Bs HemATHD (black solid line). Bs HemATHD concentration: ~10 M in 50 mM sodium phosphate (pH 8.0) and 100 mM NaCl.

PAGE 127

105 01234 0.00 0.02 0.04 0.06 0.08 Absorbance (420 nm)CO Concentration (M) Figure 4.6 : Equilibrium CO binding titration for Bs HemATHD. Absorbance changes were measured at 420 nm. Sample concen trations are those described in Figure 4.5 Preliminary results [21] on the HemAT demonstrated that the heme domain has two distinct O2 binding components. In addition, the O2 rebinding is biphasic with a dissociation constant equal to Kd = 1-2 M, associated with a rate constant kO2 = 50-80 s1, and a dissociation constant Kd = 50-100 M associated with a rate constant kO2 = 2000 s-1 for both the first phase and second phase, respectively. In contrast, the CO binding and rebinding to the heme is monophasic. The association rate constant is k’CO = 0.2-0.5 M-1 s-1 which is very close to the rate constant of the whale myoglobin k’CO = 0.51 M-1 s-1.

PAGE 128

106 Figure 4.7 and Figure 4.8 represent the single-wavele ngth kinetic trace for the O2 and CO rebinding to the heme. In order to calculate the rate constant of the O2 binding to the heme, the CO-bound form of Bs HemAT was photodissociated in presence of O2 in the solution. As the rate constant for O2 rebinding to the heme is faster than the rate constant for CO, O2 will bind first and the rate constant for O2 binding to the heme can be estimated. As O2 is in larger excess, the rate constant was assumed to be pseudo first order and the trace was fit to a single exponential decay. The rate constant of O2 binding was estimated to be 0.00264 s-1. This concludes that th e rate constant of O2 is more than six times faster than the rate constant of CO binding. As it was mentioned earlier, Phillips et al. [21] demonstrated a biphasic relaxation of O2 rebinding with two dissociation rate constant equal to kO2 = 50-80 s-1 and kO2 ~ 2000 s-1. The biphasic relaxation for the O2 rebinding the heme demonstrated by Phillips’ group, is different to the monophasic relaxation estimated in this paragraph. This divergence can be due to a difference in the preparation of the HemAT heme domain. The following Eyring’s equation was used to extract the activation energy of O2 binding. The activation of enthalpy wa s calculated equal to 8 kcal mol-1 and the entropy of energy equal to -20 kcal mol-1. Using the following Phillip’s equation, the rate constant of the O2 dissociation was estimated. k’ O2 = ((kobs k’CO* [CO]) – k’CO [CO]) / [O2] Equation 31 where kobs is the rate constant of O2 rebinding, k’CO is the rate constant of CO rebinding, [CO] is the concentration of CO in solution and [O2] is the concentration of O2 in solution. Phillip’s equation gave a rate constant of the O2 dissociation equal to k’O2 = 166.9 M-1s-1. Finally using the following Eyring’s equation: Rln(h k’O2 /kBT) = S# H#(1/T), the enthalpy of dissociation of O2 was estimated equal to 15.5 kcal mol-1 and the entropy of dissociation equal to 4.4 cal mol-1K-1 (Figure 4.9) [22,23]

PAGE 129

107 -001000001002003004 -001 000 001 002 003 004 AmplitudeTime(s) Figure 4.7 : Single wavelength transient absorption trace for O2 rebinding to Bs HemATHD. Photolysis of the CO bound Bs HemATHD results in the formation of a five coordinate high spin heme complex whic h decays back to the pre-flash CO bound complex with a pseudo first-order rate constant of 59 s1 (the corresponding second-order rate constant is 5.9 104 M1 s1 with a solution of CO con centration equal to 1 mM). This value is close to the one reported by Zhang et al. (4.3 105 M1s1) suggesting that the difference in Ka values for CO between Zhang et al and the present study is due to the values of koff (0.07 s 1 vs. 0.04 s1 from Zhang et al. and the present study, respectively) [9] In addition, the kinetic optical di fference spectrum obtained equal to ~4 s subsequent to photolysis displays a bath ochromic shift relative to the equilibrium optical difference spectrum suggesting an un-relaxed heme pocket geometry subsequent to photolysis (Figure 4.8) Since the concentration of tr ansient five coordinate heme decays mono-exponentially the un-relaxed heme pocket must persist for ~milliseconds. From the temperature dependence of the reco mbination rates the activation enthalpy and entropy were found to be 6 kcal mol1 and -28 cal mol1 K1, respectively (Figure 4.10) The PAC results, which can probe molar volume and enthalpy changes over a time scale from ~50 ns to ~20 s [8] reveals a H of -19 5 kcal mol1 and V of 4 1 mL mol1

PAGE 130

108 occurring in <50 ns subsequent to CO photolysis (Figure 4.11) No additional kinetic events were observed within th e PAC time scale. The value of for CO photolysis from the ferrous form of Bs HemATHD has previously been de termined to be 0.90 0.1 (Alam and Larsen, unpublished results). -0.010.000.010.020.030.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 AmplitudeTime(ms) Figure 4.8 : Top panel : Single wavelength transient abso rption trace for CO rebinding to Bs HemATHD. Bottom panel : Overlay of the kinetic (squares, obtained 4 ms subsequent to photolysis) and equilibrium (solid line) difference spectra (deoxy Bs HemATHD minus CO Bs HemATHD).

PAGE 131

109 3.2x10-33.3x10-33.3x10-33.4x10-33.4x10-33.5x10-33.5x10-33.6x10-3-210 -205 -200 -195 -190 Rln(hk'O2/kBT)1/T Figure 4.9 : Eyring plot for O2 recombination to deoxy Bs HemATHD. 3.3x10-33.3x10-33.4x10-33.4x10-33.5x10-33.5x10-33.6x10-3-214 -213 -212 -211 -210 -209 -208 -207 -206 Rln(hkCO/kBT)1/T Figure 4.10 : Eyring plot for CO recombination to deoxy Bs HemATHD.

PAGE 132

110 2.0x10-64.0x10-66.0x10-6-1.5x10-2-1.0x10-2-5.0x10-30.0 5.0x10-31.0x10-21.5x10-22.0x10-22.5x10-23.0x10-23.5x10-2 Normalized Acoustic AmplitudeTime (Seconds) Fe(III)4SP Reference CO-Fe(II) Bs HemAT Heme Domain2.53.03.54.04.55.05.56.06.5 70 75 80 85 90 95 100 105 110 E h (kcal mol-1)Cp (kcal mL-1) Figure 4.11 : Top panel : Overlay of photoacoustic traces for COBs HemATHD (solid line) and the Fe3+ tetrakis-(4-sulphonatophenyl)porphyrin calorimetric reference (dotted line). Bottom panel : Plot of Eh vs. ( Cp / ). which gives a intercept equal to the heat, Q returned to the solvent giving H = ( Eh Q)/ (kcal mol 1) and V = slope/ (mL mol 1).

PAGE 133

111 4.3.4. Discussion The asymmetry of monomer units observed in the X-ray crystal structure of the CN bound and free forms of Fe2+ Bs HemATHD provided the basis for an initial signaling mechanism [6] The X-ray structure reveals that th e conformation of one monomer within the asymmetric dimer contains a Tyr 70 located within the distal he me pocket and within H-bonding distance to the heme bound ligand while in the other monomer Tyr 70 is rotated ~100 out of the dist al pocket. This is accompanied by a 14 rotation of the F helix which also affects the orientation of the proximal histidine (His 123). It has been further suggested that the presence of two conformations within the homodimer accounts for the observed high and low affinity binding sites for O2 and biphasic O2 recombination kinetics with the high affinity site (faster rebinding phase) that in which Tyr 70 is located within the distal pocket and the low affinity site (slower rebinding phase) that with the solvent exposed Tyr 70 [9] In contrast, CO rebinding, subs equent to photolysis, exhibits monophasic kinetics suggesting that Tyr 70 does not significantly influence the stability of the Fe2+–CO complex (Figure 4.12)

PAGE 134

112 Figure 4.12 : X-ray crystal structure of the CN bound and free forms of Fe2+ Bs HemATHD (PDF entry 1or6). More recent resonance Raman results have al so been used to construct models for both O2 and CO binding to the heme domain of Bs HemAT that involve H-bond formation between the bound ligand and Thr 95 [10-12] The resonance Raman data indicates multiple conformations in the O2 bound form of the protein as evident by multiple Fe2+– O2 bending modes in the wild-type sensor domain (554 cm 1, 566 cm 1 and 572 cm 1). These were assigned to two ‘open’ conformations (566 cm 1 and 572 cm 1 bands) which are responsible for the lower oxygen affinity (higher O2 off-rates) and a ‘closed’ conformation (554 cm 1 band) with a higher O2 affinity. In the Thr 95A mutant only one Unliganded subunit A Unliganded subunit B Liganded subunit A Liganded subunit B

PAGE 135

113 Fe2+-O2 bending mode is observed at 569 cm 1. Examination of the X-ray structure for the deoxy form of the protein reveals that the distance between the Thr 95 hydroxy group and the heme iron is ~7.2 and this residue must move ~4 to move within the ~3 required for an H-bond with the bound ligand. Thus, in the case of O2 binding, in which H-bonding occurs between Thr 95 and th e bound oxygen molecule, a significant conformational change must take pl ace that relocates the Thr 95 residue. In contrast, resonance Raman results i ndicate that CO binding does not result in H-bonding between Thr 95 and the bound CO si nce no shifts were observed in the Fe2+– CO bending mode in the Thr 95Ala mutant. Ho wever, FTIR results by Pinakoulaki et al. [13] suggest that the CO bound form of the se nsor domain contains both non-H-bonded and strongly H-bonded conformers (as determined from C–O) in both wild-type and Thr 95Ala mutants. However, mutation of Tyr 70 to Phe perturbs the non-H-bonded C–O (shifts from 1967 cm 1 to 1962 cm 1) and produces a moderately H-bonded conformation ( C–O ~ 1931 cm 1). In a second FTIR study Pinakoulaki et al. [14] proposed a second CO binding site on the proximal side of the heme group near Tyr 133 giving rise to a C– O ~ 2069 cm 1. Photolysis results in a s light increase in intensity of this mode indicating an increase in site occupancy. It was proposed that this site can se rve as a conformational gate for ligand access to the heme. The transient difference spectrum as well as the PAC results, indicate that, photodissociation of CO results in a non-equi librium heme environment subsequent to photolysis and that no further conformationa l changes take place between ~50 ns and ~20 s. In addition, the fact that the decay of the five coordinate heme is monophasic further indicates that the non-equilibrium he me pocket persists on the ms time scale. Formation of the non-equilibrium five coor dinate conformation also gives rise to enthalpy and molar volume changes of -19 5 kcal.mol 1 and V of 4 1 mL.mol 1, respectively. The observed volume and enthal py changes may have contributions from photocleavage events local ized at the heme ( Vheme and Hheme), H-bond changes in the distal heme pocket (as described above) ( Vdistal and Hdistal), ligand migration to non-

PAGE 136

114 heme binding pockets ( VNHBP and HNHBP) and/or more global conformational changes in the heme sensing domain ( Vconf and Hconf) as follows: Vobs = Vheme + Vdistal + VNHBP + Vconf Equation 44 Hobs = Hheme + Hdistal + HNHBP + Hconf Equation 45 Upon photolysis of the Bs HemATHD the Fe2+CO bond is cleaved ( HFeC ~ 15 kcal mol 1) and the heme iron undergoes a ch ange in spin state (from lowspin to high-spin, HLS-HS ~ <1 kcal mol 1) giving an Hheme ~ 15 kcal mol 1 [8] In addition, if CO diffuses out of the heme poc ket and into the bulk solvent on this time scale an additional HCO-solv ~ 5 kcal mol 1 would also contribute putting Hheme on the order of ~20 kcal mol 1. The Hdistal contribution would likely arise from H-bond interactions between the bound CO and distal pocket residues as observed in the FTIR studies described above. The similarity in vibrational frequencies between the COMb and CO Bs HemATHD indicates similar H-bond streng ths which, for COMb, occur between the bound CO and a distal His residue [15,16] The H-bond energy between CO and the imidazole ring of His 64 in COMb has b een calculated to be ~5 kcal mol 1 making Hdistal roughly this value (i.e., for a single CO–H bond). The HNHBP term is more difficult to estimate since it is not clear to what extent the nonheme binding pocket in Bs HemATHD is occupied at room temperature and within ~20 s subsequent to photolysis. Assuming similar extinction coeffi cients for the CO stretch of the heme bound CO and the CO bound in the non-heme bi nding pocket proposed by Pinakoulaki et al. [14] the non-heme binding site represents r oughly 2% of the total protein bound CO. Upon steady-state photolysis this population increases by ~1.5 % which would not be detectable by PAC. The low occupancy of this site subsequent to photolysis suggests that a majority of the photodissociated CO leaves the heme pocket on a ~20 s time scale. Thus, the HNHBP term is likely to be negligible (i.e., outside of the PAC detection range). Using the relation Hobs = 25 kcal mol 1 + Hconf allows for the estimation of Hconf. The fact that the observe d enthalpy is -19 kcal.mol 1 indicates an additional exothermic process of ~ -44 kcal mol 1 is taking place in <50 ns, which must be due to a

PAGE 137

115 more global conformational cha nge within the heme domain and/or distal residue–ligand interactions. These results are summarized in Figure 4.13 where the thermodynamic profile demonstrated the proposed conformational change. Figure 4.13 : Thermodynamic profiles for the conforma tional change associated with the Fe(2+) Bs HemATHD photodissociation (PDB entry 1or6). Similar arguments can be made for the molar volume changes in which the estimated changes are: Vheme = VFeC + VLS-HS = ~5 mL mol 1 + ~10 mL mol 1 and for CO diffusing from the heme pocket into the solvent: VCO-solv ~ 11 mL mol 1 giving a total Vheme ~ 26 mL mol 1 [6] Once again, based upon a low occupancy for the nonheme binding site the VNHBP is assumed to be neglig ible. The contribution of Vdistal to Vobs is also difficult to estimate. As an uppe r limit the value would be that assigned to the Fe–C bond cleavage since typica l bond cleavage processes have a V of ~5 mL mol 1. However, since the H-bonding is located in the distal heme pocket rather than h < 50 ns Hconf Vconf 40 kcal mol-1Conformational Change -22 mL mol-1

PAGE 138

116 exposed to the solvent the actual contribution is likely to be mu ch smaller than this value. With this in mind Vobs = 26 mL mol 1 + Vconf and since Vobs = 4 mL mol 1, Vconf is estimated to be 22 mL mol 1 (36.5 3 / molecule). The Vconf and Hconf terms could arise from (1) a change in overall charge distribution on the protein (i .e., change in net protein dipole leading to solvent reorganization), (2) formation of one or more salt-bridge interacti ons (the release of electrostricted water molecules upon salt-bri dge formation results in volume increases) and/or (3) increases in the solvent accessi ble van der Waals volume of the protein upon photolysis. Examination of the cr ystal structures for the Fe(3+) Bs HemATHD and CN– Fe(3+) Bs HemATHD reveals that lig and binding results in a change in the solvent accessible surface (SAS) of 1282 3 (772 mL.mol 1 per dimer) (i.e., the unliganded protein has a larger SAS than that of the CN-bound protein) which is 1.8% of the total SAS of the dimer. Changes in solvation asso ciated with the change in SAS could account for the observed volume and enthalpy changes provided a similar change in SAS occurs in the COFe(2+) Bs HemATHD to Fe(2+) Bs HemATHD since surface residues are likely to experience significant changes in solvent exposure subsequent to the change in SAS. The photothermal results further demonstrat e clear differences in the dynamics of CO release from either horse heart Mb or th e heme domain of FixL. In the case of FixLH from B. japonicum photolysis of CO results in an endothermic (12 kcal mol 1), relatively small contraction (-1 mL mol 1) occurring in <50 ns that is followed by a second relaxation with ~ 150 ns and H and V of 5 kcal mol 1 and 5 mL mol 1, respectively [17] The fast phase dynamics were attributed to solvent perturbations arising from possible salt-bridge reorganiza tion (Glu 182-Arg 227) subseque nt to ligand release while the slow phase was attributed to ligand escape to the solvent. In the case of horse heart Mb, two thermodynamic phases are also observed upon photolysis with a small endothermic contraction (7 kcal mol 1 and -3 mL mol 1) occurring in <50 ns (similar to Bj FixLH) which has been suggested to aris e from solvation of charge formed by disruption of a salt-bridge between one of the heme propionate groups and Lys 45 (Arg 45 in sperm whale Mb) [18-20] A corresponding slow phase with ~ 600 ns and H and

PAGE 139

117 V of 8 kcal mol 1and 4 mL mol 1 is also observed that has be en partly attributed to the salt-bridge reformation. The extent to which conformational differen ces between ligation states participate in signal transduction is unclear. Th is is due to the fact that, to date, only oxygen activity of HemATs in whole cells (either Halobacter salinarum or B. subtillis ) have been examined due to the fact that an assay for signaling ac tivity in vitro has not been developed. Thus, the extent to which differences between the various H-bond networks involving CO and O2 actually contribute to the si gnaling initiation/propagation is unclear. Both changes in heme conformation as well as perturbations to the distal Hbonding sites upon ligand binding are likely to contribute to signal initiation/propagation. 4.3.5. Conclusion In summary, both transient absorption and PAC reveal a fast conformational transition associated with ligand release from Bs HemATHD. A conformational transition results in a non-equilibrium deoxy form of the protein that persists during the ligand rebinding phase. The thermodynamics of this transition are quite distinct from the PAS domain Bj FixL heme domain suggesting quite different signaling mechanisms between the GCS and PAS domain sensors.

PAGE 140

118 4.3.6. References [1] S. Hou, T. Freitas, R.W. Larsen, M. Pi atibratov, V. Sivozhelezov, A. Yamamoto, E.A. Meleshkevitch, M. Zimmer, G.W. Ordal and M. Alam, Proc. Natl. Acad. Sci. USA, 98 (2001), 9353–9358. [2] S. Hou, R.W. Larsen, D. Boudko, W. Riley, E. Karatan, M. Zimmer, G.W. Ordal and M. Alam, Nature 403 (2000), 540–544. [3] M.K. Chan, Curr. Opin. Chem. Biol,. 5 (2001), 216–222. [4] K.R. Rodgers, Curr. Opin. Chem. Biol,. 3 (1999), 158–167. [5] H.S. Yu, J.H. Saw, R.W. Larsen, K. Watts, M.S. Johnson, M. Zimmer, G. Ordal, B.L. Taylor and M. Alam, FEMS Microbiol. Lett., 217 (2002), 237–242. [6] W. Zhang and G.N. Phillips, Structure, 11 (2003), 1097–1110. [7] H.-J. Park, C. Suquet, J.D. Satterlee and C.-H. Kang, Biochemistry, 43 (2004), 2738– 2746. [8] R.W. Larsen and J. Miksovska, Coord. Chem. Rev. 251 (2006), 1075–1101. [9] W. Zhang, J.S. Olson and G.N. Phillips, Biophys. J., 88 (2005), 2801–2814. [10] H. Yoshimura, S. Yoshioka, K. K obayashi, T. Ohta, T. Uchida, M. Kubo, T. Kitagawa and S. Aono, Biochemistry, 45 (2006), 8301–8307. [11] T. Ohta, H. Yoshimura, S. Yoshioka, S. Aono and T. Kitagawa, J. Am. Chem. Soc., 126 (2004), 15000–15001. [12] S. Aono, T. Kato, M. Matsuki, H. Na kajima, T. Ohta, T. Uchida and T. Kitagawa, J. Biol. Chem., 277 (2002), 13528–13538. [13] E. Pinakoulaki, H. Yoshimura, S. Yoshiko, S. Aono and C. Varotsis, Biochemistry, 45 (2006), 7763–7766.

PAGE 141

119 [14] E. Pinakoulaki, H. Yoshimura, V. Dask alakis, S. Yoshioko, A. Shigetoshi and C. Varotsis, Proc. Natl. Acad. Sci. USA, 103 (2006), 14796–14801. [15] T-L. Michael, M.L. Quillin, G.N. Phillips and J.S. Olson, Biochemistry, 33 (1994), 1433–1446. [16] E. Sigfridsson and U. Ryde, J. Biol. Inorg. Chem., 4 (1999), 99–110. [17] J. Miksovska, C. Suquet, J.D. Satterlee and R.W. Larsen, Biochemistry, 44 (2005), 10028–10036. [18] J.A. Westrick, J.L. Goodman and K.A. Peters, Biochemistry, 26 (1987), 8313–8318. [19] L. Angeloni and A. Feis, Photochem. Photobiol. Sci. 2 (2003), 730–740. [20] J. Miksovska, J. Day and R.W. Larsen, J. Biol. Inorg. Chem., 8 (2003), 621–625. [21] Wei Zhang, John S. Olson, and George N. Phillips Jr., Biophys. J. 88 (2005), 28012814. [22] van Eldik R., Asano T. and le Noble W.J., Chem. Rev., 89 (1989), 549. [23] Hiromi K., Kodanshi Scien tific Books, Kodanshi, Ltd. (1979).

PAGE 142

120 4.4. Thermodynamics of conformational changes coupled to CO photodissociation from the CO-sensi ng transcriptional activator CooA 4.4.1. Introduction Heme protein sensors that ar e regulated by the binding of di atomic molecules such as O2, CO or NO are more and more studied in order to understand their mechanisms. CooA, a transcriptional ac tivator, found in the purple non-sulfur photosynthetic bacterium, Rhodospirillum rubrum was the first protein showing a heme as a prosthetic group for transcriptional activation [1] Previous studies utilized CO as a probe to study the biochemical and biophysical properties of heme protein, because it was thought to have no physiological relevance. Recently, it ha s been shown that when CO binds to the heme, it involves a physiologically activa tion of CooA. The ac tivated CooA is responsible of transcriptiona l regulation for the expression of Coo operons that allow R. rubrum to grow using CO as its sole energy source [2]

PAGE 143

121 DNA binding site C-helix Monomer A Monomer B Effector Domain Figure 4.14: Ribbon structure of CooA. Heme pocket amino acid residues, Pro 2 and His 77 are displayed (PDB entry 1ft9). Figure 4.14 represents the structure of the re duced form of CooA identified by Lanzilotta et al. [3] CooA is a homodimer with 221 amino acid residues in each subunit (24.6 kDa). Each subunit contains a protohe me that can sense CO. The amino acid sequence analysis using X-Ray crystallography confirmed that CooA is a member of the cyclic adenosine monophosphate (cAMP) r eceptor protein (CRP) and fumarate and nitrate reductate protein (FNR) transcriptional regulators family [3-5] The folding of reduced CooA is very similar to a global transcripti onal regulator in E. coli another member of the CRP family. The b-type heme containing the sensor domain (aminoterminal region) and DNA-binding domain ( carboxy-terminal region), which are linked together with a hinge region (130-1 40), form the monomer of the CooA [6,7] The DNA binding domain is a helix-turn-helix (HTH) mo tif which recognizes the binding site of the DNA target [8] Transcriptional activation occu rs when CooA binds to the DNA binding site, thereby contro lling the growth of the Rhodospirillum rubrum The transcriptional activator can be active only wh en CO binds to the heme sensor of CooA

PAGE 144

122 [6,7,9,10] The activation of CooA is unique. Lanzilotta et al. also show that the Nterminal of the proline of one subunit will bi nd with the heme of the other subunit. CooA is the first and only metalloprotein showing a proline residue bindi ng to the metal of a heme. The heme in CooA has different axial ligands in the resting and activated forms. The ferric form of CooA shows the 6-coordinate with Pro 2 in the distal side and Cys 75 in the proximal side. After reduction of the iron to the ferrous form, Cys 75 is replaced by His 77. Then, in presence of CO under anae robic conditions, CO re places Pro 2 in the distal side. The different steps during the activation of CooA are summarized in Scheme 4.3. Scheme 4.3 : The schemes for the change in the coordination structure of the heme in wildtype CooA with associated lifetimes.

PAGE 145

123 Using optical absorption spectra, Nakajima et al. show that intermediate 2 is formed in a fast process afte r reduction of intermediate 1 [11] Intermediate 3 is within 40 s after reduction of intermediate 1, formed after an elongation of the Fe-S bond (sulfur atom) from the Cys 75 binding to the heme and/ or the protonation of the thiolate residue from the Cys 75 [12,13] The ligand Cys 75 will then exchan ge with the ligand His 77 in 2.9 ms, which corresponds to the slower process and form intermediate 4. This slow step implies a large conformational change of the pr otein in order to allow the exchange of the ligand. This event was verified by X-Ray crys tallography where the cr ystal structure of intermediate 4 shows a distance of 4.8 between the Cys 75 and the iron from the heme [3] This exchange of ligand between Cys 75 a nd His 77 requires a relative flexibility of the polypeptide chain in the proximal side of the heme. This is confirmed by looking at the residues around the heme and the different possibilities to form hydrogen bonds. Only two residues can form a hydrogen bond betw een each other (Gly 43 and Met 76) in CooA, compared to an average of seve n hydrogen bonds found in Cytochrome c and Myoglobin where no ligand exchange occurs [11] It was also reported that the ligand exchange between Cys 75 and His 77 is regula ted by a hysteresis in the reduction (-320 mV) and the oxidation (-260 mV ) potentials of the heme [14,15] In presence of CO and in anaerobic conditions, Pro 2 on the heme’s distal side is replaced by CO to form intermediate 5. Photolysis of CO leads to intermediate 6. Uchida et al. demonstrate a biphasic geminate recombination of CO to th e heme within ~70 ps (~60%) and ~300 ps (~30%) after photolysis to form intermediate 7 [16] They also show that the geminate recombination of CO has a yield of ~90% in the 1.9 ns after photolysis. This hypothesis was confirmed by Rubtsov et al [17] and Kumazaki et al. [18] To compare, myoglobin shows a geminate recombination of CO of only 4% [19] This difference between CooA and myoglobin can be explained by a crowded heme pocket of CooA which operates as a barrier to the release of CO to the solvent. Finally, Pro 2 will rebind to the heme and form intermediate 8 in milliseconds after the gemi nate recombination of CO to the heme [20]

PAGE 146

124 Conformational changes associated with ligand binding/dissociation to the heme are thought to initiate signaling in heme sens or families. Therefore, the characterization of those conformational changes is essential in order to understand the signal transduction process. We have utilized time-resolved phot oacoustic calorimetry (PAC) to examine the thermodynamics and conformational dynamics following CO photodissociation from Fe(2+)CooA (Figure 4.15) Figure 4.15 : Structural diagram of the CO photodissociation from the heme domain of CooA. 4.4.2. Materials and Methods Sample preparation : Samples for PAC were prepared by diluti ng CooA into a buffer containing 25 mM MOPS and 0.1M NaCl (pH 7.4). The CObound form was obtained by saturating a solution of oxy CooA with CO resulting in a final CO concentration of 1 mM (1 atm pressure). CO will bind the heme after reduc tion of the Fe(III) form of CooA. The deoxy form of the protein was formed by placing th e oxy form of CooA in a quartz optical cuvette that was then sealed with a septum cap and purged with Ar. A freshly prepared solution of sodium dithionite was added from a buffered stock solution to give a final h CO

PAGE 147

125 concentration of protein for PAC samples of ~9.5 M. Optical spectra of the various species were obtained using a Sh imadzu UV-2401PC spectrophotometer. 4.4.3. Results and Discussion A hypothesis suggested that CooA is activated after the conformational change of the amino acids chain following the dissociation of Pro 2 due to the binding of CO to the heme. It has been demonstrated that this hypothesis is wrong by Nakajima et al. [14] and Thorsteinsson et al. [21] who used different truncated CooA mutants to show that the CooA mutants behave as an activated wild-t ype CooA. This confir ms that the proline dissociated to the heme, initia tes the CO-dependent activation but is not necessary to the activation of CooA. It has also been demonstrated by Resona nce Raman (RR) spectroscopy that the FeHis 77 bond in CooA is weak compared a Fe -His 93 in myoglobin. The weakness of this bond might explain, after CO bindi ng, the displacement of the heme in the interior of the protein into an adjacent cavity associated wi th a movement of the C-helix which might be part of the activation process of CooA [22,23] The movement of the C-helix will allow the DNA-binding site to reorie nt in order to interact with the DNA and activate the transcription. In fact, recent DFT st udies by Xu et al. demonstrated that FeC/ CO values show a weak hydrogen bond of the imidazole from the His 77, which would favor the movement of the heme [24] It has also been observed that CO bindi ng to the heme inactivated involves a reorientation of the C-helix invcluding th e entire heme pocket and the amino acid terminus [3, 25-27] Ibrahim et al. predict using UVRR spectroscopy a displacement of 2.05 of the heme inside the heme pocket [20] Yamamoto et al. confirmed using 1H NMR that the ligand Pro 2 is displaced by at least 4 from its init ial location after CO binding to the heme [28] This confirms the bent confor mation of the protein due to a large conformational change of the C-helix to a high affinity form following the CO

PAGE 148

126 binding to the heme. Coyle et al. demonstr ate that CO, when binding to the heme, interacts with residues from other subun it such as Ile 113, Gly 117 and Leu 120 and a residue from the same subunit, Leu 116 invol ving a small rotation of the two C-helices and the two hemes [22] This theory confirms the hypothe sis from Akiyama et al. which show a rotation of ~ 8 between both subunits using X-Ray spectroscopy [29] Kubo et al. show, using UVRR spectroscopy, that displacement of the C-helix following the CO binding to the heme disrupts a hydrogen bond between the charged residues Arg 118 from subunit A and Asp 72 from subunit B. Th is disruption creates a more constricted and negative electrostatic field around the pr otein. Kubo et al. also confirm the findings of Coyle et al. [22] that movement of Trp 110 inside the protein follows the conformational changes associated with the C-helix. Finally, Lanzilotta et al. [3] identified a salt-bridge between two re sidues from the hinge region, Arg 138 from subunit A and Glu 59 from subunit B. Movement of the C-helix following CO binding to the heme disrupts this salt-bridge, which was confirmed by Ibrahim et al., using UVRR spectroscopy, signifying that the hinge can be nd and facilitate the rearrangement of the C-helix in order to bind to the DNA site [20] CooA displays an optical spectrum with a Soret maximum at ~422 nm and visible bands at 540 nm ( -band) and 570 nm ( -band) in the oxidized (activated) form (Figure 4.16) The protein was reduced with sodium dithi onite in the presence of CO resulting in a Soret band bathochromic shifted at ~424 nm and visible bands at 540 nm and 570 nm.

PAGE 149

127 400 500 600 0.0 .5 1.0 1.5 2.0 2.5 3.0 3.5 AmplitudeWavelength (nm) CooA as isolated Fe(II)CooA + CO500 550 600 0.0 0.2 0. 0.6 AmplitudeWavelength (nm) Figure 4.16 : Equilibrium optical absorption spectra of CooA as isolated (black solid line) and Fe(II)CooA-CO (purple dotted line). Puranik et al. calculated the overall quantum yield following the complete photodissociation of CO and the release of CO to the solvent using the quantum yield of myoglobin [30] The quantum yield for CooA was calculated equal to 0.02 [31] PAC as discussed previously, can probe molar volume and enthalpy changes over a time scale from ~50 ns to ~20 s [32] Figure 4.17 displays an overlay of PAC traces for CO-CooA and the calorimetric refere nce compound obtained in 25 mM MOPS and 0.1M NaCl (pH 7.4) at 14C. The fact that no frequency shifts are observed between sample and reference acoustics signals indica tes no kinetic events between ~50ns and ~20 s. Moreover, the difference in amplitude s observed between the sample and the reference indicates that the photodissociation of CO involved conformational changes in the protein associated with changes in energy and volume between ~50ns and ~20 s. A plot of (S/R)Eh versus Cp / (equation 9) reveals a H and V associated with CO

PAGE 150

128 photodissociation from the Fe(II)CooA, equal to H ~ 400 30 kcal mol-1 and V ~ 11 4 mL mol-1 occurring (Figure 4.18) 60x10-680x10-6-0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 Normalized AmplitudeTime (s) Fe4SP CooA Figure 4.17 : Overlay of the acoustic waves for photolysis of CO from CooA (purple dotted line) and the reference Fe(III)4SP (black solid line).

PAGE 151

129 Figure 4.18 : Plot of (S/R)* Eh versus Cp / for CO photolysis from CooA 25 mM MOPS and 0.1M NaCl (pH 7.4) between 8C and 32 C which gives a intercept equal to the heat, Q returned to the solvent giving H = ( Eh Q)/ (kcal mol 1) and V = slope/ (mL mol 1). The observed volume and enthalpy varia tions may have contributions from photocleavage events local ized at the heme ( Vheme and Hheme), and/or conformational change of the protein: Hobs = Hheme + Hconf Equation 46 Vobs = Vheme + Vconf Equation 47 The volume and enthalpy changes for CO photodissociation from CooA were reported. PAC results for CO photodisso ciation give an average H of -400 30 kcal mol-1 associated with an average V of -11 4 mL mol-1 for CooA in 25 mM MOPS and 0.1M NaCl (pH 7.4). These results can be compar ed to volume and enthalpy changes for CO photodissociation from Fe(II) porphyrin model systems. Photolysis of CO from an 24681012141618 50 55 60 65 70 Eh (kcal mol-1)C p (kcal mL-1)V = -11 4 mL mol-1H = -400 30 kcal mol-1

PAGE 152

130 Fe(II)heme gives rise to three contributions to H and V: the cleavage of the Fe-CO bond, spin state change of the iron from the lo w spin to the high spin configuration and solvation of CO as it diffuses away from th e heme. Thus, the total changes of enthalpy and molar volume can be rewritten as: Hheme = HFe-CO + HLS-HS + HCOsolv Equation 48 Vheme = VFe-CO + VLS-HS + VCOsolv Equation 49 Upon photolysis, the variation in enthal py for a CO photodissociation from an iron porphyrin has been estimated to be H ~ 14 kcal mol-1 where HFe-CO =17 kcal mol-1 (cleavage of Fe2+CO bond), HLS-HS = <1 kcal mol-1 (heme iron going from low-spin to high-spin) and HCOsolv = -3 kcal mol-1 (diffusion of CO out of the heme pocket and into the bulk solvent) [33,34] Using the above relation Hobs = Hheme + Hconf where Hheme ~ 14 kcal mol-1 and Hobs ~ -400 kcal mol-1, Hconf can be estimated equa l to ~ -414 kcal mol-1 and indicate an additional exothermic process ta king place in <50 ns after CO photolysis. The result of Hconf must be due to a more global conf ormational change of the protein. The overall thermodynamic profile for CO photodissociation is summarized in Figure 4.19

PAGE 153

131 Figure 4.19 : Thermodynamic profiles for CO photo release from Fe(2+)CooA (PDB entry 1ft9). The variation of volume for CO photodi ssociation from an iron porphyrin can been estimated in the same way: VFe-CO = ~5 mL mol-1 (increase of the solvent accessible area), VLS-HS = ~10 mL mol-1 (increase of the electron density associated with the 3dz 2 orbital of the iron) and VCOsolv = ~20 mL mol-1 [40] The three contributions give a Vheme = ~35 mL mol-1. Once again, based on the above equation, Vobs = Vheme + Vconf where Vheme = 35mL mol 1 and Vobs = 11 mL mol-1, Vconf can be estimated to be ~ -46 mL mol-1 (73 3 / molecule). The results of Vconf and Hconf can take place first, if th ere is a change in overall charge distribution on the protein (i.e., change in net protein dipole leading to solvent reorganization), secondly if there is formation of one or more salt-bridge interactions (the

PAGE 154

132 release of electrostricted wa ter molecules upon salt-bridge fo rmation results in volume increases) and/or thirdly if there is an in crease in the solvent accessible van der Waals volume of the protein upon photolysis. Until now, the X-ray crystal structure of the Fe(2+)CooA is the only form found in the literature. As the X-ray crystal struct ure of the COFe(2+)CooA is not yet solved, the changes in solvation, asso ciated with the change in the solvent accessible solvent (SAS) between the CO liganded and unliganded pr otein, can not be calculated in order to confirm the experimental results. However, th e surface residues are likely to experience significant changes in solven t exposure subsequent to the conformational change following the photodissociation of CO to the protein. The photothermal results of horse heart myoglobin, the heme domain of FixL or the heme domain from the Bs HemAT show large difference associated with the photodissociation of CO from the protein. The results from the heme domain of FixL from Bradyrhizobium japonicum show a biphasic relaxation af ter the photodi ssociation of CO. Subsequent to the CO rel ease, the fast phase, with a H of 12 kcal mol-1 and V of 1 mL mol-1, is associated to the reorganization of the solvent following a perturbation of the salt-bridge between Glu 182 and Ar g 227. The slow phase with a lifetime ~ 150 ns, a H of 5 kcal mol-1 and V of 5 mL mol-1 is associated to the escape of the CO molecule to the solvent [35] In the case of horse heart myoglobin, a biphasic relaxation is also observed after photodissociat ion of CO. The first phase with a H of 7 kcal mol-1 and V of -3 mL mol-1 was associated to the solvation of a charge after cleavage of the salt bridge between the Lys 45 (Arg 45 in sperm whale myoglobin) and the heme propionate groups. The slow phase with a lifetime of ~ 600 ns, a H of 8 kcal mol-1 and V of 14 mL mol-1 was partly associated to the reformation of the salt-bridge [36-38] Finally, in the case of Bs HemAT, a monophasic relaxati on was observed following the photodissociation of CO. The re sults, associated with a H of -19 kcal mol-1 and V of 4 mL mol-1, are consistent with an increase of th e SAS of the protein after comparing the X-ray structure of the Bs HemAT between the CN liganded and unliganded forms of the protein [39]

PAGE 155

133 The large H and V following the photodissociation of CO to the heme domain of CooA is likely associated to the displacement of the heme into the interior of the protein with the reorientation of the C-he lix involving the entire heme pocket and the amino acid terminus. This movement will allo w the DNA-binding site to reorient in order to interact with higher affinity to the DNA. In addition to the even t, the disruption of a hydrogen bond between two charged residue s Arg 118 from the subunit A and Asp 72 from the subunit B due to the reorientation of the C-helix is also involved in the large H and V. This disruption is in accord with the contraction of the solvent around the protein. Finally, the large H and V is also associated to the disruption of the saltbridge between two residues Arg 138 from the subunit A and Glu 59 from the subunit B following the movement of the C-helix. All th ese different events upon photolysis of the CO are likely to contribute to the signal initiation/propagation a nd activation of CooA (Figure 4.20) Figure 4.20 : Detail of the two heme-binding domain of CooA showing the Pro 2, His 77, the two salt-bridges: Asp 72 / Arg 118 and Glu 59 / Arg 138 (PDB entry 1ft9).

PAGE 156

134 4.4.4. Conclusion In summary, PAC results reveal a fast c onformational change associated with the photodissociation of CO from CooA. The c onformational change results in a large enthalpy and volume change whic h were associated to a reor ientation of the C-helix in addition to the disruption of a salt-bridge. The thermodynamics resu lts of the CooA are different from the results of the globin c oupled sensor (GCS) HemAT and the PAS heme domain Bj FixL. Although the CooA is a full protein compared to the Bj FixL or HemAT heme domain protein, the results show a fast er change following th e photodissociation of CO implying different signaling mechanisms between proteins from the same family such as PAS domain, or compared to different family such as GCS.

PAGE 157

135 4.2.5. References [1] Aono S., Nakajima H., Saito K and Okada M., Biochem. Biophys. Res. Commun. 228 (1996), 752-756. [2] Kerby R. L., Ludden P. W. and Roberts G. P., J. Bacteriol 177 (1995), 2241-2244. [3] Lanzilotta W. N., Schuller D. J., Thorstein sson M. V., Kerby R. L., Roberts G. P. and Poulos T. L., Nat. Struct. Biol. 7 (2000), 876-880. [4] Shelver D., Kerby R. L., He Y. and Roberts G. P., J. Bacteriol. 177 (1995), 21572163. [5] Shelver D., Kerby R. L., He Y. and Roberts G. P., Proc. Natl. Acad. Sci. U.S.A. 94 (1997), 11216-11220. [6] Aono S., Matsuo T., Shimono T., Ohkubo K., Takasaki H. and Nakajima H., Biochem. Biophys. Res. Commun. 240 (1997), 783-786. [7] Nakajima H., Matsuo T., Tawara T. and Aono S., J. Inorg. Biochem. 78 (2000), 6368. [8] Aono S., Takasaki H., Unno H ., Kamiya T. and Nakajima H., Biochem. Biophys. Res. Commun 261 (1999), 270-275. [9] Thorsteinsson M. V., Kerby R. L. and Roberts G. P., Biochemistry 39 (2000), 82848290. [10] Thorsteinsson M. V., Kerby R. L., Youn H., Conrad M., Serate J., Staples C. R. and Roberts G. P., J. Biol. Chem. 276 (2001), 26807-26813. [11] Nakajima H., Nakagawa E., Ko bayashi K., Tagawa S-I and Aono S., J. Biol. Chem. 276 (2001), 37895-37899. [12] Dawson J. H., Andersson L. Al. and Sono M., J. Biol. Chem. 258 (1983), 1363713645.

PAGE 158

136 [13] Hanson L. K., Eaton W. A., Sligar S. G., Gunsalus I. C., Gouterman M. and Connell C. R., J. Am. Chem. Soc. 98 (1976), 2672-2674. [14] Nakajima H., Honm a Y., Tawara T., Kato T., Park S.-Y., Mikatake H., Shiro Y. and Aono S., J. Biol. Chem. 276 (2001), 7055-7061. [15] Nakajima H. and Aono S., Chem. Lett. (1999), 1233-1234. [16] Uchida T., Ishikawa H., Ishimori K., Morishima I., Nakajima H., Aono S., Mizutani Y. and Kitagawa T., Biochemistry 39 (2000), 12747-12752. [17] Rubtsov I. V., Zhang T ., Nakajima H., Aono S., Rubtsow G. I., Kumazaki S. and Yoshihara K., J. Am. Chem. Soc. 123 (2001), 10056-10062. [18] Kumazaki S., Nakajima H., Sakagushi T ., Nakagawa E., Shinohara H., Yoshihara K. and Aono S., J. Biol. Chem. 275 (2000), 38378-38383. [19] Henry E. R., Sommer J. H., Hofrichter J. and Eaton W. A., J. Mol. Biol. 166 (1983), 443-451. [20] Ibrahim M., Kuchinskas M., Youn H., Kerby R. L., Roberts G. P., Poulos T. L. and Spiro T. G., J. Inorg. Biochem. 101 (2007), 1776-1785. [21] Thorsteinsson M. V., Kerby R. L., C onrad M., Youn H., Staples C. R., Lanzilotta W., Poulos T. L., Serate J. and Roberts G. P., J. Biol. Chem. 275 (2000), 39332-39338. [22] Coyle C. M., Puranik M., Youn H., Niel sen S. B., Williams R. D., Kerby R. L. Roberts G. P. and Spiro T. G., J. Biol. Chem. 278 (2003), 35384-35393. [23] Ibrahim M., Kerby R. L., Puranik M., Wa sbotten I. H., Youn H., Roberts G. P. and Spiro T. G., J. Biol. Chem. 281 (2006), 29165-29173. [24] Xu C., Ibrahim M. and Spiro T. G., Biochemistry 47 (2008), 2379-2387. [25] Youn H., Kerby R. L., Thorsteinsson M. V., Conrad M., Staples C. R., Serate J., Beack J. and Roberts G. P., J. Biol. Chem. 276 (2001), 41603-41610.

PAGE 159

137 [26] Youn H., Kerby R. L., Thorsteinsson M. V., Clark R. W., Burstyn J. N. and Roberts G. P., J. Biol. Chem. 277 (2002), 33616-33623. [27] Youn H., Kerby R. L. Roberts G. P., J. Biol. Chem. 278 (2003), 2333-2340. [28] Yamamoto K., Ishikawa H., Takaha shi S., Ishimori K. and Morishima I., J. Biol. Chem. 276 (2001), 11473-11476. [29] Akiyama S., Fujisawa T., Ishimo ri K., Morishima I. and Aono S., J. Mol. Biol. 341 (2004), 651-668. [30] Gibson Q. H., Olson J. S., Mc kinnie R. E. and Rohlfs R. J., J. Biol. Chem. 261 (1986), 228-239. [31] Puranik M., Brondsted Nielsen S., Youn H., Hvitved A. N., Bourassa J. L., Case M. A., Tengroth C., Balakrishnan G., Thorstein sson M. V., Groves J. T., McLendon G. L., Roberts G. P., Olson J. S. and Spiro T. G., J. Biol. Chem. 279 (2004), 21096-21208. [32] Larsen R.W. and Miksovska J., Coord. Chem. Rev., 251 (2007), 1101–1127. [33] Nutt, David R. and Meuwly, M., Proc. Natl. Acad. Sci. USA, 101 (2004), 5998-6002. [34] Traylor T.G., Acc. Chem. Re., 14 (1981), 102-109. [35] Miksovska J., Suquet C., Satterlee J.D. and Larsen R.W., Biochemistry, 44 (2005), 10028–10036. [36] Angeloni L. and Feis A., Photochem. Photobiol. Sci. 2 (2003), 730–740. [37] Miksovska J., Day J. and Larsen R.W., J. Biol. Inorg. Chem. 8 (2003), 621–625. [38] Westrick J.A., Goodma n J.L. and Peters K.A., Biochemistry 26 (1987), 8313–8318. [39] Mokdad A., Nissen M., Satterlee J. D. and Larsen R. W., FEBS 581 (2007), 45124518.

PAGE 160

138 4.5. Photothermal studies of Carbon Mon oxide Ligand Photodissociated from FixL Sinorhizobium meliloti 4.5.1. Introduction In general, nitrogen is necessary for pl ant growth. The development of plants is dependent on its symbiotic relationship with Rhizobium which will activate nitrogenase and fix nitrogen. The activation of the nitroge nase is possible in restrictive conditions such as low levels of oxygen [1] In the biological oxygen sensors domain, FixL is well known because it is involved in the regulati on of nitrogen fixation gene expression in Rhizobia [2] FixL is comprised of three different domains: a transmembrane domain, a sensor domain and a kinase domain [1] The kinase domain is similar to other sensor histidine kinases and is formed by ~240 amino acid residues [3,4] On the other hand, the sensor domain is different compared to the other sensor histidine kinases. The sensor domain also called the heme domain is formed by ~150 amino acid residues [5] and activates the kinase domain by autophos phorylation of His 285 with adenosine triphosphate (ATP) after photodissociation of the oxygen from the ferrous heme (low tension of oxygen) [6,7] FixL will then control the activity of FixJ, which is the transcriptional activator, by transferring the -phosphate group. FixJ will then activate the gene nifA and fixK transcription [7] Finally, these genes are responsible for activating the proteins that will generate the production of other genes necessary to the fixation for nitrogen [7] FixL is a monomer with two distinct domains where the N-terminal PAS heme domain is the heme based oxygen sensor and the C-terminal is the histidine kinase domain [2,3,6,8] The sensor heme domain is a single iron (III) protoporphyrin IX held into place by four -helices and five -strands [6] The ligand binding center is a highspin five-coordinate in the absence of an e xogenous ligand and is low-spin six-coordinate when an exogenous ligand is bound to the heme [9] The heme is surrounded by a

PAGE 161

139 hydrophobic pocket where the imidazole from a hi stidine is the fifth ligand of the heme and the distal side is crowed by hydrophobi c groups such as Ile 209, Leu 230 and Val 232. It has been proposed that O2 association to the sensor heme domain can involve conformational changes that can be transm itted to the kinase domain. FixL form Bradyrhizobium japonicum ( Bj ) and Synorhizobium meliloti ( Sm ) also named Rhizobium meliloti ( Rm ) are the two most studied FixLs. Th e crystal structure of FixL from B. japonicum and S. meliloti are very similar [10-14] The crystal structure of FixL from B. japonicum with different ligands binding to the heme for instance CO, NO, O2 is described in the literature [15,16] In our knowledge, no crysta l structure of FixL from S. meliloti with bound ligand has been published. Even if the heme pocket between Bj FixL and Sm FixL are similar, the nomenclature is diffe rent between both of them, for instance, the histidine linked on the proximal side of the heme pocket has for nomenclature His 200 in Bj FixL and His 194 in Sm FixL (Figure 4.21)

PAGE 162

140 Figure 4.21 : Heme domain of the Sm FixL (PDB entry 1ewo). Tanaka et al suggested that the re sidues Arg 200, Arg 208, Ile 209, and Ile 210 on the salt-bridge and the hydroge n-bond network contribute to the stabilization of the O2 binding to the heme. They also demonstrated that the residue Arg 214 is involved in the regulation of the kinase activity [15] Following O2 fixation to the sensor domain, the heme goes from a high-spin five coordinate to a low-spin six coordinate which results in a decrease in the activity of the kinase. A decr ease in the activity of the kinase is also observed after the binding of CO or NO to the heme, involving a change of the iron from high-spin to low-spin six coordi nate. This implies that the ir on spin state change of the heme is involved in the kinase activity [17,18] When O2 is photodissociated from the heme sensor, the kinase domain is undergoe s changes which activate FixJ. Two different pathways have been proposed based on the cr ystal structures of the heme domains in order to activate FixJ. First, Gong et al. have observed in Bj FixL, a rearrangement of the hydrogen bond network between the heme 6,7 propionates and Arg 206 and His 214

PAGE 163

141 amino acids residues due to the flattening of the heme after the binding of the ligand [11,12] On the other hand, Miyatake et al. [19,20] have observed in Sm FixL, a very crowded distal pocket which implies that O2 can not bind to the he me without interacting with three residues: Ile 2 09, Leu 230 and Val 232. This implies then that the steric repulsion between Ile 209 and O2 causing conformational change s of the distal pocket of Sm FixL, and subsequently transmission of the si gnal and regulation of the kinase activity (Figure 4.21) Mukai et al. [21] have also suggested that Ile 209 and Ile 210 play an important role in the signal transduction be tween the heme sensor and the kinase. They showed that the conformational change s associated with the binding of O2 due to steric repulsion regulated the kinase activity. Balland et al. s howed that Arg 220 from Bj FixL is important in the signal trans duction as it interacts with O2 bound to the heme. The displacement of Arg 220 inside the heme or th e fixation of a strong li gand in that position is not responsible for the conf ormational changes of the heme but only the formation of a strong hydrogen-bonding network between Arg 220 and the ligand O2 [22] Energetic profiles for Bj FixL by Miksovska et al. demonstr ated a biphasic relaxation after photodissociation of CO from the heme. The first phase involved a contraction of the solvent associated to the change in charge di stribution after the reor ganization of the saltbridge between Glu 182 and Arg 227 or a po ssible reorientation of the Arg 206 after photodissociation of CO from the heme. The second phase with a lifetime of 150 ns was contributed to an expansion of the solvent due to the ligand released to the solvent [9] Ayers and Moffat, and Cusanovich and Meyer suggested that the si gnal originated from the heme domain can be propagated through the linker to the kinase by quaternary structural changes vi a a distortion of the -sheet [23,24] Finally, Reynolds et al. demonstrated the importance of the subunit Arg 200 (R200 for Sm FixL nomenclature) in the stabilization of the kinase inhi bition related to the oxy form of the Sm FixL [25] They showed a clear relation between the H-bond of the R200 and the 6-propionate heme group that stabilized the inactive form of Sm FixL due to O2 binding the heme (Figure 4.21) To further investigate, the importance of the H-bond between the R200 subunit and the heme 6-propionate group, time resolved photothermal methods (including

PAGE 164

142 photoacoustic calorimetry) were used to de termine energetic profiles on R200 mutants on fast (~ ns-ms) time scales [25] The thermodynamics and conformational dynamics following CO photodissociati on from different Fe(2+) Sm FixL were studied using photoacoustic calorimetry (PAC) by resolvi ng the magnitudes and timescales of molar volume and enthalpy changes associated with physiological processes. We can utilize the binding of small molecules, such as carbon monoxide, to determine the mechanism through which ligands leave the heme poc ket after photodissociation and how ligand photodissociation activates the nitrogenase thr ough transmitting the signal to the kinase by conformational changes of the protein. The binding of carbon monoxide to the reduced heme will form a six coordinate low spin iron complex. Although the CO-bound ferrous form of Sm FixL is not a physiologically active form of the enzyme, studies of CO photodissociation can provide insights into the mechanism of the transmission of the signal into the kinase including the role of relevant amino acids in the transmission. In this report, the thermodynamic profiles for th e ligand escaping from the heme active site of the heme domain in Sm FixLH, the Sm FixL wild type ( Sm FixLWT) and four different mutants (R200A(Alanine), R200Q(Glutamine), R200E(Glutamate), R200H(Histidine)) in Tris buffer is presented in order to have a range of polarities and H-bond abilities to compare to the wild-type Sm FixL. Another mutant, I209M(Methionine) was also studied as Mukai et al. showed that the conformati onal changes associated with Ile 209 subunit is involved in the kinase activity (Figure 4.21) [21] To date, no thermodynamic profile for the photo-release of CO molecule from a w ild type FixL has been described in the literature. Previous thermodynamic studies done by Miksovska et al. determined a thermodynamic profile of the heme domain of the Bj FixL after CO photodissociation but nothing on Sm FixL was done [9]

PAGE 165

143 4.5.2. Materials and Methods Sample preparation Samples for PAC were prepared by diluting Sm FixL protein into a buffer containing 20 mM Tris (pH 8.0). The deoxy form of the protein was formed by placing the oxy form of Sm FixL in a quartz optical cuvette that was then sealed with a septum cap and purged with Argon gas. A fresh dithionite solution was added from a buffered stock solution to give a final concentration of ~6 M. The CO-bound form was obtained by saturating solutions of the deoxy Sm FixL with CO resulting in a fina l solution CO concentration of 1 mM (1 atm pressure). The protein concentration for PAC samples was ~5 M while those for transient absorption were ~9 M. 4.5.3. Results The different Sm FixL wild type and mutant s (R200A, R200Q, R200E, R200H) display similar optical spectra regardless of the different mutations with a Soret maximum at ~418 nm and visible bands at ~543 nm and ~577 nm in the oxidized form (Figure 4.22, 4.23)

PAGE 166

144 300350400450500550600 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Sm FixLR200QAbsorbanceWavelenght (nm)500 550 600 00 01 02 03 AbsorbanceWavelenght (nm)300350400450500550600 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 Sm FixLWTAbsorbanceWavelenght (nm)500 550 600 00 01 02 AbsorbanceWavelenght (nm) Figure 4.22 : Equilibrium optical absorption spectra of Sm FixLWT (top panel) and Sm FixLR200Q (bottom panel) as isolated (red dash line), reduced Sm FixL (bleu dot line) and reduced CO bound Sm FixL (black solid line). Sm FixLWT and Sm FixLR200Q concentration: ~5 M in 20 mM Tris (pH 8).

PAGE 167

145 400500600 0.0 0.1 0.2 0.3 0.4 0.5 0.6 Sm FixLHAbsorbanceWavelength (nm)500550600 0.05 0.06 0.07 0.08 0.09 0.10 AbsorbanceWavelength (nm) Figure 4.23 : Equilibrium optical absorption spectra of Sm FixLH as isolated (red dash line), reduced Sm FixLH (bleu dot line) and reduced CO bound Sm FixLH (black solid line). Sm FixLH concentration: ~5 M in 20 mM Tris (pH 8). The different proteins can be reduced with sodium dithionite, resulting in a Soret band bathochromicly shifted to ~434 nm and a broad visible band at ~566 nm. The binding of CO to the ferrous enzymes result s in a Soret band at ~424 nm and visible bands at ~542 nm and ~572 nm. The mutants I209M and Sm FixL heme domain show a different optical spectrum only in the oxidized form. The muta nt I209M displays a Soret band at ~434 nm and a broad visible band centered at ~557 nm and Sm FixLH displays a Soret band at ~400 nm and a broad visible band centered at ~510 nm. The Soret and Q bands wavelengths for the wild type, R 200A, R200Q, R200E, R200H, I209M mutants and heme domain Sm FixLH are summarized in Table 4.1 The fact that the optical absorption spectra of the various forms of Sm FixLWT and mutants (R200A, R200Q,

PAGE 168

146 R200E and R200H) are independent of the natu re of the mutations associated with a range of polarities and H-bond abilities indicat es that the salt-bri dge between the heme 6,7 propionates and R200 within the distal poc ket near the heme group does not have a significant impact on the electronic structure of the heme group. Table 4.1 : Summary of the Soret and Q bands wavelengths for wild type, R200A, R200Q, R200E, R200H, I209M mutants and heme domain of Sm FixLH. Oxidized Form Reduced Form CO bound Form Soret -band -band Soret Q-band Soret -band band WT 418 543 577.2 434 566 424 542 572 R200A 418.2 544 577 434 566 424 542 572 R200Q 417.4 542.8 577 434 565 423.6 542 572 R200E 417.8 543.2 577.2 434 565 423.6 542 572.2 R200H 418.4 544 577 434.4 566 424 542.2 572 I209M 433.8 557 434.4 566 424 542 572.6 HD 395.5 510 434.4 566 423.5 542.5 571.5 Photolysis of CO bound to Sm FixLWT and the five different mutants (R200A, R200Q, R200E, R200H and I209M) probed at 450 nm results in the formation of a five coordinate high-spin heme complex whic h decays back to the pre-flash CO bound complex (Figure 4.24) with monophasic relaxation kinetic. The first-order rate constant obtained were ~2.68 x 102 s 1 for Sm FixLWT and ~2.63 x 102 s 1, ~2.74 x 102 s 1, ~2.42 x 102 s 1, ~2.47 x 102 s 1 and ~3.02 x 102 s 1 for the mutants R200A, R200Q, R200E, R200H and I209M, respectivel y (the corresponding secondorder rate constant is

PAGE 169

147 26.8 M1 s1, 26.3 M1 s1, 27.4 M1 s1, 24.2 M1 s1, 24.7 M1 s1, 30.2 M1 s1 with a solution CO concentration of 1 mM) (Table 4.2) The different rate constants associated to the CO rebinding to the Sm FixL wild type and the five different mutants are very similar to each other. The rate cons tant associated with CO rebinding to Sm FixLI209M is slower which might be associated to the f act that the methionine can form a hydrogen bond between CO molecule and th e sulfur group compare to the alkyl group in isoleucine which has a steric repulsion with CO. Figure 4.24 : Single wavelength transient absorp tion data for CO recombination to Sm FixLWT, Sm FixLR200A, Sm FixLR200Q, Sm FixLR200E, Sm FixLR200H and Sm FixLI209M at 25C. Excitation wavelength was 532 nm (<20 ps, 20 mJ/pulse, 20 Hz). Sample solution conditions are th e same as those reported in Figure 4.22 Time scale: 40ms. -0.010.000.010.020.030.04 -0.2 0.0 02 0.4 0.6 0.8 1.0 12 Dalta AbsorbanceTime (s) WTA R200A R200Q R200E R200H I209M

PAGE 170

148 Table 4.2 : Summary of the rate constant associated with CO rebinding to wild type, R200A, R200Q, R200E, R200H, I209M and heme domain of Sm FixL. FixL t1 (s) k1 (s-1) WTA 0.0268 0.0001 37.3 R200A 0.0263 0.0001 38 R200Q 0.0274 0.0001 36.5 R200E 0.0242 0.0001 41.3 R200H 0.0247 0.0004 40.5 I209M 0.0302 0.0004 33.1 Table 4.3 : Summary of the CO rebinding rate constant for Bj FixLH, truncated Bj FixLH and Sm FixLH. CO rebinding rate constants Bj FixLH140-270 10.2 0.3 s-1 Bj FixLH151-256 17.3 0.1 s-1 Sm FixLWT, R200A, R200Q, R200E, R200H and I209M Between 33 to 41s-1 Table 4.3 summarized the rate constant for CO rebinding to the heme for the Bj FixLH140-270, Bj FixLH151-256, Sm FixLWT and the five different mutants R200A, R200Q, R200E, R200H and I209M. Bj FixLH140-270 show a CO rebinding rate constant equal to 10.2 0.3 s-1, Bj FixLH151-256 equal to 17.3 0.1 s-1 and Sm FixLWT and the five different mutants between 33 to 41 s-1. The results for Sm FixLWT and the five mutants show a faster CO rebinding rate c onstant than the one observed for Bj FixLH140-270 and Bj FixLH151-256. These results indicate that the full protein Sm FixL accelerated the

PAGE 171

149 rebinding of CO and they also demonstrat e that the mutation of R200 and I209 does not affect the rebinding of CO to the he me to compare to the full protein Sm FixL wild type. 10002000300040005000 -150000 -100000 -50000 0 50000 100000 150000 200000 Sm FixLWTNormalized amplitudeTime (ns) Fe4SP Sm FixLWT Fit Residue 1000200030004000 -200000 -100000 0 100000 200000 300000 Sm FixLR200QNormalized AmplitudeTime (ns) Fe4SP R200Q Fit Residue Figure 4.25 : Overlay of the acoustic waves for the photolysis of CO from Sm FixLWT (top panel), Sm FixLR200Q (bottom panel) (cyan dotted lin e) and the reference Fe(III)4SP (black solid line).

PAGE 172

150 Figure 4.25 displays an overlay of PAC traces for COSm FixLWT, COSm FixLR200Q and the calorimetric referen ce compound obtained in 20 mM Tris buffer, pH 8.0. The fact that a frequency shift is obs erved between sample and reference acoustic signals indicates kinetic events occurring between ~50 ns and ~20 s. The deconvolution of the acoustic wave between the sample and the reference demonstrates four different phases for the Sm FixLWT and the five different mutants. The reaction volume and enthalpy changes were calculated using an of 0.86 determined by Rodgers et al. [16] The plot of (S/R)Eh versus Cp / (as per equation 9) reveals a H and V associated with CO photodi ssociation from the Fe(II) Sm FixLWT of -9 8 kcal mol-1 and 10 3 mL mol-1, respectively for the first phase, 18 16 kcal mol-1 and -18 6 mL mol1, for the second phase, -35 20 kcal mol-1 and 11 6 mL mol-1, for the third phase and 31 25 kcal mol-1 and -9 8 mL mol-1, for the fourth phase (Figure 4.26) In order to understand the energetic impact of the salt -bridge between the R200 and the heme 6,7 propionates group compared to the wild type, different mutants were synthesized by Dr Reynolds’s group. The H and V associated with CO photodissociation from the Fe(II) Sm FixLR200A are -7 6 kcal mol-1 and 15 2 mL mol-1, respectively for the first phase, 63 20 kcal mol-1 and -6 4 mL mol-1, for the second phase, 51 20 kcal mol-1 and 28 6 mL mol-1, for the third phase and 18 11 kcal mol-1 and -10 3 mL mol-1, for the fourth phase. The H and V associated with CO photodissociation from the Fe(II) Sm FixLR200Q are -26 13 kcal mol-1 and 17 2 mL mol-1, respectively for the first phase, 62 8 kcal mol-1 and -19 1 mL mol-1, for the second phase, -43 5 kcal mol-1 and 8 0.9 mL mol-1, for the third phase and 9 3 kcal mol-1 and -7.6 0.6 mL mol-1, for the fourth phase (Figure 4.26) The H and V associated with CO photodissociation from the Fe(II) Sm FixLR200E are -10 5 kcal mol-1 and 21 3 mL mol-1, respectively for the first phase, 31 14 kcal mol-1 and -21 2 mL mol-1, for the second phase, -47 6 kcal mol-1 and 10 1 mL mol-1, for the third phase and 43 10 kcal mol-1 and -5 2 mL mol-1, for the fourth phase. The H and V associated with CO photodissociation from the Fe(II) Sm FixLR200H are -39 5 kcal mol-1 and 4.7 0.7 mL mol-1, respectively for the first phase, 74 7 kcal mol-1 and -7 1.1 mL mol-1, for the

PAGE 173

151 second phase, -49 5 kcal mol-1 and 3.9 0.7 mL mol-1, for the third phase and 75 5 kcal mol-1 and -4.3 0.7 mL mol-1, for the fourth phase. 3.03.54.04.55.05.56.06.5 -150 -100 -50 0 50 100 150 Sm FixLWT 1 st phase 2 nd phase 3 rd phase 4 th phaseE h (kcal mol-1)C p (kcal mL-1) 3456789 -200 -150 -100 -50 0 50 100 150 200 250 Eh (kcal mol-1)C p (kcal mL-1) 1 st phase 2 nd phase 3 rd phase 4 th phaseSm FixLR200Q Figure 4.26 : Plot of (S/R)* Eh versus Cp / for CO photolysis from Sm FixLWT (top panel) and Sm FixLR200Q (bottom panel) in 20 mM Tris (pH 8) between 10C and 34C.

PAGE 174

152 Since it has been demonstrated that O2 can not bind to the heme without interacting with three re sidues: Ile 209, Leuv230 and Va lv232, a mutant I209M was synthesized by Dr Reynolds’s gr oup in order to have the en ergetic profiles associated with the conformational change s of the distal pocket of Sm FixL due to the changes in the interaction between I209M and O2, and understand how the transmission of the signal is affected. The H and V associated with CO photodissociation from the Fe(II) Sm FixLI209M are -54 3 kcal mol-1 and 3 0.4 mL mol-1, respectively for the first phase, 84 6 kcal mol-1 and -3.7 1 mL mol-1, for the second phase, -30 13 kcal mol-1 and 5.7 2 mL mol-1, for the third phase and 14 5 kcal mol-1 and -7.8 0.8 mL mol-1, for the fourth phase. The values of H and V, obtained for the four different phases for Sm FixLWT, mutants (R200A, R200Q, R200E, R200H and I209M) and Sm FixLH, are summarized in Table 4.4 Given that the crystal structure of Sm FixL does not exist for the full protein, Bj FixL was used in order to establish the tryptophan positions. Two tryptophanes for one subunit of Bj FixL were observed, one at residue 119 (Trp 119) a nd another at residue 178 (Trp 178). Fluorescence of Sm FixLWT and the five different mutants in the oxy and CO bound form were examined. The overlays of the Sm FixLWT and Sm FixLR200Q results for the oxy and CO bound forms are summarized in Figure 4.27 Previous work by Lakoxicz demonstrated the tryp tophan emission spectra exhibits a blue shift when the tryptophan moves from an aqueous envir onment to a hydrophobic environment, the wavelength maxima of the Sm FixL was ~330 nm [26] The overlay between the oxy and CO bound forms of the Sm FixLWT and Sm FixLR200Q does not show any blue shift but a decrease in the intensity when CO is bound to the heme in comparison to O2 binding. This indicates that the tryptophan residue does not move to a hydrophobic environment but might rotate.

PAGE 175

153 300320340360380400420440 0 50000 100000 150000 200000 250000 IntensityWavelength (nm) Sm FixLWT (oxy) Sm FixLWT (CO bound) 300320340360380400420440 0 20000 40000 60000 80000 100000 120000 140000 160000 180000 IntensityWavelength (nm) Sm FixLR200Q (oxy) Sm FixLR200Q (CO bound) Figure 4.27 : Fluorescence of Sm FixLWT and Sm FixLR200Q in the oxy and CO bound form.

PAGE 176

154 4.5.4. Discussion In order to understand the role of residues R200 and I209, different mutants of Sm FixL full protein where examined using photothermal methods and compared to Sm FixL wild type and the heme domain. Key et al. suggested, using Time-Res olved crystallography, that within 1 s after photolysis, Bj FixLH has relaxed to a conformation wh ich is identical to the deoxy form [27] They also demonstrated that the transmi ssion of the signal afte r photodissociation of CO is not restricted to a si ngle region of the heme but to an ensemble of regions for instance the movement in the FG loop or in the -sheet distal of the heme as well as the movement of the Arg 200. They also show th at after photodissociation of CO, different events are observed in the crystal structure of Bj FixL such as the doming and displacement of the heme, and the collaps e of the hydrophobic residue (Leu 236, Ile 215 and Ile 238) in the distal pocket in order to replace the space left after CO leaves the heme pocket. Furthermore, the displacemen t of the 6 propionate group and the FG loop residues Pro 212, His 213 and Ile 216 was notable as well as the conformational changes of the proximal histidine and the F -helix or the backbone atoms of the H and I -strands with the Leu 236 and Val 253 on the surface of th e protein. Due to the similar nature of Bj FixL and Sm FixL, these conformational changes may also be observed after the photocleavage of CO in Sm FixL. The difference in volum e and enthalpy observed after photodissociation of CO are summarized below. As it was explained earlier, PAC results can probe molar volume and enthalpy changes over a time scale from ~50 ns to ~20 s [28] An overlay of PAC traces for Sm FixLWT, Sm FixLR200Q and the calorimetric reference compound obtained in 20mM Tris at pH = 8 at 34C is displayed in Figure 4.24 A frequency shift is observed between sample and reference acoustic signals which, indicates different ki netic events between ~50 ns and ~20 s in addition to the events occurri ng before < 50 ns. The variation of volume and enthalpy observed after photodissociation of CO from the heme may have contributions from photocleavage events contained in the heme ( Vheme and Hheme), and/or conformational ch ange of the protein:

PAGE 177

155 Hobs = Hheme + Hconf Equation 46 Vobs = Vheme + Vconf Equation 47 PAC results associated with vol ume and enthalpy changes for CO photodissociation to Sm FixLWT, the five different mutants and the Sm FixLH show a quadriphasic relaxation. The re sults are summarized in Table 4.4 The prompt phase shows a H and V associated with CO photodi ssociation from the Fe(II) Sm FixLWT of -9 8 kcal mol-1 and 10 3 mL mol-1, respectively, 18 16 kcal mol-1 and -18 6 mL mol-1, for the second phase, -35 20 kcal mol-1 and 11 6 mL mol-1, for the third phase and 31 25 kcal mol-1 and -9 8 mL mol-1, for the fourth phase. In order to understand the energetic impact of the salt-bridge between the R200 and the heme 6,7 propionates group to compare to the wild type, di fferent mutants (R200A, R200Q, R200E and R200H) were synthesized by Dr Reynolds’s group. The H and V associated with CO photodissociation from the Fe(II) Sm FixLR200A are -7 6 kcal mol-1 and 15 2 mL mol1, respectively for the first phase, 63 20 kcal mol-1 and -6 4 mL mol-1, for the second phase, 51 20 kcal mol-1 and 28 6 mL mol-1, for the third phase and 18 11 kcal mol-1 and -10 3 mL mol-1, for the fourth phase. The H and V associated with CO photodissociation from the Fe(II) Sm FixLR200Q are -26 13 kcal mol-1 and 17 2 mL mol-1, respectively for the first phase, 62 8 kcal mol-1 and -19 1 mL mol-1, for the second phase, -43 5 kcal mol-1 and 8 0.9 mL mol-1, for the third phase and 9 3 kcal mol-1 and -7.6 0.6 mL mol-1, for the fourth phase. The H and V associated with CO photodissociation from the Fe(II) Sm FixLR200E are -10 5 kcal mol-1 and 21 3 mL mol-1, respectively for the first phase, 31 14 kcal mol-1 and -21 2 mL mol-1, for the second phase, -47 6 kcal mol-1 and 10 1 mL mol-1, for the third phase and 43 10 kcal mol-1 and -5 2 mL mol-1, for the fourth phase. The H and V associated with CO photodissociation from the Fe(II) Sm FixLR200H are -39 5 kcal mol-1 and 4.7 0.7 mL mol-1, respectively for the first phase, 74 7 kcal mol-1 and -7 1.1 mL mol-1, for the second phase, -49 5 kcal mol-1 and 3.9 0.7 mL mol-1, for the third phase and 75 5 kcal mol-1 and -4.3 0.7 mL mol-1, for the fourth phase. It has been demonstrated that O2 can not bind to the heme without interact ing with three residues: Ile 209, Leu 230 and

PAGE 178

156 Val 232, thus a mutant I209M was studied in order to have the energetic profiles associated with the changes of the distal pocket of Sm FixL due to the different interaction between Ile 209 and O2. The H and V associated with CO photodissociation from the Fe(II) Sm FixLI209M are -54 3 kcal mol-1 and 3 0.4 mL mol-1, respectively for the first phase, 84 6 kcal mol-1 and -3.7 1 mL mol-1, for the second phase, -30 13 kcal mol-1 and 5.7 2 mL mol-1, for the third phase and 14 5 kcal mol-1 and -7.8 0.8 mL mol-1, for the fourth phase.

PAGE 179

Table 4.4 : Summary of photoacoustic results for the wild type, R200A, R200Q, R200E, R200H, I 209M and heme domain of Sm FixL. 1 (ns) V1 (mL mol-1) H1 (kcal mol-1) 2 (ns) V2 (mL mol-1) H2 (kcal mol-1) 3 (ns) V3 (mL mol-1) H3 (kcal mol-1) 4 (ns) V4 (mL mol-1) H4 (kcal mol-1) WT 1 10 3 -9 8 190 -18 6 18 16 512 11 6 -35 20 1495-9 8 31 25 R200A 1 15 2 -7 6 100 -6 4 63 20 510 28 6 51 20 1510-10 3 18 11 R200Q 1 17 2 -26 13 92 -19 1 62 8 285 8 0.9 -43 5 1252-7.6 0.6 9 3 R200E 1 21 3 -10 5 120 -21 2 31 14 429 10 1 -47 6 1575-5 2 43 10 R200H 1 4.7 0.7 -39 5 95 -7 1.1 74 7 365 3.9 0.7 -49 5 1608-4.3 0.7 75 5 I209M 1 3 0.4 -54 3 115 -3.7 1 84 6 497 5.7 2 -30 13 1493-7.8 0.8 14 5

PAGE 180

158 The results can be compared to volume and enthalpy changes for CO photodissociation from Fe(II) porphyrin mode l systems. Photolysis of CO from Fe(II)heme gives rise to three contributions to H and V as it was already explain earlier: cleavage of Fe-CO bond, spin state chan ge of the iron from the low-spin to the high-spin configuration and solv ation of CO as it diffuses aw ay from the heme. Thus, the total changes of enthalpy and molar volume can be summarized as: Hheme = HFe-CO + HLS-HS + HCOsolv Equation 48 Vheme = VFe-CO + VLS-HS + VCOsolv Equation 49 Upon photolysis, the variation in enthal py for CO photodissociation from an iron porphyrin has been estimated to be H ~ 14 kcal mol-1 where HFe-CO = 17 kcal mol-1 (cleavage of Fe2+CO bond), HLS-HS = <1 kcal mol-1 (heme iron undergoes from low-spin to high spin) and HCOsolv = -3 kcal mol-1 (diffusion of CO out of the heme pocket and into the bulk solvent) [29-31] Using the above relation Hobs = Hheme + Hconf where Hheme ~ 14 kcal mol 1 and Hobs ~ -9 kcal mol-1 for Sm FixLWT, Hconf can be estimated and indicate an additional exothermic process taking pl ace in <50 ns equal to ~ -23 kcal mol-1. Using the same relation, Hconf for the five different mutants and Sm FixLH can also be estimated. For Sm FixLR200A, Hobs equal to ~ -7 kcal mol-1 which gives a Hconf ~ -21 kcal mol-1. Sm FixLR200Q shows a Hobs equal to ~ -26 kcal mol-1 which gives a Hconf ~ -40 kcal mol-1. Sm FixLR200E shows a Hobs equal to ~ -10 kcal mol-1 which gives a Hconf ~ 24 kcal mol-1. Sm FixLR200H shows a Hobs equal to ~ -39 kcal mol-1 which gives a Hconf ~ -53 kcal mol-1. Sm FixLI209M shows a Hobs equal to ~ -54 kcal mol-1 which gives a Hconf ~ -68 kcal mol-1. Finally, Sm FixLH shows a Hobs equal to ~ 8.8 kcal mol-1 (Table 4.5) which gives a Hconf ~ -5.2 kcal mol-1. These results of Hconf must be due to a more global conformational cha nge within the heme domain.

PAGE 181

159 In the same way, the variation of volume for a CO photodissociation from an iron porphyrin can been estimated using the three following contributions: VFe-CO = ~5 mL mol-1 (increase of the solvent accessible area), VLS-HS = ~10 mL mol-1 (increase of the electron density asso ciated with the 3dz 2 orbital of the iron) and VCOsolv = ~20 mL mol-1 [32] The three contributions give a Vheme = ~35 mL mol-1. Once again, based on the above equation, Vobs = Vheme + Vconf where Vheme = 35 mL mol 1 and Vobs = 10 mL mol-1, Vconf can be estimated to be ~ -25 mL mol-1 for Sm FixLWT. As for Hconf, Vconf can also be estimated for the five mutants and Sm FixLH. Sm FixLR200A shows a Vobs equal to 15 mL mol-1 which gives a Vconf ~ -20 mL mol-1. Sm FixLR200Q shows a Vobs equal to 17 mL mol-1 which gives a Vconf ~ -18 mL mol-1. Sm FixLR200E shows a Vobs equal to 21 mL mol-1 which gives a Vconf ~ -14 mL mol-1. Sm FixLR200H shows a Vobs equal to 4.7 mL mol-1 which gives a Vconf ~ -30 mL mol-1. Sm FixLI209M shows a Vobs equal to 3 mL mol-1 which gives a Vconf ~ -32 mL mol-1 (Table 4.4) Finally, Sm FixLH shows a Vobs equal to 21 mL mol-1 (Table 4.5) which gives a Vconf ~ -14 mL mol-1.

PAGE 182

160 Table 4.5 : Summary for photoacoustic results for Bj FixLH, truncated Bj FixLH and Sm FixLH. Fast phase (< 50 ns) Slow phase V (mL mol-1) H (kcal mol-1) (ns) V (mL mol-1) H (kcal mol-1) Bj FixLH140270 14 3 -1.4 0.8 150 6.0 3.5 5.3 0.7 Bj FixLH151256 25 4.0 4.9 0.4 Sm FixLH 21 6 8.8 0.9 The Vconf and Hconf results can be attributed to (1) a change in overall charge distribution on the protein (i .e., change in net protein dipole leading to solvent reorganization), (2) formation of one or more salt-bridge interacti ons (the release of electrostricted water molecules upon salt-bri dge formation results in volume increases) and/or (3) an increase in the solvent accessible van der Waals volume of the protein upon photolysis. The results from the Sm FixLH exhibit a monophasi c relaxation after the photodissociation of CO which is associated with a H of 8.8 kcal mol-1 and V of 21 mL mol-1. These results are very different fr om the heme domain of FixL from Bradyrhizobium japonicum ( Bj FixLH140-270) which show a biphasic relaxation after the photodissociation of CO. Subsequent to th e CO release, the fast phase, with a H of 14 kcal mol-1 and V of -1 mL mol-1, was associated with the reorganization of the solvent following a perturbation of the salt-bridge between Glu 182 and Arg 227. The slow phase with a lifetime ~ 150 ns, a H of 5 kcal mol-1 and V of 5 mL mol-1 was associated to the escape of the CO molecule to the solvent [35] On the other hand, the results for Sm FixLH are similar to the results from th e even further truncated heme domain Bj FixLH151-256 which was truncated at both ends of the polypeptide chain compared to

PAGE 183

161 Bj FixLH140-270, with 11 amino acid residues deleted from the N-terminus and 14 amino acid residues deleted from the C-terminus (Table 4.5) and where H is equal to 5 kcal mol-1 and V equal to 25 mL mol-1. Miksovska et al. conclude d that the truncation of 11 amino acid residues at the N-terminus and 14 amino acid residues at the C-terminus of the PAS heme domain from Bj FixL will induce changes which were associated with the protein surface that accelerates the release of ligand from the protein and/or change in the salt-bridge interactions [9] The results obtained for the variation of volume are similar between Bj FixLH140-270 and Sm FixLH but the variation of enth alpy is almost a factor of two less than both of them. This might be due to the fact that the salt-bridge does not involve the same amino acids for all of the studied proteins. The salt-bridge in Sm FixLH is between R200 and the heme propionate, while in Bj FixLH, it is between the amino acid R206 and the heme propionate. The va riation of volume and enthalpy in Bj FixLH reported by Miksovska et al. [9] are for the salt-bridge fo rmed between E182 and R227. The results for the changes in volume and enthalpy associated with Sm FixLWT, Sm FixLR200A, Sm FixLR200Q and Sm FixLR200E are very similar to each other with a H ~ -12 kcal mol-1 and V ~15 mL mol-1. This indicates that the mutation from an arginine to an alanine, a glutamine or a gl utamic acid residue does not destabilized the salt-bridge between the heme propi onate and Arg 200. On the other hand, Sm FixLR200H shows a significant difference afte r photodissociation of CO with a H of ~ -33 kcal mol1 and V of 4.7 mL mol-1. The mutation from an arginine to a histidine residue shows a volume change smaller but a higher exothermic change. This difference might be due to the fact that the amino acid histidine is shor ter and larger than the amino acid arginine producing a difference in the reorganization of the salt-bridge afte r photodissociation of CO with the 6,7 propionate group.

PAGE 184

162 The results associated with the mutant Sm FixLI209M show a H of ~ -54 kcal mol-1 and V of 3 mL mol-1which differ from Sm FixLWT. This may indicate that the steric repulsion between Ile 209 and the CO molecule is less than with Met 209. This might be due to the fact that the sulfur group on the methionine can hydrogen bond with the CO molecule instead of the repulsice in teractions between th e alkyl group from the Ile 209 and the CO molecule binding the heme. The second phase for Bj FixLH occurs around ~150 ns with a H of 5 kcal mol-1 and V of 5 mL mol-1 was associated by Miksovska et al to the CO molecule leaving the heme pocket and entering the solvent afte r CO photodissociation. The results for the variation of volume and enthalpy for the second relaxation of the Sm FixLWT and the five different mutants are summarized page 157 (Table 4.4) Sm FixLWT shows a VII equal to -18 mL mol-1 and a HII ~ 18 kcal mol-1. Sm FixLR200A shows a VII equal to -6 mL mol-1 and a HII ~ 63 kcal mol-1. Sm FixLR200Q shows a VII equal to -19 mL mol-1 and a HII ~ 62 kcal mol-1. Sm FixLR200E shows a VII equal to -21 mL mol-1 and a HII ~ 31 kcal mol-1. Sm FixLR200H shows a VII equal to -7 mL mol-1 and a HII ~ 74 kcal mol-1. Finally, Sm FixLI209M shows a VII equal to -3.7 mL mol-1 and a HII ~ 84 kcal mol-1. The results associated with the mutants Sm FixLR200A, Sm FixLR200Q, Sm FixLR200E and Sm FixLR200H show difference in volume and enthalpy changes compared to the Sm FixLWT. The change in volume is similar for Sm FixLR200Q and Sm FixLR200E to the Sm FixLWT, this indicates that the mutation does not affect the saltbridge interaction and the rel ease of the CO to the solvent. In contrast, changes for Sm FixLR200A and Sm FixLR200H are unlike Sm FixLWT. The dissimilarity between Sm FixLR200H and Sm FixLWT confirms the previous hypothesis concerning the fact that the histidine is shorter and larger than the arginine which implying different interaction for the salt-bridge between R200 and heme propionate and different conformational changes after photodissociation of CO and releas e of CO to the solvent. The results for Sm FixLR200A are more in accord with the fact that the mutation from an arginine to an alanine will not form a salt-bridge with th e heme propionate (negative charge) as an alanine is neutral to compar e to the arginine which has a positive charge. The second phase Vand H for Sm FixLI209M also confirm also th e previous hypothesis where the

PAGE 185

163 sulfur group on the methionine might fo rm a hydrogen bond causing a steric attraction instead of steric repulsion as demonstrated in Sm FixLWT consequently different conformational change are associated with the release of CO to the solvent. For each mutants, the next two phases vary from the wild type. For the Sm FixLWT, these variation of volume and enthal py may be associated to different events such as the conformational changes of the proximal histidine and the F -helix or the backbone atoms of the H and I -strands at the residue Leu 236 and Val 253 on the surface of the protein as well as the displa cement of the 6 propionate group and the FG loop residues Pro 212, His 213 and Ile 216. The re sults for the variation of volume and enthalpy for the third and fourth rela xation which occurs with a lifetime ~ 400 ns and 1500 ns for the Sm FixLWT and the five different mutants are summarized below (Table 4.4) Sm FixLWT shows a VIII equal to 11 mL mol-1 and a HIII ~ -35 kcal mol-1. Sm FixLR200A shows a VIII equal to 28 mL mol-1 and a HIII ~ 51 kcal mol-1. Sm FixLR200Q shows a VIII equal to 8 mL mol-1 and a HIII ~ -43 kcal mol-1. Sm FixLR200E shows a VIII equal to 10 mL mol-1 and a HIII ~ -47 kcal mol-1. Sm FixLR200H shows a VIII equal to 3.9 mL mol-1 and a HIII ~ -49 kcal mol-1. Sm FixLI209M shows a VIII equal to 5.7 mL mol-1 and a HIII ~ -30 kcal mol-1. Finally, the results for the variati on of volume and enthalpy for the last relaxation show a VIV equal to -9 mL mol-1 and a HIV ~ 31 kcal mol-1 for Sm FixLWT, a VIV equal to -10 mL mol-1 and a HIV ~ 18 kcal mol-1 for Sm FixLR200A, a VIV equal to -7.6 mL mol-1and a HIV ~ 9 kcal mol-1 for Sm FixLR200Q, a VIV equal to -5 mL mol-1 and a HIV ~ 43 kcal mol-1 for Sm FixLR200E, a VIV equal to -4.3 mL mol-1 and a HIV ~ 75 kcal mol-1 for Sm FixLR200H and, a VIV equal to -7.8 mL mol-1 and a HIV ~ 14 kcal mol-1 for Sm FixLI209M. The results indicate that the third and fourth phase of Sm FixLR200A and Sm FixLI209M show the largest difference from Sm FixLWT and confirm the previous hypothesis. Sm FixLR200A which does not form a salt-bridge with the heme propionate show the third phase involves an endothermi c reaction where the wild type and the other mutants show an exothermic reaction confirmi ng a different conformational change and a difference in the transmission of the signal. In addition, the cha nges in volume and enthalpy for the third phase of Sm FixLI209M provide that the steric attrac tion between

PAGE 186

164 I209 and the CO molecule instead of a ster ic repulsion result in changes in the transmission of the signal along the pr otein compared to the wild type. To conclude, each of the mutation for the residue R200 or I209 indicate different thermodynamic profiles after photodissociation of CO. The mutation that largely affected the thermodynamic profiles where Sm FixLR200A, Sm FixLR200H and Sm FixLI209M. The Sm FixLR200A demonstrated that by mutating a positive charge residue (arginine) to a neutral residue (alanine), the salt-bridge can not form and contributing to a different thermodynamic profile for the transmission of the signal. The different thermodynamic profiles for mutant Sm FixLR200H can only be explained by the fact that the new residue is shorter and larger due to the pyridine group compared the arginine but should still form a salt-bridge with the heme propionate because it has a positive. This confirms that the residue R200 is very important in the transm ission of the signal. Finally, the mutation of an isoleucine to a methionine changes the st eric repulsion with the CO molecule to an attractive interaction which alters the conf ormational change afte r photodissociation of CO. This also confirms that the residue I209 is also very important in the transmission of the signal after phot odissociation of CO. These results are consistent with observations by Reynolds et al. [25] who demonstrated using EPR and Resonance Rama n spectroscopy that th e autoxidation rates are faster for Sm FixLR200A and Sm FixLR200E compared to Sm FixLR200Q, Sm FixLR200H and Sm FixLWT. This means that the loss of the polarity in the mutant Sm FixLR200A is incapable of forming a salt-bridge with the 6-propionate group. Therefore, neutralizing the negative char ge on the heme propionate. Thereby having a different thermodynamic profile associated to a different transmission of the signal.

PAGE 187

I n Sm h f a c o t h T a s br n Figure 4.2 8 Sm FixLWT iFigure 4. 2 Previous P eart myoglo a st relaxatio n o ntraction h e cleavage o T he second r e s sociated w i r idge. In th e H = -2.2 kc a 8 the overa l s summariz e 2 8 : Thermo d P AC studie s bin indicate n occurring < V of -3 mL o f a saltb ri d e laxation oc c i th a V of 1 e case of spe r a l mol-1 and l l thermody n e d. d ynamic pro f s of CO pho t biphasic re l < 50 ns is as s mol-1 whic h d ge betwee n c urring ~70 0 1 4 mL mol-1 r m whale m y a V = -10 165 n amic profil e f iles for CO p t odissociati o l axations. In s ociated wit h h is attribute d n Arg 45 an d 0 ns after C O which is at t y oglobin th e mL mol-1. T e for CO ph o p hotoreleas e o n from bot h the case of h h a H of 7 d to the for m d the 6-propi o O photolysi s t ributed to t h e fast relaxa t T he second r o todissociat i e from Fe(2 + h sperm wha l h orse heart m kcal mol-1 a m ation of a g o nate of the s gives a H h e reformati o t ion <50 ns g elaxation w h i on from + ) Sm FixLW T l e and hors e m yoglobin, t a n d with a s m g eminate pai r heme activ e of 8 kcal m o n of the sal t g ives rise to h ich has a T e t he m all r and e site. ol-1 t a

PAGE 188

166 lifetime of ~700 ns has a H = 14.6 kcal mol-1 associated with a contraction V = 5.8 mL mol-1. The differences between these proteins are due primarily to a Lys residue in sperm whale Mb in place of Arg 45 in horse heart Mb which affects the salt-bridge with the heme propionate [33-35] In the case of Bs HemAT, a monophasic relaxation is observed following the photodissociation of CO. The results, associated with a H of 19 kcal mol-1and V of 4 mL mol-1, are consistent with an increase of the SAS of the protein after comparing the X-ray structure of the Bs HemAT between the CN liganded and unliganded forms of the protein [36] Finally, PAC results reveal a fast conformational change associated with th e photodissociat ion of CO from CooA. The conformational change results in a large enthalpy and volume change which were associated to a reorientation of the C-helix in addition to the disr uption of a salt-bridge and hydrogen bonds.

PAGE 189

167 4.5.5. Conclusion In summary, PAC reveals a quadriphasic relaxation for Sm FixLWT, and the five different mutants ( Sm FixLR200A, Sm FixLR200Q, Sm FixLR200E, Sm FixLR200H and Sm FixLI209M) associated with the photodi ssociation of CO. The thermodynamics confirm that the residue R200 and I209 are ve ry important in the transmission of the signal and a mutation involves different c onformational changes which might induce a different signal transmitted at the end. The results for the Sm FixLH show a monophasic relaxation associated with a fast disruption of the salt-bridge and release of the CO to the solvent.

PAGE 190

168 4.5.6. References [1] Rodgers K. R., Tang L., Lukas-R odgers G. S. and Wengenack N. L., Biochemistry 40 (2001), 12932-12942. [2] Gilles-Gonzalez M. A., D itta G. and Helinski D. R., Nature 350 (1991), 170-172. [3] Monson E. K., Weinstein M., D itta G. S. and Helinski D. R., Proc. Natl. Acad. Sci. USA 89 (1992), 4280-4284. [4] Lois A. F., Ditta G. S. and Helinski D. R., J. Bacteriol. 175 (1993), 1103-1109. [5] Schaller G. E., Ladd A. N., Lanahan M. B., Spanbauer J. M. and Bleeker A. B., J. Biol. Chem. 270 (1995), 12526-12530. [6] Shiro Y. and Nakamura H., International Congress Series 1233 (2002), 251-257. [7] Rodgers K. R. and Lukat-Rodgers G. S., J. Inorg. Biochem. 99 (2005), 963-977. [8] Gilles-Gonzalez M. A. and Gonzalez G., J. Biol. Chem ., 268 (1993), 16293-16297. [9] Miksovska J., Suquet C., Satterlee J. D. and Larsen R. W., Biochemistry 44 (2005), 10028-10036. [10] Taylor B. L. and Zhulin I. B., Microbiol. Mol. Biol. Rev. 63 (1999), 479-506. [11] Gong W., Hao B., Mansy S. S., Gonzalez G., Gilles-Gonzalez M. A. and Chan M. K., Proc. Natl. Acad. Sci. U.S.A. 95 (1998), 15177-15182. [12] Gong W., Hao B. and Chan M. K., Biochemistry 39 (2000), 3955-3962. [13] Perutz M. F., Paoli M. and Lesk A. M., Chem. Biol. 6 (1999), R291-R297. [14] Gilles-Gonzalez M. A., Gonzalez G., Perutz M. F., Kiger L., Marden M. and Poyart C., Biochemistry 33 (1994), 8067-8073. [15] Tanaka A., Nakamura H., Shiro Y and Fujii H., Biochemistry 45 (2006), 2515-2523.

PAGE 191

169 [16] Rodgers K. R., Lukat-Rodgers G. S. and Barron J. A., Biochemistry 35 (1996), 9539-9548. [17] Dunham C. M., Dioum E. M., Tuckerma n J. R., Gonzalez G., Scott W. G. and Gilles-Gonzalez M-A., Biochemistry 42 (2003), 7701-7708. [18] Tuckerman J. R., Gonzalez G., Di oum E. M. and Gilles-Gonzalez M-A., Biochemistry 41 (2002), 6170-6177. [19] Miyatake, H., Mukai, M., Park, S. Y., Adachi, S., Tamura, K., Nakamura, H., Nakamura, K., Tsuchiya, T., Iizuka, T. and Shiro, Y., J. Mol. Biol., 301 (2000) 415-431. [20] Miyatake H, Mukai M, Park SY, Adachi S, Tamura K, Nakamura H, Nakamura K, Tsuchiya T, Iizuka T and Shiro Y., J Mol Biol. 301 (2000), 415-31. [21] Mukai M., Nakamura K., Nakamura H., Iizuka T. and Shiro Y., Biochemistry 39 (2000), 13810–13816. [22] Balland V., Bouz hir-Sima L, Kiger L., Marden M. C., Vos M. H., Liebl U. and Mattioli T. A., J. Biol. Chem. 280 (2005), 15279-15288. [23] Ayers R. A. and Moffat K., Biochemistry 47(2008), 12078-12086. [24] Cusanovich M. A. and Meyer T. E., Biochemistry 42 (2003), 4759–4770. [25] Reynolds M., Ackley L., Blizman A ., Lutz Z., Manoff D., Miles M., Pace M., Patterson J., Pozzessere N. and Saia K., Archives of Biochemistry and Biophysics 485 (2009), 150-159. [26] Lakoxicz J.R., Principles of Fl uorescence Spectroscopy, 2nd ed., Kluwer Academic/Plenum, New York, 1999. [27] Key J., Šrajer V., Pahl R. and Moffat K., Biochemistry 46 (2007), 4706-4715. [28] Larsen R.W. and Miksovska J., Coord. Chem. Rev. 251 (2007), 1101–1127. [29] Nutt, David R. and Meuwly, M., Proc. Natl. Acad. Sci. USA, 101 (2004), 5998-6002.

PAGE 192

170 [30] Traylor T.G., Acc. Chem. Res.,. 14 (1981), 102-109. [31] Arora N. and Jayaram B., J. Comp. Chem. 18 (1997), 1245-1252. [32] Morris R. J. and Gibson Q., J. Biol. Chem. 257 (1982), 4869-4874. [33] Angeloni L.and Feis A., Photochem. Photobiol. Sci. 2 (2003), 730–740. [34] Westrick J.A., Goodma n J.L. and Peters K.A., Biochemistry, 26 (1987), 8313–8318. [35] Miksovska J., Day J. and Larsen R.W., J. Biol. Inorg. Chem., 8 (2003), 621–625. [36] Mokdad A., Nissen M., Satterlee J. D. and Larsen R. W., FEBS 581 (2007), 45124518.

PAGE 193

170 Chapter V – Model Proteins Two model proteins were undertaken in orde r to help the analys is of the variation of volume and enthalpy results thei r contribution to cell signaling. The first model protein project was the analysis of the sandbar shark hemoglobin as the hemoglobin is an allosteric protein (Figure 5.1) The binding of oxygen to one subunit will stimulate the conformational change of the subunit. As the four subunits are linked together by amino acids chain, the conformational change of one subunit will increase of the oxygen affinity of the other s ubunits. By studying an al losteric protein, the communication between proteins can be bett er understood. In th e case of the sandbar shark hemoglobin focus was gi ven to subunit interactions. Figure 5.1 : Hemoglobin (PDB entry 1bzo)

PAGE 194

171 Also two peroxidases, the horseradish pe roxidase and the soybean peroxidases, were employed to probe the energetic profile of a ligand, CO molecule, leaving the heme pocket to the solvent as per oxidases have a direct channe l between the heme pocket and the solvent (Figure 5.2) Since no energetic barrier exists for the ligand release from the pocket, the mechanism of the ligand migrati on between the heme distal pocket and the solvent will be drawn. Giving insight to the energetic mechanism of how a ligand gains access to the heme distal pocket. Figure 5.2 : Direct channel in soybean peroxidase (PDB entry 1fhf).

PAGE 195

172 5.1. Kinetic Properties of Polymorphic Hemoglobin from the Sandbar Shark Hemoglobin ( Carcharhinus plumbeus ) 5.1.1. Introduction Hnefeld discovered in 1840 the oxyge n-carrying protein called hemoglobin [1] These globular proteins transport oxygen from the lungs to the rest of the body. Usually in mammals, hemoglobins are formed by four subunits, each composed of a tightly associated chain protein wh ich contains a pocket with the iron heme group. Each hemoglobin consists of two and two subunits non-covalently bound together. The four subunits are linked together by salt-bridges, hydrogen bonds and hydrophobic interactions. When oxygen binds to the heme, it causes the heme to go from a conformation called T-state to a conformation called R-state. This transition of the heme involves the imidazole of the histidine to move and induce at the same time a strain in the protein helix, which will be transmitted to the other subunits. These conformational changes of the other subunits facilitates further the binding of oxygen by the other subunits effectively increasing their binding affinity ; hence the cooperative, behavior of the protein. CO molecule will affect the binding of oxygen to heme, as the binding affinity for CO is 200 times greater than its affinity for oxygen [2] In humans, the two and two subunits of the hemoglobin consist of 141 and 146 amino acid residues, respectively, (also called 22) [20,21] The subunits are the same size and structurally identical. The total mol ecular weight of the hemoglobin is 68 kDa. In general, fish hemoglobins have sim ilar properties as mammalian hemoglobins including molecular weights, aggregations states, positive cooperativity of ligand binding, they show a Bohr effect (Root eff ect) and exhibit hetero tropic responses to allosteric affectors [3] Fish hemoglobins have adapte d to environmental changes and have adapted their functiona l reactivity including hete rogeneity in ligand binding between and subunits, loss of cooperativity at low pH and a large Bohr effect [4-7]

PAGE 196

173 The sandbar shark (Figure 5.3) or Carcharhinus plumbeus is part of the requiem shark family Carcharhinidae. This class of shark can be found in the Atlantic Ocean and the Indo-Pacific [22] Figure 5.3 : Photographs of the sandbar shark [23] The sandbar shark is one of the biggest coas tal sharks in the world and has a very high distinguishable first dorsal fin. Sandbar sh arks are related to the dusky shark, the bignose shark, and the bull shark[22] The color of its body can be from a bluish to a brownish grey to a bronze, with a white or pale underside. Females are longer than males and can grow up between 2 and 2.5 meters compared to 1.8 meters for the male. They can swim alone or in groups which are only com posed by the same sex in shallow coastal and in deeper waters (200 meters) [22] At night, at dawn, and at dusk are usually the time of the day that they are the most active. Sandbar shark ar e found around the world in temperate to tropical waters: from Massachusetts to Brazil (western Atlantic) [22] Despite the importance of the species, very little is known on the biochemistry of the communication between the subunit of fish hemoglobin. The current study is an initial analysis of the kinetic and thermodyna mic processes of the respiratory system of the sandbar shark with a focus on hemoglobin. Like heme oxygen sensors, conformati onal changes associated with ligand binding/dissociation to hemoglobin are thought to initiate conformational change in the hemoglobin. Thus, characterizing those conforma tional changes is critically important to understand the signal transduction process. We have utilized transient absorption

PAGE 197

174 spectroscopy and time-resolved phototherma l methods to examine the thermodynamics and conformational dynamics following CO photodissociation from sandbar shark hemoglobin. Time resolved photothermal method s are particularly important since they can provide energetic profiles on fast (~ ns-ms) time scales. Here, photoacoustic calorimetry (PAC) has been used to probe the H and V profiles subsequent to CO photodissociation from sandbar shark hemogl obin. By examining the time-resolved thermodynamics associated with small molecu le dissociation, the energetic mechanism through which the signal is transmitted along the protein can be determined. Theses studies can provide insights into the mech anism of the signal migration between the different subunits which are re levant to understand how a signa l is transmitted in general inside a protein. 5.1.2. Materials and Methods Sample preparation Samples for PAC were prepared by diluti ng sandbar shark hemoglobin into a buffer containing 50 mM sodium phosphate (p H 6.0, 7.0 and 8.0). The deoxy form of the protein was formed by placing the oxy form of sandbar shark hemoglobin in a quartz optical cuvette that was then sealed with a septum cap and purged with Argon gas. A fresh dithionite solution was a dded from a buffered stock solution to give a concentration of ~13 M for the reduced from of the sandbar shark hemoglobin. The CO-bound form was obtained by saturating so lutions of the deoxy sandbar shark hemoglobin with CO, resulting in a final solution of CO concentration of 1 mM (1 atm pressure). The protein concentration for PAC samples was ~10 M while those for transient absorption were ~1.6 M.

PAGE 198

175 5.1.3. Results The recombinant heme domain of sandbar shark hemoglobin displays an optical spectrum with a Soret maximum at ~ 411 nm and visible bands at 577 nm ( -band) and 542 nm ( -band) (Figure 5.4) The hemoglobin was then reduced using a freshly prepared solution of dithionite resulting in a bathochromic shift of the Soret band to 431 nm with a broad visible band centered at 556 nm. The binding of CO to th e ferrous hemoglobin results in a Soret band at ~419 nm and visi ble bands at 568 nm and 538 nm. The liganded and deoxy forms give a spectrum of the sandbar shark hemoglobin nearly identical to the mammalian hemoglobin A (HbA) spectrum with a Soret maximum at 430 nm and a broad visible band centered at 560 nm. 300400500600700 0.0 0.5 1.0 1.5 2.0 2.5 AbsorbanceWavelength (nm) Sandbar Shark hemoglobin oxidized Sandbar Shark hemoglobin reduced Sandbar Shark hemoglobin + CO bound Figure 5.4 : Steady-state optical absorption sp ectra of sandbar shark hemoglobin (as isolated (red dotted line), deoxy sandbar shark hemoglobin (bleu dashed line) and CO sandbar shark hemoglobin (black solid line) in 50mM Tris (pH=7). [Sandbar shark hemoglobin] = ~10 M. In order to examine the effects of the pH on the photo-release of CO from sandbar shark hemoglobin, PAC data was obtained for sa ndbar shark hemoglobin with the same concentration of Tris (50 mM) but at di fferent pH = 6, 7 and 8. The values of H and

PAGE 199

176 V, obtained from deconvolution of the acoustic waves, are summarized in Table 5.1 Figure 5.5 displays an overlay of PAC traces for CO-sandbar shark hemoglobin and the calorimetric reference compound obtained in 50mM Tris, pH 7.0. The fact that a frequency shift is observed between sample and reference acoustic signals indicates kinetic events occurring between ~50 ns and ~20 s. A plot of (S/R)Eh versus Cp / (as per equation 9) reveals a tr iphasic relaxation with a H and V associated with CO photodissociation from the Fe(II)sandbar shark hemoglobin. At pH = 6, the prompt phase (<50 ns) shows a variation of volume and enthalpy equal to 6.6 mL mol-1 and 19 kcal mol-1 respectively. The intermediate phase at 0.7 s shows a volume expansion of 12.5 mL mol-1 and enthalpy of 48 kcal mol-1 and the slow phase at 2 s shows a volume contraction of -12 mL mol-1 and enthalpy of -22 kcal mol-1. At pH = 7, the prompt phase (<50 ns) shows a variation of volume and enthalpy equal to 5.5 mL mol-1 and 17 kcal mol-1 respectively. The intermediate phase at 0.7 s shows a volume expansion of 22 mL mol-1 and enthalpy of 48 kcal mol-1 and the slow phase at 2 s shows a -30 mL mol-1 and -8 kcal mol-1 (using a = 0.21 calculated using the quant um yield of myoglobin). Table 5.1 : Variation of volume and enthalpy of the CO photodissociation at pH= 6 and 7. Fast phase Intermedia te phase Slow phase 1 ( s) V1 conf (mL mol-1) H1 conf (kcal mol-1) 2 ( s) V2 conf (mL mol-1) H2 conf (kcal mol-1) 3 ( s) V3 conf (mL mol-1) H3 conf (kcal mol-1) pH = 6 <50ns 6.6 19 0.6 12.5 48 2 -12 -22 pH = 7 <50ns 5.5 17 0.6 22 48 2.4 -30 -8

PAGE 200

177 4.0x10-64.5x10-65.0x10-655x10-66.0x10-66.5x10-6-0.08 -0.06 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 Normalized AmplitudeTime (s) Fe4SP Sand bar shark hemoglobin Fit Residue Figure 5.5 : Overlay of the normalized acoustic wave of the reference Fe4SP and the sandbar shark hemoglobin in 50mM Tris at 10 C at pH=7. The absorbance of the sample and reference at the excitation w avelength (532 nm) was ~0.4. Photolysis of the CO bound sandbar shark hemoglobin at 440 nm results in the formation of a five coordinate high-spin heme complex which decays back to the preflash CO bound complex (Figure 5.6, 5.7 and 5.8) and displays a biphasic relaxation kinetic with a pseudo first-orde r rate constant of ~1.1 x 103 s1 and ~1.3 x 102 M1s1 at pH = 6, ~2 x 103 s1 and ~2.2 x 102 M1s1 at pH = 7, ~3.4 x 103 s1 and ~3.1 x 102 M1s1 at pH = 8 for the fast and slow pha ses respectively (the corresponding secondorder rate constant is ~1.1 x 106 s1 and ~1.3 x 105 M1s1 at pH = 6, ~2 x 106 s1 and ~2.2 x 105 M1s1 at pH = 7, ~3.4 x 106 s1 and ~3.1 x 105 M1s1 at pH = 8 for the fast and slow phases respectively with a solution CO concentration of 1 mM) (Table 5.2)

PAGE 201

178 -0.0050.0000.0050.0100.0150.020 -0.015 -0.010 -0.005 0.000 0.005 0.010 0.015 0.020 0.025 Delta AbsorbanceTime (s) 400410420430440450460 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Transient difference spectra UV delta absorbanceNormalized Delta AbsorbanceWavelength (nm) Figure 5.6: Top panel : Single wavelength transient abso rption trace for CO rebinding to sandbar shark hemoglobin (440nm) obtained at pH=6 in 50mM Tris. Bottom panel : Overlay of the transient kinetic in th e Soret region (black squares, obtained 15 ms subsequent to photolysis) and equ ilibrium (bleu empty round) difference spectra (deoxy sandbar shark hemoglobin minus CO sandbar shark hemoglobin). Spectra were constructed from single w avelength date obtained 5 s subsequent to photolysis.

PAGE 202

179 -0.0050.0000.0050.0100.015 -0.010 -0.005 0.000 0.005 0.010 0.015 0.020 Delta AbsorbanceTime (s) 400410420430440450460 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Transient difference spectra UV delta absorbanceNormalized Delta AbsorbanceWavelength (nm) Figure 5.7: Top panel : Single wavelength transient abso rption trace for CO rebinding to sandbar shark hemoglobin (440nm) obtained at pH=7 in 50mM Tris. Bottom panel : Overlay of the transient kinetic in th e Soret region (black squares, obtained 15 ms subsequent to photolysis) and equ ilibrium (bleu empty round) difference spectra (deoxy sandbar shark hemoglobin minus CO sandbar shark hemoglobin).

PAGE 203

180 -0.0050.0000.0050.0100.0150.020 -0.015 -0.010 -0.005 0.000 0.005 0.010 0.015 0.020 0.025 Delta AbsorbanceTime (s) 400410420430440450460 -1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Transient difference spectra UV delta absorbanceNormalized Delta AbsorbanceWavelength (nm) Figure 5.8: Top panel : Single wavelength transient abso rption trace for CO rebinding to sandbar shark hemoglobin (440nm) obtained at pH=8 in 50mM Tris. Bottom panel : Overlay of the transient kinetic in th e Soret region (black squares, obtained 15 ms subsequent to photolysis) and equ ilibrium (bleu empty round) difference spectra (deoxy sandbar shark hemoglobin minus CO sandbar shark hemoglobin).

PAGE 204

181 Table 5.2 : Variation of volume and enthalpy activation of CO photodissociation associated with their rate constant. k1 (ms-1) H 1 (kcal mol-1) S 1 (cal mol-1K-1) k2 (M-1s-1) H 2 (kcal mol-1) S 2 (cal mol-1K-1) pH = 6 1.1x103 5.6 1 -0.06 0.003 1.3 x10218 4 -0.1 0.01 pH = 7 2 x103 2.2 1.1 -0.05 0.003 2.2 x102-38 5 0.08 0.02 pH = 8 3.4 x103 -2 0.4 -0.04 0.001 3.1 x102-6.6 1 -0.02 0.003 These results are different than that of HbA which decays biphasically with a pseudo first order rate constant ~1.4 0.11 x 107 s 1 and ~1 0.4 x 104 M 1s 1 for the fast and slow phase, respectively, at pH = 7.7 [8] In the case of sandbar shark hemoglobin, the fast phase decay was attributed to the geminate recombination of CO to the originating subunit. This is slower by an order of magnitude of three compared to the CO geminate recombination of HbA. The sec ond’s slower phase decay was attributed to geminate recombination of CO by another subu nit of the same hemoglobin. This phase is also slower by an order of magnitude of two when compared to HbA. The sandbar shark hemoglobin shows a lin ear dependence between the rate and the solution pH. In addition, the kinetic opt ical difference spectrum obtained ~15 ms subsequent to photolysis displa ys a bathochromic shift relati ve to the equilibrium optical difference spectrum suggesting an un-relaxed heme pocket conformation subsequent to photolysis at pH = 6, 7 and 8 (Figure 5.6, 5.7, 5.8) The transient difference spectra of sandbar shark hemoglobin display a maximum at 439 nm and a minimum at 421 nm at pH = 6, 7 and 8, respectively which is sim ilar to the transient absorption difference spectrum of HbA which display a maximum at 436 nm and a minimum at 420 nm [8] Since the concentration of transient five c oordinate heme decays bi-exponentially, the unrelaxed heme pocket must persist for ~ milliseconds. From the temperature dependence

PAGE 205

182 of the biphasic recombination rates the activ ation enthalpy and entr opy for the fast and slow phases were found to be 5.6 1 kcal mol-1 and -0.06 0.003 cal mol-1.K-1 at pH = 6, 18 4 kcal mol-1 and -0.1 0.1 cal mol-1 K-1 at pH = 7, 2.2 1.1 kcal mol-1 and -0.05 0.003 cal mol-1 K-1 at pH = 8 for the fast phases, -38 5 kcal mol-1 and 0.08 0.02 cal mol-1 K-1 at pH = 6, -2 0.4 kcal mol-1 and -0.04 0.001 cal mol-1 K-1 at pH = 7, -6.6 1 kcal mol-1 and -0.02 0.003 cal mol-1 K-1 atpH = 8 respectively for the slow phase (Figure 5.6) The results are summarized in Table 5.2 3.203.253.303.353.403.453.50 -0.050 -0.049 -0.048 -0.047 -0.046 -0.045 -0.044 -0.043 1 st phase 2 nd phaseRln(kobsh/k B T)1/T (K-1/10-3) Figure 5.9: Eyring plot for CO recombination to deoxy sandbar shark hemoglobin at pH=6.

PAGE 206

183 3.283323.363.403.44 -0.054 -0.053 -0.052 -0.051 -0.050 -0.049 -0.048 -0.047 -0.046 -0.045 1 st phase 2 nd phaseRln(kobsh/kBT)1/T (K-1/10-3) 3203.253.303. 353.403.453.50 -0.048 -0.047 -0.046 -0.045 -0.044 -0.043 1st phase 2nd phaseRln(kobsh/k B T)1/T (K-1/10-3) Figure 5.10: Eyring plot for CO recombination to deoxy sandbar shark hemoglobin at pH = 7 (top panel) and 8 (bottom panel).

PAGE 207

184 Figure 5.11 corresponds to the photolysis of CO bound sandbar shark hemoglobin at 440 nm at different temper atures (17C and 33C) and pH (6, 7 and 8). The results show that rate constant increases with incr eased of temperature as well as decreased pH (Table 5.2) 0.0000.0050.0100.015 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Delta Absorbance (440nm)Time (s) pH = 6 pH = 7 pH = 8Temperature 17C 0.0000.0050.0100.015 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 Delta Absorbance (440nm)Time (s) pH = 6 pH = 7 pH = 8Temperature 33C Figure 5.11: Single wavelength transient absorption data for CO recombination to sandbar shark hemoglobin at 17C (Top panel) and 33 C (Bottom panel). Excitation wavelength was 532nm (<20 ps, 20 mJ/pulse, 20 Hz).

PAGE 208

185 Kinetic optical difference spectra relativ e to the equilibrium optical difference spectrum obtained ~15 s subsequent to photolysis at pH = 6 and 7. Figure 5.12 displays a bathochromic shift. Changes in this di fference spectra as a function of tim give indication of the relaxation stat e of the heme pocket. From this, we know that the five coordinate heme does not relax to the un-rela xed heme pocket within the time frame of the measurement and must persist for at least 1 ms after phot olysis of the CO.

PAGE 209

186 400410420430440450460 -1.0 -0.5 0.0 0.5 1.0 UV 50s 100s 250s 500s 750s 1msNormalized Delta AbsorbancceWavelength (nm) 400410420430440450460 -0.020 -0.015 -0.010 -0.005 0.000 0.005 0.010 0.015 0.020 UV 50s 100s 250s 500s 750s 1msNormalized De lta AbsorbanceWavelength (nm) Figure 5.12 : Transient difference spectrum overlaid on equilibrium deoxy minus CO-bound difference spectrum at pH = 6 and 7.

PAGE 210

187 5.1.4. Discussion By understanding the thermodynamics followi ng CO binding to fish hemoglobin and the different energetic profiles associated w ith the transmission of the signals along the different subunits, we might be able to do a correlation with human hemoglobin. In the present work, we study the kinetic and ther modynamic profiles for the dissociation and rebinding of CO to the heme of the sa ndbar shark hemoglobin. The CO recombination results show a much lower affinity for the sandbar shark hemoglobin for CO than HbA. This is consistent with results exhibiti ng a liganded T-state conformation which are similar to the previous studies on different fish hemoglobins [9-11] Previous studies have shown that after the photodissociation of CO to the heme (Scheme 5.1) different time relaxations are observed (Table 5.3) The first relaxation is a geminate recombination occurring at ~ 50-70 ns. The second relaxation ~ 0.7-0.8 s corresponds to the geminate recombination of CO which es capes to the heme pocket and rebinds to another subunit of the protein in addition to the R to T re laxation of one subunit of the protein called tertiary relaxa tion. The third relaxation ~ 20 s corresponds to the R to T relaxation of the hemoglobin called quaternar y relaxation. The fourth relaxation ~ 190 s was assigned to the bimolecular rebinding of CO molecule to the R forms of the hemoglobin. The last relaxation ~ 3.8 ms was designated to the bimolecular rebinding of CO molecule to the T forms of the hemoglobin.

PAGE 211

188 Scheme 5.1 : R and T allosteric states of the hemoglobin. Table 5.3 : Different time relaxations of human hemoglobin. I II III IV V Relaxation 40 ns 0.7-0.8 s 20 s 190 s 3.8 ms The transient difference spectra for pH = 6, 7 and 8 indicate that photodissociation of CO results in a non-equilibrium heme environment between the T and R forms subsequent to CO photolysis. In addition, the f act that the decay of the five coordinate heme is biphasic and the transient abso rption difference spectra is red-shift (Figure 5.6, 5.7 and 5.8) indicate that the non-equilibrium heme pocket persists on the milliseconds time scale.

PAGE 212

189 The PAC results can probe molar volume and enthalpy changes over a time scale from ~50 ns to ~20 s [12] An overlay of PAC traces for CO-sandbar shark hemoglobin and the calorimetric reference compound obtaine d in 50 mM Tris at pH = 6 and 7 at 10C (Figure 5.5) reveal a frequency shift between sa mple and reference acoustic signals which indicating different kinetic events between ~50 ns and ~20 s in addition to the events occurring before < 50 ns. The variati on of volume and enthal py observed after the photodissociation of CO from the heme ma y have contributions from photocleavage events localized in the heme ( Vheme and Hheme), and/or conformational change of the protein: Hobs = Hheme + Hconf Equation 46 Vobs = Vheme + Vconf Equation 47 PAC results (Table 5.2) for CO photodissociation to sandbar shark hemoglobin give an average of the variation of volume and enthalpy at pH= 6 and 7 associated with a triphasic relaxation. At pH 6, the prompt phase (<50 ns) shows a average volume expansion and enthalpy equal to 6.6 mL mol-1 and 19 kcal mol-1 respectively. The intermediate phase at 0.7 s shows a 12.5 mL mol-1 expansion and 48 kcal mol-1 and the slow phase at 2 s shows a contractio n of -12 mL mol-1 and -22 kcal mol-1. At pH = 7, the prompt phase (<50 ns) shows a variation of volume and enthalpy equal to an expansion of 5.5 mL mol-1 and 17 kcal mol-1 respectively. The intermediate phase at 0.7 s shows a 22 mL mol-1 and 48 kcal mol-1 and the slow phase at 2 s shows a -30 mL mol-1 and -8 kcal mol-1. These results can be compared to volume and enthalpy changes for CO photodissociation from Fe(II) porphyrin mode l systems. Photolysis of CO from Fe(II)heme gives rise to three contributions to H and V: cleavage of Fe-CO bond, spin state change of the iron from the low-spin to the high-spin configur ation and solvation of CO as it diffuses away from the heme.

PAGE 213

190 Thus, the total changes of enthalpy a nd molar volume can be summarized as: Hheme = HFe-CO + HLS-HS + HCOsolv Equation 48 Vheme = VFe-CO + VLS-HS + VCOsolv Equation 49 Upon photolysis, the variation in enthal py for a CO photodissociation from an iron porphyrin has been estimated to be H ~ 14 kcal mol-1 where HFe-CO = 17 kcal mol1 (cleavage of Fe2+CO bond), HLS-HS = <1 kcal mol-1 (heme iron undergoes from lowspin to high spin) and HCOsolv = -3 kcal mol-1 (diffusion of CO out of the heme pocket and into the bulk solvent) [13-15] Using the above relation Hobs = Hheme + Hconf where Hheme ~ 14 kcal mol 1 and Hobs ~ 19 kcal mol-1 at pH = 6 and 17 kcal mol-1 at pH = 7, Hconf can be estimated and indicate an additional endothermic proce ss taking place in <50 ns equal to ~ 5 kcal mol-1 at pH = 6 and ~ 3 kcal mol-1 at pH = 7. The result of Hconf must be due to a more global conformational change within the heme domain. The variation of volume for a CO phot odissociation from an iron porphyrin can be estimated in the same way: VFe-CO = ~5 mL mol-1 (increase of the solvent accessible area), VLS-HS = ~10 mL mol-1 (increase of the electron density associated with the 3dz 2 orbital of the iron) and VCOsolv = ~20 mL mol-1 [5] The three contributions give a Vheme = ~35 mL mol-1. Once again, based on the above equation, Vobs = Vheme + Vconf where Vheme = 35mL mol 1 and Vobs = 6.6 mL mol-1, Vconf can be estimated to be ~ -28.4 mL mol-1 at pH = 6 and at pH = 7 Vobs = 5.5 mL mol-1 which correspond to Vconf ~ -29.5 mL mol-1. The results of Vconf and Hconf can be thougth to arise from (1) a change in overall charge distribution on th e protein (i.e., change in ne t protein dipole leading to solvent reorganization), (2) form ation of one or more salt-bri dge interactions (the release of electrostricted water mol ecules upon salt-bridge formatio n results in volume increases) and/or (3) an increase in the solvent accessible van der Waals volume of the protein upon

PAGE 214

191 photolysis. Hofrichter et al studied human hemoglobin us ing nanosecond absorption spectra [16] They explained that the relaxation I (< 50 ns) is also accompanied by a spectra change which indicates a structural change of the hemoglobin which can result by a collapse of the globin due to the escape of the CO to the solvent. This observation can be compared to the sandbar shark hemogl obin. The variation of volume and enthalpy ( Vconf = ~ -30 mL mol-1 and Hconf = ~ 3 kcal mol-1 at pH 6 and 7) related to the relaxation I is associated with the CO gemina te recombination to the heme from which it was photodissociated but also with the CO solvation into the bulk solvent which will imply a contraction of the bulk solvent due to the hole leaving by CO from protein matrix. The second relaxation which corresponds to the relaxation II which occurs between 0.7 and 0.9 s was associated by Hofrichter et al. to a pure tertiary change in addition to the CO geminate recombination to another subunit other than the one from which it was photodissociated [16] The results show a Vconf II = ~ 12.5 mL mol 1 at pH 6 and Vconf II = ~ 22 mL mol 1 at pH 7 and Hconf II = ~ 48 kcal mol 1 at pH 6 and 7 corresponding to the variation of volume and enthalpy of the conformational change from R to T-state of the subunit. The difference in Vconf between pH 6 and 7 observed can be due to the fact that the electrostriction of the hemoglobin is not the same between pH 6 and 7. Previous studies demonstrate that the relaxation from the R-state to the T-state of a subunit is associated with the movement of the Fe-His bond The tilted Fe-His configuration relaxation is a ssociated with the moving in plane of the iron. Moreover, previous studies show that th e strain associated with CO binding show no movement out of plane of the iron which indicates a rapid transfer of the binding strain to the new conformation of the Fe-His in terface. To compare, the c onformation associated with O2 binding to the heme show a rema ining tilted conformation of the chain. The third phase which corresponds to th e relaxation III and occurs around ~ 20 s was associated by Hofrichter et al. to a pure quaternary ch ange from R to T state in

PAGE 215

192 addition to a very small CO geminate recomb ination to another subunit than the one from which it was photodissociated [16] In shark hemoglobin, this re laxation was faster by an order of 10 to compare to the HbA. This conformational change was associated with a Vconf III = ~ -12 mL mol 1 and Hconf III = ~ -22 kcal mol-1 at pH 6 and Vconf III = ~ -30 mL mol 1 and Hconf III = ~ -8 kcal mol-1 at pH 7. These results show that the relaxation R to T-state quaternary transition is associated to a contraction of the solvent in addition to an endothermic reaction. In Figure 5.13 the overall thermodynamic profile for CO photodissociation at pH 6 and 7 is summarized.

PAGE 216

193 Figure 5.13 : Thermodynamic profile for CO photolysis from (Fe2+)sandbar shark hemoglobin at pH = 6 and 7.

PAGE 217

194 To compare with the horse heart Mb, tw o thermodynamic phases were observed upon photolysis occurring in <50 ns. The prompt phase coupled with a small endothermic contraction with H = 7 kcal mol 1 and V = -3 mL mol 1 has been related to a solvation of charge formed by disruption of a salt-b ridge between one of the heme propionate groups and Lys 45 (Arg 45 in sperm whale Mb) [17-19] The slow phase with a lifetime around ~ 600 ns was associated to a H = 8 kcal mol 1 and V = 14 mL mol 1 and partly attributed to the salt-bridge reformation. 5.1.5. Conclusion In summary, transient absorption displa ys a biphasic relaxation kinetic which correspond to the CO geminate recombination to the same subunit and to another subunit. PAC reveals a triphasic relaxation associated with the CO geminate recombination, a tertiary transition for one subunit and the R to T quaternary transition of the hemoglobin. The thermodynamics of this transition are similar to the transition in the human hemoglobin with the exception of the third relaxation which is faster for the shark hemoglobin.

PAGE 218

195 5.1.6. References [1] Hnefeld F.L., Die Chemismus in der thierischen Organization Leipzig, (1840). [2] Guyton A C, Medical Physiology 11ed. 2005, page 509. [3] De Young A., Kwiatkowski L. D. and Noble R. W., Fish Hemoglobins, Methods in Enzymology 31 (1994), 124-150. [4] Pennelly R. R., Riggs A. and Noble R. W., Biochim. Biophys. Acta 533 (1978), 120129. [5] Morris R. J. and Gibson Q., J. Biol. Chem. 257 (1982), 4869-4874. [6] Tan A. L., Noble R. W. and Gibson Q. H., J. Biol. Chem. 248 (1973), 2880-2888. [7] Noble R. W., Kwiatkowski L. D., De Young A., Davis B. J., Haedrich R. L., Tam J-T. and Riggs A. F., Biochim. Biophys. Acta 870 (1986), 552-563. [8] Larsen R. W., Bell J. R. and McAuliffe W. J., J. Biochem. Mol. Biol. And Biophys. 3 (1999), 203-210. [9] Binotti I., Giovenco S., Giardina B., Antonini E ., Brunori M. and Wymann J., Arch. Biochem. Biophysic. 142 (1971), 274-280. [10] Bunn H. F. and Forget B. G., Hemoglobin: Molecular, Genetic, and Clinical Aspects, Phildelphia: W. B. Saunders, Co., (1986), 126-167. [11] Friedman J. M., Campbe ll B. F. and Noble R. W., Biophys. Chem. 37 (1990), 43-59. [12] Larsen R.W.and Miksovska J., Coord. Chem. Rev. 251 (2006), 1075–1101. [13] Nutt, David R. and Meuwly, M., Proc. Natl. Acad. Sci. USA, 101 (2004), 5998-6002. [14] Traylor T.G., Acc. Chem. Res. 14 (1981), 102-109. [15] Arora N. and Jayaram B., J. Comp. Chem. 18 (1997), 1245-1252.

PAGE 219

196 [16] Hofrichter J., Sommer J. H ., Henry E. R. and Eaton W. A., Proc. Natl. Acad. Sci. 80 (1983), 2235-2239. [17] Angeloni L. and Feis A., Photochem. Photobiol. Sci 2 (2003), 730–740. [18] Miksovska J., Day J. and Larsen R.W., J. Biol. Inorg. Chem. 8 (2003), 621–625. [19] Westrick J.A., Goodma n J.L. and Peters K.A., Biochemistry 26 (1987), 8313–8318. [20] Hemoglobin: structure, function, evolution, and pa thology by Richard Earl Dickerson, Irving Geis (1983) 176 pages. [21] Human anatomy & physiology by Elaine Nicpon Marieb, Katja Hoehn (2007) 1159 pages. [22] Ferrari A. and A., Sharks (2002) New York: Firefly Books. [23] http://en.wikipedia.o rg/wiki/Sandbar_shark.

PAGE 220

197 5.2. Photothermal Studies of CO Photod issociation from Peroxidases from Horseradish and Soybean 5.2.1. Introduction Plant peroxidases are one of the most diverse families of heme proteins in the biological world. This class of proteins participates in the oxidation of a large array of organic and inorganic compounds vi a heme-catalyzed reduction of H2O2 to water [1] The general reaction for the catalytic cycle of plant pero xidases involving the degradation of various peroxides is described in Scheme 5.2 where AH and A• represent the reduced substrate and its oxidized radical product respectively and compounds I and II are a high valent Fe4+=O porphyrin -cation radical and Fe4+=O heme intermediaries, respectively [1-3] Scheme 5.2: General reaction for the cataly tic cycle of plant peroxidases Plant peroxidases are divided into three families. Class I peroxidases consist of the intracellular peroxidases, examples of which include yeast cyto chrome c peroxidase (CCP) and ascorbate peroxidase (AP) while class II peroxidases include the secretory fungal peroxidases, which are involved in the degradation of lignin (e.g., lignin peroxidases (LiPs) and manganese-dependent peroxidase (MnPs)). Finally, class III peroxidases consist of the secretory plant peroxidases including horseradish and soybean peroxidase [3,4] Fe 3 + porp + H2O2 Compound I + H2O AH Compound II+ A Fe 3 + porp + A + H2O AH

PAGE 221

198 Figure 5.14 : Ribbon structures of horseradish peroxidase (PDB entry 1hch) (top) and soybean peroxidase (PDB entry 1fhf) (bottom). Key distal heme pocket amino acid residues, Arg 38, His 42 and Phe 41 are displayed on the left.

PAGE 222

199 Crystal structures are now available for a nu mber of plant peroxida ses revealing similar structural architectures. Cla ss III plant peroxidases contain three domains (distal heme binding B domain, proximal heme binding F dom ain and the D domain) derived from a single glycosylated polypeptide chain of a pproximately 300 residues, a single iron (III) protoporphyrin IX active site a nd two bound calcium ions (see Figure 5.14 ). The role of the two calcium ions is to stabilize the protein structur e and modulate the catalytic activity of the enzyme [4,5] Within the different members of the peroxidase family, the secondary structure is predominantly -helical, containing on av erage ten helices (A-J). The penta-coordinate heme active site also contains a central ir on exhibiting a quantummechanical admixture of spin states (h igh-spin S = 5/2 and low-spin S = 3/2) [6,7] The heme-binding pocket of these proteins c ontain five highly conserved amino acids consisting of Arg 38, Phe 41 and His 42 on th e distal side and Asn 247 and His 142, on the proximal side with His 142 serving as the fifth ligand of the Fe(III) protoporphyrin IX [8] Finally a common feature of the peroxidase family is the presence of a direct access channel from the heme active site to the su rrounding solvent. Thus, small molecules such as the H2O2 substrate can readily diffuse into the heme active site. Horseradish peroxidase (HRP) and soybean peroxidase (SBP) are two members of the class III peroxidases. HRP is not only im portant in the physiol ogical function of the root system of horseradish (i nvolved in the degradation of H2O2) but also has biotechnology applications in cluding immunoassay diagnostics chemical synthesis and bioremediation [9] SBP has important biotechnologica l applications as well including biosensors and biocatalysts [10] HRP and SBP are structurally related with ~ 60% sequence similarity and share analogous catalytic mechanisms [3,2,8,10] The structures of HRP and SBP share the class III structural topology of 13 -helices and two small sheet regions [2,4] However, SBP exhibits a much hi gher thermal stability and a more open access channel to the heme distal pocke t. The ligand access channel of SBP also contains a binding site for Tris molecules which resembles to the secondary substrate binding site associated with HRP [2]

PAGE 223

200 Comparison of the distal pocket of HRP, SBP and the myoglobins (sperm whale and horse heart) reveals an act ive site of HRP and SBP that is larger (i.e., void volume) than the active site of the myoglobins consiste nt with the fact that the heme pocket of plant peroxidases can bind a va riety of relatively large or ganic substrates. The heme pockets of both peroxidases and globins also contain a distal His residue (His E7 in globins, His 42 in HRP and SBP) that can modul ate the binding of diatomic ligands to the heme iron. Unlike globins, peroxidases contai n an Arg residue (Arg 38) in place of the distal phenylalanine found in globins. Both Hi s 42 and Arg 38 have been shown to play a key role in peroxidase enzymatic activity [12,22] The role of Arg 38 in modulating CO binding/release in ferrous peroxidase was demo nstrated using an R38L mutant of HRP, which gives CO association/di ssociation rates similar to t hose observed for the globins. This is partly due to differences in distal pocket polarity between globins and peroxidases as well as differences in hydrogen bonding [23] While the heme bound CO of myoglobin can only form one hydrogen bond (between CO and His 42), the corresponding CO in HRP can form a hydrogen bond with either the positively charged guanidinium group of Arg 38 or to His 42 [22,18] It has also been observed that the proximal imidazole ligand has a strong hydrogen bond with Asp 243 whic h alters the imid azolate character [21] This, in turn, alters the electronic structure of bound ligands such as CO as evident by a 1932 cm-1 CO stretching band and a 490 cm-1 Fe-C stretching mode. The hydrogen bonding interactions between CO and His 42 gi ve rise to pH dependent dissociation kinetics. At low pH, the rate constant for CO dissociation decreases due to a protonation of His 42, while the rate constant fo r CO association is pH independent [24,25] Of the many time-resolved optical methods typically employed in the study of heme protein dynamics, time resolved photothermal methods are particularly important since they can provide energetic profiles on fast (~ ns-ms) time scales. Here, photoacoustic calorimetry (PAC) has been used to probe the H and V profiles subsequent to CO photodissociation from Fe(II)HRP and Fe(II)SBP. By examining the time-resolved thermodynamics associated with small molecu le dissociation, the energetic mechanism through which ligands gain access to the heme distal pocket can be determined. Although the CO-bound ferrous forms of HRP or SBP are not physiologically relevant, theses

PAGE 224

201 studies can provide insights into the mechan ism of ligand migration between the heme distal pocket and the bulk solvent which are re levant to physiological substrate entry into the enzyme’s active site. 5.2.2. Materials and Methods Sample Preparation: Horseradish peroxidase and soybean peroxi dase were purchased from Sigma-Aldrich and Fe(III)tetrakis(4-sulf onatophenyl)porphine (Fe4SP) was obtained from FrontierScientific, Inc. HRP and SBP were used w ithout further purificat ion. HRP and SBP were solubilized in either 50 mM, 100 mM, 150 mM Tris (pH 7.0), 50 mM Tris/100 mM NaCl (pH 8.0), 50 mM (pH 7.0) or 0.2 M (pH 7.0) phos phate buffer. Samples for PAC studies were placed in a 1cm quartz cuvette sealed with a septum cap and subsequently purged with argon. The proteins were then reduced w ith a freshly prepared solution of sodium dithionite and purged with CO to obtain the HRP-CO or SBP-CO samples. Optical spectra of the various species were obtained using a Shimadzu UV-2401PC spectrophotometer. 5.2.3. Results Both HRP and SBP display similar optical spectra regardless of buffer conditions with a Soret maximum at ~407 nm and a broa d visible band centered at ~505 nm in the oxidized (resting) form (Figure 5.15) Both proteins can be reduced with sodium dithionite, resulting in a Sore t band bathochromicly shifted to 423 nm and a visible band at 555 nm. The binding of CO to the ferrous enzymes results in a Soret band at ~423 nm and visible bands at 550 nm and ~575 nm. The fact that the optical absorption spectra of the various forms of HRP and SBP are indepe ndent of the nature of the buffer system employed indicates that Tris binding within the distal pocket near the heme group of SBP does not have a significant impact on the electronic structure of the heme group.

PAGE 225

202 350400450500550600 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 AbsorbanceWavelenght (nm)500550600 0.0 0.2 AbsorbanceWavelenght (nm)350400450500550600 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 AbsorbanceWavelenght (nm)500550600 0.0 0.2 AbsorbanceWavelenght (nm) Figure 5.15 : Top panel : Equilibrium optical absorption spectra of horseradish peroxidase as isolated (red solid line), reduced HRP (bleu dashed line) and reduced CO bound HRP (black dotted line). HRP in 50 mM Tris and 100 mM NaCl (pH 8.0). Bottom panel : Equilibrium optical absorption spectra of the soybean peroxidase as isolated (red solid line), reduced SBP (bleu dashed line) and reduced CO bound SBP (black dotted line). SBP in 50 mM Tris and 100 mM NaCl (pH 8.0).

PAGE 226

203 Figure 5.16 displays an overlay of PAC tr aces for CO-HRP, CO-SBP and the calorimetric reference compound obtained in 0.2 M phosphate buffer, pH 8.0. The fact that no frequency shifts are observed between sample and referen ce acoustics signals indicates no kinetic events occu rring between ~50 ns and ~20 s. However, differences in amplitudes indicate photolytic events occu rring in < ~50 ns. A plot of (S/R)Eh versus Cp / (as per equation 9) reveals H and V associated with CO photodissociation from the Fe(II)HRP and Fe(II)SBP of 18 7 kcal mol-1/6 1 mL mol-1 for HRP and 20 9 kcal mol-1/2.4 0.6 mL mol-1 for SBP (using a = 0.7 for both HRP and SBP [28] ). In order to examine the effects of a bound Tris molecule on the photo-re lease of a CO from SBP, PAC data was obtained for SBP and HRP w ith differing concentrations of Tris (50 mM, 100 mM and 150 mM). A buffer system consisting of 50 mM Tris and 100 mM NaCl was also utilized in order to probe for the effects of electros tatic interactions upon CO photodissociation. In all cases the acous tic waves of the CO-HRP and CO-SBP overlap in frequency with the reference indicat ing only fast events occurring within the PAC time scale (i.e., < ~50 ns). The values of H and V, obtained from amplitude differences, are summarized in Table 5.4 and 5.5

PAGE 227

204 Figure 5.16 : Overlay of the acoustic waves for the photolysis of CO from HRP (top panel) (bleu dotted line), SBP (bottom panel) (green dotted line) and the reference Fe(III)4SP (black solid line). HRP and SBP in 50 mM Tris and 100 mM NaCl (pH 8.0) at 22C. 2.0x10-64.0x10-66.0x10-68.0x10-6-0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 AmplitudeTime (s)2.0x10-64.0x10-66.0x10-6-0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 AmplitudeTime (s)

PAGE 228

205 The H and V values associated with CO -HRP photodissociation are quite similar regardless of the nature of the buffe r system with the average values of the enthalpy and molar volume changes of H of 16 6 kcal mol-1 and a V of 8.2 0.6 mL mol-1 (~13-3 / molecule). In contrast, the molar volume and enthalpy changes observed for CO-SBP photodissociation are dependant upon the solution buffer conditions. In 50 mM phosphate buffer, pH 7.0, H and V are found to be -9 kcal mol-1 and 4 mL mol-1, respectively, while in 200 mM phosphate buffe r the enthalpy value increases to 20 kcal mol-1. In Tris buffer with concentrations ranging from 50 mM to 150 mM the average values of H and V for SBP are found to be 6 5 kcal mol-1 and 5.7 0.3 mL mol-1 (~8-3 / molecule), respectively. For solutions containing 50 mM Tr is and 100 mM NaCl a H 20 5 kcal mol-1 and a V to 9.1 0.3 mL mol-1 are observed similar to what is observed in 0.2 M phosphate (Figure 5.17) 05101520253035 50 100 150 200 250 300 350 E h (kcal mol-1) C p kcal mL-1) Figure 5.17 : Plot of (S/R)* Eh versus Cp / for CO photolysis from HRP (bleu square) and SBP (green dot) in 50 mM Tris (pH 7.0) between 6C and 34C.

PAGE 229

206 5.2.4. Discussion One of the key features associated with heme proteins is the mechanism through which local heme pocket dynamics modulates both the affinity of the heme-protein complex for a specific ligand and ligand specificity [29-37] Examination of the crystal structures of a wide range of heme proteins wi th structurally variable distal heme pockets reveals complex networks of open pockets and access channels through which the gaseous ligands must traverse in order to bind to the heme iron. The thermodynamic profiles associated with CO photo-release from heme proteins can provide key mechanistic insights into this key proce ss by providing detailed energetic maps associated with ligand exit/entry into to the protein matrix. COHRP Photodissociation In the case of COHRP, photodissoci ation gives rise to an average H of 16 6 kcal mol-1 associated with an average V of 8.2 0.6 mL mol-1 regardless of buffers conditions (phosphate and Tris) occurring in < ~ 50 ns. The observed H and V values can be divided into contributions due to heme-CO dissociation and to the subsequent protein/solvent response. Photol ysis of CO from an Fe(II)hem e in the absence of protein involves cleavage of the Fe-CO bond, change in spin state of the heme iron from the lowspin to high-spin configuration and solvation of CO as it diffuses into the bulk solvent. Thus, the total changes in en thalpy and molar volume for this process can be expressed as: Hheme = HFe-CO + HLS-HS + HCOsolv Equation 48 Vheme = VFe-CO + VLS-HS + VCOsolv Equation 49 where HFe-CO ~ 17 kcal mol-1, HLS-HS ~ 0 kcal mol-1 and HCOsolv ~ -3 kcal mol-1 giving a Hheme ~ 14 kcal mol-1 for photolysis [38,39] In the case of CO-HRP, enthalpy changes associated with the protein response subsequent to photolysis also contribute to the total H giving HTot = Hstr + Hheme where Hstr represents any protein

PAGE 230

207 conformational/solvation changes associ ated with the CO photodissociation and Hheme represents the variation in enthalpy for photolysis of CO from the heme group. From the results in Table 5.4 Hstr = Hobs – Hheme ~ 2 kcal mol-1 which is within the uncertainty of the PAC measurements. The corresponding variation in molar volume can also be expressed as: V = Vstr + Vheme Equation 50 where Vheme is composed of the same three c ontributions as for the variation in enthalpy. Previous studies of various mode l compounds have provided estimates for the changes in volume for these processes as follows: VFe-CO ~ 5 mL mol-1 (the increase is due to changes in the solvent accessible area of the newly created molecules), VLS-HS ~ 10 mL.mol-1 (repulsion of solvent ‘ligands’ due to an increase in electron density associated with the heme iron 3dz 2 orbital) and VCOsolv ~ 20 mL mol-1 give a value for Vheme ~ 35 mL mol-1. Again, using the PAC results provided in Table 5.4 gives, Vstr ~ 8 mL mol-1 – 35 mL mol-1 = -27 mL mol-1 [40] Assuming the protein matrix can accommodate the structural changes associated with the low-spin to high-spin transition as well as the cleavage of the Fe-CO bond Vstr reduces to ~ -12 mL mol-1. The changes in enthalpy support a mechanism through which CO migrates out of the distal heme pocket and into the bulk so lvent subsequent to photo-cleavage in ~ <50 ns. Such rapid ligand escape is consistent with the presence of a direct channel linking the heme group to the bulk solvent. These resu lts further indicate th at this ligand access channel does not contain any significant kinetic barriers for CO escape. The corresponding changes in molar volume, Vstr ~ -12 mL mol-1, cannot arise simply from CO migration into the bulk solvent. In gene ral volume contractions may arise from 1) electrostriction in which a charged amino aci d becomes solvent exposed subsequent to CO release, 2) an overall decrease in the van der Waals volume of the protein and/or 3) input of a solvent molecule(s ) into a protein cavity. Examination of the results in Table 5.4 reveals no significan t changes in either V or H as a function of solution ionic strength ruling out electros triction contributions to Vstr. The magnitude of the volume decrease is, however, consistent with the input of a single wa ter molecule into the protein

PAGE 231

208 subsequent to CO migration to the bulk solv ent resulting in a decrease in solvent molar volume by ~ -18 mL mol-1 (i.e. VH2O ~ 18 mL mol-1). Although there is no data concerning the changes in van der Waals volumes of HRP upon ligand association/release in solution which precludes any estimates of the protein contribution to Vstr the similarity between the molar volume of water and the value of Vstr suggests water input as the major contribution to th e molar volume change upon CO photorelease. Table 5.4 : Summary of PAC results for CO photo-release from ferrous HRP. HRP Sample V (mL mol-1) H (kcal mol-1) 50mM Tris 9.3 0.4 16 6 100mM Tris 9.6 0.6 11 6 150mM Tris 8.9 0.3 21 3 50mM Tris + 100mM NaCl 7.7 0.6 14 8 0.2M Phosphate 6 1 18 7 Average 8.2 0.6 16 6

PAGE 232

209 Table 5.5 : Summary of PAC results for CO photo-release from ferrous SBP. SBP Sample V (mL mol-1) H (kcal mol-1) 0.05 M Phosphate 4 0.7 -9 3 0.2M Phosphate 2.4 0.6 20 9 50mM Tris 5.6 0.3 10 5 100mM Tris 6.4 0.3 4 5 150mM Tris 5.1 0.3 4 4 Average 5.7 0.3 6 5 50mM Tris + 100mM NaCl 9.1 0.3 20 5 Entry of a water molecule into the protein would also be expect ed to contribute to Hstr. The simplest estimate of HH2O-protein is the enthalpy of vaporization for water which arises from a water molecule leaving th e bulk solvent to the vapor phase. For water Hvap = 10.6 kcal mol-1 which does not appear in Hstr estimated in the previous section [41] However, Hvap assumes a water molecule breaking hydrogen bonds with neighboring water molecules in the bulk solv ent and entering an environment in which the water molecule is isolated (i.e., vapor phase). In the contex t of the protein matrix this would be equivalent to a water molecule leaving the bulk and entering a hydrophobic pocket which is thermodynamically unfavorable In contrast, a water molecule breaking hydrogen bonds between itself and neighboring water molecules from the bulk water,

PAGE 233

210 entering the protein and forming new hydrogen bonds with amino acids would result in a negligible change in enthalpy, as observed. Ser 35 Gln176 His 170 Ser 35 Gln176 His 170 Figure 5.18 : Illustration of the differences in heme structure between CO bound (shown in yellow) and deoxy HRP. Examination of the X-ray structures fo r the ferrous and ferrous CO bound forms of HRP reveal very simila r heme structures (see Figure 5.18 ) with slight differences in heme propionate positions. Thus, from the X-ra y structures it is not clear what heme conformational dynamics drive the input of a wa ter molecule into the entrance channel to the heme distal pocket. Either these changes are quite subtle or there are additional dynamics which occur in solution that are not observed in the crysta ls. The structures do reveal the addition of a water molecule near the Gln 176 in the ferrous form consistent with the photothermal results.

PAGE 234

211 Photodissociation of CO from ferrous HRP has also been previously studied using PAC by Feis and Angeloni [42] These authors also obtained a monophasic relaxation within the prompt heat phase subsequent to photolysis. However, the change in volume associated with this relaxation was found to be V= 29.6 mL mol-1 and the enthalpy change was H = 35 kcal mol-1. Although the values of H and V obtained previously are nearly twice those obtained in the presen t study, similar conclusi ons were drawn, i.e., that the observed change in volume is due pr imarily to the displacement of CO to the surrounding solvent. Previous PAC Studies of COHRP Photolysis Feis and Angeloni have also been previ ously studied the photodissociation of CO from ferrous HRP using PAC [42] The results, they obtained, are also a monophasic relaxation within the prompt heat phase subseq uent to photolysis, although the change in volume and enthalpy associated with this relaxation was calculated equal to be V= 30 mL mol-1 and H = 42 kcal mol-1, respectively. Feis and Angeloni also concluded a CO migration into the bulk solvent with no signifi cant energetic barriers. The origin of the disparities in the change of volume and enthal pies (almost a factor of three) observed our studies and the studies by Feis and Angeloni is not clear. Fe is and Angeloni explain their large change in enthalpy (42 kcal mol-1) to the cleavage of Fe-C bond energy. The theoretical value of the cleavage of Fe -C bond energy is equal to 35 kcal mol-1 which was obtained for a gas phase (CO)(imidazole)Feporphyrin model system using molecular dynamics coupled with density functional theory (MD-DFT) [46] More recent studies on the CO-Myoglobin using MD-DFT suggest a Fe -C bond cleavage energy of less than 10 kcal mol-1 [47] Moreover, previous studies in our laboratory demonstrated a photodissociation of CO from Fe(II)tetrasulphona tophenyl porphyrin with both a water or imidazole as a proximal ligand in addition to the study of the photodi ssociation of CO to the ferrous microperoxidase11 equal to 12-17 kcal mol-1 for Fe-C bond cleavage [48] These results are in the ra nge of the change in enthalpy presented in the study.

PAGE 235

212 One of the differences between Feis and A ngeloni studies and these studies is the PAC instrumentation used to generate these results. Feis and Angelino used two different PAC detectors [42] One detector holds the sample; the other one holds the reference. The laser is divided in two in orde r to excite both sample and re ference at the same time. The PAC instrument used in this study employed only one detector. The sample which has to be anaerobic in this case is studied first. Then the reference will replace the sample and the acoustic wave as function of temperature will be collected in the same way than the sample. The sample and the reference are st udied in the same detector holder as the instrument response factor K must be removed in order to take the ratio of the sample and reference signals as demonstrated in equation 9. In this only case, the interface between the sample or reference and the detector holde r is the same which imply that the factor K is the same for both sample and reference sign als. By using two diffe rent detectors holder for the sample and the reference as Feis a nd Angeloni studies, the instrument response factor is not the same which imply that the ratio between the sample and the reference is incorrect. This can explain the difference by a lmost a factor of three between Feis and Angeloni results and the results presented here. COSBP Photodissociation The thermodynamics of CO photodissociation from SBP (Table 5.4 and 5.5) are distinct from those observed for CO-HRP and are dependent upon the solution conditions (Table 5.6 and 5.7) In the presence of 50 mM phosphate buffer, pH 8 Hstr is found to be ~ -23 kcal mol-1 with a corresponding Vstr of -16 mL mol-1 while at higher ionic strength (0.2 M phosphate) Hstr ~ 6 kcal mol-1 and Vstr ~-18 mL mol-1 As with HRP, the changes in molar volume are consistent w ith the input of a water molecule into the heme distal cavity subsequent to CO photol ysis regardless of solution ionic strength. However, in the case of HRP, Hstr ~ 0 presumably due to th e formation of additional hydrogen bonds between the water molecule and amino acid(s) within protein that compensate, energetically, for the loss of hydrogen bonding between the entering water molecule and neighboring water molecules in the bulk solution. This is clearly not the

PAGE 236

213 case in SBP solubilized in phosphate buffer for which Hstr 0 suggesting distinct hydrogen bonding interactions within SBP. In addition, the differences in enthalpy between the high and low ionic strengths would suggest that the hydrogen bonding between the entering water molecule and th e protein matrix is influenced by local charges. At higher ionic streng th the value of the enthalpy ch ange is only slightly lower than Hvap for water indicating minimal H-bond forma tion while at lower ionic strengths Hstr is quite exothermic. Examination of the crys tal structures reveal three Arg residues which are likely involved in stabilization of water clusters within the heme distal pocket (Arg 31) and/or the ligand access channel ne ar the heme edge (Arg 173 and Arg 175). Since these residues are likely to be charged within the pH range used here (pH 7 8) changes in solution ionic strength could significantly impact hydrogen bond formation between the Arg side chains and nearby water molecules. Table 5.6 : Summary of Hstr and Vstr associated with CO photo-release from ferrous HRP. HRP Sample Vstr (mL mol-1) Hstr (kcal mol-1) 50mM Tris -11 2 100mM Tris -10 -3 150mM Tris -11 7 50mM Tris + 100mM NaCl -12 0 0.2M phosphate -14 4 Average -12 2

PAGE 237

214 Table 5.7 : Summary of Hstr and Vstr associated with CO photo-release from ferrous SBP. SBP Sample Vstr (mL mol-1) Hstr (kcal mol-1) 0.05 M phosphate -16 -23 0.2M phosphate -17 6 50mM Tris -14 -4 100mM Tris -14 -10 150mM Tris -15 -10 Average -14 -8 50mM Tris + 100mM NaCl -11 6 The presence of Tris also affects the observed thermodynamics associated with CO release in SBP. The values of Hstr and Vstr for CO-SBP in Tris and Tris + NaCl are -8 kcal mol-1/-14 mL mol-1 and 6 kcal mol-1/-11 mL mol-1, respectively. In both cases the observed volume changes are contractions with magnitudes within th e range of the molar volume of water molecule similar to both HR P and SBP solubilized in phosphate buffer while the corresponding changes in enthal py are distinct. Although the values of Hstr for SBP solubilized in Tris buffer are dependent upon solution ionic st rength suggesting that the hydrogen bonding interactions involving water uptake are dependent upon charged groups they are distinct from those measured for SBP solubilized in phosphate buffer in the absence of Tris. This i ndicates that the pr esence of Tris also affects the hydrogen bonding interactions between bound water molecules and the protein matrix.

PAGE 238

215 Examination of the X-ray structure of SBP (Figure 5.19) reveals a water cluster in the distal heme pocket near a Tris molecule. Sin ce the experiments were carried out in Tris buffer at pH=8.0 and pKa of Tris is 8.1, ~ 50% of the Tris molecules are protonated (RNH3 +) resulting in a charge center near the water cluster. At higher ionic strengths it is likely that protonated Tris molecules will form ion pairs with the Clions resulting in charge neutralization of the bound Tris. Si nce hydrogen bonding is an electrostatic interaction, changes in charge density near hydrogen bonding sites may have a significant influence on the enthalpy of H-bond forma tion accounting for the difference in enthalpy between the SBP in Tris and Tris + 100 mM NaCl. Tris Arg31 Arg175 Arg173 Tris Arg31 Arg175 Arg173 Figure 5.19 : Illustration of the heme active site of ferric SBP showing the position of the bound Tris molecule.

PAGE 239

216 It is of interest to note that Hstr is the same for SBP photolysis in the presence of either Tris 100 mM NaCl or 0.2 M phospha te buffer. Since there is no evidence of phosphate binding to SBP this would suggest that water uptake upon CO photolysis for SBP is not influenced by the presence of a ne utral Tris molecule but rather the presence of a charge in distal heme poc ket. In addition, the fact that Hstr is 6 kcal mol-1 for SBP and ~ 0 kcal mol-1 for HRP suggests a different h ydrogen bonding network for the bound water molecules between the tw o proteins. This is somewhat consistent with the X-ray structures for the oxidized forms of both proteins which indicate differences in the solvent access channel with that of SBP be ing more open than that for HRP. The thermodynamic profiles of HRP and SBP are summarized in Figure 5.20

PAGE 240

217 Figure 5.20 : Thermodynamic profiles for CO pho torelease from ferrous HRP and SBP under the various solvent conditions. HRP 0.2M NaPhos SBP 0.2 M Phos./ 50 mMTris+ 100 mMNaCl SBP 50 mMTris 5 kcal mol-1 SBP 50 mMNaPhos HRP 0.2M NaPhos SBP 0.2 M Phos./ 50 mMTris+ 100 mMNaCl SBP 50 mMTris 5 kcal mol-1 SBP 50 mMNaPhos HRP 0.2 M NaPhos SBP 50 mMTris+ 100 mMNaCl SBP 50 mMTris 2 mLmol-1 SBP 0.2 M Phos. SBP 0.05 M Phos. HRP 0.2 M NaPhos SBP 50 mMTris+ 100 mMNaCl SBP 50 mMTris 2 mLmol-1 SBP 0.2 M Phos. SBP 0.05 M Phos. h + CO h

PAGE 241

218 Miksovska et al. have studied CO photodi ssociation from globin type peroxidase, dehaloperoxidase from A. ornata using photoacoustic calorimetry [43] The relatively small molar volume and enthalpy changes asso ciated with CO photorelease from this enzyme, V = 9.4 0.6 mL mol-1 and H = 8 3 kcal mol-1, were attributed to the high flexibility of His 55 in the distal heme bindi ng site and associated with the alteration of hydrogen bonding networks. Interestingly, liga nd escape to the surr ounding solvent was reported to occur within 50 ns upon photodisso ciation suggesting th at a rapid ligand escape to the surrounding solvent is also a characteristic for enzymes with peroxidase activity. Comparison with Myoglobin The PAC results for CO-HRP and CO-SBP are also quite distinct from those observed for the globin class of proteins. Pr evious PAC studies of CO photodissociation from both sperm whale and horse heart myogl obin indicate biphasic relaxations while both HRP and SBP display monophasic relaxati ons. In the case of horse heart myoglobin, the fast relaxation occurring < 50 ns is associated with a H of 7 kcal mol-1 associated with a small contraction V of -3 mL mol-1 which is attributed to the formation of a geminate pair and the cleavage of a salt-b ridge between Arg 45 a nd the 6-propionate of the heme active site. The second relaxation oc curring ~700 ns after CO photolysis gives a H of 8 kcal mol-1 associated with a V of 14 mL mol-1 which is attributed to the reformation of the salt-bridge. [44,45] In the case of sperm whale myoglobin the fast relaxation <50 ns gives rise to a H = -2.2 kcal mol-1 and a V = -10 mL mol-1. The second relaxation which has a lifetime of ~700 ns has a H = 14.6 kcal mol-1 associated with a contraction V = 5.8 mL mol-1 [44,45] The differences between these proteins are due primarily to a Lys residue in sperm whal e Mb in place of Arg 45 in horse heart Mb which affects the salt-bridge with the heme propionate. Transient docking sites, energy barriers betw een transient states and even kinetics of ligand migration were studied by computati on in order to offer the pathways of the ligand entry or exit into sperm whale Mb [29-34] For example, room temperature

PAGE 242

219 molecular dynamics (MD) simulations were used by Ruscio et al. [35] to probe not only the trajectories for CO diffusion from the heme group to the bulk solvent but also trajectories for the diffusion of CO from th e bulk solvent into the protein matrix and approaching the heme active site. The re sults of theses MD studies by Ruscio demonstrated in terms of ligand escape s ubsequent to Fe-CO bond cleavage, that the pathways is primarily involving two Xe atom docking sites, namely Xe4 and Xe1 which is consistent with previous time resolved X-ray studies of the photodissociation of CO from the ferrous protein [36,37] Ligand migration pathways, in the case of the peroxidases, have not yet been examined either computationally or usi ng X-ray crystallography to probe Xe binding sites comparable to those observed in M b. Only three populations of CO were found using a recent step-scan FTIR study of the CO stretching frequency subsequent to CO photolysis [49] One of these populations reveals a tr ansient ‘surface bound’ state and the other two reveal different conformers of th e heme bound CO. The FTIR signals show that the decay for these three populations has the same rate constants with no evidence of transiently docked CO to sites within the pr otein matrix even if the time resolution was only ~ 15 ms. Using the CASTp algorithm, th e interior cavities of both SBP and HRP were studied and demonstrated a large internal ca vities (1,332 3 in HRP and 1,569 3 in SBP) surrounding the heme active site which ar e in direct contact with the bulk solvent (see Figure 5.21 ) [50] Additional internal cavities were also found in both HRP and SBP but these cavities do not appear to have a di rect connection with the distal heme pocket which implies that it is not likely to be acces sible to CO after photolysis. In contrast, sperm whale Mb demonstrated several inte rnal cavities (the largest is 1,182 3) which have directly access to the distal heme pocke t. These cavities provide not only transient docking sites but also ligand escape channels similar to those identified in the MD simulations discussed earlier. These analyses demonstrated that unl ike Mb, the photolysis of both COHRP and COSBP results in a rapi d migration of CO from the heme pocket into the bulk solvent which in the same time allowing one or more water molecules to enter through the same heme access channel (Figure 5.22)

PAGE 243

220 Figure 5.21 : Illustration showing the largest intern al cavities in horseradish peroxidase and sperm whale Mb using CASTp.

PAGE 244

221 Figure 5.22 : Summary of the pathway of CO after pho tolysis in HRP and SBP with the view of the heme pocket for Soybean peroxidase do wn the direct channel. CO is bound to the heme iron (a). Then CO is photodissociated from the heme iron (b). In the first fifty nanoseconds, CO leaves the heme pocket to th e surrounding solvent by the direct channel and two water molecules are input of into the ligand access channel (c). c < 50ns a b h

PAGE 245

222 5.2.5. Conclusion In summary, the results presented here demonstrate unique thermodynamics for CO release from heme peroxidases that refl ect the open access channel between the bulk solvent and the heme active site. The results are consistent with water molecules being input into the substrate ch annel upon CO release with varying degrees of hydrogen bonding depending upon the nature of the protein. In the case of HRP, the results suggest a single water molecule being taken up upon CO release and this uptake is not affected by the nature of the solvent. In contrast, CO photorelease from SBP is dependent upon the presence of Tris docked within the heme di stal pocket as well as the solution ionic strength. The Tris molecule affects the hydr ogen bonding network of the associated water molecules. In addition, the data suggest that it is the charge on the Tris molecule that is most influential. These results are clearly di stinct from the globin class of proteins, e.g., myoglobin, highlighting the role of the protei n matrix in modulating the energy barriers to ligand access to the distal heme pocket.

PAGE 246

223 5.2.6. References [1] Zelent B., Kaposi A. D., Nucci N. V., Shar p K. A., Dalosto S., D ., Wright W. W. and Vanderkooi J. M., J. Phys. Chem., 108 (2004), 10317-10324. [2] Henriksen A., Mirza O., Indiani C., Teilum K., Smulevich G., Welinder K. G. and Gajhede M., Protein Sci., 10 (2001), 108-115. [3] Gajhede M., Schuller D. J., Henrikse n A., Smith A. T. and Poulos T. L., Nat. Struct. Biol., 4 (1997), 1032-1038. [4] Hiraga S., Sasaki K., Ito H., Ohashi Y and Matsui H., Plant Cell Physiol., 42 (2001), 462-468. [5] Veitch N. C., Phytochemistry, 65 (2004), 249-259. [6] Indiani C., Feis A., Howes B. D., Marzocchi M. P. and Smulevich G., J. Inorg. Biochem., 79 (2000), 269-274. [7] Howes B. D., Schiodt C. B., Welinder K. G., Marzocchi M. P., Ma J.G., Zhang J., Shelnutt J. A and Smulevich G., Biophys. J ., 77 (1999), 478-492. [8] Kamal J.K. A. and Behere D. V., J. Inorg. Biochem ., 94 (2002), 236-242. [9] Veitch N. C., Phytochemistry Reviews, 3 (2004), 3-18. [10] Ryan B. J., Carolan N. and O’Fagain C., Trends in Biotechnology, 24 (2006), 355363. [11] Bedard P,.and Mabrouk P. A., Biochem. Biol. Res. Comm., 240 (1997), 65-67. [12] Finkelstein I. J., Ishikawa H., Ki m S., Massari A. M. and Fayer M. D., PNAS, 104 (2007), 2637-2642. [13] Kaposi A.D., Vanderkooi J. M. and Stavrov S. S., Biophys. J., 91 (2006), 41914200.

PAGE 247

224 [14] Kaposi A.D., Fidy J., Manas E. S., Vanderkooi J.M., Wright W.W., Biochimica et Biophysica Acta, 1435 (1999), 41-50. [15] Kaposi A.D., Prabhu N.V., Dalosto S.D ., Sharp K.A., Wright W.W., Stavrov S.S. and Vanderkooi J.M., Biophysical Chemistry, 106 (2003), 1-14. [16] Kaposi A.D., vanderkooi J.M., Wri ght W.W., Fidy J. and Stavrov S.S., Biophys. J., 81 (2001), 3472-3482. [17] Henriksen A., Schuller D. J., Meno K., Welinder K.G., Smith A. T. and Gajhede M., Biochemistry, 37 (1998), 8054-8060. [18] Carlsson G.H., Nicholls P., Svis tunenko D., Berglund G.I. and Hajdu J., Biochemistry, 44 (2005), 635-642. [19] Khajehpour M., Troxler T and Vanderkooi J.M., Biochemistry 42 (2003), 26722679. [20] Ingledew W.J. and Rich P.R., Biochemical Society Transactions, 33 (2005), 886889. [21] Evengelista-Kirkup R., Smulevich G. and Spiro T. G., Biochemistry, 25 (1986), 4420-4425. [22] Feis A., Rodriguez-Lopez J. N., Thorneley R. N. F. and Smulevich G., Biochemistry, 37 (1998), 13575-13581. [23] Meunier B., Rodriguez-Lop ez J. N., Smith A. T., Thorneley R. N. F. and Rich P. R., Biochemistry, 34 (1995), 14687-14692. [24] Coletta M., Ascoli F., Brunori M. and Traylor T. G, J. Biol. Chem ., 261 (1986), 9811-9814. [25] Smulevich G., Paoli M., De Sanctis G ., Mantini A. R., Ascoli F. and Coletta M., Biochemistry, 36 (1997), 640-649. [26] Larsen R. W. and Miksovska J., Coord. Chem. Rev ., 251 (2007), 1101-1127.

PAGE 248

225 [27] Gensch T. and Viappiani C., Photochem. Photobiol. Sci ., 2 (2003), 699-721. [28] Brunori, M., Giacometti, G.M ., Antonini, E., and Wyman, J., Proc Natl Acad Sci U S A ., 70 (1973), 3141-3144. [29] Bossa, C., Amadei, A., Daidone, I., Anse lmi, M., Vallone, B., Brunori, M. and Di Nola, A., Biophys. J., 89 (2005), 465-474. [30] Nutt, David R. and Meuwly, M., Proc. Natl. Acad. Sci. USA, 101 (2004), 5998-6002. [31] Cohen, J. and Schulten, K., Biophys. J., 93 (2007), 3591-3600. [32] Bossa, C., Anselmi, M., Roccatano, D., Amadei, A., Vallone, B., Brunori, M., and Di Nola, A., Biophys. J. ,86 (2004), 3855-3862. [33] Plattner, N. and Meuwly, M., Biophys. J., 94 (2008), 2505-2515. [34] Cohen, J., Arkhipov, A., Braun, R. and Schulten, K., Biophys. J., 91 (2006), 18441857. [35] Ruscio, J.Z., Kumar, D., Shukla, M., Pr isant, M.G., Murali, T.M., and Onufriev, A. V., Proc. Natl. Acad. Sci. USA, 105 (2008), 9204-9209. [36] Srajer, V., Teng, T-y., Ursby, T., Praderva nd, C., Ren, Z., Adachi, S-i., Schildkamp, W., Bourgeois, D., Wulff, M. and Moffat, K., Science, 274 (1996), 1726-1729. [37] Della Longa, S., Arcovito, A., Benfa tto, M., Congiu-Castella no, A., Girasole, M., Hazemann, J.L., and Lo Bosco, A., Proc. Natl. Acad. Sci. USA, 100 (2003), 8704-8709. [38] Norris C. L. and Peters K. S., Biophysic. J., 65 (1993), 1660-1665. [39] Traylor T.G., Acc. Chem. Res., 14 (1981), 102-109. [40] R. van Eldik, A. Asano, and W. J. le Noble, Chemical Rev., 89 (1989), 549-688. [41] Murphy, D.M., and Koop, T., Q. J. R. Meteorol. Soc., 131 (2005), 1539-1565. [42] Feis A. and Angeloni L., J. Phys. Chem., 105 (2001), 2638-2643.

PAGE 249

226 [43] Miksovska J., Horsa, S., Davis M.F., and Franzen S., Biochemistry 47 (2008), 11510-11517. [44] Angeloni L. and Feis A., Photochem. Photobiol. Sci., 2 (2003), 730-740. [45] Westrick J. A., Goodman J. L. and Peters K. A., Biochemistry, 26 (1987), 83138318. [46] Rovira, C., Kunc, K., Hutter, J., Ballone, P., and Parrinello, M., J. Phys. Chem. A, 101 (1997), 8914-8925. [47] Strickland, N., Mulholla nd, A.J., and Harvey, J.N., Biophys. J., 81 (2006), L27-L29. [48] Miksovska, J., Norstrom, J. and Larsen, R.W., Inorg. Chem., 44 (2005), 1006-1014. [49] Marechal, A., Inglew ood, W. J. and Rich, P.R., Biochem. Soc. Trans., 36 (2008), 1165-1168. [50] Dundas, J., Ouyang, Z., Tseng, J., Bi nkowski, A., Turpaz, Y., and Liang, J. Nucl. Acids Res., 34 (2006), 116-118.

PAGE 250

227 Chapter VI – Conclusion 6.1. Summary and Conclusion The research presented in this disserta tion is mainly focused on the understanding of the transmission of the signal in signali ng proteins using time-resolved thermodynamic methods. Cell signaling is a complex system of communication and is involved in the development and coordination of the cell. The ability of the cell to sense and respond to a signal will be implicated in the development and the immunity of the cell. Some diseases like cancer, diabetes or autoimmunity can incr ease if an error appears in the transmission of the signal. Thus, by understanding the m echanism of cells signaling, more efficient treatments for diseases might be developed. As the signaling pathway in large organisms such as eukaryotes is more complex than small organisms, for instance, bacteria, the study of bacterial systems is ve ry attractive. The communication in bacteria is similar to eukaryote cells (extra/intracellu lar communication with others cells or organelles and/or adaptation to environmental changes). Thus the study of the si gnal, activation and transmission of bacteria will be similar but simpler. Different types of signals (amino acids, different pH or temperat ures, or gas molecules) can bi nd to the sensor domain in bacteria, involve a cascade of phosphorylati on and transmit the signal to the flagella which will then reply to the sign al and will tumble if the signal is repellent or go forward if the signal is attractant. The different ligands that can bind the heme sensor and activate the different steps involved in the transmission of the signal, are gas molecules such as O2, CO and NO. The study of soluble heme oxygen sensor will give an opportunity of understanding in detail the mechanism of small molecule binding to a sensing domain and the transmission of the signal to the effector domain. Time-resolved photothermal methods as well as transient optical techniques were used to obtain th ermodynamic profiles for ligand binding/release in heme based signaling prot eins including HemAT from Bacillus subtilis FixL from

PAGE 251

228 Bradyrhizobium japonicum and Sinorhizobium meliloti and CooA from Rhodospirillum rubrum Furthermore, a number of model system s were also observed to understand the underlying thermodynamic processes in heme ligation. To summarize, this dissertation has contributed to the following: The first project presented in the model systems is the study of the mepepy. The results for the trans to cis isomerization of the mepepy reveal one energy minimum for the cis form and two energy mini ma separated by 9 kcal mol-1 for the trans form which correspond to two different conformations of the trans form of the mepepy. Computational studies demonstrated that th e dipole moments between the trans and the cis form of the mepepy are very similar which illustrate a very restraint change in the electronic structure of mepepy to compare to azobenzene molecules or other push-pull molecules. Finally, it has been demonstrated that the mepepy pyridine group can form an H-bond with the solvent which is lost after isomerization of the mepepy from the trans to cis form. The second project for the model systems de monstrate several key features of the trans to cis isomerization of mepepy and spin crossover of the iron in the Fe(III)(salten)(mepepy) complex. Foremost, a volume change equal to V = ~ 0.7 0.3 mL mol 1 was calculated for the mepepy ligand a nd was associated to the loss of H-bond between a water and the mepepy pyridyl group. The enthalpy change associated to the isomerization to the trans to cis mepepy in the mixture acetonitrile/water is equal to H = ~ 33 10 kcal mol 1 and associated to the cleavage of a positive charge-assisted H-bond between the water and/or acetonitrile and the protonated mepepy pyridinium group in addition to the loss of H-bond between a wa ter and mepepy pyridyl group. PAC results demonstrated that the volume change ( V = ~ 0.9 0.3 mL mol 1) and the enthalpy change ( H = ~ 37 10 kcal mol 1) of the Fe(III)(salten)(mepepy) complex are related to the spin crossover of the iron associated with an expansion of the Fe(III)(salten)(mepepy) complex into the solvent and to the cleavage of a Fe…N bond with the formation of a

PAGE 252

229 hydrogen bond between a water molecule and th e secondary amine. The spin change of the iron is estimated to be equal to V = ~ 0 mL mol 1 and H = ~ 0 kcal mol 1. The last model system project was the us e of the Debye-Hckel equation in order to calculate the solvent excluded volume of molecules, complexes and/or proteins. The theory was validated using two ruthenium complexes. These results confirm a charge transfer between the ligand and the metal for Ru(bpy)3 and Ru(phen)3 close to one. As well, the excluded volume change was calcula ted and equal to ~ 2.1 and -1.9 mL mol-1 for the prompt phase and ~1.7 and 1.9 mL mol-1 for the slow phase, for both Ru(bpy)3 and Ru(phen)3 respectively. Different signaling projects were studied in order to understand the transmission of a signal along a protein. The firs t signaling project is the study of Bs HemAT from Bacillus subtilis a GCS domain sensor and an aero tactic transducer. In summary, both transient absorption and PAC data reveal a fa st conformational transition associated with ligand release from Bs HemATHD. The conformational transition results in a nonequilibrium deoxy form of the protein whic h persists during the ligand rebinding phase. The thermodynamics of this transition ar e quite distinct from the PAS domain Bj FixL heme domain suggesting quite different si gnaling mechanisms between the GCS and PAS domain sensors. The second signaling project is the study of CooA from Rhodospirillum rubrum CooA is a transcriptional activator. PAC re sults reveal a fast conformational change associated with the photodissociation of CO from CooA. The c onformational change demonstrated a large enthalpy and volume change which were associated to a reorientation of the C-helix in addition to the disruption of a salt-bridge. The last signaling project is th e study of full domain wild-type Sm FixL associated with five different mutants from two different critical amino acids (R200 and I209M) which are supposed to be involved in the tran smission of the signal and also the study of

PAGE 253

230 Sm FixL heme domain. Sm FixL from Sinorhizobium meliloti is involved in the regulation of the nitrogen fixation gene expression in Rhizobi PAC results reveal a quadriphasic relaxation for the different Sm FixLWT, and the five different mutants associated with the photodissociation of CO. These thermodynamic results confirm that the residues R200 and I209 are very important in the transmission of the signal and if a mutation is taking place, it might involve different conforma tional changes which might induce the transmission of a different signal. The results for Sm FixLHD reveal a monophasic relaxation which is different to the results from Bj FixLHD and suggest a fast release of the ligand to the solvent. Finally, two different model protein syst ems were studied in order to help understand of the transmission of the signal along a protein. The first model protein system is the study of the sandbar shark hemoglobin from Carcharhinus plumbeus as hemoglobin is an allosteric protein. The tr ansient absorption result displays a biphasic relaxation kinetic. The first relaxation was attributed to the CO geminate recombination to the same subunit and the second relaxation to the recombination to another subunit. PAC demonstrated a triphasic relaxation wher e the first phase was associated to CO geminate recombination, the second phase to th e tertiary transition from R to T for one subunit and the third phase to th e R to T quaternary transition of the globin. The rate of the relaxation are similar to the human hem oglobin with the exception of the third phase which is faster in the shark hemoglobin. The second model protein project was the study of two peroxidases: the horseradish and soybean peroxida se which participates in the oxidation of a large array of organic and inorganic compounds vi a heme catalyzed reduction of H2O2 to water. Unique thermodynamics for CO release from heme peroxidases which reflect the open access channel between the solvent and the heme activ e site are revealed w ith the PAC results. These results are consistent with the input of water mol ecules into the channel upon CO release. In the case of HRP, a single water mo lecule is being input into the channel upon

PAGE 254

231 CO release which is not affected by the nature of the solvent. In contrast, in the case of SBP, the CO photorelease is dependent upon the presence of a Tris molecule docked within the heme distal pocket as well as the solution ionic strength. These results are clearly distinct from the globin class of prot eins, e.g., myoglobin, highl ighting the role of the protein matrix in modulating the energy ba rriers to ligand access to the distal heme pocket. The different results which are presented in Table 6.1 reveal that the thermodynamic profiles are different for each pr otein. The results for the PAS domain are different to the GCS domain data. The resu lts collected for the FixL protein which belongs to the PAS domain family, shown one two or four different phases depending on the type of FixL ( Bj FixL or Sm FixL) and on the type of Sm FixL mutation ( Sm FixLR200A...). These results are different to compare to HemAT protein which belong to the GCS domain family and show n only a monophasic relaxation. CooA from the CooA domain is different to the results obtained for the GCS (HemAT) or the PAS heme domain ( Bj FixL or Sm FixL). Although CooA is a full protein to compare to the Bj FixL or HemAT heme domain protein, the results shown faster change following the photodissociation of CO implying different sign aling mechanisms with the protein from the PAS domain or GCS domain. The results presented with Sm FixL which is a full protein showed four different phases to comp are to CooA which was also the study of a full protein demonstrating once again a diffe rence of the transm ission of the signal between the proteins. Bs HemAT which is a myoglobin like domain should show similar thermodynamic profile to compare to myoglobin as the two heme domain present similarity. The results show very differ ent thermodynamic profile with a monophasic relaxation obtained with Bs HemAT and a biphasic relaxation with myoglobin. To conclude, inside the PA S domain family, the result s presented are different, which demonstrated that even inside the same family and with the same gas molecule CO, the transmission of the signal is diffe rent and each protein replies to a signal differently. All the data conclude that each protein is unique, has a unique way to respond

PAGE 255

232 to a signal and its own thermodynamic profile following the transmission of the signal even if the protein belongs to the same family. The results summarized in the Table 6.1 were all collected in Dr Larsen’s laboratory and the results in blue were collected by the Author.

PAGE 256

Table 6.1 : Summarized of the variation of volume and enth alpy for different heme-based sensor proteins. <50 ns Intermediary Phase 1 Intermediary Phase 2 Slow Phase Sample H (kcal mol-1) V (mL mol-1) (ns) H (kcal mol-1) V (mL mol-1) (ns) H (kcal mol-1) V (mL mol-1) (ns) H (kcal mol-1) V (mL mol-1) Fe4SP-2MeIm 17 3 21 0.7 FixLH 12 3 -1 0.5 150 5 3 5 0.5 tFixL 21 0.7 4 0.3 Sm FixL (full protein) -9 8 10 3 190 18 16 -18 6 512 -35 20 11 6 1495 31 25 -9 8 Sm FixLR200A -7 6 15 2 100 63 20 -6 4 510 51 20 28 6 1510 18 11 -10 3 Sm FixLR200Q -26 13 17 2 92 62 8 -19 1 285 -43 5 8 0.9 1252 9 3 -7.6 0.6 Sm FixLR200E -10 5 21 3 120 31 14 -21 2 429 -47 6 10 1 1575 43 10 -5 2 Sm FixLR200H -39 5 4.7 0.7 95 74 7 -7 1.1 365 -49 5 3.9 0.7 1608 75 5 -4.3 0.7 Sm FixLI209M -54 3 3 0.4 115 84 6 -3.7 1 497 -30 13 5.7 2 1493 14 5 -7.8 0.8 Sm FixLH 8.8 0.9 21 6 EcDos 11 0.6 -2 1 150 -28 0.9 9 0. 2 2000 -89 9 -18 2 CooA (full protein) -414 -46 HemAT-Bs -20 5 4 1 HH Mb 7 3 -1 0.6 607 18 1 Sandbar Shark Hb (pH=6) 19 6.6 0.6 48 12.5 2 -22 -12 Sandbar Shark Hb (pH=7) 17 5.5 0.6 48 22 2.4 -8 -30

PAGE 257

234 6.2. Future Directions In order to better understand signaling prot eins in general and how a signal is transmitted along the protein, the author propos es different directions for the signaling protein projects and for the model system proj ects which will help the understanding of the thermodynamic processes in heme ligation. Model system project: In order to confirm the V and H of the spin crossover of the iron, another system should be tried, for instance the Fe3+ tetrakis-(4-sulphonatophenyl) porphyrin (Fe4SP) with the mepepy lig and linked in its fifth position. In order to confirm the re sults obtained for the Debye -Hckel equation, the results should be checked by a computational method. The Debye-Hckel equation should be al so tested with another ruthenium complex than the Ru(bpy)3 and Ru(phen)3, for instance, the Ruthenium(II)bis(2,2‘-bipyridine )(4,4‘-dicarboxy-2,2‘-bipyridine) (Ru(bpy)2(dcbpy)) in order to confirm the hypothesis stating that the excluded volume change can be calculated using the Debye-Hckel equation. This ruthenium complex will also confirm that the V is ionic strength dependent but the H is independent and that the excl uded volume change does not dependent on the size of the ruthenium complex. Myoglobin should be examined with the Debye-Hckel equation and should confirm that the Debye-Hckel equation can be used to find the excluded volume change of proteins, as myogl obin is well studied and known. Another model with a total charge separation should be tested instead of the ruthenium complexes which have a partial charge separation after photoexcitation.

PAGE 258

235 Signaling protein project: Bs HemAT protein should be analyzed afte r attaching the linker to the heme domain in order to draw a thermodynamic profile of the transmission of the signal until the linker. Then the kinase domain should be attached to the heme domain and the linker to draw the complete th ermodynamic profile of the full protein. HemAT protein from Halobacterium salinarum should also be studied to compare both proteins Bs HemAT and Hs HemAT and examine if the thermodynamic profile is similar or diffe rent in the same family. Bs HemAT should also be studied for signa l transmission using NO as the signal activator. As Bs HemAT was already studied with the gas molecule O2 and CO, these results can be compared with the same protein, and dem onstrate the effect that different gas molecules have on the thermodynamic profiles. A NO sensor protein should also be examined in order to be able to compare the thermodynamic profiles between the four different heme-based sensors (PAC, GCS, CO sensor with CooA and NO sens or with the HNOB family). For instance, GCs, a second messenger, mononucleotid e activator or Rsp2043_Rhsp, a second messenger, dinucleotide activat or would be a good start. Sm FixL protein should be analyzed using O2 gas molecule to compare the thermodynamic profiles between CO and O2 for signal transmission. As FixL is inhibited when O2 binds the heme domain, the thermodynamic profiles of the transmission of the signa l between an activator an d inhibitor signal can be compared. Finally, the thermodynamic profile of m yoglobin (Mb) with different mutations should also be examined, for instance, mutation of the Lys 45 for the horse heart Mb or Arg 45 for the sperm whale Mb, in order to observe the difference in the thermodynamic profiles when CO is released to the solvent.

PAGE 259

About the Author Audrey Mokdad received her Bachelor’s de gree in Chemistry from the University of Caen, France in 2000. In 2001, she finished her Master’s degree in Chemistry with a major in organic chemistry and in 2003, her Master’s degree in chemical engineering. While, in the Bachelor and Mast er’s program in France, she did several internships with domestic and international companies such as Rhodia, GlaxoSmithKline or L’Oreal. In 2004, Audrey was admitted to the Ph.D. program at the University of South Florida and joined Dr. Randy W. Larsen’s research group. Her research focuses on the time resolved thermodynamic studies in heme signaling protei ns and model systems. She is co-authored of three scientific publications and is currently preparing several other manuscripts. She has presented her research at local, regional and national conferences, including Annual Raymond N. Castle Student Research Conference, Florida Annual Meeting and Exposition of the American Chemical Society (FAME), Florida Inorganic Mini-Symposium (FIMS), Southeast Regional Meeting of the Ameri can Chemical Society (SERMACS), Annual Meeting of the Biophysical Society conference and American Chemical Society National Meeting (ACS).