Citation
A Novel photodissociative probe of water in serum albumin binding sites

Material Information

Title:
A Novel photodissociative probe of water in serum albumin binding sites
Creator:
Di, Qiao-Qing
Publication Date:
Language:
English
Physical Description:
xvi, 198 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Albumins ( jstor )
Binding sites ( jstor )
Emission spectra ( jstor )
Fluorescence ( jstor )
Health savings accounts ( jstor )
Ligands ( jstor )
Molecules ( jstor )
Protons ( jstor )
Pyridines ( jstor )
Solvents ( jstor )
Binding Sites ( mesh )
Department of Medicinal Chemistry thesis Ph.D ( mesh )
Dissertations, Academic -- College of Pharmacy -- Department of Medicinal Chemistry -- UF ( mesh )
Fluorescent Dyes ( mesh )
Molecular Probe Techniques ( mesh )
Molecular Probes ( mesh )
Protons ( mesh )
Research ( mesh )
Serum Albumin -- chemistry ( mesh )
Water ( mesh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1998.
Bibliography:
Bibliography: leaves 191-197.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Qiao-qing Di.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
029898009 ( ALEPH )
51589734 ( OCLC )

Downloads

This item has the following downloads:


Full Text









A NOVEL PHOTODISSOCIATIVE PROBE OF WATER IN SERUM ALBUMIN BINDING SITES













By

QIAO-QING DI


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1998

























Copyright 1998

by

Qiao-qing Di






























I would like to dedicate this dissertation to my family, all of them contributed in their own unique way to make this possible.














ACKNOWLEDGMENTS

I would like to thank Dr. Stephen. G. Schulman for his guidance, support and encouragement in my academic and professional development. I would also like to acknowledge my committee members Dr. John H. Perrin, Dr. Kenneth B. Sloan, Dr. James D. Winefordner, and Dr. Jeffrey Hughes for their support and encouragement throughout my stay at the University of Florida. Finally, but not least, I would like to acknowledge Dr. Otto S. Wolfbeis and his group members, Dr. James F. Preston III and his group member, and Dr. Lei Wei for their part in helping me to finish this work.















TABLE OF CONTENTS
page


ACKNOW LEDGM ENTS ............................................................................... . ........... . iv

LIST OF TABLES .......................................................................................................... viii

LIST OF FIGURES .......................................................................................................... ix

CHAPTERS

1 INTRODUCTION ........................................................................................................1

2 BACKGROUND ........................................................................................................ 11

The Structure of W ater................................................................................................... 11
Prototropic Conductivity................................................................................................12
Prototropic Reactivity in Electronically Excited States.................................................18
Properties of Acids and Bases in the Lowest Excited Singlet State ...........................18
The Effect of State of Protonation on Electronic Spectra...........................................19
The Effect of Solvents on Electronic Spectra.............................................................20
Steady-State Kinetics of Excited State Proton Transfer Reaction..............................24
Protein Binding...............................................................................................................29
Binding Studies - Experim ental M ethods..................................................................30
Binding Equation - Derivation...................................................................................31
Binding Studies - Treatm ent of Data..........................................................................34
Determination of Complex Stoichiometry - The Method of Continuous
Variation ................................................................................................................35
Identification of the Binding Sites..............................................................................37
Characterization of the Binding Sites on Albumins - Fluorescence Quenching.........38 Characterization of the Binding Sites on Albumins - Circular Dichroism ................41

3 EXPERIM ENTAL......................................................................................................47

M aterials .........................................................................................................................47
Instrum ental ....................................................................................................................49
M ethods...........................................................................................................................49
Ground and Excited State Ionization Constants Determinations (Titration
Procedure)..............................................................................................................49
Solvent Studies ...........................................................................................................50









Qualitative Spectrophotometric Examination of 2-Naphthol-8-SulfonateAlbumin
C om plex Interaction..............................................................................................51
Fluorometric Determination of Compound-Albumin Complex StoichiometryJob's M ethod .........................................................................................................52
2-Naphthol-8-Sulfonate-Albumin Titration Experiments-Determination of
Binding Constants and Number of Binding Sites.................................................53
Medium Effects on the Affinity of the Site in Albumin for the 2-Naphthol-8Sulfonate ................................................................................................................54
Displacement of 2-Naphthol-8-Sulfonate from Albumin...........................................55
Fluorescence Quenching of Albumin Bound 2-Naphthol-8-Sulfonate.......................56
Circular Dichroism Experiments................................................................................58

4 R E SU L T S ....................................................................................................................60

Photophsicochemical Properties of 2-Naphthol-8-Sulfonate.........................................60
Absorption Spectra of 2-Naphthol-8-Sulfonate..........................................................60
Fluorescence Excitation Spectra of 2-Naphthol-8-Sulfonate......................................60
Fluorescence Emission Spectra of 2-Naphthol-8-Sulfonate.......................................60
Determination of Ground and Excited State Ionization Constants.............................61
Steady-State Kinetics of Excited State Proton Transfer Reaction of 2N aphthol-8-Sulfonate............................................................................................61
Solvent Studies ...........................................................................................................63
B inding Studies...............................................................................................................66
Qualitative Spectrophotometric Examination of 2-Naphthol-8-SulfonateAlbumin Complex Interaction ..............................................................................66
Fluorometric Determination of 2-Naphthol-8-Sulfonate -Albumin Complex
Stoichiom etry ........................................................................................................68
2-Naphthol-8-Sulfonate-Albumin Titration Experiments-Determination of
B inding C onstants.................................................................................................69
Medium Effects on the Affinity of the Site in Albumin for the 2-Naphthol-8Sulfonate ................................................................................................................72
Displacement of 2-Naphthol-8-Sulfonate from Albumin...........................................73
Fluorescence Quenching of Albumin-Bound 2-Naphthol-8-Sulfonate ......................74
Circular Dichroism Experiments ................................................................................76

5 D ISC U SSIO N ............................................................................................................ 15 1

The Effect of pH ...........................................................................................................151
Steady-State Kinetics of Excited State Proton Transfer Reaction of 2N aphthol-8-Sulfonate..........................................................................................152
Solvent Studies ............................................................................................................ 153
The Effects of Solvent on the Spectral Properties of 2-Naphthol-8-Sulfonate..........153
The Effect of Solvent on Fluorescence Quantum Yields..........................................159
The Effects of Solvent on the Excited State Proton Transfer ...................................159








Binding Studies............................................................................................................162
Identification and Characterization of the Binding Sites on Albumin......................162
Fluorescence Quenching Experiments......................................................................168
Circular Dichroism Experiment................................................................................172
M odel for Binding of 2-Naphthol-8-Sulfonate to Albumin......................................175
The Effect of Binding on the Spectral Properties of 2-Naphthol-8-Sulfonate..........177
The Effect of Binding on Excited State Proton Transfer..........................................180
The Effects of pH and Ionic Strength on the Binding of 2-Naphthol-8Sulfonate to Albumin..........................................................................................186

6 CONCLUSIONS.......................................................................................................189

REFERENCES ................................................................................................................191

BIOGRAPHICAL SKETCH ..........................................................................................198














LIST OF TABLES


Table page 4-1. Photophysicochemical properties of 2-naphthol-8-sulfonate ..............................63

4-2. Solvent studies - absorption and fluorescence data..............................................65

4-3. Results of Job study - number of binding sites ...................................................69

4-4. Results of Scatchard plot - binding constants and number of binding sites .........72














LIST OF FIGURES


Figure page 2-1. A molecular mechanism for prototropic mobility ................................................16

4-1. Absorption spectra of 2-naphthol-8-sulfonate at different H0 or pH values ..........77

4-2. Fluorescence excitation spectra of 2-naphthol-8-sulfonate in different pH
values .....................................................................................................................7 8

4-3. Fluorescence emission spectra of 2-naphthol-8-sulfonate in different Ho or pH
v alues .....................................................................................................................79

4-4. The variation of absorbances with H0 or pH of 2-naphthol-8-sulfonate
monoanion and its conjugate base dianion ...........................................................80

4-5. The variation of relative fluorescence intensities with H0 or pH of 2-naphthol8-sulfonate monoanion and its conjugate base dianion ........................................81

4-6. Absorption spectra of 2-naphthol-8-sulfonate monoanion in different solvents ...82 4-7. Absorption spectra of 2-naphthol-8-sulfonate dianion in different solvents .........83

4-8. Fluorescence emission spectra of 2-naphthol-8-sulfonate monoanion in
different solvents...................................................................................................84

4-9. Fluorescence emission spectra of 2-naphthol-8-sulfonate dianion in different
solvents ..................................................................................................................85

4-10. Relative fluorescence intensities of 2-naphthol-8-sulfonate as a function of
ethanol concentration ............................................................................................86

4-11. Relative fluorescence intensities of 2-naphthol-8-sulfonate as a function of
1,4-dioxane concentration.....................................................................................87

4-12. Absorption spectra of free, BSA-bound and HSA-bound 2-naphthol-8sulfonate at pH 3.04 ..............................................................................................88








4-13. Absorption spectra of free, BSA-bound and HSA-bound 2-naphthol-8sulfonate at pH 4.03 and jI 0.1 ..............................................................................89

4-14. Absorption spectra of free, BSA-bound and HSA-bound 2-naphthol-8sulfonate at pH 5.01 and g 0.1 ..............................................................................90

4-15. Absorption spectra of free, BSA-bound and HSA-bound 2-naphthol-8sulfonate at pH 5.93 and I 0.1 ..............................................................................91

4-16. Absorption spectra of free, BSA-bound and HSA-bound 2-naphthol-8sulfonate at pH 6.93 and g 0.1 ..............................................................................92

4-17. Absorption spectra of free, BSA-bound and HSA-bound 2-naphthol-8sulfonate at pH 8.09 and I 0.1 ..............................................................................93

4-18. Absorption spectra of free, BSA-bound and HSA-bound 2-naphthol-8sulfonate at pH 9.50 and I 0.1..............................................................................94

4-19. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 3.04........................................................................95

4-20. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 4.04 and g 0.001 .....................................................96

4-21. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 5.07 and p 0.001 .....................................................97

4-22. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 6.07 and I 0.001 ....................................................98

4-23. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate as well as BSA and HSA at pH 7.05 and i 0.001 ............99

4-24. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 8.12 and i 0.001..................................................100

4-25. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 8.94 and i 0.001 ..................................................101

4-26. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 10.01 and i 0.001................................................102

4-27. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 4.03 and p 0.1 ......................................................103








4-28. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 4.98 and IX 0.1 ......................................................104

4-29. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 5.93 and p 0.1 ......................................................105

4-30. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate as well as BSA and HSA at pH 6.93 and p 0.1.............106

4-31. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 8.09 and p 0.1 ......................................................107

4-32. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 9.02 and p 0.1 ......................................................108

4-33. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 9.99 and p 0.1 ......................................................109

4-34. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 4.03 and p 1.0......................................................110

4-35. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 4.95 and g 1.0......................................................111

4-36. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate as well as BSA and HSA at pH 6.93 and g 1.0.............112

4-37. Fluorescence emission spectra of free, BSA-bound and HSA-bound
2-naphthol-8-sulfonate at pH 8.92 and g 1.0 ......................................................113

4-38. Fluorescence emission spectra of BSA-2-naphthol-8-sulfonate complex
at different molar ratios of 2-naphthol-8-sulfonate to BSA................................114

4-39. Fluorescence emission spectra of HSA-2-naphthol-8-sulfonate complex
at different molar ratios of 2-naphthol-8-sulfonate to HSA................................115

4-40. Job's plot of BSA-2-naphthol-8-sulfonate monoanion at pH 7.4 and p 0.1........116

4-41. Job's plot of BSA 2-naphthol-8-sulfonate dianion at pH 7.4 and p 0.1 ............117

4-42. Plots of relative fluorescence intensity as a function of total 2-naphthol-8sulfonate concentration for the BSA-2-naphthol-8-sulfonate titrations with
a constant amount of the protein at pH 7.4 and p 0.1 .........................................118








4-43. Plots of relative fluorescence intensity as a function of total 2-naphthol-8sulfonate concentration for the HSA-2-naphthol-8-sulfonate titrations with
a constant amount of the protein at pH 7.4 and p 0.1 .........................................119

4-44. Scatchard plot of r/[L] versus r for the bovine serum albumin-2-naphthol8-sulfonate...........................................................................................................120

4-45. Scatchard plot of r/[L] versus r for the human serum albumin-2-naphthol8-sulfonate...........................................................................................................12 1

4-46. The fluorescence emission ratio of BSA-2-naphthol-8-sulfonate complex
as a function of pH at different ionic strengths...................................................122

4-47. The fluorescence emission ratio of HSA-2-naphthol-8-sulfonate complex
as a function ofpH at different ionic strengths...................................................123

4-48. Marker-induced changes in fluorescence of 2-naphthol-8-sulfonate bound
to H SA ................................................................................................................124

4-49. Marker-induced changes in fluorescence of 2-naphthol-8-sulfonate bound
to B SA .................................................................................................................. 125

4-50. Absorption spectra of pyridine, 2-naphthol-8-sulfonate in the presence of
pyridine at pH 5.01 and t 0.1 .............................................................................126

4-51. The fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence
of pyridine at pH 5.0 and [t 0.1 ..........................................................................127

4-52. The fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence
of pyridine at pH 9.0 and p 0.1 ..........................................................................128

4-53. The fluorescence emission spectra of HSA-2-naphthol-8-sulfonate in the
presence of pyridine at pH 5.0 and p 0.1 with constant amount of 2-naphthol8-sulfonate (10 ptM) and HSA (7 paM)................................................................129

4-54. The fluorescence emission spectra of HSA-2-naphthol-8-sulfonate in the
presence of pyridine at pH 9.0 and t 0.1 with constant amount of 2-naphthol8-sulfonate (10 pM) and HSA (7 pM)................................................................130

4-55. The fluorescence emission spectra of HSA-2-naphthol-8-sulfonate in the
presence of pyridine at pH 5.0 and p 0.1 with constant amount of 2-naphthol8-sulfonate (10 pM) and HSA (50 pM)..............................................................131








4-56. The fluorescence emission spectra of HSA-2-naphthol-8-sulfonate in the
presence of pyridine at pH 9.0 and 1t 0.1 with constant amount of 2-naphthol8-sulfonate (10 pM ) and HSA (50 pM )...............................................................132

4-57. The fluorescence emission spectra of BSA-2-naphthol-8-sulfonate in the
presence of pyridine at pH 5.0 and p 0.1 with constant amount of 2-naphthol8-sulfonate (10 pM ) and BSA (7 pM ) ...............................................................133

4-58. The fluorescence emission spectra of BSA-2-naphthol-8-sulfonate in the
presence of pyridine at pH 9.0 and p 0.1 with constant amount of 2-naphthol8-sulfonate (10 pM ) and BSA (7 jLM ) ...............................................................134

4-59. The fluorescence emission spectra of BSA-2-naphthol-8-sulfonate in the
presence of pyridine at pH 5.0 and ji 0.1 with constant amount of 2-naphthol8-sulfonate (10 tM ) and BSA (50 pM ) ..............................................................135

4-60. The fluorescence emission spectra of BSA-2-naphthol-8-sulfonate in the
presence of pyridine at pH 9.0 and [t 0.1 with constant amount of 2-naphthol8-sulfonate (10 RM ) and BSA (50 gM ) .............................................................136

4-61. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of
pyridine in ethanol .............................................................................................137

4-62. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of
pyridine in 75% (V /V ) ethanol ..........................................................................138

4-63. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of
pyridine in 50% (V/V ) ethanol ..........................................................................139

4-64. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of
pyridine in 1,4-dioxane ......................................................................................140

4-65. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of
pyridine in 75% (VNV) 1,4-dioxane ...................................................................141

4-66. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of
pyridine in 50% (V/V ) 1,4-dioxane ...................................................................142

4-67. Plots for fluorescence quenching of 2-naphthol-8-sulfonate in the absence and
presence of albumin as well as albumin with pyridine at pH 5 and P 0.1 ...........143

4-68. Plots for fluorescence quenching of 2-naphthol-8-sulfonate in the absence and
presence of albumin as well as albumin with pyridine at pH 9 and p 0.1 ..........144








4-69. CD spectra of HSA, HSA-2-naphthol-8-sulfonate at pH 9.0 and pH 5.0
in the far-U V region............................................................................................145

4-70. CD spectra of HSA; HSA-2-naphthol-8-sulfonate at pH 9.0 and pH 5.0
in the near-U V region ........................................................................................146

4-71. CD spectra of HSA; HSA-2-naphthol-8-sulfonate at pH 9.0 and pH 5.0
in the wavelength region 300 - 450 nm.. ............................................................147

4-72. CD spectra of BAS and BSA-2-naphthol-8-sulfonate at pH 5.0
in the far-U V region............................................................................................148

4-73. CD spectra of BAS and BSA-2-naphthol-8-sulfonate at pH 5.0
in the near-U V region ........................................................................................149

4-74. CD spectra of BAS and BSA-2-naphthol-8-sulfonate at pH 5.0
in the wavelength region 300 - 450 nm .............................................................150

5-1. A simplified schematic representation of energy levels (A) attained by a
molecule in solution in the course of light absorption and emission (B)............154

5-2. Potential energy cross sections drawn along two coordinates Q, and Qr
for the ground and excited states.........................................................................157

5-3. Amino acid sequence of HSA in an arrangement reflecting the heart-shape
structure................................................................................................................164

5-4. Schematic representation of a protein molecule, with a layer or two of
strongly associated water, suspended in aqueous solution .................................182














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

A NOVEL PHOTODISSOCIATIVE PROBE OF WATER IN SERUM ALBUMIN BINDING SITES

By

Qiao-qing Di

December, 1998

Chairman: Dr. Stephen G. Schulman
Major Department: Medicinal Chemistry

Ligand-protein interactions play a key role in the distribution and transport of small molecules in biological systems. Understanding the molecular basis of these interactions is important in the rational design of new and more efficient therapeutic agents that can recognize and bind to specific biological targets for improved drug activity. Water plays a fundamental role in the interactions of proteins with their ligands. However, the properties of water molecules in the vicinity of a biomolecule differ appreciably from those of bulk water. For the sake of understanding of the structure and dynamics of the aqueous environment surrounding the concerned biomolecule, a novel photodissociative probe of water in protein binding sites based on the proton dissociation reaction of a hydroxyaromatic in the excited state was discovered and characterized. In this study, the steady-state fluorescence and lifetime measurements have been employed to determine the rate of proton dissociation from the excited probe molecule in albumin








binding sites. The identification and characterization of the binding sites for the probe in albumin were performed through fluorescence quenching experiments, displacement experiments, and circular dichroism methods. The results indicated (1) for surface-bound 2-naphthol-8-sulfonate, the reduced proton dissociation rate is due to the protein-water interactions which cause an overall reduction of the reorientational rate of water; for the internally bound 2-naphthol-8-sulfonate, the proton transfer reaction takes place via the hydrogen-bonded chains; (2) the primary factor involved in the 2-naphthol-8-sulfonatealbumin complex formation is the existence of the structured water hydration around the albumin molecule; and (3) 2-naphthol-8-sulfonate can be used to detect surface differences between proteins that perform the same functions in different species, more specifically, differences in the surface properties of proteins differing only in a few amino acids. This study has demonstrated that the kinetic studies of excited singlet state proton transfer can disclose the properties of the immediate environment of the proton emitter, and the applicability of 2-naphthol-8-sulfonate as a useful fluorescence probe to study the role of water in the binding sites on protein.














CHAPTER 1
INTRODUCTION

Ligand-protein interactions play a key role in the distribution and transport of small molecules in biological systems. Understanding the molecular basis of these interactions is important in the rational design of new and more efficient therapeutic agents that can recognize and bind to specific biological targets for improved drug activity (1). Water plays a fundamental role in the interactions of proteins with their ligands. The binding process generally involves an entropically favored displacement of solvent molecules from the protein and ligand surfaces and an enthalpically favored reorganization of these solvent molecules (2). For example, approximately 65 water molecules appear to be released on binding of glucose to hexokinase (3). Some solvent molecules may, however, be trapped at the protein-ligand interface. These may make an enthalpic contribution to the ligand binding free energy by, for example, mediating hydrogen bond bridges between the ligand and the protein. Because of their mobility relative to the protein and their ability to both accept and donate hydrogen bonds, water molecules are adaptable liganding partners that are able to fill empty space, modulate the binding specificity of the protein, and play a role in its function. Numerous examples have been reported of the structural and functional importance of water molecules, as they are associated with proteins and in many cases, have a direct and crucial function in molecular recognition and catalysis (4). Examples of protein structures in which water








molecules have been observed to mediate protein-small molecule interactions including those of the complexes of cholesterol oxidase with a steroid substrate (5), retinol-binding protein with retinol (6), adipocyte lipid-binding protein with arachidonic acid (7), adipocyte lipid-binding protein with palmitate and with hexadecanesulfonic acid (8), Larabinose-binding protein with L-arabinose, D-fucose, and D-galactose (9, 10), and ribulose-1,5-bisphosphate carboxylase oxygenase with 2-carboxyarabinitol bisphosphate

(11). Often, understanding the role of water in protein-ligand interaction offers promise for the rational design of new and more efficient therapeutic agents that can recognize and bind to specific biological targets for improved drug activity. For example, one of these ordered water molecules, seen in complexes of the HIV-protease with peptide ligands, has guided the design of a novel tightly-bound inhibitor (12, 13).

Many biochemical processes proceed at the interface of water with the rigid

structures of proteins. Of the total surface of the protein, there is only a very small area called the binding site or active site, in which binding or catalysis takes place. Such a site is a very special environment. It is a cleft in the low dielectric matrix of the protein, spotted with hydrophilic and charged amino acid residues that surround a microscopic space some 10 - 20 A in its longest dimension. The combination of charged moieties, dielectric discontinuity and microscopic dimension tends to alter the energy levels of molecules that enter it, which is the primary step of binding or catalysis. For a complete knowledge of the function of such systems, an understanding of the structure and dynamics of the aqueous environment surrounding the concerned biomolecule is thus, essential. The properties of water molecules in the vicinity of a biomolecule differ








appreciably from those of bulk water (4, 14-18). It is now known that the hydration shell surrounding a protein molecule comprises different types of water. Comparison of the water structure surrounding the surface of the protein indicates that the water molecules observed can be classified into three categories. The first contains those that are observed at the same site, making the same interactions with the protein surface in every independently solved structure in crystallographical experiment. These water molecules have the residence times in the range 10-9 - 10' s and are well ordered. The second category consists of a number of water molecules that are observed at a particular site in only one structure, but not in others. These water molecules have an average residence time in the range 10 - 50 ps and are slightly disordered. Finally, there are water molecules that must be present at the surface of the protein, but that are not observed crystallographically at all because they are always disordered. The dynamics and structure of ordered water near proteins, DNA, and in reverse micelles have been the subject of intense research over several decades (19). Our understanding of ordered water at the atomic level comes primarily from several approaches: single crystal diffraction (X-ray and neutron), and nuclear magnetic resonance spectroscopy (NMR) are two experimental methods and the most extensively used techniques to understand the interaction of water with proteins, and molecular dynamics (MD) simulations and empirical analyses of known structures are complementary theoretical methods. The investigation of bound water molecules by NMR techniques has been impeded by technical problems associated with distinguishing small numbers of discrete, bound water protons from the enormous number of protons in bulk solvent, although advances in water suppression techniques allowed some very important studies of non-isotopically








labeled proteins (20, 21), and recent studies have increasingly employed selective detection of protons bound to heteronuclei ("3C or "5N) to improve suppression of bulk water signals (22, 23). Another limitation of this technique is that some potential water sites are not accessible to study because they are too far away from a proton (e.g. water molecules near some carbonyls or carboxylates). X-ray and neutron crystallography reveal the favored average positions occupied in the crystal by water molecules. These water molecules, which are surrounded by backbone atoms and satisfy the main-chain hydrogen bonding capacity of peptide groups not involved in hydrogen bonds with other peptide groups, appear to be an integral part of the folded protein and are highly ordered (24, 25). Since this structure of protein, as determined by crystal diffraction methods, is of the form favored by the crystallization system, not necessarily its active or liganded form, which usually involves surface semi ordered water molecules. For this reason it is advisable to rely on dynamic measurements and through them to gain some structural information.

In our studies, we use the method based on a well-defined probing reaction, namely, proton transfer of excited hydroxyaromatics. The proton dissociation of an excited hydroxyaromatic (26, 27) is given as follows: k12
OOH* + H20 -k21 (O-* + H30+

k(r + nr) Ihvl hv2 k'(r + nr)

DOH + H20 k (O- + H30+


(1-1)








The excited protonated form OOH* can either decay to the ground state by a

radiative plus a nonradiative process (1, o r)) emitting at a characteristic wavelength (hvi) or it might dissociate before decaying to the ground state anion (OO). It is well established that hydroxyaromatics are stronger acids in their lowest excited singlet states than in the ground state (pK* < pK) (28). Thus the emission of the protonated form of hydroxyaromatics is at a shorter wavelength than that of the deprotonated form (v, > v2). A measurable dissociation of ~OH* will take place only if it is faster than the fluorescence decay k12 > k +,nr). Otherwise, the molecule will reach the ground state before the proton is emitted from the excited state. Hydroxyaromatics with large pK shifts are not only effective proton emitters but are also convenient compounds to measure the lifetime of each excited species. The spectral shift between the emission bands is large enough so that each form can be independently measured. Steady-state fluorescence and lifetime measurements can be employed to determine the rate of proton dissociation from these excited hydroxyaromatics. The mechanism of proton transfer for reactants in solutions is relatively well understood (29). The rate constant of proton transfer is extremely sensitive to its environment. Except for very strong acids (pK < 0), no dissociation will take place unless water molecules are in the immediate vicinity to act as proton acceptors (30-32). Thus, the rate of proton transfer can be interpreted in terms of a single parameter, the probability of successful proton transfer to water within the lifetime of the excited state.

With such opportunities at hand, the advantage of a short observation period is self-evident. If we can limit the monitoring to a very short-time frame, after a proton has








been released in a defined site, the physical information obtained by the analysis will reflect only that space which the proton could probe during the observation time. Thus, the temporal resolution is transformed into spatial resolution. Under proper conditions, microspace as small as the hydration layer of a protein or the specific site on a protein can be studied, totally insensitive to the huge bulk volume in which the sample is suspended. Therefore, on measuring the dynamics of proton transfer, the kinetic analysis of the reaction can quantitate the properties of the immediate environment, such as the physical and chemical properties of the water in the ligand binding site of a protein or in an active site on an enzyme, density of immobile binding sites, or electrostatic interaction.

2-naphthol-8-sulfonate (its structure is given below) is a novel photodissociative probe of water in protein binding sites because of its unusual behavior compared with its isomers such as 2-naphthol-6-sulfonate (its structure is given below). 2-naphthol-6sulfonate in aqueous solutions undergoes excited state proton transfer from the hydroxyl group to a water molecule, which is manifested by appearance of the fluorescence band of the protonated form (monoanion) accompanied by the fluorescence band of the deprotonated form (dianion) while only the protonated form is directly excited. When adsorbed on bovine serum albumin only the fluorescence of the protonated form was observed (33) which indicates that excited state proton transfer in this case is much slower than emission of fluorescence. On the contrary, we found that the rate of excited state proton transfer of albumin-bound 2-naphthol-8-sulfonate is still comparable to or greater than the rate of fluorescence. Thus, it can be used to characterize the processes and microenvironments of ligand protein binding sites such as the role of water in the binding sites of protein, and enzyme catalytic sites.








Animal models are often used in protein binding and pharmacokinetic studies, and the results obtained are then extrapolated to humans. Considering the high degree of amino acid sequence similarity that can be found in serum albumins from different species, it has been anticipated that other serum albumins had binding sites analogous to

03
SOH OH / L OH

SO3

2-naphthol-8-sulfonate 2-naphthol-6-sulfonate

the sites identified on human serum albumin. A characteristic of serum albumin is the ability to transport a multitude of ligands such as fatty acids, amino acids, steroids, metal ions, and drugs (34). The binding involves hydrophobic, hydrophilic, cationic and anionic substances. The ability of albumin to act as an important extracellular antioxidant (35) or impart protection from free radicals, and other harmful chemical agents (36) agrees well with the increased susceptibility of analbuminemic rats to cancer (37). It is widely accepted in the pharmaceutical industry that the overall distribution, metabolism, and efficacy of many drugs can be altered based on their affinity for serum albumin. In addition, many promising new drugs are rendered ineffective because of their unusually high affinity for this abundant protein. Obviously, an understanding of the chemistry of the various classes of pharmaceutical interactions with albumin can suggest new approaches to drug therapy and design. Consistent with this biological role and a high affinity for a variety of ligands, they are one of the most extensively studied and applied classes of proteins in biochemistry. However, as a protein, albumin is far from typical,








and the widespread interest in and application of albumin have not been balanced by an understanding of its molecular structure as a high-resolution crystal structure is not yet available for serum albumin. Since the 2.8 A resolution of the available crystal structure of the highly homologous human serum albumin (38) does not allow individual water molecules to be located, little is known about the role of water in the ligand-albumin interactions. For this reason the binding site of both human and bovine serum albumin has been chosen as a model to study the role of water in the binding site of protein.

Albumin is characterized by a low content of tryptophan and methionine, and a high content of cystine and charged amino acids such as aspartic, glutamic, lysine and arginine. The main reason for the high solubility of albumin is its high total charge. Human serum albumin (HSA) is a protein of 69000 Da and consists of 585 amino acids with one Trp residue at position 214. The three-dimensional structure of HSA, as determined by X-ray crystallography, consists of 67% a-helical structure. The rest of the residues form random coils and extended chains (38). The overall topology of the molecule is heart-shaped, with three repeating helical domains labeled I, II, and III, and each domain was identified to consist of two subdomains A and B. The primary ligand binding sites on HSA were shown to be located in subdomains IIA and liIA, although numerous other low-affinity sites also exist. Bovine serum albumin (BSA) shares a high degree of homology with HSA and was suggested to have a structure similar to that of HSA, with minor differences. For example, BSA has two Trp residues, one near the surface of the protein and the other in the interior, in contrast to the single tryptophan present in HSA (39). BSA also has a high degree of a-helical content. BSA can be








readily crystallized for purification purposes, but large crystals suitable for X-ray diffraction studies have not yet been obtained. The ability of both albumins to undergo a major reversible conformational isomerization with changes in pH was observed several decades ago (40, 41) and several forms were classified: the N form, or normal form, is predominant at neutral pH; the B form, or the basic form, occurs above pH 8.0; the F form, or fast migrating form, produces abruptly at pH values less than 4.0; the E form, or expanded form, appears at pH less than 3.5; and the A form, or aged form, occurs with time at pH values greater than 8.0. At the present time, little structural information is known about the B form. There is a decrease in helical content of the E, F, and B forms. There is evidence that two major and structurally selective organic ligand binding sites exist on human albumin (42) and Sudlow et al. have termed these sites I and II. Both sites were originally distinguished on the basis of differences in the pattern of ligand binding and displacement. The ligands that absorb to site I are mainly weak acids with bulky structures with a negative charge in the center of the molecule (43). Azapropazone, phenylbutazone and warfarin are typical examples of the ligands which interact with site I and this locus is often referred to as the warfarin binding site. Ligands with an elongated shape, with or without a negative charge, bind to site II and diazepam together with octanoic acid and L-tryptophan are typical ligands that selectively interact with this site. Thus, this binding locus is known as the benzodiazepine/indole binding site (43). Spectroscopic investigations, binding to albumin fragments and affinity-labeling of amino acid residues, such as tryptophan 214 and tyrosine 411, suggested site I is located in domain II whereas site II is found in domain III (44). The locations of these sites has been confirmed by X-ray crystallography and He and Carter (38) have shown ligand








binding sites I and II proposed by Sudlow et al are formed within subdomains IIA and IIIA, respectively.

To use 2-naphthol-8-sulfonate as a protein-binding site probe, it is necessary to carry out binding studies about the 2-naphthol-8-sulfonate - protein binding and determine the binding parameters associated with this process: for example, influence of medium composition on the binding, the binding constant, the number of binding sites, the location of binding, etc. In this content, a systematic investigation of the binding behavior of 2-naphthol-8-sulfonate to BSA and HSA was also undertaken.















CHAPTER 2
BACKGROUND

An understanding of proton transfer in the lowest excited singlet state in aqueous solutions cannot be adequate without first possessing knowledge of the physical chemistry of water. Therefore, this subject is briefly reviewed. A review of prototropic conductivity in water and dynamic aspects of proton transfer in the lowest excited singlet state is also included. The material for these reviews was taken from references 45, 46, 47 and the references contained therein.

The Structure of Water


Due to electronic hybridization of molecular orbitals, the water molecule has the shape of a V, with an angle (in the liquid phase) of 105 between OH arms. Two of the electron pairs of oxygen are non bonded and the other two electrons are shared with the two atoms of hydrogen. As a consequence, there is an excess of negative charge in one side of the molecule and an excess of the positive charge in the other side (the side of the hydrogen atoms). The resulting dipole moment is very high. It is then usual to represent the water molecule by a central core including the oxygen atom and four point charges distributed around it with tetrahedral symmetry. Two are positive and correspond to the hydrogen atoms and the two negative are the lone pairs. Water is not a free rotor, but rather tightly involved in hydrogen bonding. The local structure around any given water








molecule tends to be tetrahedral. There are roughly four water molecules in the first solvation shell, two donating and two accepting hydrogen bonds from the central water. X-ray diffraction yields a clear description of the oxygen-oxygen radial distribution function (48). Its peaks correspond to nearest-neighbors and next-nearest-neighbors in tetrahedral symmetry. These peaks diminish in amplitude with increasing temperature, indicating a decreasing hydrogen-bonding content. There is evidence for the persistence of even larger ice-structure motifs in liquid water (49). Likewise, Raman experiments

(50) and molecular-dynamics simulations (51) show that in liquid water most molecules are involved in either three or four hydrogen bonds, with the enhancement of tetrahedral symmetry at lower temperatures. On the other hand, water molecule rotation does occur. Current interpretations of Raman-induced Kerr effect (52), Rayleigh light scattering (53) and inelastic neutron scattering (54) suggest that water molecule reorientation takes 1-2 ps at room temperature.

Prototropic Conductivity


Proton conductivity in water is abnormally high. At room temperature, its limiting ionic conductance is about seven times that of a sodium cation, or approximately five times that of K'.

In solution, the proton is not adequately described in terms of the bare ion. Using spectroscopic techniques, it has been demonstrated that the proton in water is localized in the form of relatively long-lived simple ion: the approximately symmetrical and planar H3O+ (hydronium) cation. Unlike water, which is involved in four hydrogen bonds as both a donor and an acceptor, the positive H3O does not allow hydrogen-bonding to its








oxygen. Using the detailed microscopic information obtained from their simulations, Tuckerman et al. verified that H3O is preferentially solvated as H5,02 and H904' complexes (55, 56). The three protons in H30O are equivalent and form hydrogen bonds to three H20 molecules, thus making up the H9,04 complex that is embedded in the hydrogen bond network. In the HsO2 complex, the two water molecules that bind the excess proton between them, each form hydrogen bonds to a pair of H20 molecules, embedding this complex in the hydrogen bond network. The structure of H502' and H904' complexes can be analyzed by determining the radial distributions of O and H atoms with respect to the oxygen atom O* binding to the excess proton. Consistent with the ionic nature of H3O, the O*O distance in H904+ complex ro.0 = 2.5 A is shorter than roo = 2.8 A in pure water while the O*H bond length ro.H ,1.1 A is slightly longer (rOH , 1.02 A for pure water) and weaker than in pure water. The absence of the 1.9 A OH maximum in the O*H radial distribution confirms that H3O' behaves like a cation that only donates protons to hydrogen bonds with the coordinated H20 molecules, but itself receiving none. One of the O*O distances in H502 complex, i.e., O*O*, is shorter than the other two, and ro.H = 1.3 A. The H502+ and H904+ complexes might at first appear to be clearly distinct structures. However, analysis of results of radial distribution functions for both complexes show that the environment of the O* atoms associated with the excess proton are remarkably similar. Two of the three ligand H20's in the H904+ structure are hydrogen-bonded to three second neighbors of the H30+ ion. This, together with the H30+ ion itself, yields the normal fourfold coordination in water. The third H20 bonded to H30 has one less hydrogen bond. This undercoordinated molecule frequently (but not always)








turns out to have been the partner in the former H502 complex state. A small shift of the proton to the left or the right converts the symmetric H5,02 into an asymmetric H904' conformation.

Gas-phase mass-spectrometric data (57) show that water ligands in the first

solvation shell are more tightly bound than those in the second and subsequent shells. Although these data cannot be directly transferred to bulk solution, it is clear that hydrogen bonding to H30 is considerably stronger than in bulk water. The extra strong hydrogen bonding correlates with shorter OH ...0O distances, 2.55 A compared with 2.8 A in bulk water. A proton conductivity mechanism that involves the rupture of even one of these three bonds will lead to an activation energy exceeding the observed 2-3 kcal/mol value. Thus, the first solvation shell of H3O must remain intact during the proton transfer act (58).

Proton conductivity is incoherent as shown by experiments that proton

conductivity in ice is about a factor of two slower than in super-cooled water of equal temperature (59, 60). The evidence for proton conductivity in water is unequivocal, it increases with increasing temperature and pressure (61). That the fraction of hydrogenbonded water decreases with increasing temperature is clear intuitively as well as from simulations (51) and experiment (52). Thus proton conductivity increases as hydrogen bonds weaken.

The deuterium isotope effect on prototropic mobility is rather small, 1.4 at room temperature (62). As suggested by Westheimer and Melander (63) kinetic isotope effects are maximal for the symmetric, AG = 0, reaction. Proton hopping between two water








molecules is expected to correspond to this symmetric limit. The maximal kinetic isotope effect for reactions involving proton transfer is 6-10 (64). The small isotope effect on proton conductivity suggests that the reaction coordinate does not involve proton motion. The 1.4 deuterium isotope effect on proton conductivity (62) can also be interpreted as arising from water reorientation that leads to hydrogen bond cleavage. Rotation around the heavy oxygen is a nearly pure hydrogen motion, so that the isotope effect should be given by the mass ratio, mD/mH = 2. The OD ...0O versus OHO 0 enthalpy difference is thus expected to be AAH = '2RT In2 = 0.2 kcal/mol at room temperature. This has been recently verified experimentally (52). The possibility that proton motion within clusters is rate-limiting may be ruled out by the small deuterium isotope effect.

Since the hydrogen bond content of water is so high, every molecule participating in 3-4 hydrogen bonds (51, 52), the whole body of water may considered as one huge cluster. The formation of new hydrogen bonds at the periphery of water clusters is not the rate limiting step. Proton conductivity correlates with hydrogen bond cleavage rather than with its formation, as suggested by the fact that it increases with decreasing hydrogen bond content of water, i.e. at higher temperatures and pressures (61). Proton hopping times as obtained by NMR and "70 resonance (65), rp & 1.5 ps agree with single molecule reorientation times (52-54), a process requiring hydrogen bond cleavage.

With the above mentioned restrictions in mind, a revised Grotthuss mechanism of proton conductivity in water is proposed by Agmon (45, 66) as follows.

Consider a triply coordinated H30, i.e. the H904 cation, as the reactant state.

Since the rate limiting step involves cleavage of a hydrogen bond, but not one in the first








solvation shell, it is natural to consider hydrogen bond cleavage in the second solvation shell as the rate limiting step. This leads to isomerization of the H904' cation into the H502 cation. This is depicted in Figure 2-1. For clarity, four of the oxygen atoms are identified by corresponding letters.


a





(b)


--a


- d


Figure 2-1. A molecular mechanism for prototropic mobility. Taken from reference 45.



Initially, the proton is located on Ob and H20c is one of the three water ligands in the first solvation shell, Figure 2-la. Thus, Ob H'"Oc is an extra strong bond which cannot easily break. H20, has at least one hydrogen bond donating water ligand, H2Od, which is








in the second solvation shell with respect to Ob and thus, of ordinary strength. The cleavage of the O---HOd bond is postulated to be the rate limiting step, Figure 2-1lb. The reorientation of H2Od may be quite small, so that just a single hydrogen bond is broken. This may cost 2-3 kcal/mol and take around 1 ps at room temperature. Following bond cleavage, one can expect ultrafast (fs) readjustment of bond angles and bond lengths to form H2Ob-...H --OcH2. For example, the Ob"0Oc distance should shrink by about 0.15 A, from 2.55 A in H904' to the 2.4 A distance of H502' in vacua (67). At the same time, the distance between Ob and the remaining two oxygens are expected to increase from 2.55 A to below 2.8 A. Subsequently, fast fluctuations of surrounding water dipoles momentarily stabilize Ob or Oc, leading to large proton polarizabilities within this complex (68, 69). Eventually, one of these fluctuations couples to reorientation of the H20a water, which donates the fourth hydrogen bond to Ob. The proton is now located on Oc, Figure 2-1c. The H502' has reisomerized to a H904' cation, centered on a neighboring water molecule. The proton has effectively hopped incoherently across the O-O distance in H904".

Tuckerman et al. (55, 56) conducted ab initio molecular dynamics simulations on a proton in a 32-water molecule cluster combining a density functional description of electronic structure and finite temperature dynamics. For H30' ion they found a dynamic solvation complex, which continuously fluctuates between a H502' and a H904' structure as a result of proton transfer. H30' and its associated complexes integrate naturally into the hydrogen bond network, enabling proton transfer without substantial activation and rearrangement of the solvent. According to the results of the simulation they suggested that the rate limiting step for the migration of the excess proton is the concerted dynamics of the second solvation shell hydrogen bonded to the ligand H20. The quantum dynamics








and energetics of an excess proton in water studied computationally by Lobaugh et al.

(70) indicated that the mechanism of proton mobility depends critically on the dynamics of the water molecules in the second solvation shell of the proton transfer complex and the oxygen-oxygen distance fluctuations of water molecules that hydrogen bond to the complex. Such computational work is qualitatively consistent with the mechanism proposed by Agmon (45, 66).

Prototropic Reactivity in Electronically Excited States


The excitation of a molecule from its ground state (S,) to its lowest excited singlet state (S,) causes a change in electronic dipole moment of the excited molecule. The electronic charge distribution of the excited molecule is generally quite different from that of the molecule in its So state. As a result, the chemical properties of the molecule in S, state often are significantly different from those in the ground state. Since the lifetimes of the lowest excited singlet state are typically of the order 10-" - 107 s, certain types of chemical reactions such as proton transfer can occur within the lifetime of the excited state. Whether or not the proton transfer in S, state can take place depends on several factors, one of those is the properties of acids and bases in S, state. Properties of Acids and Bases in the Lowest Excited Singlet State

Upon excitation to the S, state, the electronic distribution of an aromatic acid or

base changes. Acidic and basic functional groups may, depending on their nature, become either enriched or deficient in electronic charge when the molecule is excited. This may lead to substantial differences between the acidities of the same functional groups in So and S, as reflected by the respective equilibrium constants pK and pK*. Usually, only








those functional groups bonded directly to an aromatic ring will experience changes in charge distribution sufficient to cause detectable differences between pK and pK*

Electron-donating groups have lone electron pairs which upon excitation, may be transferred to the lowest-lying vacant rt* orbitals of the aromatic system. Groups of this kind include hydroxyl, sulfhydryl, and amino groups. Upon excitation, the electronic charge density at these groups decreases and consequently, a proton may be more readily lost from or with more difficulty added to the group in S, than in So. Consequently, pK* < pK, and these groups are thus more acidic (or less basic) in S, than they are in So.

Functional groups having low-lying vacant nt orbitals will, as a result of electronic excitation, accept electronic charge from the aromatic system. Examples include carbonyl, carboxyl, carboxylate, and amide groups. The increase in electronic charge density resulting from excitation makes it more difficult to remove a proton (or easier to add a proton). As a result, pK* > pK, and these groups are, therefore, less acidic (or more basic) in S, than they are in So.

The Effect of State of Protonation on Electronic Spectra

Protonation of a functional group which is attached to an aromatic system may have a profound effect on the absorption and fluorescence spectra of that molecule. This is the result of the electronic charge stabilization incurred by the presence of the proton at the functional group.

In molecules containing an electron-withdrawing group, the excitation of those molecules from the So to the S, state results in the movement of electronic charge to the electron withdrawing group. This causes the S, state to be stabilized to a greater degree








than is the So state by protonation at the electron withdrawing group and decreases the energy difference between So and S, state in the protonated molecule. The spectral result is the shifting of the longest-wavelength absorption bands and the fluorescence spectrum to longer wavelengths upon protonation and to shorter wavelengths upon deprotonation.

In molecules containing an electron-donating group, the loss of electronic charge from the functional group which accompanies excitation from the So to the Si state is inhibited by protonation and facilitated by deprotonation. This causes the energy difference between So and S, to be greater in the conjugate species having the higher state of protonation. As a result, the longest wavelength absorption and fluorescence bands of these molecules shift to shorter wavelengths upon protonation and to longer wavelengths upon deprotonation.

The displacement of the long wavelength absorption and fluorescence bands of molecules possessing electron acceptor groups, to longer wavelengths upon protonation and to shorter wavelengths upon deprotonation, is related to the increase in basicity and decrease in acidity upon going from the So to the S, state. In molecules having electron donor groups, the movement of the long wavelength absorption and fluorescence bands, to shorter wavelengths upon protonation and to longer wavelengths upon deprotonation, is indicative of a decrease in basicity and an increase in acidity upon going from the So to the S , state.

The Effect of Solvents on Electronic Spectra

The effect of solvents on electronic spectral bands depends upon the nature and relative strength of the interaction of solvent molecule with the ground and excited states








of the solute molecule. These interactions are predominately electrostatic in nature. All solvent effects may be reduced to a comparison of whether the ground or excited state is more stabilized relative to each other. In general, if the excited state is more polarizable than the ground state, then interactions with polar solvents lead to stabilization of the excited state relative to the ground state and transitions shift to longer wavelengths.

For nonpolar solutes dissolved in nonpolar solvents, only the dispersion forces are expected to contribute significantly. These forces are relatively weak; hence, the absorption and fluorescence wavelength maxima of most nonpolar molecules are not strongly solvent-dependent. This leads to small shifts of absorption and fluorescence spectra with increasing dielectric constant of the solvent.

When a polar solute is dissolved in a polar solvent, dipole-dipole interactions are dominant. In many but not all solute molecules, electronic excitation increases the degree of charge separation in a molecule, and thus, increases the dipole moment. In such a case, the energy of the excited state should be decreased to a greater extent than that of the ground state by increasing the "polarity" of the solvent. Thus, increasing the "polarity" of solvent should cause both So -+ S, absorption and SI - So fluorescence to shift to lower energy (longer wavelength).

There is an ever-present possibility that solute molecule containing polar

functional groups may engage in specific chemical interactions such as hydrogen bonding with some solvents. Such specific interactions, if present, usually are the dominant factors affecting the absorption and fluorescence wavelengths of the solute. If a nonbonding pair on a solute molecule is coordinated by a hydrogen-atom of the solvent, the hydrogen








bonding interaction lowers the energy of the ground state as well as that of the n, n* state of the solute. However, because the ground state molecule has two electrons in the nonbonding orbital and the excited state has only one, the stabilization of the ground state is greatest. As a result, the energies of n -- n* absorptions increase (the spectra shift to higher frequencies or shorter wavelengths) with increasing solvent hydrogen-bond donor capacity. Hydrogen-bonding solvents also have a marked effect upon intramolecular charge-transfer absorption spectra. Hydrogen-bond donor solvents interacting with unshared valence electron pairs on functional groups which are charge-transfer acceptors in the excited state (e.g. -COOH) enhance charge-transfer by introducing a partial positive charge into the functional group. This interaction stabilizes the charge-transfer excited state relative to the ground state so that the absorption spectra tend to shift to lower energies with increasing hydrogen-bond donor capacity of the solvent. An increase of hydrogen-bond donor capacity of the solvent tends to produce shifts to higher energies when interacting with unshared valence electron-pair on functional groups which are charge-transfer electron donor in the excited state (e.g. -OH, -NH2). Hydrogen-bond acceptor solvents tend to produce shifts to longer wavelengths when solvating hydrogen atoms on functional groups which are charge-transfer electron donors in the excited state. This is a result of the partial withdrawal of the positively charged proton from the functional group which facilitates transfer of electronic charge away from the functional group. Solvation of hydrogen atoms on functional groups which are charge-transfer electron acceptors in the excited state inhibits charge-transfer by leaving a residual








negative charge on the fimunctional group. Thus this interaction tends to result in shifting of the absorption spectrum to shorter wavelengths.

Solvent polarity and hydrogen bonding effects upon fluorescence spectra are qualitatively similar to those upon absorption spectra. In many cases, however, the fluorescence shifts of a giving solute in a giving series of solvents may not parallel, even qualitatively, the absorption shifts in the same series of solvents. This is understandable because there is a possible alteration of hydrogen bonding sites in the solute when the solvent cage is changed from the S, configuration to the So configuration.

Up to the present, attention has been focused on the effects of solvent properties on the spectral positions of the electronic spectra bands. However, the intensities of fluorescence and, to a lesser extent, absorption spectra also depend on the nature of the solvent. In general, solvation which interferes with electronic interaction between aromatic ring and functional group tends to diminish the molar absorptivity as in the arylamines in highly protic solvents or carboxylic acids in hydrogen bond acceptor solvents.

The intensities of fluorescence spectra are extremely sensitive to solvent polarity and hydrogen bonding properties. In general, in molecules with a n, n* state as the lowest excited singlet state, in the isolated molecule strong fluorescence is favored by high polarity and high proticity of the solvent whereas weak or no fluorescence is favored by aproticity and low polarity of the solvent. In many molecules whose lowest excited singlet states are of the nr, 7t* or intramolecular charge transfer types, increases in solvent hydrogen bonding capacity frequently decreases fluorescence quantum yields. In








numerous cases, hydrogen bonding of a fluorescent molecule with solvent causes decreases in fluorescence quantum yields because of a substantial increase in the rate of internal conversion, at the expense of fluorescence (71, 72). Steady-State Kinetics of Excited State Proton Transfer Reaction

The steady-state approach to the determination of proton transfer rate constants of acids or bases in the S, state is based on the assumption of the attainment of a steady-state involving the various photophysical and photochemical processes deactivating S,. In the absence of buffer species (i.e. proton is the only protonating species and water is the only proton acceptor present in appreciable quantity) in aqueous solutions, proton dissociation of an excited hydroxyaromatics (26, 27) is given in equation (1-1): k12
OOH* + H20 k21 OO-* + H30+

k(r + nr) hv hV2 k'(r + nr)

DOH + H20 ( O- + H30+

(1-1)

The variations of the relative fluorescence quantum yields of hydroxyaromatics and their conjugate bases with pH are known to depend upon the kinetics of proton transfer in the lowest excited singlet state (26, 27). Under the assumption, the relative fluorescence quantum yields of the excited hydroxyaromatics have been shown to vary, approximately according to

I1 + k2To'[H30+] (2-1)
- O (2-1)
0 1 + k12 T0 + k21,to'[H3O]

while those of the excited conjugate base hydroxylate species vary according to








1 k12 0O (2-2)
0' 1 + k12To0 + k21 0'[H3O](

where [H3O ] is the molar concentration of hydrogen ions, to and ro' are the lifetimes of excited acid and conjugate base in the absence of proton transfer (i.e., in the low and high pH limits, respectively), and kl2 and k21 are the rate constants for dissociation of the excited hydroxyl species and protonation of the excited hydroxylate species, respectively. More refined treatments have corrected equations (2-1) and (2-2) for transient reprotonation of the hydroxylate species prior to the establishment of the steady state and for the effect of ionic strength on the "equilibrium" between the reactants and the activated complex in the transition state (73, 74).

Dividing equation (2-1) by equation (2-2) yields

o 1 k21 0
+ 1 '� [H3.+] (2-3) '/o' k,20 + kl2t

A plot of ( 0/0 )/( 0'/o' ) versus [H30'] should yield a straight line of slope k21T0' / kl2t0 and an ordinate intercept of 1/ kl2t0, so that if the lifetimes of the excited state acid and conjugate base can be measured or estimated, the rate constants of proton transfer, kl2 and k21, can be calculated.

In equation (2-3) k21 is often typical of diffusion-controlled reactions (k21 5 5 x 1010 M s') and generally To' 1 x 10-7 s, so that k21TO' < 10'. Consequently, at pH > 5, k21To0'[H30+] / kl2o -> 0 and 0/o, O'0', and therefore ( '/40' ) / ( /40) become independent of pH; that is,

I
(-) 1 (2-4)
0 c�onst --1 + k12 0o








S)o k120 (2-5)
( )onst 1+ k12T0


0 1
- - kl270o (2-6)

0 0 const= 2T(-6
001 00


which permits a quick determination of k 12. For excited conjugate bases with To' l x 10. s and / or with k21 < 5 x 100 M-' s-', [H30'] can be greater than 10-5 M and still the acid and its conjugate base may demonstrate ( #'/o0' ) / ( /0) independent of pH. In other words, the smaller k21T0', the lower is the pH at which the reprotonation of the excited conjugate base, as reflected by the dependence of /00 and '/I40' on [HaO], will be observed.

It is important to note that in equations (2-1) and (2-2), kl2 and k21 are somewhat dependent upon the composition of the reaction medium (i.e., they general vary with [H30']). The true rate constants of the reaction kl2(0) and k21 (0) are the rate constants in the pure reference solvent, in this case, water, at infinite dilution of solutes (i.e., when [H3O] = aH , where aH' is the activity of the hydrogen ion). In relatively dilute acidic solutions of pH 1-4 the correction of equation (2-3) for medium effects has made use of the Bronsted kinetic activity factor F (75), based upon the Debye-Hickel type ionic screemning treatment.


-log F = 2AIZ4f (2-7)
1 + aB(

where A and B are constants of the solution dependent on temperature (T) and dielectric


constant (s),








A = 1.826 x 106 (ET)-3/2 (2-8) B = 5.031 x 109 (ET) -/2 (2-9)

t is the ionic strength of the solution, and a is, to a first approximation, taken to be a mean ionic size parameter for all participants in the reaction, Z is the charge of the dianion. The factor F corrects the rate constants at each value of g to the value corresponding to reaction at zero ionic strength. A plot of ( /0 )/( '/4o' ) against F[H30O] then should yield a straight line of slope k21T'0' / kl2T0 and an ordinate intercept of 1/ k2'r0.

The fluorescence lifetime measurement methods can be classified into two main categories, namely indirect and direct methods.

Indirect methods of fluorescence lifetime measurement are listed as follows:

(a) based on the relationship that the fluorescence quantum yield (4) is related to the fluorescence lifetime (T):

S= T / tsN (2-10)

where r is the actual lifetime and is proportional to the relative fluorescence intensity measurement and rN is the natural or radiative lifetime of the excited state (the lifetime that would be measured if fluorescence was the only process originating from S,).

(b) based on the relationship described as:

1/rN = 2900n2 02sdfV (2-11)

io in pm'" is the wavenumber corresponding to the maximum absorption, - is the molar absorptivity in 1 mol-' cm-', and n is the refractive index of the solvent. For most








materials, the experimental values of half band widths are ca. 0.3 cm', and assigning n2

2 and the region of interest as 2.5 pm', equation (2-9) is reduced to

1/tN 104 m (2-12)

(c) based on the quenching of fluorescence by using the Stern-Volmer relation (76):

l0/ I= 1 + K,[Q] (2-13)

where I0 and I are fluorescence intensities observed in the absence and presence of a concentration [Q] of quencher. KQ, the quenching constant, is related to the bimolecular reaction rate, kFQ, by the equation

KQ = kFQTO (2-14)

Therefore, by estimating the K0 using the fluorescence quenching data and assuming that kFQ has a value of about 10o 1 mol' s' corresponding to a diffusion-controlled reaction, a value for To can be obtained.

The above indirect methods provide an estimate of the fluorescence lifetime, but are tedious and are not favored over the other approaches which allow direct lifetime measurement.

At least two direct experimental methods of measuring lifetime are available, namely time resolved (pulse excitation) and phase resolved decay approaches.

(a) In the time resolved method the sample is excited with a special high speed flash-lamp or pulsed laser and the time-dependent decay of fluorescence intensity is measured. The rate of decay of the initially excited population is

dN(t)
dt -(7 + k) N(t) (2-15)

where N(t) is the number of excited molecules at time t following excitation, y is the








emissive rate, and k is the rate of nonradiative decay. At t = 0, N(t) = No (initial population of molecules in the S, state), integration of equation (2-15) yields

N(t) = No e -t,, (2-16)

where t0 = ( y + k )- and is the lifetime of the excited state. Since the fluorescence intensity I(t) = kN(t), thus

I(t) = 10 e -t (2-17)

where I0 is the intensity of light incident upon the absorber. Equation 2-17 represents the exponential decay with x0 being the time for the fluorescence to decay to 1/e of its initial value.

(b) In the phase resolved method, the sample is excited with sinusoidally modulated light. The phase shift and demodulation of the emission, relative to the incident light, is used to calculate the lifetime. For simple exponential decays, the fluorescence lifetime is related to the phase-shift, E, by the following relationship

To = tan � / 2tf (2-18) wherefis the frequency of modulation.

Protein Binding


A characteristic of serum albumin is the capability of extensively binding a

variety of different substances in a reversible manner. For this reason serum albumin has been widely used as a model compound to study the factors involved in the interaction between macromolecules and low molecular weight compounds. Studies on the binding behavior of serum albumin have also been stimulated by the physiological role which this protein serves as a vehicle in the exchange of fatty acids and bilirubin, between the








tissues. Furthermore, many drugs are bound to serum albumin, thus, ensuring a more protracted pharmacological effect of these substances.

We initially observed that the excited state proton dissociation of albumin-bound 2-naphthol-8-sulfonate is slowed down compared with that of the naphthol derivative dissolved in aqueous solutions. This kinetic parameter can be used to probe the microenvironment of the binding sites of serum albumin at which 2-naphthol-8-sulfonate is bound. As mentioned in previous chapter, to use 2-naphthol-8-sulfonate as a proteinbinding site probe, it is necessary to carry out binding studies about the 2-naphthol-8sulfonate - protein binding and determine the binding parameters associated with this process: influence of medium composition on the binding, the binding constant, the number of binding sites, etc.

Binding Studies - Experimental Methods

Equilibrium dialysis (77), ultrafiltration (78), and HPLC (79) have been widely used to determine the binding affinity and site specificity of small molecules for serum albumin. Equilibrium dialysis and ultrafiltration are labor-intensive and, hence, not wellsuited for rapid analysis, whereas HPLC assays yield only a "percent bound" instead of the actual affinity constants.

The binding of ligands to proteins may induce a change in a spectral property of the protein or even the ligand. Advantage can be taken of the spectral perturbation to monitor the interaction of ligand and protein. Changes in absorbance, fluorescence, magnetic resonance and optical rotary dispersion have been utilized to measure the interactions of ligands to proteins. Fluorescence spectroscopy is arguably one of the most








versatile techniques for studying ligand protein interactions. The interaction may affect a number of fluorescence parameters such as the fluorescence intensity of the protein, the fluorescence intensity of the ligand and the fluorescence polarization of the ligand. Moreover, many ligands which do not fluoresce can be studied by this method if they either quench the native fluorescence of the protein or displace a fluorescent probe attached to the protein. Fluorescence spectroscopy is more rapid than equilibrium dialysis for studying ligand protein interactions, and fluorescence is much more sensitive than many other techniques such as absorption spectroscopy (80). For example, the fluorescence method makes use of the considerable increase in intrinsic emission that occurs when warfarin (81) and other fluorescent compounds are transferred from the polar, aqueous environment to the nonpolar, hydrophobic binding site of albumin. Albumin itself shows a very modest, although measurable change in its intrinsic fluorescence upon binding of any ligand (82, 83), whereas the fluorescence quantum yield of warfarin increases as much as seven times upon binding to HSA (81). Similarly, large increases in fluorescence accompanying binding of dansyl- I -sulfonate and dansylsarcrosine (82, 83) to albumin were observed. Binding Equation - Derivation

For the vast majority of cases, ligand binding to proteins may be considered to be a reversible interaction which obeys the law of mass action.

The simplest case to consider is:








P + L -+ PL (2-19)

where a protein P can bind, reversibly, a ligand molecule L at a single site to form a protein-ligand complex, PL. A association constant (k) can be defined:

k = [PL]/[P][L] (2-20) and the protein conservation equation written as:

[P,] = [P] + [PL] (2-21) where [P,] is the total concentration of the protein. Then, [P] = [P,] - [PL] and

k = [PL]/([L] [P,] - [PL][L]) (2-22)

The ratio [PL]/ [P,] represents the moles of ligand bound per mole of protein and is given the symbol r. Equation (2-22) can be rearranged to

[PL]/ [P,] + [PL][L]k/[P,] = k[L] (2-23) and substituting r for [PL]/ [P,], equation (2-23) becomes

r + rk[L] = k[L] or r = k[L]/(1 + k[L]) (2-24)

Equation (2-24), then, represents the equation for the special case of one binding site on the protein molecule. If there are a number (N) of independent, identical binding sites, the equation has the following form

r, + r2 + ... r = total = Nk[L]/(1 + k[L]) (2-25)

The free or unbound ligand concentration [L] present at equilibrium is determined by the appropriate analytical method. Then equation (2-25) can be applied to calculate the binding parameters k and N. It is not uncommon for more than one type of site to exist on








a given protein. Each site has its own association constant, and for this reason the following more general form of the binding equation has developed (84):

rota, = Nik[L]/(1 + ki[L]) + N2k2[L]/(1 + k2[L]) + ... N. k. [L]/(1 + k. [L]) (2-26)

For ligands like 2-naphthol-8-sulfonate, whose acidic form fluoresces much more intensely when bound than as the free ligand, the fraction of ligand bound, a, is usually determined from the spectrofluorometric titration by using the equation (85):

a = (FP - Ff)/(Fb- Ff) (2-27)

where FP and Ff are the fluorescence intensities of a given concentration of ligand in solutions of low protein concentration and in solutions without any protein, respectively; and Fb is the fluorescence intensity of the same concentration of the fully bound ligand. The latter is taken to be the fluorescence intensity of the ligand in the presence of excess protein. Similarly, the fraction of ligand free, P3, is determined by using the equation:

13 = (Fb - Fp)/(Fb- Ff) (2-28)

The total ligand concentration, [L,], is equal to the sum of the unbound and bound concentrations.

[L,] = [L] + [PL] (2-29)

then, 03 = (Fb- Fp)/(Fb - Ff) = [L]/[L,] or

[L] = P3[L,] = [L,](Fb - F,)/(Fb - Ff) (2-30)

The value of r can be calculated from the fraction of ligand bound by using the following relationship:

r = a[L,]/[P,] (2-31)








Binding Studies - Treatment of Data

Binding data may be plotted in several ways in order to determine the values of N (number of binding sites) and k (the association constant). A direct plot of r, the number of moles of ligand bound per mole of protein, as a function of free ligand concentration is usually hyperbolic and does not yield accurate values of N and k. The reciprocal plot, 1/ r vs. 1/[L], from equation (2-25), enables the direct determination of N and k from the graph since the plot should produce a straight line with intercept 1/N and slope 1/Nk. However, this plot spreads the low values of 1/ r poorly so that the values for N, and therefore k, are not reliable.

Scatchard (86) used equation (2-25) in the rearranged form

r/[L] = kN - k r (2-32)

to develop an alternative method of plotting the binding study data. The Scatchard plot is simply a plot of r /[L] vs. r. Such a plot yields a straight line, and values for N and kN may be determined from the intercepts on the abscissa and ordinate, respectively, when the N binding sites are noninteracting and have the same binding affinity. It is advisable to obtain the intercept and slope values by means of regression analysis rather by finding the best fit using a ruler. The level of the significance of the fitted curve and thus, the coefficient of regression, improves as the fit approaches the ideal value. The Scatchard plot gives a more even distribution of the data points than does the previously mentioned reciprocal plot of 1/ r vs. 1/[L]. However, if the individual binding sites have different binding affinities, due either to the distinct nature of the sites or to the interaction between the bound ligands at different sites, the Scatchard plot gives a downward curve








instead of a straight line. As the primary binding site becomes saturated with the ligand, these secondary sites become more important. If the binding strength of the individual binding sites are significantly different from each other, the Scatchard plot may still be used to estimate binding parameters for the individual binding site by appropriately analyzing the curve in relation to the equation (2-26) using a trial and error process of fitting values to the experimental curve (86). Computer software is now available for computing nonlinear fitting of binding data.

Rosenthal (87) and Klotz and Hunston (88) have suggested convenient methods to graphically represent the binding data in a complex system. Briefly, the curve describing the relationship between r /[L] and r should take into consideration the data of the high and low affinity curves. The description of Rosenthal (87) is particularly useful for this purpose.

Determination of Complex Stoichiometry - The Method of Continuous Variation

A spectrophotometric method for obtaining the stoichiometry of metal-ligand complexes was first described by Job (89) and later expanded by Vosburgh and Cooper

(90) to include cases in which a given pair of components form more than one compound. Because of the obvious similarity between the metal-ligand reaction and the proteinligand reaction, it was felt that this method of continuous variation (or Job study as it will be referred to hereafter) could be employed to determine the stoichiometry of the 2naphthol-8-sulfonate-albumin complex.

Job's method involves measuring some intensive property in a series of solutions of constant total molarity, but of varying metal-to-ligand ratio. The measured property is








generally absorbance or fluorescence intensity. In practice, two equimolar stock solutions, one of metal and the other of ligand, are prepared. A set of working solutions is then obtained by mixing VL ml of the stock ligand solution with (VT - VL) ml of the stock metal solution, where VT is a fixed total volume and VL is a variable, 0 < VL < VT. The absorbances or fluorescence intensities of these solutions are then measured at a fixed wavelength, and plotted as a function of mole fraction of ligand, (VL / VT), or of metal, {(VT - VL) / VT}. The position of maximum absorbance or fluorescence intensity on this plot, in relation to the mole-fraction axis, gives the stoichiometry of the complex.

To determine N, the maximum number of binding sites, solutions of equimolar concentrations of albumin and 2-naphthol-8-sulfonate are mixed in varying proportions. Letting (x) be the mole fraction of one of the components, (1 - x) would be the mole fraction of the other component. A suitable property of the resulting solution such as absorbance or fluorescence is measured. The difference (A) between the observed fluorescence intensity and the corresponding value in the absence of either 2-naphthol-8sulfonate or albumin is plotted against the mole fraction x of one of the components. The resulting curve should have a maximum (or minimum) if the fluorescence intensity measured is larger or smaller than that for either albumin or 2-naphthol-8-sulfonate alone. The mole fraction x at which A is a maximum (or minimum) is related to the maximum number of binding sites N, by the simple relationship

N = x / (1 -x) (2-33)

The values of x = 0.5, 0.67, 0.75 at which A is a maximum (or minimum) correspond to the maximum numbers of binding sites, N = 1, 2, 3 respectively. This treatment is








applicable when the quantum yields of fluorescence, and therefore, the fluorescence intensities of the bound ligand, are constant and independent of the stoichiometry of the complex in which the ligand resides.

Identification of the Binding Sites

Two specific binding sites have been established on the HSA: site I and site II

(42). X-ray studies of crystalline human albumin (38) support this view and indicate that site I and site II are located within specialized cavities in subdomain IIA and liIA, respectively. Matsushita et al. (91) and Panjehshahin el al. (92) reported that BSA has binding sites with similar properties to those of site I and site II on HSA. In addition, the crystallographic structure of equine albumin has also been determined, and the data suggests that there are also two specific ligand binding sites on this molecule (93) as well. All of these indicate that, with respect to binding sites, mammalian albumins are analogous to human albumin, considering the structural similarities between the molecules. To identify the binding sites for a ligand on HSA and BSA, competitive binding studies can be performed using typical site I and site II binding maker ligands such as warfarin, and diazepam, respectively.

The site-oriented approach to ligand albumin binding was first methodically studied by Sudlow et al. (42, 83). They characterized the relative strength of ligand at sites on albumins by monitoring the ability of the ligands to decrease the fluorescence of site-specific probes. The fluorescence quantum yield of the ligand bound to albumin was measured before and after the addition of the site-specific probes. The measured








fluorescence value corresponds to a relative fluorescence intensity expressed as a percentage of the initial fluorescence:
I -I
x 100%
II

or

I2
- x 100%
II

where 1, and 12 represent the fluorescence intensities of the ligand plus albumin without and with the probe, respectively.

A decrease in the fluorescence intensity of the complex (ligand-albumin) can be interpreted as a displacement of the ligand from its binding site by the added probe, probably through a competitive mechanism. The displacer and the ligand may be bound to the same site.

Characterization of the Binding Sites on Albumins - Fluorescence Quenching

Solute accessibility to fluorescent probes or groups attached to protein molecules is often used to monitor conformational aspects of these macromolecules. Solute accessibility is most often determined with fluorescence quenchers by measuring and comparing the specific rate of quenching of the fluorophore free in solution, kFQ, with its rate attached to the protein, kBQ, or by comparing values of kB, for two or more conditions that may affect the conformation of the protein. It is usually assumed that a decrease in quenching rate for the bound fluororophore indicates a decreased accessibility due to geometrical masking factors in the protein. However, elementary considerations indicate that for rapid, diffusion-controlled reactions that characterize the fluorescence quenching








process, kBQ should depend not only on masking factors but also on the translational and rotational mobilities of the liganded macromolecules as well as on orientational constraints imposed by the association of ligands with macromolecules. The lower the rotational mobility, the lower is the probability that a bound ligand, exposed on a small fraction of the surface area of the macromolecule, will meet a quencher molecule during its excited state lifetime.

In the fluorescence quenching experiment, the fluorescence intensities of the fluorophore attached to protein or free in solution are monitored as a function of the quencher concentration. The intensities are then analyzed with the Stemrn-Volmer equation

(76):

10/I= 1 + K9[Q] (2-13)

where 10 and I are fluorescence intensities observed in the absence and presence of a concentration [Q] of quencher. K9, the quenching constant, is related to the bimolecular reaction rate, kFQ, by the equation

KQ = kFQro (2-14) where to is the fluorescence lifetime measured in the absence of quencher.

In the absence of static quenching, plots of 10 / I vs. [Q] are expected to follow a

straight line, the slope of which may be used to determine K9. The parameter of interest is kFQ, which may be determined by measurement of ;o and from the slope of the SternVolmer plot.

Many hydroxyaromatics exhibit fluorescence from two species in aqueous

systems. When only the protonated form of the hydroxyaromatic exists in the ground








state, the dual emission results from the formation of the ionized hydroxyaromatics by excited state proton transfer. This reaction occurs because the hydroxyaromatics are much stronger acids in the excited state. Therefore, in these cases, the fluorescence of the protonated and deprotonated forms should be measured separately. The formula representing the relative fluorescence intensities of protonated and deprotonated forms, respectively are derived easily employing the following equation (94).

kl2
OOH* + H20 k - OH + H30+ k21
kf kq k fkq'


DOH + hv DOH O0- + hv' OO(2-34)

where k' = k21 [H30+].

Then, for the protonated form,

I / 10 = ( 1 + ac'[Q] ) / ( 1 + C [Q] + 13 [Q]2) (2-35) and for the deprotonated form,

I' / Io' = 1 / ( 1 + C [Q] + 13 [Q]2 ) (2-36) Therefore,

( I / I0 ) / ( I' / I0' ) = 1 + a' [Q] (2-37)

The meanings of constants ao, a', and 13 are as follows,

= {kq To ( l+ k') + kq' to'( 1 + kl2 to )}/ (1 + kl2 t0 + k' ro')

(2-38)

ac' = kq' To' / ( 1 + k' To') (2-39)








P = kqkq' To To' / ( 1 + k21 To + k' To') (2-40)

where to and to' are the lifetimes of excited protonated and deprotonated forms, respectively. When k' - 0, i.e., [H3O ] is small, the above equations become as follows.

I/Io= 1/(1+ {kqto/(1+k2 t0) }[Q]) (2-41) (I / I0 ) / (I' / 10')= 1 + k,' to' [Q] (2-42)

Using the experimentally determined values of kl2, t0, and To', kq and kq' are evaluated.

Commonly used quenchers, such as I-, Ag', were examined in this study and it was found that these quenchers were ineffective in this case. As reported (95), at room temperature, the fluorescence of naphthalene derivatives, including naphthols, is efficiently quenched by alkylamines. However, in the case of 2-naphthol-8-sulfonate, these alkylamines such as 1,4-diazabicyclo-(2. 2. 2) octane, N, N'-dimethylpiperazine, ethanolamine, tris(hydroxymethyl)-aminomethane, 2-bromoethylamine hydrobromide, showed little quenching efficiency at pH 5. Since these alkylamines are protonated at pH 5, therefore, they lose the ability to form hydrogen-bonded complexes with naphthol, which is thought to be necessary for quenching to occur (95). Thus, the weak organic base - pyridine - was chosen for the present study. Characterization of the Binding Sites on Albumins - Circular Dichroism

In the wavelength region where the optically active chromophore absorbs light

there is also an unequal absorption of the right and left circularly polarized components of plane polarized light, i.e., E L # R, where E L and e Rare the molar absorptivities for the left and right circularly polarized components. This phenomenon gives rise to circular








dichroism (CD). Circular dichroism is usually expressed as a differential dichroic absorption (E L " eR ), usually written A . The combination of unequal absorption (circular dichroism) and unequal velocity of transmission (optical rotation) of left and right circularly polarized light in the region in which optically active absorption bands are observed is a phenomenon called the" Cotton effect". When the differential dichroic absorption of a simple Cotton effect is plotted as a function of wavelength, the CD maximum and the absorption maximum will occurr at the same wavelength. Since CD results from a difference between E Land e R, it follows that this phenomenon is not observed outside of the wavelength region where the optically active chromophore absorbs light. Thus, the basic information which can be deduced from circular dichroism curves is obtained most easily in the immediate vicinity of the spectral region of maximal absorption.

Optically active chromophores may be divided into two extreme types: (a) the inherently asymmetric chromophore, and (b) the inherently symmetric chromophore which is asymmetrically perturbed. The former type includes such compounds as hexahelicene and certain substituted biphenyls and allenes in which the chromophore itself is asymmetric. These structures are seldom encountered in biological systems. By far, the large majority of chromophores are symmetrical and can only become optically active when perturbed by an asymmetric center or locus. When an asymmetric center induces optical activity in a chromophore which is part of the same molecule, then that molecule is said to be intrinsically optically active. Intrinsic optical activity can also occur if the chromophore has an asymmetric arrangement in space, e.g. a carbonyl group








in the a-helical regions of proteins. Extrinsic optical activity may be observed when a symmetric molecule binds to a macromolecule. Such optical activity usually results from an interaction between the ligand and an asymmetric binding site, but may also occur when binding results in an asymmetric spatial arrangement e.g. the helical arrangement of the basic dye molecules bound to poly-L-glutamic acid (96). Since the extrinsic optical activity is induced in the chromophore by its environment, rather than being inherent (as in the first type), the magnitude of the associated Cotton effect is often considerably smaller than in the first type of chromophore.

The sign and magnitude of an induced Cotton effect (extrinsic or intrinsic)

depends upon the spatial relationship between the asymmetric center and the perturbed chromophore (97). The space around a chromophore is divided into regions of positive and negative contribution to a CD signal according to well-defined symmetry rules. A given asymmetric center may therefore generate either a positive or a negative CD signal, depending upon its spatial relationship to the perturbed chromophore. The CD signal for each transition reflects the extent to which the transition contributes to the phenomenon of optical activity. As a rule, all CD bands corresponding to the same group of transitions should have the same sign upon alterations induced by effects changing the molecular symmetry. The magnitude of a CD signal will increase as the distance between the asymmetric center and the perturbed chromophore decreases (97).

Circular dichroism may be expressed as the differential dichroic absorption, Ae or as molar ellipticity [0] (deg. cm2 decimole') calculated from the formula








[0] = 0 x 100/ c (2-43)

where 0 is observed ellipticity (deg.), I cell pathlength (cm), and c concentration (moles/liter). Molar ellipticity and differential dichroic absorption are related by the expression

[0] = 3300 AE (2-44)

Circular dichroism is particularly important in the study of biological systems and macromolecules. It is used to yield information about a range of properties including:

(a) the structure of macromolecules, their secondary and tertiary structural content, and the degree of mobility of certain components;

(b) structural transitions, e.g. helix-coil transitions and the processes of unfolding and refolding of native and denatured structures;

(c) conformational changes of macromolecules induced by binding small molecules, e.g. substrates, ligands, coenzymes, effectors;

(d) symmetry properties of small chromophoric molecules in which optical activity is induced upon binding to macromolecules.

CD bands of proteins occur in two spectral regions. The far-UV or amide region (170-250 nm) is dominated by contributions of the peptide bonds, whereas CD bands in the near-UV region (250-300 nm) originate from the aromatic amino acids. In addition, disulfide bonds give rise to several CD bands. The two spectral regions give different kinds of information about protein structure. The CD in the amide region reports on the backbone (i.e. the secondary) structure of a protein and is used to characterize the secondary structure and changes therein, in particular the a-helix. CD bands in the near-








UV region are observed when, in a folded protein, aromatic side chains are immobilized in an asymmetric environment. The near-UV CD spectrum represents a highly sensitive criterion for the native state of a protein, thus, it can be used as a fingerprint of the correctly folded conformation. Examples of the conformational changes of macromolecules induced by binding small molecules, e.g. substrates, ligands, coenzymes, effectors have been observed. Trynda-Lamiesz et al. (98) found that the binding of adriamycin to HSA lowers the helicity of the native protein of ca. 15% by means of the changes in the far-UV region of CD spectra. Xu et al. (99) found that the a-helix of human erythrocyte membrane and BSA increased after the addition of croton alkaloids

(CA) (antitumor drugs) by measuring CD between 200-240 nm. Another antitumor drug cisplatin was also found effective in altering membrane a-helix content. These results combined with other experimental results indicate that the conformation change in membrane protein may account for the pharmacological effect of CA and cisplatin.

It has been found that many small symmetric molecules become optically active on binding to proteins and other macromolecules. In a majority of these examples optical activity results not from a special spatial arrangement of the ligand but from a perturbation of the ligand chromophore by an asymmetric locus at the binding site. Since extrinsic rotational activity (CD signal) reflects the three-dimensional characteristics of specific binding sites on a macromolecule, it offers an experimental means of exploring such sites. Examples of small symmetric molecules become optically active on binding to proteins and other macromolecules have been observed including those of the complexes of phenylbutazone and its analogues with HSA (100), benzodiazepines with HSA (101)








and with BSA (102). Chignell (103) found that the binding of flufenamic acid to porcine, equine, and bovine serum albumins generated biphasic CD signals similar to those just described for human serum albumin. This suggests that these proteins have similar ligand binding sites. In contrast, the binding of flufenamic acid to canine or ovine serum albumins generated a single positive CD signal, indicating that the ligand binding sites on these proteins are different from those on human serum albumin. More recently, Fleury et al. (104) found that pronounced differences in the interactions of monomeric (lactone and carboxylate) and the J-type self-aggregated form of camptothecin (CTP), an inhibitor of DNA topoisomerase I, with HSA and BSA were observed in CD spectroscopy. This can provide the possibility of identifying individual amino acid residues which play a key role in CTP/HSA interactions.














CHAPTER 3
EXPERIMENTAL


Materials


2-Naphthol-8-sulfonic acid potassium salt (2-naphthol-8-sulfonate) was obtained from TCI America, Portland, Oregon, USA. Bovine serum albumin (Lot No. 16H9310; 126H19307) as essentially fatty acid free (below 0.005%), prepared from fraction V, and human serum albumin (Lot No. 24H9314; 46H9319) supplied as 1 x crystallized and lyophilized, prepared from fraction V, were purchased from Sigma Chemical Company, St. Louis, Missouri, USA. Sulfuric acid (certified A. C. S.), perchloric acid (69-72%, TraceMetal grade), sodium hydroxide 2N solution (certified), methanol (HPLC grade, UV cutoff 205 nm), ethanol (HPLC grade, UV cutoff 205 nm), tertiary-butanol (certified), acetonitrile (HPLC grade, UV cutoff 190 nm), dimethyl sulfoxide (certified, spectranalyzed UV cutoff 262 nm), 1,4-dioxane (99 Mol % pure, certified), glycerin (certified, spectranalyzed), pyridine (certified A.C.S.), silver nitrate (laboratory grade), potassium iodide (U.S.P.), potassium phosphate primary standard monobasic (certified), sodium phosphate dibasic (certified A.C.S), sodium thiosulfate pentahydrate (certified A.C.S.), were obtained from Fisher Scientific, Fair Lawn, New Jersey, USA. N,N'dimethylpiperazine (98%), ethanolamine (99%), tris(hydroxymethyl)-aminomethane (99.9+%), phosphoric acid (A.C.S.) were obtained from Aldrich Chemical Company,








Inc., Milwaukee, Wisconsin, USA. Salicylic acid (U.S.P.), sodium phosphate tribasic (analytic reagent grade) were obtained from Mallinckrodt Chemical Works, St. Louis, Missouri, USA. 1,4-Diazabicyclo-(2.2.2)octane, 2-bromoethylamine hydrobromide were obtained from Eastman Kodak Company, Rochester, New York, USA. 2-Propanol (distilled in glass, UV cutoff 204 nm) was obtained from Burdick & Jackson Laboratories Inc., Muskegon, Michigan, USA. Ibuprofen was obtained from Research Laboratories of the Upjohn Company, Kalamazoo, Michigan, USA. Diazepam was obtained from Hoffmann-La Roche Inc., Nutley, New Jersey, USA. Phenylbutazone was received as a gift. All chemicals were used as received with no further purification. All studies except the pK and pK* titrations were conducted in deionized water buffered to desired pH with phosphate. The decision to use a phosphate buffer in this project was based upon the results of an investigation of the extent of complex formation between BSA and several common buffer ions by Klotz and Urquhart (105). Phthalate and veronal were found to be particularly effective in displacing bound ions from their protein complexes. Minimal effects were shown by glycine and phosphate while acetate, citrate and bicarbonate were intermediate in effectiveness. Pronounced competitive effects in binding were observed with nonbuffering anions such as nitrate and chloride. The protein solutions were prepared in phosphate buffer before experiment. Protein concentration was determined assuming a molecular weight of 69,000. The 2-naphthol-8-sulfonate concentration of the solutions used in the studies were chosen such that the absorbance at the isosbestic point, which was used as the excitation wavelength in the fluorescence measurements, would be less than 0.02 thereby reducing the probability of nonlinear fluorescence.








Instrumental


Steady state fluorescence measurements were made on a Perkin-Elmer LS-5

Fluorescence spectrophotometer. UV-visible measurements were made employing either a Shimadzu UV-2501PC UV-VIS recording spectrophotometer or Shimadzu UV160U UV-VIS recording spectrophotometer. CD measurements were made using JASCO J500C spectropolarimeter. The pH measurements were made with a Fisher Scientific Accumet 950 pH/ion Meter.

Fisher Scientific Vortex Genie Mixer was used in this study.

Socorex positive displacement micropipetter (1-5 l), Eppendorf adjustable micropipetter (2-10gl), Gilson adjustable pipetter (1-200PItl), Eppendorf adjustable pipetter (100-1000 gl), and Oxford adjustable pipetter (1000-5000 tl) were used in this study.

Methods


Ground and Excited State Ionization Constants Determinations (Titration Procedure)

Two series of solutions whose pH are below 3 were prepared with reagent grade sulfuric acid and tracemetal grade perchloric acid diluted with water. The exact molarity of the concentrated acid was determined by titration with standardized sodium hydroxide solution. The corrected Hammett acidity scale (106) was used to calibrate the concentrated sulfuric acid and perchloric acid. Solutions with various ionic strength in the pH range 3 - 11 were phosphate buffers. Solutions whose pH are higher than 11 were made by sodium hydroxide solution and the pH values were assigned according to








"Reagent chemicals and standards" (107). Deionized water was used throughout. Each sulfuric acid or perchloric acid solution, buffer solution or sodium hydroxide solution in a

5 ml volumetric flask was injected with 50 gl of 1.0 x 10-' M stock solution of 2naphthol-8-sulfonate in deionized water immediately prior to the measurements to minimize any possible decomposition errors. After each injection a Fisher Scientific Vortex Genie Mixer was used to mix the solution before reading. Values for the pK's and pK*'s were then obtained from the inflection points in the UV absorbance vs. pH and fluorescence intensity vs. pH plots, respectively. All UV and fluorometric determinations were performed at least three times, and the reported results are the mean of the data. Solvent Studies

As discussed in the chapters 1 and 2 of this dissertation, studies of 2-naphthol-8sulfonate in a variety of solvents have supported the concepts that

(a) except for very strong acids (pK < 0), no dissociation will take place unless water molecules are in the immediate vicinity to act as proton acceptors;

(b) observed changes in the fluorescence emission maximum are related to the polarity of the environment around the fluorescent species.

Both absorption and fluorescence spectra were obtained for 2-naphthol-8sulfonate and/or its conjugate base in the following solvents (if the solubility was sufficiently high): water or buffer solutions, methanol, ethanol (neat, 75% and 50%), 2propanol, tert-butanol, acetonitrile, dimethyl sulfoxide, 1,4-dioxane (neat, 75% and 50%) and glycerin. The purpose of this study was to evaluate the relationship between solvent polarity or hydrogen bonding strength and the fluorescence properties of 2-naphthol-8-








sulfonate and then to relate this to the behavior of the compound in the presence of albumin.

For these solvent studies, a concentration of 2-naphthol-8-sulfonate of 1 x 104 M and 1 x 10- M was used for the measurements of absorption and fluorescence spectra, respectively.

Qualitative Spectrophotometric Examination of 2-Naphthol-8-Sulfonate-Albumin
Complex Interaction

This initial experiment was conducted to evaluate qualitatively in terms of

spectral changes following the addition of an albumin solution to 2-naphthol-8-sulfonate dissolved in buffer solutions. At first, both UV and fluorescence spectra of equimolar solution (1 x 10-' M) of albumin and 2-naphthol-8-sulfonate were compared to that ofa 1 x 10-s M solution of 2-naphthol-8-sulfonate alone at the same pH. Gross differences between the spectra with and without albumin, specifically fluorescence intensity and /or UV absorbance changes and peak shifting, were noted. Since more than one isosbestic point in the UV spectrum was present, the excitation wavelength which resulted in the least fluorescence of albumin alone was chosen in the fluorescence study. After the initial experiment, the fluorescence emission spectra of three sets of solutions containing different molar ratio of 2-naphthol-8-sulfonate to albumin, i.e., 1:1, 1:5 and 1:10, were recorded. The fluorescence spectra for the latter two sets consisting different molar ratios of 2-naphthol-8-sulfonate to albumin were found to be essentially identical and did not change with increased concentration of albumin. Therefore, molar ratio of 2-naphthol-8sulfonate to albumin in the complex of 1:5 was chosen throughout this project, ensuring








that the intensities at monitored wavelengths are independent of the 2-naphthol-8sulfonate/albumin ratio.

The absorption and fluorescence spectra for albumin-bound 2-naphthol-8sulfonate at pH range 3-10 were recorded. Two different lots of albumin from each species were examined by recording the fluorescence emission spectrum. Fluorometric Determination of Compound-Albumin Complex Stoichiometry-Job's
Method

Job's method was employed to determine 2-naphthol-8-sulfonate-albumin

complex stoichiometry. For the Job study, a series of solutions ranging from 2-naphthol8-sulfonate alone (l x 10 M) to albumin alone (1 x 10s M) was prepared in buffer solutions. The composition of the intermediate 2-naphthol-8-sulfonate-albumin solutions was varied in such a manner that the total 2-naphthol-8-sulfonate plus albumin concentration in each solution remained constant (1 x 10- M). The fluorescence intensities were monitored for each solution using an excitation wavelength which corresponded to an isosbestic point resulted in the least fluorescence of albumin alone. The Job data were plotted in the usual manner, i.e., the differences between the observed fluorescence intensity of 2-naphthol-8-sulfonate-albumin solutions and fluorescence intensity at the wavelength of maximum emission of free 2-naphthol-8-sulfonate against the mole fraction of 2-naphthol-8-sulfonate, and from this plot the stoichiometry of 2naphthol-8-sulfonate-albumin complex could be determined. All fluorometric determinations were performed at least three times, and the reported results are the mean of the data.








2-Naphthol-8-Sulfonate-Albumin Titration Experiments-Determination of Binding
Constants and Number of Binding Sites

The fluorometric titrations were carried out as follows: 2.0 ml of albumin solution of appropriate concentration in a 1 cm quartz cell were titrated by successive additions of jpl volumes of ax 1 0-' M solution of 2-naphthol-8-sulfonate. Titration by 2-naphthol-8sulfonate with increasing albumin concentration was performed until a point was reached at which two successive titrations showed identical or very similar increases in fluorescence intensity throughout the titration. This was necessary to ensure that the proper albumin concentration level to fully bind 2-naphthol-8-sulfonate had been reached. Then, 2.0 ml of phosphate buffer solution in a 1 cm quartz cell were titrated by successive additions of pl volumes of a I x 10-' M solution of 2-naphthol-8-sulfonate to obtain the fluorescence of free 2-naphthol-8-sulfonate. For these titration experiments, the same excitation wavelength was used as in the Job study, and the fluorescence intensity was monitored at the emission maximum of the 2-naphthol-8-sulfonate-albumin complex. After each addition of 2-naphthol-8-sulfonate solution, a Fisher Scientific Vortex Genie Mixer was used to mix the solution before reading.

According to the relationship

F = I0[2.3A - (2.3A)2/2! + (2.3A)3/3! - (2.3A)4/4! + ..] (3-1)

where F is the relative fluorescence intensity, 4 is the quantum yield of the emitting species, 10 is the intensity of the exciting radiation, and A is the absorbance, the linearity of fluorescence intensity with the concentration of the emitting species can only be taken for granted at very low absorbance (< 0.02) at the excitation wavelength, in which case the higher power terms in equation (3-1) become negligible by comparison with the first








term. For example, when A > 0.02, the deviation from linearity is greater than 2%. At absorbances of 0.02 < A < 0.15, the second term in equation (3-1) results in 115 x A% deviation from linearity at any point in the fluorometric titration. At higher absorbances (A > 0.15), the third term in equation (3-1) generally also becomes large enough to have to be considered. Thus, determination of the absorbance of free 2-naphthol-8-sulfonate and albumin solution after successive addition of 2-naphthol-8-sulfonate was carried out as follows: 2.0 ml of albumin solution of appropriate concentration in a 1 cm UV quartz cell were titrated by successive additions of pl volumes of ax 1 0' M solution of 2naphthol-8-sulfonate. The absorbance was monitored at the wavelength of excitation after each addition of 2-naphthol-8-sulfonate stock solution. Same procedure was used to determine the absorbance of free 2-naphthol-8-sulfonate solution. After each addition of 2-naphthol-8-sulfonate solution, a Fisher Scientific Vortex Genie Mixer was used to mix the solution before reading. All UV and fluorometric determinations were performed at least three times, and the reported results are the mean of the data. Medium Effects on the Affinity of the Site in Albumin for the 2-Naphthol-8-Sulfonate

The effect of pH on the binding to albumin was measured as follows. The albumin and 2-naphthol-8-sulfonate ( at a constant ratio and concentration) were brought to different pH values with constant ionic strength in the range 5-9. Then the fluorescence intensities of the bound 2-naphthol-8-sulfonate and its conjugate base were measured.

With varying ionic strength and constant pH, a similar procedure was used for the measurement of the effect of ionic strength on the binding to albumin.








For these experiments, the same excitation wavelength was used as in the Job study, and the fluorescence intensity was monitored at the emission maximum of the 2naphthol-8-sulfonate-albumin complex. All UV and fluorometric determinations were performed at least three times, and the reported results are the mean of the data. Displacement of 2-Naphthol-8-Sulfonate from Albumin

The displacement of 2-naphthol-8-sulfonate from albumin binding site by typical site I and site II binding maker such as phenylbutazone, salicylic acid, and diazepam, ibuprofen, respectively, was carried out as follows: 2.0 ml of the albumin solution of concentration 1 x 10-' M dissolved in pH 5.0 phosphate buffer, placed in a 1 cm quartz cell, were injected with 2 pl of 1 x 10.2 M stock solution of 2-naphthol-8-sulfonate. The molar ratio of 2-naphthol-8-sulfonate to albumin was kept at 1:1 in order to keep the nonspecific binding of 2-naphthol-8-sulfonate to a minimum. The fluorescence intensity was monitored at the emission maximum of 2-naphthol-8-sulfonate-albumin complex. Then, above solution was titrated by successive additions of 5 Ptl of appropriate concentration stock solution of site binding maker dissolved in methanol. After each addition of site binding maker, the fluorescence intensity was monitored. The titration was performed until a point was reached at which the molar ratio of site binding maker to albumin was 5:1. Fluorescence values were corrected for dilution resulting from the addition of the site binding maker stock solution. The measured fluorescence value was expressed as a percentage of the initial fluorescence. In order to ensure that the site binding maker emitted no fluorescence at the wavelengths used for the displacement experiments, the fluorescence intensities of all site binding makers used in the displacement experiments








including typical site I binding makers such as warfarin and phenyprocoumon were monitored after addition to albumin solution. It was confirmed that all site binding makers except warfarin and phenyprocoumon emitted no or negligible fluorescence at the wavelengths used for the displacement experiments. For the displacement experiments, the same excitation wavelength was used as in the Job study. After each addition of 2naphthol-8-sulfonate solution, a Fisher Scientific Vortex Genie Mixer was used to mix the solution before reading. All fluorometric determinations were performed at least three times, and the reported results are the mean of the data. Fluorescence Quenching of Albumin Bound 2-Naphthol-8-Sulfonate

Fluorescence quenching experiment was carried out as follows:

(1) 2.0 ml of the albumin solution of concentration either 5 x 10.' M or I x 10- M dissolved in either pH 5.0 or pH 9.0 phosphate buffer, placed in a 1 cm quartz cell, were injected with 2 tl of 1 x 10.2 M stock solution of 2-naphthol-8-sulfonate. The fluorescence emission spectrum was recorded. Then, above solution was titrated by successive additions of 5 pl of appropriate concentration stock solution of pyridine dissolved in either pH 5.0 or pH 9.0 phosphate buffer. After each addition of quencher the fluorescence emission spectrum was recorded. Same procedures were used for recording the fluorescence emission spectra of free 2-naphthol-8-sulfonate in buffer solutions in the absence and presence of pyridine. In order to determine the quenching effect of pyridine on albumin, the fluorescence emission spectra of albumin solutions in the absence and presence of pyridine were recorded. Same procedures were also used for recording the fluorescence emission spectra of 2-naphthol-8-sulfonate in ethanol (neat, 75%(v/v),








50%(v/v)), and 1,4-dioxane (neat, 75%(v/v), 50%(v/v)). The purpose of this study was to evaluate the relationship between solvent polarity or hydrogen bonding strength and the fluorescence quenching properties of 2-naphthol-8-sulfonate and then to relate this to the behavior of the compound in the presence of albumin. In order to evaluate the hydrogen bonding in the ground state, the absorption spectrum of 2-naphthol-8-sulfonate in the presence of pyridine at pH 5.0 was also recorded.

(2) 2.0 ml of the albumin solution of concentration 5 x 10-' M dissolved in either pH 5.0 or pH 9.0 phosphate buffer, placed in a I cm quartz cell, were injected with 2 pl of 1 x 10.2 M stock solution of 2-naphthol-8-sulfonate. The fluorescence intensity was monitored at the emission maximum of 2-naphthol-8-sulfonate-albumin complex. Then, above solution was titrated by successive additions of 5 [l of appropriate concentration stock solution of pyridine. After each addition of quencher, the fluorescence intensity was monitored. Same procedures were used for monitoring the fluorescence intensity of free 2-naphthol-8-sulfonate in buffer solutions, and albumin solutions in the absence and presence of pyridine.

For the fluorescence quenching experiments, the same excitation wavelength was used as in the Job study. After each addition of 2-naphthol-8-sulfonate solution, a Fisher Scientific Vortex Genie Mixer was used to mix the solution before reading. Fluorescence values were corrected for dilution resulting from the addition of the quencher stock solution. All fluorometric determinations were performed at least three times, and the reported results are the mean of the data.








Circular Dichroism Experiments

As discussed in the chapter 2 to this dissertation, studies of the CD spectra of

ligand-protein complex can (a) probe changes in the conformation of the macromolecule, and (b) probe macromolecule interaction with small molecules, especially achiral ones whose induced CD is due solely to their interaction with the macromolecule. Since the wavelength of the extremum of a circular dichroism curve almost coincides with the position of the maximum in the absorption spectroscopy. Proper choices of protein concentration, pathlength, and solvent are essential for obtaining good CD spectra and for avoiding artifacts. Since the CD instrument measures very small differences in the transmitted light, the total absorbance of the sample in the desired spectral region is of utmost importance. A good signal-to-noise ratio is achieved when the absorbance is around 1.0. Therefore, the CD experiment was carried out as follows:

(1) absorption spectra of the samples and buffer at pH values of 5.0 and 9.0 in the desired spectral region were recorded prior to the CD experiment to determine the protein concentration and the ligand to protein ratio, and to select the optimal conditions for the CD measurement at different spectral region.

(2) the albumin, and albumin and 2-naphthol-8-sulfonate ( at a constant ratio and concentration, more specifically, albumin concentration is 4.0 x 10-' M, 2.0 x 10-' M, and

7.5 x 10' M in the wavelength region 200-250 nm, 250-300 nm, and 300-400/450 nm, respectively, and molar ratio of 2-naphthol-8-sulfonate to albumin is 1:1, and 3:1 for BSA and HSA in the region below 300 nm and 2.7:1 above 300nm, respectively) were brought to pH value either 5.0 or 9.0. Then, the baseline of the buffer and the CD spectra of






59


albumin and the 2-naphthol-8-sulfonate-albumin complex were recorded successively under identical instrumental settings in the same CD cell (i. e. 1 cm circular quartz cell).

Results are expressed as molar ellipticity calculated with reference to the albumin concentration, using a molecular weight of 66,500. Each CD spectrum reported is the average of three scans for short or long wavelength.















CHAPTER 4
RESULTS


Photophsicochemical Properties of 2-Naphthol-8-Sulfonate Absorption Spectra of 2-Naphthol-8-Sulfonate

Figure 4-1 shows the absorption spectra of 2-naphthol-8-sulfonate in the Hammett acidity - pH range -1.97 - 14.0. The respective maxima are 333 nm for monoanion (protonated form) and 360.4 nm for dianion (deprotonated form) in the wavelength range 300 nm - 400 nm, and two isosbestic points in the same wavelength range were clearly shown at 309 nm and 337 nm, which indicated a ground state acid - base reaction. Fluorescence Excitation Spectra of 2-Naphthol-8-Sulfonate

The fluorescence excitation spectra of 2-naphthol-8-sulfonate are depicted in

Figure 4-2. The excitation spectral properties are similar to those shown in the absorption spectra.

Fluorescence Emission Spectra of 2-Naphthol-8-Sulfonate

The recorded fluorescence emission spectra of 2-naphthol-8-sulfonate are shown in Figure 4-3. The respective maxima are 379 nm and 456 nm. An isoemissive point indicates an excited state acid - base reaction. At Ph of about 5, the hydroxyl group in the ground state is fully protonated, but not so in the lowest excited singlet state (see below). The excited molecules dissociate and majority of the emission is at the wavelength of the









excited 2-naphthol-8-sulfonate dianion (456 nm). The dissociation can be prevented if 2naphthol-8-sulfonate is dissolved in acid solution, pH < pK* , such as 2.88 M perchloric acid. Under such conditions we observe the emission of 2-naphthol-8-sulfonate monoanion with maximum at 379 nm. This reveals that 2-naphthol-8-sulfonate in aqueous solutions undergoes excited state proton transfer from the hydroxyl group to a water molecule, which is manifested by appearance of the fluorescence band of the protonated form accompanied by fluorescence emission band of the deprotonated form while only the protonated form is directly excited. Determination of Ground and Excited State Ionization Constants

Titration experiments in the Hammett acidity - pH range -1.97 - 14.0 for 2naphthol-8-sulfonate were performed, and the absorbances and the relative fluorescence intensities of 2-naphthol-8-sulfonate as a function of pH are shown in Figures 4-4 and 4-5.

The values of pK and pK* at ionic strength 0.1 were determined graphically to be 9.7 and 1.6, respectively. These titration results of 2-naphthol-8-sulfonate indicated that 2-naphthol-8-sulfonate monoanion is the predominant species present in the ground state and the dianion is the predominant species present in the excited state at pH < pK, say,

7.4.

Steady-State Kinetics of Excited State Proton Transfer Reaction of 2-Naphthol-8Sulfonate

The rate constants for excited state proton transfer reaction (1-1) were determined by two methods:








kl2 H
OOH* + H2 k21 OO-* + H30+
k(r + nr) hv, hv21 k'(r + nr)

OOH + H20 OO- + H30+

(1-1)

(a) from the spectrofluorimetric titration at low hydrogen ion concentration (3 < pH < 9),

(b) from the spectrofluorimetric titration in acidic pH range.

In method (a), the following equation


- - k12To (2-6)

0 0 const

was used. Equation (2-6) after rearrangement yields kl2To value from which k12 can be obtained using t0 value of 7.1 x 10-9 s obtained from Professor Wolfbeis's laboratory.

In method (b), the following modified equation (2-3)

0/0o 1 k21T'0 F[H3]
'40-= T+ o F[HzO']
- kl2 0 k,2t0

was used.

The modified equation (2-7) is used for calculation of the Bronsted kinetic activity factor F and is given as follows:

01_ 2AZ . Z _ pf
- logF = 2AjZjZj[V4
1 + V/where A = 0.51, I Z+ I = 1 and I Z- =2 for H30O and 2-naphthol-8-sulfonate dianion, respectively, under acidic conditions using perchloric acid only, p = [H30].








Using linear dependence of the ratio of relative fluorescence quantum yields

/ on F[ H3O'], from the ordinate intercept the value of 1/ kl2T0 and from the slope the value of k21T0' / kl2T0 were obtained. Then the excited state proton transfer rate constants kl2 and k21 were calculated using the lifetimes to and ro', 7.1 x 10-9 s and 13.3 x 10.' s, respectively, and the results are presented in Table 4-1.

The value of pK* of 2-naphthol-8-sulfonate was calculated from the ratio of excited state proton transfer rate constants kl2 and k21 and the result is presented along with the graphically determined pK* value in Table 4-1.


Table 4-1: Photophysicochemical properties of 2-naphthol-8-sulfonate.
k12 k2I pK* 2.7 x 10' � 0.38 s-' 9.2 x 1010 � 0.75 M' s' 1.53 Method (a) (calculated)
6.2 x 10'9 � 0.13 s-' 2.1 x 10" � 0.75 M-' s-' 1.6 Method (b) (graphically determined)


Solvent Studies

Binding of chromophores to protein affects their electronic transitions and hence generates difference absorption and fluorescence spectra. Since the chromophores are generally surrounded by a less polar environment upon binding to protein, similar effects are expected when the chromophores are transferred from a purely aqueous solvent to an organic solvent. Thus we first investigated solvent perturbation on absorption and fluorescence spectra of 2-naphthol-8-sulfonate.








For the sake of comparison, the absorption spectra of 2-naphthol-8-sulfonate

monoanion and its conjugate base dianion dissolved in aqueous and organic solvents are shown in Figures 4-6 and 4-7. The maximal wavelengths of 2-naphthol-8-sulfonate in various solvents are presented along with solvent dielectric constant values, where available, in Table 4-2. The peaks of 2-naphthol-8-sulfonate in organic solvents corresponded to red shifts of the original bands with respect to those in aqueous solution.

The fluorescence emission spectra of 2-naphthol-8-sulfonate monoanion and its

conjugate base dianion in aqueous and organic solvents are shown in Figures 4-8 and 4-9, respectively. The maximal wavelengths of 2-naphthol-8-sulfonate in various solvents are presented in Table 4-2. The fluorescence peaks of 2-naphthol-8-sulfonate, especially monoanion, in organic solvents shifted to shorter wavelengths compared with that in aqueous solution.

As shown in Figure 4-10, the fluorescence emission spectra of 2-naphthol-8sulfonate in ethanol-water mixtures strongly depend on the concentration of water. A similar set of the emission spectra was also found for 2-naphthol-8-sulfonate in 1,4dioxane-water mixtures as depicted in Figure 4-11. In pure organic solvents, only a single emission band which is attributed to the protonated form (monoanion) of the excited 2naphthol-8-sulfonate is observed at - 368 nm. However, upon addition of water at constant solute concentration, the fluorescence intensity of the protonated band decreases gradually and a new band appears at - 455 nm. The new band is attributed to the conjugate base of the excited 2-naphthol-8-sulfonate dianion. This behavior is accounted for by deprotonation of the excited 2-naphthol-8-sulfonate since the hydroxyl group of 2naphthol-8-sulfonate behaves like a stronger acid in the excited state than in the ground








state. The shift of the isoemissive point to the longer wavelength region where water

concentration is 50% and 25% by volume or higher in the 1,4-dioxane-water and ethanolTable 4-2: Solvent studies - absorption and fluorescence data.
Wavelength of Wavelength of Absorption Maxima Fluorescence Solvent Dielectric (nm) Maxima Constant (nm) Mono- Dianion Mono- Dianion anion anion
Water 78.38 333.0 N* 379 456
Perchloric Acid NA** 332.2 N 379 N
(2.88 M)
Sodium Hydroxide NA N 360.4 N 456 (1 N)
Methanol 32.66 336.0 361.6 369.5 455
Ethanol 24.3 336.0 364.0 369.5 455 75% (V/V) Ethanol 32.8 < c < 43.4 - 371 Broad 449
50% (V/V) Ethanol 43.4< s < 55.0 - - 372 456 1,4-Dioxane 2.209 336.6 360.8 367 445 75% (V/V) 12.1< c < 27.5 - - 370 N
1,4-Dioxane
50% (V/V) 27.5< s < 44.4 - - 371 Broad 1,4-Dioxane - 449 Dimethyl Sulfoxide 44.6 338.2 - 366.5 473 Acetonitrile 37.5 334.0 368.0 < 360 Broad ~ 463
Glycerol 42.5 336.8 362.0 373 Broad S460
Tertiary Butanol 10.9 336.0 - 370 Bovine Serum NA -337 -366 376 445
Albumin
Human Serum NA -334 N 376 453
Albumin
* The letter "N" indicates that nothing is observed.
** The letters "NA" means that the data is not available.
***The symbol "--" means that either the experiment is not performed or the result cannot be obtained.








water mixtures, respectively, might be caused by the change in solvent structure in the organic-water mixtures. These results also reveal that no dissociation will take place unless water molecules are in the immediate vicinity to act as proton acceptors.

The observations of such an isoemissive point in both aqueous solution and

organic-water mixtures demonstrates the presence of only protonated (monoanion) and deprotonated (dianion) forms of 2-naphthol-8-sulfonate in the excited state.

Binding Studies


Qualitative Spectrophotometric Examination of 2-Naphthol-8-Sulfonate-Albumin
Complex Interaction

Upon ligation to albumin, the spectroscopic properties of 2-naphthol-8-sulfonate were changed, which reflect the perturbation caused by changing the environment of 2naphthol-8-sulfonate from aqueous solution to protein.

The absorption and fluorescence emission spectra for 2-naphthol-8-sulfonatealbumin complex in the pH range 3-10 at various ionic strengths were recorded and summarized in Figures 4-12 - 4-37. The absorption spectra of 2-naphthol-8-sulfonate bound to albumin were obtained as a difference spectra (2-naphthol-8-sulfonate-albumin versus albumin). The respective maxima for the albumin bound 2-naphthol-8-sulfonate complex are presented in Table 4-2. As can be seen, the absorption spectra of BSA-2naphthol-8-sulfonate complex show maxima at 336 nm and 363.5 nm, and the maximal absorption bands for HSA-2-naphthol-8-sulfonate complex show little change compared to 2-naphthol-8-sulfonate alone in buffer. A peak was observed at 334 nm in the difference absorption spectra of 2-naphthol-8-sulfonate produced by HSA at molar ratio








of 2-naphthol-8-sulfonate to HSA 1:5. This maximum was positive and corresponded to a small red shift of the original absorption peak.

The fluorescence of BSA- and HSA- bound 2-naphthol-8-sulfonate still consists of two bands, typical for the dianion (445 nm and 453 nm, for BSA- and HSA- bound 2naphthol-8-sulfonate, respectively,) and monoanion (376 nm), but shifted to shorted wavelengths. BSA-2-naphthol-8-sulfonate complex are much more intensely fluorescent but varying with pH while HSA-2-naphthol-8-sulfonate complex shows less profound increase in intensity but constant with pH.

The results show that the absorption bands and fluorescence bands of 2-naphthol8-sulfonate are perturbed to produce the difference spectral maxima upon its binding to albumins. Similarities between the albumin- and organic solvent-generated difference spectra indicate that 2-naphthol-8-sulfonate binds to the binding site with less polar environment in the albumin molecule. In this system, the perturbant is the albumin molecule, or more precisely the less polar environment of the binding site.

To ensure that the intensities at monitored wavelengths are independent of the 2naphthol-8-sulfonate / albumin ratio, the fluorescence emission spectra of three sets of solutions containing different molar ratio of 2-naphthol-8-sulfonate to albumin, i.e., 1:1, 1:5 and 1:10, were recorded and are depicted in Figures 4-38 and 4-39 for BSA- and HSA-bound 2-naphthol-8-sulfonate, respectively. The fluorescence spectra for the latter two sets consisting different molar ratios of 2-naphthol-8-sulfonate to albumin were found to be essentially identical and did not change with increased concentration of albumin.








Two different lots of albumin from each species were examined by recording the fluorescence emission spectrum. And it was found that the emission peaks of albumin bound 2-naphthol-8-sulfonate are identical for the lots derived from the same species. These differences in the spectra of albumin bound 2-naphthol-8-sulfonate are due to species differences in the albumins tested in this study. Fluorometric Determination of 2-Naphthol-8-Sulfonate -Albumin Complex Stoichiometry

The results of the Job study are presented in Table 4-3. Two of the Job plots for albumin bound 2-naphthol-8-sulfonate monoanion and dianion are shown in Figures 4-40 and 4-41. The rounded appearance of the maximum of the curve indicates that the complex formed is somewhat dissociated. Therefore the extrapolations of these Job curves can only be considered to approximate the complex stoichiometry. Extrapolation along the more linear portions of the curves monitored at wavelengths 445 nm and 453 nm gave a value of x m 0.5 which corresponds to N = 1 for BSA- and HSA-bound 2naphthol-8-sulfonate dianion, respectively. However, extrapolation along the more linear portions of the curves monitored at wavelength 376 nm which corresponds to albumin bound 2-naphthol-8-sulfonate monoanion gave different values of x for BSA- and HSA2-naphthol-8-sulfonate under various pH and ionic strength conditions. Since all x values for HSA- 2-naphthol-8-sulfonate are around 0.75, which corresponds to a N value of 3, it is reasonably certain that each HSA molecule binds roughly three 2-naphthol-8-sulfonate monoanion molecules in the pH range 5-9 under the experimental conditions. For BSA-2naphthol-8-sulfonate, with increasing ionic strength from 0.001 to 0.3 in the pH range 5-9, the x values decrease from 0.67 to 0.5 corresponding to a N value from 2 to 1. It is








clear that the number of binding sites on BSA is smaller than that on HSA for 2-naphthol8-sulfonate.

2-Naphthol-8-Sulfonate-Albumin Titration Experiments-Determination of Binding
Constants

The fluorometric titrations of albumin with 2-naphthol-8-sulfonate are shown in

Figures 4-42 and 4-43. For compounds like 2-naphthol-8-sulfonate, which show

increased fluorescence intensity upon binding, the fraction of ligand bound, a, is usually

calculated by using the following equation:

ca = (Fp - Ff)/(Fb - Ff) (2-27)


Table 4-3: Results of Job Study - number of binding sites.
Ionic pH 5.0 pH 7.4 pH 9.0 Strength
Monoanion Dianion Monoanion Dianion Monoanion Dianion
BSA
0.001 x = 0.66 x = 0.5 x = 0.69 x = 0.5 x = 0.69 x = 0.5
N = 1.9 N = 1 N= 2.2 N= 1 N=2.2 N = 1
0.01 x = 0.62 x = 0.5 x = 0.6 x = 0.5 x = 0.58 x = 0.5
N = 1.6 N= 1 N= 1.5 N = 1 N= 1.4 N= 1
0.1 x = 0.56 x = 0.5 x = 0.59 x = 0.5 x = 0.57 x = 0.5
N =1.3 N= 1 N= 1.4 N = 1 N= 1.3 N= 1
0.3 x = 0.51 x = 0.5 x = 0.56 x = 0.5 x = 0.57 x = 0.5
N=1 N = 1 N= 1.3 N= 1 N= 1.3 N= 1
HSA
0.001 x = 0.74 x = 0.5 x = 0.78 x = 0.5 x = 0.78 x = 0.5
N=2.8 N= 1 N=3.5 N= 1 N=3.5 N= 1
0.1 x = 0.74 x = 0.5 x = 0.77 x = 0.5 x = 0.75 x = 0.5
N=2.8 N= I N=3.3 N= I N=3 N= 1


In order for the above equation to yield good values of a, the fluorescence

intensity of the bound ligand must be linear function of its concentration. As mentioned

in chapter 3, fluorescence intensity is related to absorbance in a power series:








F = )I0[2.3A - (2.3A)2/2! + (2.3A)3/3! - (2.3A) /4! + ***] (3-1)

Only when the absorbance at the exciting wavelength is very low (< 0.02) does fluorescence intensity show a direct linear relationship to ligand concentration. When the absorbance lies between 0.02 and 0.15, the second term in the equation (3-1) becomes significant, and each point in the titration curve must be corrected for this value. At higher absorbances, the third term in the equation (3-1) may have to be considered.

The molar absorptivity for the albumin-2-naphthol-8-sulfonate complex at the

excitation wavelength of 337 nm was determined to be 2.2 x 10' liter mole' (2-naphthol8-sulfonate) cm-'. Therefore, a 2-naphthol-8-sulfonate concentration greater than 9 x 10' M would result in an absorbance value greater than 0.02, and a deviation from linearity would result unless the second term in equation (3-1) is considered. Because the concentration of 2-naphthol-8-sulfonate in the albumin titration experiments exceeded this 9 x 10' M limit, it was inappropriate to use equation (2-27) to calculate the concentration of bound 2-naphthol-8-sulfonate. Consequently, the observed fluorescence intensity at each point during the titrations with high protein concentration was corrected for the second term in equation (4-3). Curve (0) in Figures 4-42 and 4-43 was plotted after applying the corrected factor, a straight line plot was obtained. The fact that the corrected values yielded a straight line verified that the deviation from linearity was indeed due to the absorbance effect.

The curve (A) in Figure 4-42 approached a plateau indicating the saturation of BSA binding sites by 2-naphthol-8-sulfonate.








In the HSA titration experiment, the high HSA concentration titrations were initially performed using HSA concentrations of 1 x 10' M and I x 10-5 M with a 2naphthol-8-sulfonate stock solution concentration of 1 x 10-2 M. The results from these experiments showed the maximum intensity was obtained in the 1 x 10-' M HSA titration. Lower intensity was obtained in the higher HSA concentration experiment (1 x 104 M) due to nonlinearity of HSA fluorescence at this concentration level. Titrations were then performed at HSA concentrations of 1 x 10-' M and 2.5 x 10-6 M. In these albumin titration experiments, the maximum volume of ligand solution which could be added to the 2.0 ml of albumin without encountering an error due to sample volume change was placed at 30 pil. Thus, in the 2.5 x 10' M HSA titration, the titration curve had not completely reached a plateau after the addition of 30 pl of the 2-naphthol-8-sulfonate solution.

A common method of treating binding data makes use of the Scatchard equation

(29), which was discussed in the background to this dissertation. A basic assumption of the Scatchard treatment is that the binding sites are independent, noninteracting and have the same binding affinity for the ligand. Such assumptions are not necessary when N = 1, and in that case, a Scatchard plot of r /[L] versus r should yield a straight line. However, as the number of binding sites increases, both the likelihood of the existence of different types of binding sites and of electrostatic interaction between the sites, even if they are similar, increases and the Scatchard-type plots often deviate from linearity. Participation of multiple classes of binding sites is easily apparent from the Scatchard plot since there is a marked concave curvature of the plot. In such case, extrapolation along the linear portions of the curve should yield a set of slopes and intercepts, which correspond to








different classes of binding sites. Scatchard plots of r/[L] versus r for the BSA- and HSA2-naphthol-8-sulfonate titration with the protein concentration of 2.5 x 10' M at pH 7.4 and p 0.1 are shown in Figures 4-44 and 4-45, respectively. Since the slope of a Scatchard plot corresponds to the negative k value and y-intercept corresponds to Nk, the value of N may be determined simply by dividing the y-intercept value by slope of the line (times -1). Scatchard analysis of the fluorescence data showed a non-linear curve, suggesting the presence of at least two classes of sites for the binding of 2-naphthol-8sulfonate to albumins. The best fitting values for the binding parameters obtained by the fluorescence method are shown in Table 4-4.


Table 4-4: Results of Scatchard plot - binding constants and number of binding sites. Protein Concentration (M) pH 7.4 BSA
2.5 x 10-6 k, = 3.9 x 105, N, = 1.6 (r2 = 0.9947) k2 = 2.5 x 10', N2 = 2.2 (r2 = 0.9972)
HSA
2.5 x 106 k, = 2.3 x 106, N, = 1.4 (r2 = 0.5877) k2 = 1.2 x 10', N2 = 6.2 (r2 = 0.9665)



Medium Effects on the Affinity of the Site in Albumin for the 2-Naphthol-8-Sulfonate

The fraction of the bound ligand was measured by the fluorescence intensity of excited 2-naphthol-8-sulfonate monoanion and dianion. The fluorescence of excited 2naphthol-8-sulfonate monoanion increases upon binding, therefore any dissociation of the ligand from the site will lower the emission at 376 nm and increase the emission at 445 nm and 453 nm for BSA and HSA bound 2-naphthol-8-sulfonate, respectively. Thus dissociation will lower the ratio of emission I376/I445 and 1376/1453 for BSA and HSA bound








2-naphthol-8-sulfonate, respectively. The effects of pH and ionic strength on the fluorescence emission ratio of 2-naphthol-8-sulfonate in the presence of albumin were depicted in Figures 4-46 and 4-47. As seen in Figures 4-46 and 4-47, the fluorescence emission ratio of 2-naphthol-8-sulfonate in the presence of BSA but not HSA decreases with rising pH in the pH range 5-9 at a constant ionic strength, and the fluorescence emission ratio of 2-naphthol-8-sulfonate in the presence of BSA but not HSA increases with rising ionic strength at a fixed pH in the pH range 5-9. Displacement of 2-Naphthol-8-Sulfonate from Albumin

To determine the binding sites for 2-naphthol-8-sulfonate on HSA and BSA,

displacement studies were carried out using the typical site I binding markers for human albumin, namely, phenylbutazone and salicylic acid, and site II binding markers, diazepam and ibuprofen. The displacement results are shown in Figures 4-48 and 4-49 for HSA- and BSA-bound 2-naphthol-8-sulfonate, respectively. A decrease in the fluorescence intensity of the albumin-2-naphthol-8-sulfonate complex can be interpreted as a displacement of 2-naphthol-8-sulfonate from its binding site by the added specific site binding marker, probably through a competitive mechanism. As can be seen in Figure 4-48, the fluorescence intensity of 2-naphthol-8-sulfonate bound to HSA was remarkably decreased by one of the site II binding markers, namely, ibuprofen, but not by diazepam, a specific site II binding marker and site I binding markers. On the other hand, it is clear from the Figure 4-49 that the fluorescence intensity of 2-naphthol-8-sulfonate bound to BSA was more extensively decreased by site II binding markers but not by site I binding markers.








Fluorescence Quenching of Albumin-Bound 2-Naphthol-8-Sulfonate

Many parameters can be obtained by fluorescence spectroscopic methods to

provide insights into the environment, structure, and dynamics of a fluorescent probe that is either covalently bound or liganded to a biological molecule. One important, commonly used method is the addition of a quenching agent to reduce the fluorescence of the probe. By comparing the quenching efficiency, the environment of the probe, and thus a specific region of the biomolecule, can be characterized in terms of solvent accessibility. Therefore, quenching studies were carried out using pyridine.

The absorption spectra of pyridine and pyridine-containing solution at pH 5.0 are shown in Figure 4-50. Compared with that of 2-naphthol-8-sulfonate in pyridine-free solution, the changes in the absorption spectra upon addition of pyridine are very minor.

Unlike the absorption spectrum, the emission bands of unbound 2-naphthol-8sulfonate observed in pH 5.0 and 9.0 buffer solutions decrease in intensity upon pyridine titration as depicted in Figures 4-51 and 4-52, the position and shape of the band was not changed. These results are consistent with those observed by Mataga (94).

The effects of pyridine on the fluorescence emission of albumin-2-naphthol-8sulfonate complex were shown in Figures 4-53 - 4-60. At low molar ratio of albumin to 2-naphthol-8-sulfonate, upon addition of pyridine at both pH values, the intensities for albumin bound 2-naphthol-8-sulfonate monoanion and dianion decreased but to a lesser degree as compared to the unbound 2-naphthol-8-sulfonate. However, at high molar ratio of albumin to 2-naphthol-8-sulfonate, the emission bands of HSA-bound 2-naphthol-8sulfonate shows little change at pH 5, while at pH 9, the intensity of HSA-bound 2-








naphthol-8-sulfonate dianion decreases with increasing concentration of pyridine. In all cases, the reduction in intensity of the HSA-bound 2-naphthol-8-sulfonate by pyridine is considerably less than that found for free 2-naphthol-8-sulfonate. For BSA-bound 2naphthol-8-sulfonate, the intensity of monoanion decreases with increasing concentration of pyridine at both pH values concomitant with increasing in intensity of dianion at pH 5. In no case did the intensity increase for the long-wavelength emission match the intensity lost from the monoanionic emission.

It is reported by several groups (95, 108) that the effects of alkylamine on the fluorescence emission of hydroxyaromatics such as naphthol in organic solvent were dramatic. Either an increase in emission intensity or a new peak at the long-wavelength which corresponds to the naphtholate anion accompanying with the quenching of the neutral form was observed upon addition of alkylamine. In order to investigate the abnormal quenching behavior of BSA-bound 2-naphthol-8-sulfonate, the quenching experiments of unbound 2-naphthol-8-sulfonate in organic solvents and organic-aqueous mixtures were conducted and the results are depicted in Figures 4-61 - 4-66. It is clear from these figures that in all cases the intensities of 2-naphthol-8-sulfonate monoanion and dianion decrease with increasing concentration of pyridine. Therefore, the increase in intensity of BSA-bound 2-naphthol-8-sulfonate dianion might be caused by the alteration of structure or affinity of BSA to 2-naphthol-8-sulfonate upon addition of pyridine.

The emission intensities of albumin bound and unbound 2-naphthol-8-sulfonate in pyridine-containing solutions relative to the intensities in pyridine-free solutions had been determined at a number of pyridine concentrations. The emission intensities of unliganded albumins in the absence and presence of pyridine had been also determined.








Since the emission intensities of unliganded albumins at the monitored wavelengths are very low, the quenching effect of pyridine on albumins can only be regarded as an estimate. All results are shown in Figures 4-67 and 4-68. As depicted in these figures, the fluorescence intensities of unliganded albumins are slightly affected by the addition of pyridine, especially for unliganded BSA.

Circular Dichroism Experiments

As mentioned in chapter 2, the CD spectrum is very useful to study the biological system because it can provide information about structural transitions and conformational changes of macromolecules induced by binding small molecules. Thus, the binding of 2naphthol-8-sulfonate to albumin was quantitatively monitored by circular dichroism spectra at pH 5 and 9 in the wavelength range 200 nm - 450 nm.

The CD spectra of HSA and BSA in the far-UV region are typical for a-helical proteins as shown in Figures 4-69 - 4-74. The CD spectra of the albumin-2-naphthol-8sulfonate complex were found to be practically identical to the spectra of unliganded albumin. Similarly, CD bands in the near-UV region of albumin-2-naphthol-8-sulfonate complex did not differ from those of unliganded albumin. These indicated that the CD spectra of albumin below 300 nm did not appear to be affected by the binding of 2naphthol-8-sulfonate. A notable positive Cotton effect at 332 nm in the CD spectra for the HSA-2-naphthol-8-sulfonate complex was observed at both pH values as shown in Figure 4-74, which is attributed to the induced Cotton effect of 2-naphthol-8-sulfonate upon binding to HSA. No significant changes in the same wavelength region of the CD spectra of BSA-2-naphthol-8-sulfonate were observed.


















5.000







S2.475
C


210.0 355.0 500.0 wavelength (nm)
Figure 4-1. Absorption spectra of 2-naphthol-8-sulfonate at different Ho0 or pH values: 1, -1.97 - 7.7; 2, 8.6; 3, 9.6 - 10.28; 4, 11.8; 5, 12.9.


















































EX 269 20 38 320 34 36 386
EM 429 wavelength (nm)



Figure 4-2. Fluorescence excitation spectra of 2-naphthol-8-sulfonate in different pH values: 1, 4.08 - 6.92; 2, 8.02; 3, 8.99; 4, 9.54; 5, 10.07.









































- 11o n
3 8




6 7
7
6
go
� = I l I I l '






8614 10 11 3

92 5 1 2 1 1z

EM 360 30 40 42 440 460 4gg " 2 S4 566 S" 600
EN 3 wavelength (mn)





Figure 4-3. Fluorescence emission spectra of 2-naphthol-8-sulfonate in different Ho or pH values: 1, -1.97; 2, -1.25; 3, -0.63;4, 0.02; 5, 0.63; 6, 1.04; 7, 1.31; 8, 2.06;
9, 3.04; 10, 4.03 - 9.02; 11, 9.99; 12, 11.8 - 14.0.



































0.1


0.0 --'
-4 -2 0 2 4 6 8 10 12 14 16 H0 or pH




Figure 4-4. The variation of absorbances with H0 or pH of 2-naphthol-8-sulfonate monoanion (0) and its conjugate base dianion (0).









































-2 0 2 4 6 8 10 12 14 H, or pH






Figure 4-5. The variation of relative fluorescence intensities with Ho or pH of 2-naphthol8-sulfonate monoanion (*) and its conjugate base dianion (0).






















0.4

0.3I

0,,
I'









0.0
0.1 \



0.0
280 300 320 340 360 380
wavelength (nm)






Figure 4-6. Absorption spectra of 2-naphthol-8-sulfonate monoanion in different solvents: - 2.88 M perchloric acid; - - methanol; - - - ethanol; - -- 1,4-dioxane;
- - - DMSO; ........ t-butanol; - - acetonitrile.








83























3.800








3. 000,










A 2.00
b
a









1.00
� i








%\






0.0 e- - - g r250.0 300.0 350.0 400.0 450.0 Wavelength (na.)








Figure 4-7. Absorption spectra of 2-naphthol-8-sulfonate dianion in different solvents:

IN sodium hydroxide; - - - ethanol; -. methanol; -. - 1,4-dioxane;

- . glycerol; ........ acetonitrile.







84




















100 SO


2




6- 3

4





6



7










EN 360 386 400 420 446 460 468 588 Z2 546 568 598 68
EN 337 Wavelength (Man)






Figure 4-8. Fluorescence emission spectra of 2-naphthol-8-sulfonate monoanion in different solvents: 1, 2.88M perchloric acid; 2, DMSO; 3, 1,4-dioxane; 4, methanol; 5, ethanol; 6, acetonitrile; 2, glycerol.




Full Text

PAGE 1

A NOVEL PHOTODISSOCIATIVE PROBE OF WATER IN SERUM ALBUMIN BINDING SITES By QIAO-QING DI A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1998

PAGE 2

Copyright 1998 by Qiao-qing Di

PAGE 3

I would like to dedicate this dissertation to my family, all of them contributed in their own unique way to make this possible.

PAGE 4

ACKNOWLEDGMENTS I would like to thank Dr. Stephen. G. Schulman for his guidance, support and encouragement in my academic and professional development. I would also like to acknowledge my committee members Dr. John H. Perrin, Dr. Kenneth B. Sloan, Dr. James D. Winefordner, and Dr. Jeffrey Hughes for their support and encouragement throughout my stay at the University of Florida. Finally, but not least, I would like to acknowledge Dr. Otto S. Wolfbeis and his group members. Dr. James F. Preston III and his group member, and Dr. Lei Wei for their part in helping me to finish this work. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS iv LIST OF TABLES viii LIST OF FIGURES ix CHAPTERS 1 INTRODUCTION 1 2 BACKGROUND H The Structure of Water 11 Prototropic Conductivity 12 Prototropic Reactivity in Electronically Excited States 18 Properties of Acids and Bases in the Lowest Excited Singlet State 18 The Effect of State of Protonation on Electronic Spectra 19 The Effect of Solvents on Electronic Spectra 20 Steady-State Kinetics of Excited State Proton Transfer Reaction 24 Protein Binding 29 Binding Studies Experimental Methods 30 Binding Equation Derivation 31 Binding Studies Treatment of Data 34 Determination of Complex Stoichiometry The Method of Continuous Variation 35 Identification of the Binding Sites 37 Characterization of the Binding Sites on Albumins Fluorescence Quenching 38 Characterization of the Binding Sites on Albumins Circular Dichroism 41 3 EXPERIMENTAL 47 Materials 47 Instrumental 49 Methods 49 Ground and Excited State Ionization Constants Determinations (Titration Procedure) 49 Solvent Studies 50 V

PAGE 6

Qualitative Spectrophotometric Examination of 2-Naphthol-8-SulfonateAlbumin Complex Interaction 51 Fluorometric Determination of Compound-Albumin Complex StoichiometryJob's Method 52 2-Naphthol-8-Sulfonate-Albumin Titration Experiments-Determination of Binding Constants and Number of Binding Sites 53 Medium Effects on the Affinity of the Site in Albumin for the 2-Naphthol-8Sulfbnate 54 Displacement of 2-Naphthol-8-Sulfonate from Albumin 55 Fluorescence Quenching of Albumin Bound 2-Naphthol-8-Sulfonate 56 Circular Dichroism Experiments 58 4 RESULTS 60 Photophsicochemical Properties of 2-Naphthol-8-Sulfonate 60 Absorption Spectra of 2-Naphthol-8-Sulfonate 60 Fluorescence Excitation Spectra of 2-Naphthol-8-Sulfonate 60 Fluorescence Emission Spectra of 2-Naphthol-8-Sulfonate 60 Determination of Ground and Excited State Ionization Constants 61 Steady-State Kinetics of Excited State Proton Transfer Reaction of 2Naphthol-8-Sulfonate 61 Solvent Studies 63 Binding Studies 66 Qualitative Spectrophotometric Examination of 2-Naphthol-8-SulfonateAlbumin Complex Interaction 66 Fluorometric Determination of 2-Naphthol-8-Sulfonate -Albumin Complex Stoichiometry 68 2-Naphthol-8-Sulfonate-Albumin Titration Experiments-Determination of Binding Constants 69 Medium Effects on the Affinity of the Site in Albumin for the 2-Naphthol-8Sulfbnate 72 Displacement of 2-Naphthol-8-Sulfonate from Albumin 73 Fluorescence Quenching of Albumin-Bound 2-Naphthol-8-Sulfonate 74 Circular Dichroism Experiments 76 5 DISCUSSION 151 The Effect ofpH 151 Steady-State Kinetics of Excited State Proton Transfer Reaction of 2Naphthol-8-Sulfonate 152 Solvent Studies 153 The Effects of Solvent on the Spectral Properties of 2-Naphthol-8-Sulfonate 153 The Effect of Solvent on Fluorescence Quantum Yields 159 The Effects of Solvent on the Excited State Proton Transfer 159 vi

PAGE 7

Binding Studies Identification and Characterization of the Binding Sites on Albumin 162 Fluorescence Quenching Experiments 168 Circular Dichroism Experiment 172 Model for Binding of 2-Naphthol-8-Sulfonate to Albumin 175 The Effect of Binding on the Spectral Properties of 2-Naphthol-8-Sulfonate 177 The Effect of Binding on Excited State Proton Transfer 180 The Effects of pH and Ionic Strength on the Binding of 2-Naphthol-8Sulfonate to Albumin 186 6 CONCLUSIONS 189 REFERENCES 191 BIOGRAPHICAL SKETCH 198 vii

PAGE 8

T ICF rM7 XAT3T CC Lib I Ur lArSLxio Table page 4-1. 63 4-2. 65 4-3. 69 4-4. Results of Scatchard plot binding constants and number of binding sites .. 7? ) "J viii

PAGE 9

LIST OF FIGURES Figure page 2-1 . A molecular mechanism for prototropic mobility 16 4-1 . Absorption spectra of 2-naphthol-8-sulfonate at different Hq or pH values 77 4-2. Fluorescence excitation spectra of 2-naphthol-8-sulfonate in different pH values ^8 4-3. Fluorescence emission spectra of 2-naphthol-8-sulfonate in different Hq or pH values 4-4. The variation of absorbances with Hq or pH of 2-naphthol-8-sulfonate monoanion and its conjugate base dianion 80 4-5. The variation of relative fluorescence intensities with Hq or pH of 2-naphthol8-sulfonate monoanion and its conjugate base dianion 81 4-6. Absorption spectra of 2-naphthol-8-sulfonate monoanion in different solvents ...82 4-7. Absorption spectra of 2-naphthol-8-sulfonate dianion in different solvents 83 4-8. Fluorescence emission spectra of 2-naphthol-8-sulfonate monoanion in different solvents 84 4-9. Fluorescence emission spectra of 2-naphthol-8-sulfonate dianion in different solvents 85 4-10. Relative fluorescence intensities of 2-naphthol-8-sulfonate as a function of ethanol concentration 86 4-11. Relative fluorescence intensities of 2-naphthol-8-sulfonate as a ftmction of 1,4-dioxane concentration 87 4-12. Absorption spectra of free, BSA-bound and HSA-bound 2-naphthol-8sulfonate at pH 3.04 88 ix

PAGE 10

4-13. Absorption spectra of free, BSA-bound and HSA-bound 2-naphthol-8sulfonate at pH 4.03 and |a 0.1 89 4-14. Absorption spectra of free, BSA-bound and HSA-bound 2-naphthol-8sulfonate at pH 5.01 and ^ 0.1 90 4-15. Absorption spectra of free, BSA-bound and HSA-bound 2-naphthol-8sulfonate at pH 5.93 and ^ 0.1 91 4-16. Absorption spectra of free, BSA-bound and HSA-bound 2-naphthol-8sulfonate at pH 6.93 and ^ 0.1 92 4-17. Absorption spectra of free, BSA-bound and HSA-bound 2-naphthol-8sulfonate at pH 8.09 and ^ 0.1 93 4-18. Absorption spectra of free, BSA-bound and HSA-bound 2-naphthol-8sulfonate at pH 9.50 and \i 0.1 94 4-19. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate at pH 3.04 95 4-20. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate at pH 4.04 and |x 0.001 96 4-2 1 . Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate at pH 5.07 and \i 0.001 97 4-22. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-suIfonate at pH 6.07 and ^ 0.001 98 4-23. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate as well as BSA and HSA at pH 7.05 and \i 0.001 99 4-24. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate at pH 8.12 and ^ 0.001 100 4-25. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate at pH 8.94 and \i 0.001 101 4-26. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate at pH 10.01 and [i 0.001 102 4-27. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate at pH 4.03 and ^ 0.1 103 X

PAGE 11

4-28. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate at pH 4.98 and n 0.1 104 4-29. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate at pH 5.93 and 0.1 105 4-30. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate as well as BSA and HSA at pH 6.93 and ^ 0.1 106 4-3 1 . Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate at pH 8.09 and \i 0.1 107 4-32. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate at pH 9.02 and ^ 0.1 108 4-33. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate at pH 9.99 and \i 0.1 109 4-34. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate at pH 4.03 and ^ 1.0 110 4-35. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate at pH 4.95 and |j 1 .0 1 1 1 4-36. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate as well as BSA and HSA at pH 6.93 and \i 1.0 112 4-37. Fluorescence emission spectra of free, BSA-bound and HSA-bound 2-naphthol-8-sulfonate at pH 8.92 and ^ 1 .0 113 4-38. Fluorescence emission spectra of BSA-2-naphthol-8-sulfonate complex at different molar ratios of 2-naphthol-8-sulfonate to BSA 1 14 4-39. Fluorescence emission spectra of HSA-2-naphthol-8-sulfonate complex at different molar ratios of 2-naphthol-8-sulfonate to HSA 115 4-40. Job's plot of BSA-2-naphthol-8-sulfonate monoanion at pH 7.4 and |a 0. 1 116 4-41. Job's plot of BSA 2-naphthol-8-sulfonate dianion at pH 7.4 and ^ 0.1 117 4-42. Plots of relative fluorescence intensity as a function of total 2-naphthol-8sulfonate concentration for the BSA-2-naphthol-8-sulfonate titrations with a constant amount of the protein at pH 7.4 and |a 0. 1 118 xi

PAGE 12

4-43. Plots of relative fluorescence intensity as a function of total 2-naphthol-8sulfonate concentration for the HSA-2-naphthol-8-sulfonate titrations with a constant amount of the protein at pH 7.4 and ^ 0. 1 119 4-44. Scatchard plot of r/[L] versus r for the bovine serum albumin-2-naphthol8-sulfbnate 120 4-45. Scatchard plot of r/[L] versus r for the human serum albumin-2-naphthol8-sulfonate 121 4-46. The fluorescence emission ratio of BSA-2-naphthol-8-sulfonate complex as a function of pH at different ionic strengths 122 4-47. The fluorescence emission ratio of HSA-2-naphthol-8-sulfonate complex as a function of pH at different ionic strengths 123 4-48. Marker-induced changes in fluorescence of 2-naphthol-8-sulfonate bound to HSA 124 4-49. Marker-induced changes in fluorescence of 2-naphthol-8-sulfonate bound to BSA 125 4-50. Absorption spectra of pyridine, 2-naphthol-8-sulfonate in the presence of pyridine at pH 5.01 and )i 0.1 126 4-5 1 . The fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of pyridine at pH 5.0 and |J 0.1 127 4-52. The fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of pyridine at pH 9.0 and ^ 0.1 128 4-53. The fluorescence emission spectra of HSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 5.0 and |^ 0.1 with constant amount of 2-naphthol8-sulfonate (10 ^iM) and HSA (7 ^M) 129 4-54. The fluorescence emission spectra of HSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 9.0 and \i 0.1 with constant amount of 2-naphthol8-sulfonate (10 ^iM) and HSA (7 ^M) 130 4-55. The fluorescence emission spectra of HSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 5.0 and \\. 0.1 with constant amount of 2-naphthol8-sulfonate (10 ^M) and HSA (50 ^M) 131 xii

PAGE 13

4-56. The fluorescence emission spectra of HSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 9.0 and |i 0.1 with constant amount of 2-naphthol8-sulfonate (10 ^iM) and HSA (50 ^iM) 132 4-57. The fluorescence emission spectra of BSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 5.0 and |i 0.1 with constant amount of 2-naphthol8-sulfonate (10 ^M) and BSA (7 ^iM) 133 4-58. The fluorescence emission spectra of BSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 9.0 and \i 0.1 with constant amount of 2-naphthol8-sulfonate (10 ^iM) and BSA (7 ^M) 134 4-59. The fluorescence emission spectra of BSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 5.0 and [i 0.1 with constant amount of 2-naphthol8-sulfonate (10 ^M) and BSA (50 ^iM) 135 4-60. The fluorescence emission spectra of BSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 9.0 and 0.1 with constant amount of 2-naphthol8-sulfonate (10 ^iM) and BSA (50 ^iM) 136 4-61 . Fluorescence emission spectra of 2-naphthol-8-suIfonate in the presence of pyridine in ethanol 137 4-62. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of pyridine in 75% (VA^) ethanol 138 4-63. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of pyridine in 50% (VA^) ethanol 139 4-64. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of pyridine in 1,4-dioxane 140 4-65. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of pyridine in 75% (V/V) 1,4-dioxane 141 4-66. Fluorescence emission spectra of 2-naphthol-8-suIfonate in the presence of pyridine in 50% (VA^) 1,4-dioxane 142 4-67. Plots for fluorescence quenching of 2-naphthol-8-sulfonate in the absence and presence of albumin as well as albumin with pyridine at pH 5 and ^ 0.1 143 4-68. Plots for fluorescence quenching of 2-naphthol-8-sulfonate in the absence and presence of albumin as well as albumin with pyridine at pH 9 and \i 0.1 144 xiii

PAGE 14

4-69. CD spectra of HSA, HSA-2-naphthol-8-sulfonate at pH 9.0 and pH 5.0 in the far-UV region 4-70. CD spectra of HSA; HSA-2-naphthol-8-sulfonate at pH 9.0 and pH 5.0 in the near-UV region 4-71. CD spectra of HSA; HSA-2-naphthol-8-sulfonate at pH 9.0 and pH 5.0 in the wavelength region 300 450 nm 147 4-72 CD spectra of BAS and BSA-2-naphthol-8-sulfonate at pH 5.0 in the far-UV region 4-73. CD spectra of BAS and BSA-2-naphthol-8-sulfonate at pH 5.0 in the near-UV region 474. CD spectra of BAS and BSA-2-naphthol-8-sulfonate at pH 5.0 in the wavelength region 300 450 nm 150 51 . A simplified schematic representation of energy levels (A) attained by a molecule in solution in the course of light absorption and emission (B) 154 5-2. Potential energy cross sections drawn along two coordinates and for the ground and excited states 157 5-3. Amino acid sequence of HSA in an arrangement reflecting the heart-shape structure 1^ 5-4. Schematic representation of a protein molecule, with a layer or two of strongly associated water, suspended in aqueous solution 182 xiv

PAGE 15

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy A NOVEL PHOTODISSOCIATIVE PROBE OF WATER IN SERUM ALBUMIN BINDING SITES By Qiao-qing Di December, 1998 Chairman: Dr. Stephen G. Schulman Major Department: Medicinal Chemistry Ligand-protein interactions play a key role in the distribution and transport of small molecules in biological systems. Understanding the molecular basis of these interactions is important in the rational design of new and more efficient therapeutic agents that can recognize and bind to specific biological targets for improved drug activity. Water plays a fundamental role in the interactions of proteins with their ligands. However, the properties of water molecules in the vicinity of a biomolecule differ appreciably from those of bulk water. For the sake of understanding of the structure and dynamics of the aqueous environment surrounding the concerned biomolecule, a novel photodissociative probe of water in protein binding sites based on the proton dissociation reaction of a hydroxyaromatic in the excited state was discovered and characterized. In this study, the steady-state fluorescence and lifetime measurements have been employed to determine the rate of proton dissociation from the excited probe molecule in albumin XV

PAGE 16

binding sites. The identification and characterization of the binding sites for the probe in albumin were performed through fluorescence quenching experiments, displacement experiments, and circular dichroism methods. The results indicated (1) for surface-bound 2-naphthol-8-sulfonate, the reduced proton dissociation rate is due to the protein-water interactions which cause an overall reduction of the reorientational rate of water; for the internally bound 2-naphthol-8-sulfonate, the proton transfer reaction takes place via the hydrogen-bonded chains; (2) the primary factor involved in the 2-naphthol-8-sulfonatealbumin complex formation is the existence of the structured water hydration around the albumin molecule; and (3) 2-naphthol-8-sulfonate can be used to detect surface differences between proteins that perform the same functions in different species, more specifically, differences in the surface properties of proteins differing only in a few amino acids. This study has demonstrated that the kinetic studies of excited singlet state proton transfer can disclose the properties of the immediate environment of the proton emitter, and the applicability of 2-naphthol-8-sulfonate as a useful fluorescence probe to study the role of water in the binding sites on protein. xvi

PAGE 17

CHAPTER 1 INTRODUCTION Ligand-protein interactions play a key role in the distribution and transport of small molecules in biological systems. Understanding the molecular basis of these interactions is important in the rational design of new and more efficient therapeutic agents that can recognize and bind to specific biological targets for improved drug activity (1). Water plays a fundamental role in the interactions of proteins with their ligands. The binding process generally involves an entropically favored displacement of solvent molecules from the protein and ligand surfaces and an enthalpically favored reorganization of these solvent molecules (2). For example, approximately 65 water molecules appear to be released on binding of glucose to hexokinase (3). Some solvent molecules may, however, be trapped at the protein-ligand interface. These may make an enthalpic contribution to the ligand binding free energy by, for example, mediating hydrogen bond bridges between the ligand and the protein. Because of their mobility relative to the protein and their ability to both accept and donate hydrogen bonds, water molecules are adaptable liganding partners that are able to fill empty space, modulate the binding specificity of the protein, and play a role in its function. Numerous examples have been reported of the structural and functional importance of water molecules, as they are associated with proteins and in many cases, have a direct and crucial function in molecular recognition and catalysis (4). Examples of protein structures in which water 1

PAGE 18

molecules have been observed to mediate protein-small molecule interactions including those of the complexes of cholesterol oxidase with a steroid substrate (5), retinol-binding protein with retinol (6), adipocyte lipid-binding protein with arachidonic acid (7), adipocyte lipid-binding protein with palmitate and with hexadecanesulfonic acid (8), Larabinose-binding protein with L-arabinose, D-fucose, and D-galactose (9, 10), and ribulose-l,5-bisphosphate carboxylase oxygenase with 2-carboxyarabinitol bisphosphate (11). Often, understanding the role of water in protein-ligand interaction offers promise for the rational design of new and more efficient therapeutic agents that can recognize and bind to specific biological targets for improved drug activity. For example, one of these ordered water molecules, seen in complexes of the HIV-protease with peptide ligands, has guided the design of a novel tightly-bound inhibitor (12, 13). Many biochemical processes proceed at the interface of water with the rigid structures of proteins. Of the total surface of the protein, there is only a very small area called the binding site or active site, in which binding or catalysis takes place. Such a site is a very special environment. It is a cleft in the low dielectric matrix of the protein, spotted with hydrophilic and charged amino acid residues that surround a microscopic space some 10 20 A in its longest dimension. The combination of charged moieties, dielectric discontinuity and microscopic dimension tends to alter the energy levels of molecules that enter it, which is the primary step of binding or catalysis. For a complete knowledge of the fimction of such systems, an understanding of the structure and dynamics of the aqueous environment surrounding the concerned biomolecule is thus, essential. The properties of water molecules in the vicinity of a biomolecule differ

PAGE 19

appreciably from those of bulk water (4, 14-18). It is now known that the hydration shell surrounding a protein molecule comprises different types of water. Comparison of the water structure surrounding the surface of the protein indicates that the water molecules observed can be classified into three categories. The first contains those that are observed at the same site, making the same interactions with the protein surface in every independently solved structure in crystallographical experiment. These water molecules have the residence times in the range 10"' 10"^ s and are well ordered. The second category consists of a number of water molecules that are observed at a particular site in only one structure, but not in others. These water molecules have an average residence time in the range 10 50 ps and are slightly disordered. Finally, there are water molecules that must be present at the surface of the protein, but that are not observed crystallographically at all because they are always disordered. The dynamics and structure of ordered water near proteins, DNA, and in reverse micelles have been the subject of intense research over several decades (19). Our understanding of ordered water at the atomic level comes primarily from several approaches: single crystal diffraction (X-ray and neutron), and nuclear magnetic resonance spectroscopy (NMR) are two experimental methods and the most extensively used techniques to understand the interaction of water with proteins, and molecular dynamics (MD) simulations and empirical analyses of known structures are complementary theoretical methods. The investigation of bound water molecules by NMR techniques has been impeded by technical problems associated with distinguishing small numbers of discrete, bound water protons from the enormous number of protons in bulk solvent, although advances in water suppression techniques allowed some very important studies of non-isotopically

PAGE 20

4 labeled proteins (20, 21), and recent studies have increasingly employed selective detection of protons bound to heteronuclei ('^C or "N) to improve suppression of bulk water signals (22, 23). Another limitation of this technique is that some potential water sites are not accessible to study because they are too far away from a proton (e.g. water molecules near some carbonyls or carboxylates). X-ray and neutron crystallography reveal the favored average positions occupied in the crystal by water molecules. These water molecules, which are surrounded by backbone atoms and satisfy the main-chain hydrogen bonding capacity of peptide groups not involved in hydrogen bonds with other peptide groups, appear to be an integral part of the folded protein and are highly ordered (24, 25). Since this structure of protein, as determined by crystal diffraction methods, is of the form favored by the crystallization system, not necessarily its active or liganded form, which usually involves surface semi ordered water molecules. For this reason it is advisable to rely on dynamic measurements and through them to gain some structural information. In our studies, we use the method based on a well-defined probing reaction, namely, proton transfer of excited hydroxyaromatics. The proton dissociation of an excited hydroxyaromatic (26, 27) is given as follows: hv2 k'(r + nr) T OO+ H3O+ 00-* + H3O+ f OOH + H2O kd ^ (1-1)

PAGE 21

5 The excited protonated form OOH* can either decay to the ground state by a radiative plus a nonradiative process (k(, + „,)) emitting at a characteristic wavelength (/zv,) or it might dissociate before decaying to the ground state anion (OO ). It is well established that hydroxyaromatics are stronger acids in their lowest excited singlet states than in the ground state (pK* < pK) (28). Thus the emission of the protonated form of hydroxyaromatics is at a shorter wavelength than that of the deprotonated form (v, > V2). A measurable dissociation of OOH* will take place only if it is faster than the fluorescence decay k.j > k^^ + n,y Otherwise, the molecule will reach the ground state before the proton is emitted from the excited state. Hydroxyaromatics with large pK shifts are not only effective proton emitters but are also convenient compounds to measure the lifetime of each excited species. The spectral shift between the emission bands is large enough so that each form can be independently measured. Steady-state fluorescence and lifetime measurements can be employed to determine the rate of proton dissociation from these excited hydroxyaromatics. The mechanism of proton transfer for reactants in solutions is relatively well understood (29). The rate constant of proton transfer is extremely sensitive to its envirorunent. Except for very strong acids (pK < 0), no dissociation will take place unless water molecules are in the immediate vicinity to act as proton acceptors (30-32). Thus, the rate of proton transfer can be interpreted in terms of a single parameter, the probability of successful proton transfer to water within the lifetime of the excited state. With such opportunities at hand, the advantage of a short observation period is self-evident. If we can limit the monitoring to a very short-time frame, after a proton has

PAGE 22

been released in a defined site, the physical information obtained by the analysis will reflect only that space which the proton could probe during the observation time. Thus, the temporal resolution is transformed into spatial resolution. Under proper conditions, microspace as small as the hydration layer of a protein or the specific site on a protein can be studied, totally insensitive to the huge bulk volume in which the sample is suspended. Therefore, on measuring the dynamics of proton transfer, the kinetic analysis of the reaction can quantitate the properties of the immediate environment, such as the physical and chemical properties of the water in the ligand binding site of a protein or in an active site on an enzyme, density of immobile binding sites, or electrostatic interaction. 2-naphthol-8-sulfonate (its structure is given below) is a novel photodissociative probe of water in protein binding sites because of its unusual behavior compared with its isomers such as 2-naphthol-6-sulfonate (its structure is given below). 2-naphthol-6sulfonate in aqueous solutions undergoes excited state proton transfer from the hydroxyl group to a water molecule, which is manifested by appearance of the fluorescence band of the protonated form (monoanion) accompanied by the fluorescence band of the deprotonated form (dianion) while only the protonated form is directly excited. When adsorbed on bovine serum albumin only the fluorescence of the protonated form was observed (33) which indicates that excited state proton transfer in this case is much slower than emission of fluorescence. On the contrary, we found that the rate of excited state proton transfer of albumin-bound 2-naphthol-8-sulfonate is still comparable to or greater than the rate of fluorescence. Thus, it can be used to characterize the processes and microenvironments of ligand protein binding sites such as the role of water in the binding sites of protein, and enzyme catalytic sites.

PAGE 23

7 Animal models are often used in protein binding and pharmacokinetic studies, and the results obtained are then extrapolated to humans. Considering the high degree of amino acid sequence similarity that can be found in serum albumins from different species, it has been anticipated that other serum albumins had binding sites analogous to ?03OH SO32-naphthol-8-sulfonate 2-naphthol-6-sulfonate the sites identified on human serum albumin. A characteristic of serum albumin is the ability to transport a multitude of ligands such as fatty acids, amino acids, steroids, metal ions, and drugs (34). The binding involves hydrophobic, hydrophilic, cationic and anionic substances. The ability of albumin to act as an important extracellular antioxidant (35) or impart protection from free radicals, and other harmful chemical agents (36) agrees well with the increased susceptibility of analbuminemic rats to cancer (37). It is widely accepted in the pharmaceutical industry that the overall distribution, metabolism, and efficacy of many drugs can be altered based on their affinity for serum albumin. In addition, many promising new drugs are rendered ineffective because of their unusually high affinity for this abundant protein. Obviously, an understanding of the chemistry of the various classes of pharmaceutical interactions with albumin can suggest new approaches to drug therapy and design. Consistent with this biological role and a high affinity for a variety of ligands, they are one of the most extensively studied and applied classes of proteins in biochemistry. However, as a protein, albumin is far from typical.

PAGE 24

8 and the widespread interest in and application of albumin have not been balanced by an understanding of its molecular structure as a high-resolution crystal structure is not yet available for serum albumin. Since the 2.8 A resolution of the available crystal structure of the highly homologous human serum albumin (38) does not allow individual water molecules to be located, little is known about the role of water in the ligand-albumin interactions. For this reason the binding site of both human and bovine serum albumin has been chosen as a model to study the role of water in the binding site of protein. Albumin is characterized by a low content of tryptophan and methionine, and a high content of cystine and charged amino acids such as aspartic, glutamic, lysine and arginine. The main reason for the high solubility of albumin is its high total charge. Human serum albumin (HSA) is a protein of 69000 Da and consists of 585 amino acids with one Trp residue at position 214. The three-dimensional structure of HSA, as determined by X-ray crystallography, consists of 67% a-helical structure. The rest of the residues form random coils and extended chains (38). The overall topology of the molecule is heart-shaped, with three repeating helical domains labeled 1, 11, and III, and each domain was identified to consist of two subdomains A and B. The primary ligand binding sites on HSA were shown to be located in subdomains IIA and IIIA, although numerous other low-affmity sites also exist. Bovine serum albumin (BSA) shares a high degree of homology with HSA and was suggested to have a structure similar to that of HSA, with minor differences. For example, BSA has two Trp residues, one near the surface of the protein and the other in the interior, in contrast to the single tryptophan present in HSA (39). BSA also has a high degree of a-helical content. BSA can be

PAGE 25

9 readily crystallized for purification purposes, but large crystals suitable for X-ray diffraction studies have not yet been obtained. The ability of both albumins to undergo a major reversible conformational isomerization with changes in pH was observed several decades ago (40, 41) and several forms were classified: the N form, or normal form, is predominant at neutral pH; the B form, or the basic form, occurs above pH 8.0; the F form, or fast migrating form, produces abruptly at pH values less than 4.0; the E form, or expanded form, appears at pH less than 3.5; and the A form, or aged form, occurs with time at pH values greater than 8.0. At the present time, little structural information is known about the B form. There is a decrease in helical content of the E, F, and B forms. There is evidence that two major and structurally selective organic ligand binding sites exist on human albumin (42) and Sudlow et al. have termed these sites I and II. Both sites were originally distinguished on the basis of differences in the pattern of ligand binding and displacement. The ligands that absorb to site I are mainly weak acids with bulky structures with a negative charge in the center of the molecule (43). Azapropazone, phenylbutazone and warfarin are typical examples of the ligands which interact with site I and this locus is often referred to as the warfarin binding site. Ligands with an elongated shape, with or without a negative charge, bind to site II and diazepam together with octanoic acid and L-tryptophan are typical ligands that selectively interact with this site. Thus, this binding locus is known as the benzodiazepine/indole binding site (43). Spectroscopic investigations, binding to albumin fragments and affinity-labeling of amino acid residues, such as tryptophan 214 and tyrosine 411, suggested site I is located in domain II whereas site II is found in domain III (44). The locations of these sites has been confirmed by X-ray crystallography and He and Carter (38) have shown ligand

PAGE 26

10 binding sites I and II proposed by Sudlow et al are formed within subdomains IIA and III A, respectively. To use 2-naphthol-8-suIfonate as a protein-binding site probe, it is necessary to carry out binding studies about the 2-naphthol-8-sulfonate protein binding and determine the binding parameters associated with this process: for example, influence of medium composition on the binding, the binding constant, the number of binding sites, the location of binding, etc. In this content, a systematic investigation of the binding behavior of 2-naphthol-8-sulfonate to BSA and HSA was also undertaken.

PAGE 27

CHAPTER 2 BACKGROUND An understanding of proton transfer in the lowest excited singlet state in aqueous solutions cannot be adequate without first possessing knowledge of the physical chemistry of water. Therefore, this subject is briefly reviewed. A review of prototropic conductivity in water and dynamic aspects of proton transfer in the lowest excited singlet state is also included. The material for these reviews was taken from references 45, 46, 47 and the references contained therein. The Structure of Water Due to electronic hybridization of molecular orbitals, the water molecule has the shape of a V, with an angle (in the liquid phase) of 105 "between OH arms. Two of the electron pairs of oxygen are non bonded and the other two electrons are shared with the two atoms of hydrogen. As a consequence, there is an excess of negative charge in one side of the molecule and an excess of the positive charge in the other side (the side of the hydrogen atoms). The resulting dipole moment is very high. It is then usual to represent the water molecule by a central core including the oxygen atom and four point charges distributed around it with tetrahedral symmetry. Two are positive and correspond to the hydrogen atoms and the two negative are the lone pairs. Water is not a free rotor, but rather tightly involved in hydrogen bonding. The local structure around any given water 11

PAGE 28

12 molecule tends to be tetrahedral. There are roughly four water molecules in the first solvation shell, two donating and two accepting hydrogen bonds from the central water. X-ray diffraction yields a clear description of the oxygen-oxygen radial distribution function (48). Its peaks correspond to nearest-neighbors and next-nearest-neighbors in tetrahedral symmetry. These peaks diminish in amplitude with increasing temperature, indicating a decreasing hydrogen-bonding content. There is evidence for the persistence of even larger ice-structure motifs in liquid water (49). Likewise, Raman experiments (50) and molecular-dynamics simulations (51) show that in liquid water most molecules are involved in either three or four hydrogen bonds, with the enhancement of tetrahedral symmetry at lower temperatures. On the other hand, water molecule rotation does occur. Current interpretations of Raman-induced Kerr effect (52), Rayleigh light scattering (53) and inelastic neutron scattering (54) suggest that water molecule reorientation takes 1-2 ps at room temperature. Prototropic Conductivity Proton conductivity in water is abnormally high. At room temperature, its limiting ionic conductance is about seven times that of a sodium cation, or approximately five times that of K*. In solution, the proton is not adequately described in terms of the bare ion. Using spectroscopic techniques, it has been demonstrated that the proton in water is localized in the form of relatively long-lived simple ion: the approximately symmetrical and planar HjO"^ (hydronium) cation. Unlike water, which is involved in four hydrogen bonds as both a donor and an acceptor, the positive HjO"^ does not allow hydrogen-bonding to its

PAGE 29

13 oxygen. Using the detailed microscopic information obtained from their simulations, Tuckerman et al. verified that HjO^ is preferentially solvated as H5O2* and 11,04^ complexes (55, 56). The three protons in H^O^ are equivalent and form hydrogen bonds to three HjO molecules, thus making up the H904^ complex that is embedded in the hydrogen bond network. In the HjOj^ complex, the two water molecules that bind the excess proton between them, each form hydrogen bonds to a pair of HjO molecules, embedding this complex in the hydrogen bond network. The structure of H5O2* and 11904"^ complexes can be analyzed by determining the radial distributions of O and H atoms with respect to the oxygen atom O* binding to the excess proton. Consistent with the ionic nature of H30^ the 0*0 distance in H904^ complex ro-o = 2.5 A is shorter than Too = 2.8 A in pure water while the 0*H bond length r^.^ .1 A is slightly longer (roH « 1 .02 A for pure water) and weaker than in pure water. The absence of the 1 .9 A OH maximum in the 0*H radial distribution confirms that Hfi* behaves like a cation that only donates protons to hydrogen bonds with the coordinated HjO molecules, but itself receiving none. One of the 0*0 distances in H502^ complex, i.e., 0*0*, is shorter than the other two, and ro.H = 1.3 A. The HjOj"^ and UgO/ complexes might at first appear to be clearly distinct structures. However, analysis of results of radial distribution functions for both complexes show that the environment of the O* atoms associated with the excess proton are remarkably similar. Two of the three ligand HjO's in the H9O4* structure are hydrogen-bonded to three second neighbors of the H30^ ion. This, together with the H3O* ion itself, yields the normal fourfold coordination in water. The third HjO bonded to HjO^ has one less hydrogen bond. This undercoordinated molecule frequently (but not always)

PAGE 30

14 turns out to have been the partner in the former HjOj^ complex state. A small shift of the proton to the left or the right converts the symmetric Hfii^ into an asymmetric UgO^* conformation. Gas-phase mass-spectrometric data (57) show that water ligands in the first solvation shell are more tightly bound than those in the second and subsequent shells. Although these data cannot be directly transferred to bulk solution, it is clear that hydrogen bonding to HjO^ is considerably stronger than in bulk water. The extra strong hydrogen bonding correlates with shorter OH-0 distances, 2.55 A compared with 2.8 A in bulk water. A proton conductivity mechanism that involves the rupture of even one of these three bonds will lead to an activation energy exceeding the observed 2-3 kcal/mol value. Thus, the first solvation shell of HjO^must remain intact during the proton transfer act (58). Proton conductivity is incoherent as shown by experiments that proton conductivity in ice is about a factor of two slower than in super-cooled water of equal temperature (59, 60). The evidence for proton conductivity in water is unequivocal, it increases with increasing temperature and pressure (61). That the fraction of hydrogenbonded water decreases with increasing temperature is clear intuitively as well as from simulations (51) and experiment (52). Thus proton conductivity increases as hydrogen bonds weaken. The deuterium isotope effect on prototropic mobility is rather small, 1 .4 at room temperature (62). As suggested by Westheimer and Melander (63) kinetic isotope effects are maximal for the symmetric, AG 0, reaction. Proton hopping between two water

PAGE 31

15 molecules is expected to correspond to this symmetric limit. The maximal kinetic isotope effect for reactions involving proton transfer is 6-10 (64). The small isotope effect on proton conductivity suggests that the reaction coordinate does not involve proton motion. The 1.4 deuterium isotope effect on proton conductivity (62) can also be interpreted as arising from water reorientation that leads to hydrogen bond cleavage. Rotation around the heavy oxygen is a nearly pure hydrogen motion, so that the isotope effect should be given by the mass ratio, yfmjm^ = V2 . The OD-0 versus OH-0 enthalpy difference is thus expected to be AAH = '/zRT ln2 = 0.2 kcal/mol at room temperature. This has been recently verified experimentally (52). The possibility that proton motion within clusters is rate-limiting may be ruled out by the small deuterium isotope effect. Since the hydrogen bond content of water is so high, every molecule participating in 3-4 hydrogen bonds (51, 52), the whole body of water may considered as one huge cluster. The formation of new hydrogen bonds at the periphery of water clusters is not the rate limiting step. Proton conductivity correlates with hydrogen bond cleavage rather than with its formation, as suggested by the fact that it increases with decreasing hydrogen bond content of water, i.e. at higher temperatures and pressures (61). Proton hopping times as obtained by NMR and '^O resonance (65), Xp « 1.5 ps agree with single molecule reorientation times (52-54), a process requiring hydrogen bond cleavage. With the above mentioned restrictions in mind, a revised Grotthuss mechanism of proton conductivity in water is proposed by Agmon (45, 66) as follows. Consider a triply coordinated H3O*, i.e. the H9O4* cation, as the reactant state. Since the rate limiting step involves cleavage of a hydrogen bond, but not one in the first

PAGE 32

16 solvation shell, it is natural to consider hydrogen bond cleavage in the second solvation shell as the rate limiting step. This leads to isomerization of the cation into the HjOj^ cation. This is depicted in Figure 2-1. For clarity, four of the oxygen atoms are identified by corresponding letters. Figure 2-1. A molecular mechanism for prototropic mobility. Taken from reference 45. Initially, the proton is located on O,, and ^20^ is one of the three water ligands in the first solvation shell, Figure 2-la. Thus, ObH-O, is an extra strong bond which cannot easily break. HjO,. has at least one hydrogen bond donating water ligand, HjOj, which is

PAGE 33

17 in the second solvation shell with respect to Ob and thus, of ordinary strength. The cleavage of the OjHOj bond is postulated to be the rate limiting step, Figure 2lb. The reorientation of HjOj may be quite small, so that just a single hydrogen bond is broken. This may cost 2-3 kcal/mol and take around 1 ps at room temperature. Following bond cleavage, one can expect ultrafast (fs) readjustment of bond angles and bond lengths to form HjOb" •H*" 0<.H2. For example, the 0^-Oc distance should shrink by about 0.15 A, from 2.55 A in H904'^ to the 2.4 A distance of HjOj^ in vacua (67). At the same time, the distance between and the remaining two oxygens are expected to increase from 2.55 A to below 2.8 A. Subsequently, fast fluctuations of surrounding water dipoles momentarily stabilize or O^, leading to large proton polarizabilities within this complex (68, 69). Eventually, one of these fluctuations couples to reorientation of the HjO^ water, which donates the fourth hydrogen bond to O^. The proton is now located on O^., Figure 2-1 c. The 1^502"^ has reisomerized to a 1^904^ cation, centered on a neighboring water molecule. The proton has effectively hopped incoherently across the 0-0 distance in H9O4*. Tuckerman et al. (55, 56) conducted ab initio molecular dynamics simulations on a proton in a 32-water molecule cluster combining a density functional description of electronic structure and finite temperature dynamics. For H30^ ion they found a dynamic solvation complex, which continuously fluctuates between a H5O2* and a H904^ structure as a result of proton transfer. HjO"^ and its associated complexes integrate naturally into the hydrogen bond network, enabling proton transfer without substantial activation and rearrangement of the solvent. According to the results of the simulation they suggested that the rate limiting step for the migration of the excess proton is the concerted dynamics of the second solvation shell hydrogen bonded to the ligand HjO. The quantum dynamics

PAGE 34

.IT ' J,.
PAGE 35

19 those functional groups bonded directly to an aromatic ring will experience changes in charge distribution sufficient to cause detectable differences between pA^ and pK* Electron-donating groups have lone electron pairs which upon excitation, may be transferred to the lowest-lying vacant n* orbitals of the aromatic system. Groups of this kind include hydroxyl, sulfhydryl, and amino groups. Upon excitation, the electronic charge density at these groups decreases and consequently, a proton may be more readily lost from or with more difficulty added to the group in S, than in Sq. Consequently, pK* < pK, and these groups are thus more acidic (or less basic) in S, than they are in Sq. Functional groups having low-lying vacant n orbitals will, as a resuh of electronic excitation, accept electronic charge from the aromatic system. Examples include carbonyl, carboxyl, carboxylate, and amide groups. The increase in electronic charge density resulting from excitation makes it more difficult to remove a proton (or easier to add a proton). As a result, pK* > pK, and these groups are, therefore, less acidic (or more basic) in S, than they are in Sq. The Effect of State of Protonation on Electronic Spectra Protonation of a functional group which is attached to an aromatic system may have a profound effect on the absorption and fluorescence spectra of that molecule. This is the result of the electronic charge stabilization incurred by the presence of the proton at the ftinctional group. In molecules containing an electron-withdrawing group, the excitation of those molecules from the Sq to the S, state results in the movement of electronic charge to the electron withdrawing group. This causes the S, state to be stabilized to a greater degree

PAGE 36

20 than is the Sq state by protonation at the electron withdrawing group and decreases the energy difference between So and S, state in the protonated molecule. The spectral result is the shifting of the longestwavelength absorption bands and the fluorescence spectrum to longer wavelengths upon protonation and to shorter wavelengths upon deprotonation. In molecules containing an electron-donating group, the loss of electronic charge from the functional group which accompanies excitation from the Sq to the S, state is inhibited by protonation and facilitated by deprotonation. This causes the energy difference between Sq and S, to be greater in the conjugate species having the higher state of protonation. As a result, the longest wavelength absorption and fluorescence bands of these molecules shift to shorter wavelengths upon protonation and to longer wavelengths upon deprotonation. The displacement of the long wavelength absorption and fluorescence bands of molecules possessing electron acceptor groups, to longer wavelengths upon protonation and to shorter wavelengths upon deprotonation, is related to the increase in basicity and decrease in acidity upon going from the Sq to the S, state. In molecules having electron donor groups, the movement of the long wavelength absorption and fluorescence bands, to shorter wavelengths upon protonation and to longer wavelengths upon deprotonation, is indicative of a decrease in basicity and an increase in acidity upon going from the Sq to the S, state. The Effect of Solvents on Electronic Spectra The effect of solvents on electronic spectral bands depends upon the nature and relative strength of the interaction of solvent molecule with the ground and excited states

PAGE 37

21 of the solute molecule. These interactions are predominately electrostatic in nature. All solvent effects may be reduced to a comparison of whether the ground or excited state is more stabilized relative to each other. In general, if the excited state is more polarizable than the ground state, then interactions with polar solvents lead to stabilization of the excited state relative to the ground state and transitions shift to longer wavelengths. For nonpolar solutes dissolved in nonpolar solvents, only the dispersion forces are expected to contribute significantly. These forces are relatively weak; hence, the absorption and fluorescence wavelength maxima of most nonpolar molecules are not strongly solvent-dependent. This leads to small shifts of absorption and fluorescence spectra with increasing dielectric constant of the solvent. When a polar solute is dissolved in a polar solvent, dipole-dipole interactions are dominant. In many but not all solute molecules, electronic excitation increases the degree of charge separation in a molecule, and thus, increases the dipole moment. In such a case, the energy of the excited state should be decreased to a greater extent than that of the ground state by increasing the "polarity" of the solvent. Thus, increasing the "polarity" of solvent should cause both So -> S, absorption and S, -> Sq fluorescence to shift to lower energy (longer wavelength). There is an ever-present possibility that solute molecule containing polar ftinctional groups may engage in specific chemical interactions such as hydrogen bonding with some solvents. Such specific interactions, if present, usually are the dominant factors affecting the absorption and fluorescence wavelengths of the solute. If a nonbonding pair on a solute molecule is coordinated by a hydrogen-atom of the solvent, the hydrogen

PAGE 38

22 bonding interaction lowers the energy of the ground state as well as that of the n,n* state of the solute. However, because the ground state molecule has two electrons in the nonbonding orbital and the excited state has only one, the stabilization of the ground state is greatest. As a result, the energies of n -> 7t* absorptions increase (the spectra shift to higher frequencies or shorter wavelengths) with increasing solvent hydrogen-bond donor capacity. Hydrogen-bonding solvents also have a marked effect upon intramolecular charge-transfer absorption spectra. Hydrogen-bond donor solvents interacting with unshared valence electron pairs on functional groups which are charge-transfer acceptors in the excited state (e.g. -COOH) enhance charge-transfer by introducing a partial positive charge into the functional group. This interaction stabilizes the charge-transfer excited state relative to the ground state so that the absorption spectra tend to shift to lower energies with increasing hydrogen-bond donor capacity of the solvent. An increase of hydrogen-bond donor capacity of the solvent tends to produce shifts to higher energies when interacting with unshared valence electron-pair on functional groups which are charge-transfer electron donor in the excited state (e.g. -OH, -NH2). Hydrogen-bond acceptor solvents tend to produce shifts to longer wavelengths when solvating hydrogen atoms on functional groups which are charge-transfer electron donors in the excited state. This is a result of the partial withdrawal of the positively charged proton from the functional group which facilitates transfer of electronic charge away from the ftinctional group. Solvation of hydrogen atoms on functional groups which are charge-transfer electron acceptors in the excited state inhibits charge-transfer by leaving a residual

PAGE 39

23 negative charge on the functional group. Thus this interaction tends to resuU in shifting of the absorption spectrum to shorter wavelengths. Solvent polarity and hydrogen bonding effects upon fluorescence spectra are qualitatively similar to those upon absorption spectra. In many cases, however, the fluorescence shifts of a giving solute in a giving series of solvents may not parallel, even qualitatively, the absorption shifts in the same series of solvents. This is understandable because there is a possible alteration of hydrogen bonding sites in the solute when the solvent cage is changed from the S, configuration to the Sq configuration. Up to the present, attention has been focused on the effects of solvent properties on the spectral positions of the electronic spectra bands. However, the intensities of fluorescence and, to a lesser extent, absorption spectra also depend on the nature of the solvent. In general, solvation which interferes with electronic interaction between aromatic ring and functional group tends to diminish the molar absorptivity as in the arylamines in highly protic solvents or carboxylic acids in hydrogen bond acceptor solvents. The intensities of fluorescence spectra are extremely sensitive to solvent polarity and hydrogen bonding properties. In general, in molecules with a n, n* state as the lowest excited singlet state, in the isolated molecule strong fluorescence is favored by high polarity and high proticity of the solvent whereas weak or no fluorescence is favored by aproticity and low polarity of the solvent. In many molecules whose lowest excited singlet states are of the n, n* or intramolecular charge transfer types, increases in solvent hydrogen bonding capacity frequently decreases fluorescence quantum yields. In

PAGE 40

24 numerous cases, hydrogen bonding of a fluorescent molecule with solvent causes decreases in fluorescence quantum yields because of a substantial increase in the rate of internal conversion, at the expense of fluorescence (71, 72). Steady-State Kinetics of Excited State Proton Transfer Reaction The steady-state approach to the determination of proton transfer rate constants of acids or bases in the S, state is based on the assumption of the attairmient of a steady-state involving the various photophysical and photochemical processes deactivating S,. In the absence of buffer species (i.e. proton is the only protonating species and water is the only proton acceptor present in appreciable quantity) in aqueous solutions, proton dissociation of an excited hydroxyaromatics (26, 27) is given in equation (1-1): ki2 ^ OOH* + H2O , 00-* + H3O+ k(r + nr) k2l hvi hv2 k'(r + nr) OOH + H2O 00+ H3O+ (1-1) The variations of the relative fluorescence quantum yields of hydroxyaromatics and their conjugate bases with pH are known to depend upon the kinetics of proton transfer in the lowest excited singlet state (26, 27). Under the assumption, the relative fluorescence quantum yields of the excited hydroxyaromatics have been shown to vary, approximately according to «!> _ 1 + k„To'[H30-] i 1 + k„To + k^ToTH.O^] ^ " ^ while those of the excited conjugate base hydroxylate species vary according to

PAGE 41

25 (2-2) (|)o' 1 + k„To + k„To'[H,0^] where [HjO"^] is the molar concentration of hydrogen ions, Xq and Tq are the Hfetimes of excited acid and conjugate base in the absence of proton transfer (i.e., in the low and high pH limits, respectively), and k|2 and k2, are the rate constants for dissociation of the excited hydroxyl species and protonation of the excited hydroxylate species , respectively. More refined treatments have corrected equations (2-1) and (2-2) for transient reprotonation of the hydroxylate species prior to the establishment of the steady state and for the effect of ionic strength on the "equilibrium" between the reactants and the activated complex in the transition state (73, 74). Dividing equation (2-1) by equation (2-2) yields ^ +^^[H30"] (2-3) 5, k2iXo'[H30*] / k,2Xo ^ 0 and ^/^o, ^'/^q', and therefore ( f /(|)o' ) / ( ^/^o) become independent of pH; that is.

PAGE 42

26 I / const (2-5) I const const 12 ''0 (2-6) which permits a quick determination of k,2. For excited conjugate bases with Xq' < 1 x 10 s and / or with kjj < 5 x 10'" M' s"', [HjO^] can be greater than 10'' M and still the acid and its conjugate base may demonstrate ( ^'/^q ) / ( ^/^o) independent of pH. In other words, the smaller kjiXo', the lower is the pH at which the reprotonation of the excited conjugate base, as reflected by the dependence of ^/^q and (j)'/(j)o' on [HjO^], will be observed. It is important to note that in equations (2-1) and (2-2), k|2 and kj, are somewhat dependent upon the composition of the reaction medium (i.e., they general vary with [H3O*]). The true rate constants of the reaction k|2(0) and k2i(0) are the rate constants in the pure reference solvent, in this case, water, at infinite dilution of solutes (i.e., when [1130"^] = aH^, where is the activity of the hydrogen ion). In relatively dilute acidic solutions of pH 1-4 the correction of equation (2-3) for medium effects has made use of the Br0nsted kinetic activity factor F (75), based upon the Debye-Hiickel type ionic screening treatment. -logF = (2-7) where A and B are constants of the solution dependent on temperature (T) and dielectric constant (e).

PAGE 43

27 A = 1.826 X 10' (sT)-'/' (2-8) B = 5.031 X 10' (eT) '/' (2-9) ^ is the ionic strength of the solution, and a is, to a first approximation, taken to be a mean ionic size parameter for all participants in the reaction, Z is the charge of the dianion. The factor F corrects the rate constants at each value of \i to the value corresponding to reaction at zero ionic strength. A plot of ( ^I^q )/( ^'l^o ) against ^[HjO^] then should yield a straight line of slope k2|To' / k,2To and an ordinate intercept of 1/ ki2To. The fluorescence lifetime measurement methods can be classified into two main categories, namely indirect and direct methods. Indirect methods of fluorescence lifetime measurement are listed as follows: (a) based on the relationship that the fluorescence quantum yield ((j)) is related to the fluorescence lifetime (x): (|) = t/Tn (2-10) where x is the actual lifetime and is proportional to the relative fluorescence intensity measurement and x^ is the natural or radiative lifetime of the excited state (the lifetime that would be measured if fluorescence was the only process originating from S,). (b) based on the relationship described as : l/xN = 2900rt^Vo^ Wv (2-11) V 0 in |j,m"' is the wavenumber corresponding to the maximum absorption, e is the molar absorptivity in 1 mol'' cm ', and n is the refractive index of the solvent. For most

PAGE 44

28 materials, the experimental values of half band widths are ca. 0.3 cm"', and assigning = 2 and the region of interest as 2.5 |am"', equation (2-9) is reduced to l/TN«10^^.ax (2-12) (c) based on the quenching of fluorescence by using the Stem-Volmer relation (76): Io/I-1+Kq[Q] (2-13) where Iq and I are fluorescence intensities observed in the absence and presence of a concentration [Q] of quencher. Kq, the quenching constant, is related to the bimolecular reaction rate, kpQ, by the equation KQ = kpQTo (2-14) Therefore, by estimating the Kq using the fluorescence quenching data and assuming that kpQ has a value of about 10'° 1 mol' s ' corresponding to a diffusion-controlled reaction, a value for Tq can be obtained. The above indirect methods provide an estimate of the fluorescence lifetime, but are tedious and are not favored over the other approaches which allow direct lifetime measurement. At least two direct experimental methods of measuring lifetime are available, namely time resolved (pulse excitation) and phase resolved decay approaches, (a) In the time resolved method the sample is excited with a special high speed flash-lamp or pulsed laser and the time-dependent decay of fluorescence intensity is measured. The rate of decay of the initially excited population is dN{t) -~^ =
PAGE 45

29 emissive rate, and k is the rate of nonradiative decay. At r = 0, N(t) = (initial population of molecules in the S, state), integration of equation (2-15) yields N(t) = Noe-("" (2-16) where Xq = ( y + A: )"' and is the lifetime of the excited state. Since the fluorescence intensity 1(0 = kN{t), thus 1(0 = Io^-^'^" (2-17) where Iq is the intensity of light incident upon the absorber. Equation 2-17 represents the exponential decay with Tq being the time for the fluorescence to decay to \/e of its initial value. (b) In the phase resolved method, the sample is excited with sinusoidally modulated light. The phase shift and demodulation of the emission, relative to the incident light, is used to calculate the lifetime. For simple exponential decays, the fluorescence lifetime is related to the phase-shift, 0, by the following relationship To = tan 0/271/ (2-18) where /is the frequency of modulation. Protein Binding A characteristic of serum albumin is the capability of extensively binding a variety of different substances in a reversible manner. For this reason serum albumin has been widely used as a model compound to study the factors involved in the interaction between macromolecules and low molecular weight compounds. Studies on the binding behavior of serum albumin have also been stimulated by the physiological role which this protein serves as a vehicle in the exchange of fatty acids and bilirubin, between the

PAGE 46

30 tissues. Furthermore, many drugs are bound to serum albumin, thus, ensuring a more protracted pharmacological effect of these substances. We initially observed that the excited state proton dissociation of albumin-bound 2-naphthol-8-sulfonate is slowed down compared with that of the naphthol derivative dissolved in aqueous solutions. This kinetic parameter can be used to probe the microenvironment of the binding sites of serum albumin at which 2-naphthol-8-sulfonate is bound. As mentioned in previous chapter, to use 2-naphthol-8-sulfonate as a proteinbinding site probe, it is necessary to carry out binding studies about the 2-naphthol-8sulfonate protein binding and determine the binding parameters associated with this process: influence of medium composition on the binding, the binding constant, the number of binding sites, etc. Binding Studies Experimental Methods Equilibrium dialysis (77), ultrafiltration (78), and HPLC (79) have been widely used to determine the binding affinity and site specificity of small molecules for serum albumin. Equilibrium dialysis and ultrafiltration are labor-intensive and, hence, not wellsuited for rapid analysis, whereas HPLC assays yield only a "percent bound" instead of the actual affinity constants. The binding of ligands to proteins may induce a change in a spectral property of the protein or even the ligand. Advantage can be taken of the spectral perturbation to monitor the interaction of ligand and protein. Changes in absorbance, fluorescence, magnetic resonance and optical rotary dispersion have been utilized to measure the interactions of ligands to proteins. Fluorescence spectroscopy is arguably one of the most

PAGE 47

31 versatile techniques for studying ligand protein interactions. The interaction may affect a number of fluorescence parameters such as the fluorescence intensity of the protein, the fluorescence intensity of the ligand and the fluorescence polarization of the ligand. Moreover, many ligands which do not fluoresce can be studied by this method if they either quench the native fluorescence of the protein or displace a fluorescent probe attached to the protein. Fluorescence spectroscopy is more rapid than equilibrium dialysis for studying ligand protein interactions, and fluorescence is much more sensitive than many other techniques such as absorption spectroscopy (80). For example, the fluorescence method makes use of the considerable increase in intrinsic emission that occurs when warfarin (8 1 ) and other fluorescent compounds are transferred from the polar, aqueous environment to the nonpolar, hydrophobic binding site of albumin. Albumin itself shows a very modest, although measurable change in its intrinsic fluorescence upon binding of any ligand (82, 83), whereas the fluorescence quantum yield of warfarin increases as much as seven times upon binding to HSA (81). Similarly, large increases in fluorescence accompanying binding of dansyl-1 -sulfonate and dansylsarcrosine (82, 83) to albumin were observed. Binding Equation Derivation For the vast majority of cases, ligand binding to proteins may be considered to be a reversible interaction which obeys the law of mass action. The simplest case to consider is:

PAGE 48

32 P + L->PL (2-19) where a protein P can bind, reversibly, a ligand molecule L at a single site to form a protein-Iigand complex, PL. A association constant (k) can be defined: k=[PL]/[P][L] (2-20) and the protein conservation equation written as: [P,] = [P] + [PL] (2-21) where [P,] is the total concentration of the protein. Then, [P] = [P,] [PL] and k = [PL]/([L][P,]-[PL][L]) (2-22) The ratio [PL]/ [P,] represents the moles of ligand bound per mole of protein and is given the symbol r. Equation (2-22) can be rearranged to [PL]/ [P,] + [PL][L]k/[P,] = k[L] (2-23) and substituting r for [PL]/ [P,], equation (2-23) becomes r + /-k[L] = k[L] or r = k[L]/( 1 + k[L]) (2-24) Equation (2-24), then, represents the equation for the special case of one binding site on the protein molecule. If there are a number (N) of independent, identical binding sites, the equation has the following form r, + + ... r„ = r,„,„ = Nk[L]/(l + k[L]) (2-25) The free or unbound ligand concentration [L] present at equilibrium is determined by the appropriate analytical method. Then equation (2-25) can be applied to calculate the binding parameters k and N. It is not uncommon for more than one type of site to exist on

PAGE 49

33 a given protein. Each site has its own association constant, and for this reason the following more general form of the binding equation has developed (84): r„3, = N,k,[L]/(l + k,[L]) + N,k,[L]/(l + k,[L]) + ... N„ k„ [L]/(l + k„ [L]) (2-26) For ligands like 2-naphthoI-8-sulfonate, whose acidic form fluoresces much more intensely when bound than as the free ligand, the fraction of ligand bound, a, is usually determined from the spectrofluorometric titration by using the equation (85): a = (Fp Ff)/(F, Ff) (2-27) where Fp and Ff are the fluorescence intensities of a given concentration of ligand in solutions of low protein concentration and in solutions without any protein, respectively; and F|, is the fluorescence intensity of the same concentration of the fiilly bound ligand. The latter is taken to be the fluorescence intensity of the ligand in the presence of excess protein. Similarly, the fraction of ligand free, p, is determined by using the equation: p = (F,-Fp)/(F,-Ff) (2-28) The total ligand concentration, [LJ, is equal to the sum of the unbound and bound concentrations. [L,] = [L] + [PL] (2-29) then, P = (F,-Fp)/(F,-Ff) = [L]/[L,] or [L] p[L,] = [L,](F, Fp)/(F, Ff) (2-30) The value of r can be calculated from the fraction of ligand bound by using the following relationship: r = a[L,]/[P,] (2-31)

PAGE 50

34 Binding Studies Treatment of Data Binding data may be plotted in several ways in order to determine the values of N (number of binding sites) and k (the association constant). A direct plot of r, the nimiber of moles of ligand bound per mole of protein, as a function of free ligand concentration is usually hyperbolic and does not yield accurate values of N and k. The reciprocal plot, \l r vs. 1/[L], from equation (2-25), enables the direct determination of N and k from the graph since the plot should produce a straight line with intercept 1/N and slope 1/Nk. However, this plot spreads the low values of 1/ r poorly so that the values for N, and therefore k, are not reliable. Scatchard (86) used equation (2-25) in the rearranged form r /[L] = kN k r (2-32) to develop an alternative method of plotting the binding study data. The Scatchard plot is simply a plot of r /[L] vs. r. Such a plot yields a straight line, and values for N and kN may be determined from the intercepts on the abscissa and ordinate, respectively, when the N binding sites are noninteracting and have the same binding affinity. It is advisable to obtain the intercept and slope values by means of regression analysis rather by finding the best fit using a ruler. The level of the significance of the fitted curve and thus, the coefficient of regression, improves as the fit approaches the ideal value. The Scatchard plot gives a more even distribution of the data points than does the previously mentioned reciprocal plot of 1/ r vs. 1/[L]. However, if the individual binding sites have different binding affinities, due either to the distinct nature of the sites or to the interaction between the bound ligands at different sites, the Scatchard plot gives a downward curve

PAGE 51

35 instead of a straight line. As the primary binding site becomes saturated with the Ugand, these secondary sites become more important. If the binding strength of the individual binding sites are significantly different from each other, the Scatchard plot may still be used to estimate binding parameters for the individual binding site by appropriately analyzing the curve in relation to the equation (2-26) using a trial and error process of fitting values to the experimental curve (86). Computer software is now available for computing nonlinear fitting of binding data. Rosenthal (87) and Klotz and Hunston (88) have suggested convenient methods to graphically represent the binding data in a complex system. Briefly, the curve describing the relationship between r /[L] and r should take into consideration the data of the high and low affinity curves. The description of Rosenthal (87) is particularly useful for this purpose. Determination of Complex Stoichiometry The Method of Continuous Variation A spectrophotometric method for obtaining the stoichiometry of metal-ligand complexes was first described by Job (89) and later expanded by Vosburgh and Cooper (90) to include cases in which a given pair of components form more than one compound. Because of the obvious similarity between the metal-ligand reaction and the proteinligand reaction, it was felt that this method of continuous variation (or Job study as it will be referred to hereafter) could be employed to determine the stoichiometry of the 2naphthol-8-sulfonate-albumin complex. Job's method involves measuring some intensive property in a series of solutions of constant total molarity, but of varying metal-to-ligand ratio. The measured property is

PAGE 52

36 generally absorbance or fluorescence intensity. In practice, two equimolar stock solutions, one of metal and the other of ligand, are prepared. A set of working solutions is then obtained by mixing Vl ml of the stock ligand solution with (Vtml of the stock metal solution, where is a fixed total volume and Vl is a variable, 0 ^ Vl ^ V^. The absorbances or fluorescence intensities of these solutions are then measured at a fixed wavelength, and plotted as a function of mole fraction of ligand, (Vl / V-r), or of metal, {(Vy Vl) / Vy}. The position of maximum absorbance or fluorescence intensity on this plot, in relation to the mole-fraction axis, gives the stoichiometry of the complex. To determine N, the maximum number of binding sites, solutions of equimolar concentrations of albumin and 2-naphthol-8-sulfonate are mixed in varying proportions. Letting (x) be the mole fraction of one of the components, (1 x) would be the mole fraction of the other component. A suitable property of the resulting solution such as absorbance or fluorescence is measured. The difference (A) between the observed fluorescence intensity and the corresponding value in the absence of either 2-naphthol-8sulfonate or albumin is plotted against the mole fraction x of one of the components. The resulting curve should have a maximum (or minimum) if the fluorescence intensity measured is larger or smaller than that for either albumin or 2-naphthol-8-sulfonate alone. The mole fraction x at which A is a maximum (or minimum) is related to the maximum number of binding sites N, by the simple relationship N = x/(l-x) (2-33) The values ofx = 0.5, 0.67, 0.75 at which A is a maximum (or minimum) correspond to the maximum numbers of binding sites, N = 1, 2, 3 respectively. This treatment is

PAGE 53

37 applicable when the quantum yields of fluorescence, and therefore, the fluorescence intensities of the bound ligand, are constant and independent of the stoichiometry of the complex in which the ligand resides. Identification of the Binding Sites Two specific binding sites have been established on the HSA: site I and site II (42). X-ray studies of crystalline human albumin (38) support this view and indicate that site I and site II are located within specialized cavities in subdomain IIA and IIIA, respectively. Matsushita et al. (91) and Panjehshahin el al. (92) reported that BSA has binding sites with similar properties to those of site I and site II on HSA. In addition, the crystallographic structure of equine albumin has also been determined, and the data suggests that there are also two specific ligand binding sites on this molecule (93) as well. All of these indicate that, with respect to binding sites, mammalian albumins are analogous to human albumin, considering the structural similarities between the molecules. To identify the binding sites for a ligand on HSA and BSA, competitive binding studies can be performed using typical site I and site II binding maker ligands such as warfarin, and diazepam, respectively. The site-oriented approach to ligand albumin binding was first methodically studied by Sudlow et al. (42, 83). They characterized the relative strength of ligand at sites on albumins by monitoring the ability of the ligands to decrease the fluorescence of site-specific probes. The fluorescence quantum yield of the ligand bound to albumin was measured before and after the addition of the site-specific probes. The measured

PAGE 54

38 fluorescence value corresponds to a relative fluorescence intensity expressed as a percentage of the initial fluorescence: ' ^ xlOQ% ^1 or -p-xlOO% where I, and I2 represent the fluorescence intensities of the ligand plus albumin without and with the probe, respectively. A decrease in the fluorescence intensity of the complex (ligand-albumin) can be interpreted as a displacement of the ligand from its binding site by the added probe, probably through a competitive mechanism. The displacer and the ligand may be bound to the same site. Characterization of the Binding Sites on Albumins Fluorescence Quenching Solute accessibility to fluorescent probes or groups attached to protein molecules is often used to monitor conformational aspects of these macromolecules. Solute accessibility is most often determined with fluorescence quenchers by measuring and comparing the specific rate of quenching of the fluorophore free in solution, kpg, with its rate attached to the protein, k^Q, or by comparing values of k^g for two or more conditions that may affect the conformation of the protein. It is usually assumed that a decrease in quenching rate for the bound fluororophore indicates a decreased accessibility due to geometrical masking factors in the protein. However, elementary considerations indicate that for rapid, diffusion-controlled reactions that characterize the fluorescence quenching

PAGE 55

39 process, Icbq should depend not only on masking factors but also on the translational and rotational mobilities of the liganded macromolecules as well as on orientational constraints imposed by the association of ligands with macromolecules. The lower the rotational mobility, the lower is the probability that a bound ligand, exposed on a small fraction of the surface area of the macromolecule, will meet a quencher molecule during its excited state lifetime. In the fluorescence quenching experiment, the fluorescence intensities of the fluorophore attached to protein or free in solution are monitored as a function of the quencher concentration. The intensities are then analyzed with the Stem-Volmer equation where Iq and I are fluorescence intensities observed in the absence and presence of a concentration [Q] of quencher. Kq, the quenching constant, is related to the bimolecular reaction rate, kpQ, by the equation where Tq is the fluorescence lifetime measured in the absence of quencher. In the absence of static quenching, plots of Iq/ I vs. [Q] are expected to follow a straight line, the slope of which may be used to determine Kq. The parameter of interest is kpQ, which may be determined by measurement of Tq and from the slope of the StemVolmer plot. Many hydroxyaromatics exhibit fluorescence from two species in aqueous systems. When only the protonated form of the hydroxyaromatic exists in the ground (76): Io/I=1+Kq[Q] (2-13) (2-14)

PAGE 56

40 state, the dual emission results from the formation of the ionized hydroxyaromatics by excited state proton transfer. This reaction occurs because the hydroxyaromatics are much stronger acids in the excited state. Therefore, in these cases, the fluorescence of the protonated and deprotonated forms should be measured separately. The formula representing the relative fluorescence intensities of protonated and deprotonated forms, respectively are derived easily employing the following equation (94). OOH* + H2O =5 kf / \ k, O0H + /2V OOH CDO+ /IV' OO(2-34) where k' = k2,[H30"]. Then, for the protonated form, I / lo = ( 1 + a'[Q] ) / ( 1 + a [Q] + p [Q]^ ) (2-35) and for the deprotonated form, r/Io' = l/(l+a[Q] + p[Q]^) (2-36) Therefore, (I/Io)/(r/Io')=l+a'[Q] (2-37) The meanings of constants a, a', and P are as follows, a={^„To(l+k') + V V(l+k,2To)}/(l+k„To + k' V) (2-38) a' = Vxo'/(l+k'To') (2-39)

PAGE 57

41 P k^' To To' / ( 1 + k^, To + k' To') (2-40) where Tq and To' are the lifetimes of excited protonated and deprotonated forms, respectively. When k' ~ 0, i.e., [U^O*] is small, the above equations become as follows. I/Io=l/(l + {^,To/(l+k„To)}[Q]) (2-41) (I / lo ) / ( r / lo' ) = 1 + To' [Q] (2-42) Using the experimentally determined values of k.j, Tq, and To', and k^ are evaluated. Commonly used quenchers, such as T, Ag*, were examined in this study and it was found that these quenchers were ineffective in this case. As reported (95), at room temperature, the fluorescence of naphthalene derivatives, including naphthols, is efficiently quenched by alkylamines. However, in the case of 2-naphthol-8-sulfonate, these alkylamines such as l,4-diazabicyclo-(2. 2. 2) octane, N, N'-dimethylpiperazine, ethanolamine, tris(hydroxymethyl)-aminomethane, 2-bromoethylamine hydrobromide, showed little quenching efficiency at pH 5. Since these alkylamines are protonated at pH 5, therefore, they lose the ability to form hydrogen-bonded complexes with naphthol, which is thought to be necessary for quenching to occur (95). Thus, the weak organic base pyridine was chosen for the present study. Characterization of the Binding Sites on Albumins Circular Dichroism In the wavelength region where the optically active chromophore absorbs light there is also an unequal absorption of the right and left circularly polarized components of plane polarized light, i.e., where 6 Land e Rare the molar absorptivities for the left and right circularly polarized components. This phenomenon gives rise to circular

PAGE 58

42 dichroism (CD). Circular dichroism is usually expressed as a differential dichroic absorption (Gl^r)» usually written Ae. The combination of unequal absorption (circular dichroism) and unequal velocity of transmission (optical rotation) of left and right circularly polarized light in the region in which optically active absorption bands are observed is a phenomenon called the " Cotton effect". When the differential dichroic absorption of a simple Cotton effect is plotted as a function of wavelength, the CD maximum and the absorption maximum will occurr at the same wavelength. Since CD results from a difference between €Land e^, it follows that this phenomenon is not observed outside of the wavelength region where the optically active chromophore absorbs light. Thus, the basic information which can be deduced from circular dichroism curves is obtained most easily in the immediate vicinity of the spectral region of maximal absorption. Optically active chromophores may be divided into two extreme types: (a) the inherently asymmetric chromophore, and (b) the inherently symmetric chromophore which is asymmetrically perturbed. The former type includes such compounds as hexahelicene and certain substituted biphenyls and allenes in which the chromophore itself is asymmetric. These structures are seldom encountered in biological systems. By far, the large majority of chromophores are symmetrical and can only become optically active when perturbed by an asymmetric center or locus. When an asymmetric center induces optical activity in a chromophore which is part of the same molecule, then that molecule is said to be intrinsically optically active. Intrinsic optical activity can also occur if the chromophore has an asymmetric arrangement in space, e.g. a carbonyl group

PAGE 59

43 in the a-helical regions of proteins. Extrinsic optical activity may be observed when a symmetric molecule binds to a macromolecule. Such optical activity usually results from an interaction between the ligand and an asymmetric binding site, but may also occur when binding results in an asymmetric spatial arrangement e.g. the helical arrangement of the basic dye molecules bound to poly-L-glutamic acid (96). Since the extrinsic optical activity is induced in the chromophore by its environment, rather than being inherent (as in the first type), the magnitude of the associated Cotton effect is often considerably smaller than in the first type of chromophore. The sign and magnitude of an induced Cotton effect (extrinsic or intrinsic) depends upon the spatial relationship between the asymmetric center and the perturbed chromophore (97). The space around a chromophore is divided into regions of positive and negative contribution to a CD signal according to well-defined symmetry rules. A given asymmetric center may therefore generate either a positive or a negative CD signal, depending upon its spatial relationship to the perturbed chromophore. The CD signal for each transition reflects the extent to which the transition contributes to the phenomenon of optical activity. As a rule, all CD bands corresponding to the same group of transitions should have the same sign upon alterations induced by effects changing the molecular symmetry. The magnitude of a CD signal will increase as the distance between the asymmetric center and the perturbed chromophore decreases (97). Circular dichroism may be expressed as the differential dichroic absorption, Ae or as molar ellipticity [0] (deg. cm^ decimole ') calculated from the formula

PAGE 60

44 [e] = exioo//c (2-43) where 9 is observed ellipticity (deg.), / cell pathlength (cm), and c concentration (moles/liter). Molar ellipticity and differential dichroic absorption are related by the expression [9] = 3300 Ae (2-44) Circular dichroism is particularly important in the study of biological systems and macromolecules. It is used to yield information about a range of properties including: (a) the structure of macromolecules, their secondary and tertiary structural content, and the degree of mobility of certain components; (b) structural transitions, e.g. helix-coil transitions and the processes of unfolding and refolding of native and denatured structures; (c) conformational changes of macromolecules induced by binding small molecules, e.g. substrates, ligands, coenzymes, effectors; (d) symmetry properties of small chromophoric molecules in which optical activity is induced upon binding to macromolecules. CD bands of proteins occur in two spectral regions. The far-UV or amide region (170-250 nm) is dominated by contributions of the peptide bonds, whereas CD bands in the near-UV region (250-300 nm) originate from the aromatic amino acids. In addition, disulfide bonds give rise to several CD bands. The two spectral regions give different kinds of information about protein structure. The CD in the amide region reports on the backbone (i.e. the secondary) structure of a protein and is used to characterize the secondary structure and changes therein, in particular the a-helix. CD bands in the near-

PAGE 61

45 UV region are observed when, in a folded protein, aromatic side chains are immobilized in an asymmetric environment. The near-UV CD spectrum represents a highly sensitive criterion for the native state of a protein, thus, it can be used as a fingerprint of the correctly folded conformation. Examples of the conformational changes of macromolecules induced by binding small molecules, e.g. substrates, ligands, coenzymes, effectors have been observed. Trynda-Lamiesz et al. (98) found that the binding of adriamycin to HSA lowers the helicity of the native protein of ca. 15% by means of the changes in the far-UV region of CD spectra. Xu et al. (99) found that the a-helix of human erythrocyte membrane and BSA increased after the addition of croton alkaloids (CA) (antitumor drugs) by measuring CD between 200-240 nm. Another antitumor drug cisplatin was also found effective in altering membrane a-helix content. These results combined with other experimental results indicate that the conformation change in membrane protein may account for the pharmacological effect of CA and cisplatin. It has been found that many small symmetric molecules become optically active on binding to proteins and other macromolecules. In a majority of these examples optical activity results not from a special spatial arrangement of the ligand but from a perturbation of the ligand chromophore by an asymmetric locus at the binding site. Since extrinsic rotational activity (CD signal) reflects the three-dimensional characteristics of specific binding sites on a macromolecule, it offers an experimental means of exploring such sites. Examples of small symmetric molecules become optically active on binding to proteins and other macromolecules have been observed including those of the complexes of phenylbutazone and its analogues with HSA (100), benzodiazepines with HSA (101)

PAGE 62

46 and with BSA (102). Chignell (103) found that the binding of flufenamic acid to porcine, equine, and bovine serum albumins generated biphasic CD signals similar to those just described for human serum albumin. This suggests that these proteins have similar ligand binding sites. In contrast, the binding of flufenamic acid to canine or ovine serum albumins generated a single positive CD signal, indicating that the ligand binding sites on these proteins are different from those on human serum albumin. More recently, Fleury et al. (104) found that pronounced differences in the interactions of monomeric (lactone and carboxylate) and the J-type self-aggregated form of camptothecin (CTP), an inhibitor of DNA topoisomerase I, with HSA and BSA were observed in CD spectroscopy. This can provide the possibility of identifying individual amino acid residues which play a key role in CTP/HSA interactions.

PAGE 63

CHAPTER 3 EXPERIMENTAL Materials 2-Naphthol-8-sulfonic acid potassium salt (2-naphthol-8-sulfonate) was obtained from TCI America, Portland, Oregon, USA. Bovine serum albumin (Lot No. 16H9310; 126H9307) as essentially fatty acid free (below 0.005%), prepared from fraction V, and human serum albumin (Lot No. 24H9314; 46H9319) supplied as 1 x crystallized and lyophilized, prepared from fraction V, were purchased from Sigma Chemical Company, St. Louis, Missouri, USA. Sulfiiric acid (certified A. C. S.), perchloric acid (69-72%, TraceMetal grade), sodium hydroxide 2N solution (certified), methanol (HPLC grade, UV cutoff 205 nm), ethanol (HPLC grade, UV cutoff 205 nm), tertiary-butanol (certified), acetonitrile (HPLC grade, UV cutoff 190 nm), dimethyl sulfoxide (certified, spectranalyzed UV cutoff 262 nm), 1 ,4-dioxane (99 Mol % pure, certified), glycerin (certified, spectranalyzed), pyridine (certified A.C.S.), silver nitrate (laboratory grade), potassium iodide (U.S.P.), potassium phosphate primary standard monobasic (certified), sodium phosphate dibasic (certified A.C.S), sodium thiosulfate pentahydrate (certified A.C.S.), were obtained from Fisher Scientific, Fair Lawn, New Jersey, USA. N,N'dimethylpiperazine (98%), ethanolamine (99%), tris(hydroxymethyl)-aminomethane (99.9+%), phosphoric acid (A.C.S.) were obtained from Aldrich Chemical Company, 47

PAGE 64

48 Inc., Milwaukee, Wisconsin, USA. Salicylic acid (U.S.P.), sodium phosphate tribasic (analytic reagent grade) were obtained from Mallinckrodt Chemical Works, St. Louis, Missouri, USA. l,4-Diazabicyclo-(2.2.2)octane, 2-bromoethylamine hydrobromide were obtained from Eastman Kodak Company, Rochester, New York, USA. 2-Propanol (distilled in glass, UV cutoff 204 nm) was obtained from Burdick & Jackson Laboratories Inc., Muskegon, Michigan, USA. Ibuprofen was obtained from Research Laboratories of the Upjohn Company, Kalamazoo, Michigan, USA. Diazepam was obtained from Hoffmann-La Roche Inc., Nutley, New Jersey, USA. Phenylbutazone was received as a gift. All chemicals were used as received with no fiarther purification. All studies except the pA^ and pK* titrations were conducted in deionized water buffered to desired pH with phosphate. The decision to use a phosphate buffer in this project was based upon the resuhs of an investigation of the extent of complex formation between BSA and several common buffer ions by Klotz and Urquhart (105). Phthalate and veronal were found to be particularly effective in displacing bound ions from their protein complexes. Minimal effects were shown by glycine and phosphate while acetate, citrate and bicarbonate were intermediate in effectiveness. Pronounced competitive effects in binding were observed with nonbuffering anions such as nitrate and chloride. The protein solutions were prepared in phosphate buffer before experiment. Protein concentration was determined assuming a molecular weight of 69,000. The 2-naphthol-8-sulfonate concentration of the solutions used in the studies were chosen such that the absorbance at the isosbestic point, which was used as the excitation wavelength in the fluorescence measurements, would be less than 0.02 thereby reducing the probability of nonlinear fluorescence.

PAGE 65

49 Instrumental Steady state fluorescence measurements were made on a Perkin-Elmer LS-5 Fluorescence spectrophotometer. UV-visible measurements were made employing either a Shimadzu UV-2501PC UV-VIS recording spectrophotometer or Shimadzu UV160U UV-VIS recording spectrophotometer. CD measurements were made using JASCO J500C spectropolarimeter. The pH measurements were made with a Fisher Scientific Accumet 950 pH/ion Meter. Fisher Scientific Vortex Genie Mixer was used in this study. Socorex positive displacement micropipetter (1-5 ^1), Eppendorf adjustable micropipetter (2-10|il), Gilson adjustable pipetter (l-200|il), Eppendorf adjustable pipetter (100-1000 ^1), and Oxford adjustable pipetter (1000-5000 nl) were used in this study. Methods Ground and Excited State Ionization Constants Determinations (Titration Procedure) Two series of solutions whose pH are below 3 were prepared with reagent grade sulfuric acid and tracemetal grade perchloric acid diluted with water. The exact molarity of the concentrated acid was determined by titration with standardized sodium hydroxide solution. The corrected Hammett acidity scale (106) was used to calibrate the concentrated sulfuric acid and perchloric acid. Solutions with various ionic strength in the pH range 3-11 were phosphate buffers. Solutions whose pH are higher than 1 1 were made by sodium hydroxide solution and the pH values were assigned according to

PAGE 66

50 "Reagent chemicals and standards" (107). Deionized water was used throughout. Each sulfuric acid or perchloric acid solution, buffer solution or sodium hydroxide solution in a 5 ml volumetric flask was injected with 50 ^il of 1.0 x 10"^ M stock solution of 2naphthol-8-sulfonate in deionized water immediately prior to the measurements to minimize any possible decomposition errors. After each injection a Fisher Scientific Vortex Genie Mixer was used to mix the solution before reading. Values for the pfCs and pA^*'s were then obtained from the inflection points in the UV absorbance vs. pH and fluorescence intensity vs. pH plots, respectively. All UV and fluorometric determinations were performed at least three times, and the reported results are the mean of the data. Solvent Studies As discussed in the chapters 1 and 2 of this dissertation, studies of 2-naphthol-8sulfonate in a variety of solvents have supported the concepts that (a) except for very strong acids (pK < 0), no dissociation will take place imless water molecules are in the immediate vicinity to act as proton acceptors; (b) observed changes in the fluorescence emission maximum are related to the polarity of the environment around the fluorescent species. Both absorption and fluorescence spectra were obtained for 2-naphthol-8sulfonate and/or its conjugate base in the following solvents (if the solubility was sufficiently high): water or buffer solutions, methanol, ethanol (neat, 75% and 50%), 2propanol, tert-butanol, acetonitrile, dimethyl sulfoxide, 1,4-dioxane (neat, 75% and 50%) and glycerin. The purpose of this study was to evaluate the relationship between solvent polarity or hydrogen bonding strength and the fluorescence properties of 2-naphthol-8-

PAGE 67

51 sulfonate and then to relate this to the behavior of the compound in the presence of albumin. For these solvent studies, a concentration of 2-naphthol-8-sulfonate of 1 x 10"^ M and 1 X 10"' M was used for the measurements of absorption and fluorescence spectra, respectively. Qualitative Spectrophotometric Examination of 2-Naphthol-8-Sulfonate-Albumin Complex Interaction This initial experiment was conducted to evaluate qualitatively in terms of spectral changes following the addition of an albumin solution to 2-naphthol-8-sulfonate dissolved in buffer solutions. At first, both UV and fluorescence spectra of equimolar solution (1 X 10"' M) of albumin and 2-naphthol-8-sulfonate were compared to that of a I X 10"' M solution of 2-naphthol-8-sulfonate alone at the same pH. Gross differences between the spectra with and without albumin, specifically fluorescence intensity and /or UV absorbance changes and peak shifting, were noted. Since more than one isosbestic point in the UV spectrum was present, the excitation wavelength which resulted in the least fluorescence of albumin alone was chosen in the fluorescence study. After the initial experiment, the fluorescence emission spectra of three sets of solutions containing different molar ratio of 2-naphthol-8-sulfonate to albumin, i.e., 1:1,1:5 and 1:10, were recorded. The fluorescence spectra for the latter two sets consisting different molar ratios of 2-naphthol-8-sulfonate to albumin were found to be essentially identical and did not change with increased concentration of albumin. Therefore, molar ratio of 2-naphthol-8sulfonate to albumin in the complex of 1 :5 was chosen throughout this project, ensuring

PAGE 68

52 that the intensities at monitored wavelengths are independent of the 2-naphthol-8sulfonate/albumin ratio. The absorption and fluorescence spectra for albumin-bound 2-naphthol-8sulfonate at pH range 3-10 were recorded. Two different lots of albumin from each species were examined by recording the fluorescence emission spectrum. Fluorometric Determination of Compound-Albumin Complex StoichiometryJob's Method Job's method was employed to determine 2-naphthol-8-sulfonate-albumin complex stoichiometry. For the Job study, a series of solutions ranging from 2-naphthol8-sulfonate alone (1 x 10"' M) to albumin alone (1 x 10"' M) was prepared in buffer solutions. The composition of the intermediate 2-naphthol-8-sulfonate-albumin solutions was varied in such a marmer that the total 2-naphthol-8-sulfonate plus albumin concentration in each solution remained constant (1 x 10"' M). The fluorescence intensities were monitored for each solution using an excitation wavelength which corresponded to an isosbestic point resulted in the least fluorescence of albumin alone. The Job data were plotted in the usual manner, i.e., the differences between the observed fluorescence intensity of 2-naphthol-8-sulfonate-albumin solutions and fluorescence intensity at the wavelength of maximum emission of free 2-naphthol-8-sulfonate against the mole fraction of 2-naphthol-8-sulfonate, and from this plot the stoichiometry of 2naphthol-8-sulfonate-albumin complex could be determined. All fluorometric determinations were performed at least three times, and the reported results are the mean of the data.

PAGE 69

53 2-Naphthol-8-Sulfonate-Albumin Titration Experiments-Determination of Binding Constants and Number of Binding Sites The fluorometric titrations were carried out as follows: 2.0 ml of albumin solution of appropriate concentration in a 1 cm quartz cell were titrated by successive additions of ^1 volumes of a 1 x 10"^ M solution of 2-naphthol-8-sulfonate. Titration by 2-naphthol-8sulfonate with increasing albumin concentration was performed until a point was reached at which two successive titrations showed identical or very similar increases in fluorescence intensity throughout the titration. This was necessary to ensure that the proper albumin concentration level to fully bind 2-naphthol-8-sulfonate had been reached. Then, 2.0 ml of phosphate buffer solution in a 1 cm quartz cell were titrated by successive additions of ^1 volumes of a 1 x 1 0'^ M solution of 2-naphthol-8-sulfonate to obtain the fluorescence of free 2-naphthol-8-sulfonate. For these titration experiments, the same excitation wavelength was used as in the Job study, and the fluorescence intensity was monitored at the emission maximum of the 2-naphthol-8-sulfonate-albumin complex. After each addition of 2-naphthol-8-sulfonate solution, a Fisher Scientific Vortex Genie Mixer was used to mix the solution before reading. According to the relationship F = (t)Io[2.3A (2.3A)V2! + (2.3A)V3! (2.3A)V4! + •••] (3-1) where F is the relative fluorescence intensity, (j) is the quantum yield of the emitting species, !„ is the intensity of the exciting radiation, and A is the absorbance, the linearity of fluorescence intensity with the concentration of the emitting species can only be taken for granted at very low absorbance (< 0.02) at the excitation wavelength, in which case the higher power terms in equation (3-1) become negligible by comparison with the first

PAGE 70

54 term. For example, when A > 0.02, the deviation from linearity is greater than 2%. At absorbances of 0.02 < A < 0.15, the second term in equation (3-1) results in 1 15 x A% deviation from linearity at any point in the fluorometric titration. At higher absorbances (A > 0.15), the third term in equation (3-1) generally also becomes large enough to have to be considered. Thus, determination of the absorbance of free 2-naphthol-8-sulfonate and albumin solution after successive addition of 2-naphthol-8-sulfonate was carried out as follows: 2.0 ml of albumin solution of appropriate concentration in a 1 cm UV quartz cell were titrated by successive additions of |xl volumes of a 1 x 1 0"' M solution of 2naphthol-8-sulfonate. The absorbance was monitored at the wavelength of excitation after each addition of 2-naphthol-8-sulfonate stock solution. Same procedure was used to determine the absorbance of free 2-naphthol-8-sulfonate solution. After each addition of 2-naphthol-8-sulfonate solution, a Fisher Scientific Vortex Genie Mixer was used to mix the solution before reading. All UV and fluorometric determinations were performed at least three times, and the reported results are the mean of the data. Medium Effects on the Affinitv of the Site in Albumin for the 2-Naphthol-8-Sulfonate The effect of pH on the binding to albumin was measured as follows. The albumin and 2-naphthol-8-sulfonate ( at a constant ratio and concentration) were brought to different pH values with constant ionic strength in the range 5-9. Then the fluorescence intensities of the bound 2-naphthol-8-sulfonate and its conjugate base were measured. With varying ionic strength and constant pH, a similar procedure was used for the measurement of the effect of ionic strength on the binding to albumin.

PAGE 71

55 For these experiments, the same excitation wavelength was used as in the Job study, and the fluorescence intensity was monitored at the emission maximum of the 2naphthol-8-sulfonate-albumin complex. All UV and fluorometric determinations were performed at least three times, and the reported results are the mean of the data. Displacement of 2-Naphthol-8-Sulfonate from Albumin The displacement of 2-naphthol-8-sulfonate from albumin binding site by typical site I and site II binding maker such as phenylbutazone, salicylic acid, and diazepam, ibuprofen, respectively, was carried out as follows: 2.0 ml of the albumin solution of concentration 1 x 10'^ M dissolved in pH 5.0 phosphate buffer, placed in a 1 cm quartz cell, were injected with 2 ^l of 1 x 10'^ M stock solution of 2-naphthoI-8-sulfonate. The molar ratio of 2-naphthol-8-sulfonate to albumin was kept at 1 : 1 in order to keep the nonspecific binding of 2-naphthoI-8-sulfonate to a minimum. The fluorescence intensity was monitored at the emission maximum of 2-naphthol-8-sulfonate-albumin complex. Then, above solution was titrated by successive additions of 5 [il of appropriate concentration stock solution of site binding maker dissolved in methanol. After each addition of site binding maker, the fluorescence intensity was monitored. The titration was performed until a point was reached at which the molar ratio of site binding maker to albumin was 5:1. Fluorescence values were corrected for dilution resulting from the addition of the site binding maker stock solution. The measured fluorescence value was expressed as a percentage of the initial fluorescence. In order to ensure that the site binding maker emitted no fluorescence at the wavelengths used for the displacement experiments, the fluorescence intensities of all site binding makers used in the displacement experiments

PAGE 72

56 including typical site I binding makers such as warfarin and phenyprocoumon were monitored after addition to albumin solution. It was confirmed that all site binding makers except warfarin and phenyprocoumon emitted no or negligible fluorescence at the wavelengths used for the displacement experiments. For the displacement experiments, the same excitation wavelength was used as in the Job study. After each addition of 2naphthol-8-sulfonate solution, a Fisher Scientific Vortex Genie Mixer was used to mix the solution before reading. All fluorometric determinations were performed at least three times, and the reported results are the mean of the data. Fluorescence Quenching of Albumin Bound 2-Naphthol-8-Sulfonate Fluorescence quenching experiment was carried out as follows: (1) 2.0 ml of the albumin solution of concentration either 5 x 10 ' M or 1 x 10'' M dissolved in either pH 5.0 or pH 9.0 phosphate buffer, placed in a 1 cm quartz cell, were injected with 2 ^1 of 1 x 10"^ M stock solution of 2-naphthol-8-sulfonate. The fluorescence emission spectrum was recorded. Then, above solution was titrated by successive additions of 5 ^1 of appropriate concentration stock solution of pyridine dissolved in either pH 5.0 or pH 9.0 phosphate buffer. After each addition of quencher the fluorescence emission spectrum was recorded. Same procedures were used for recording the fluorescence emission spectra of free 2-naphthol-8-sulfonate in buffer solutions in the absence and presence of pyridine. In order to determine the quenching effect of pyridine on albumin, the fluorescence emission spectra of albumin solutions in the absence and presence of pyridine were recorded. Same procedures were also used for recording the fluorescence emission spectra of 2-naphthol-8-sulfonate in ethanol (neat, 75%(v/v),

PAGE 73

57 50%(v/v)), and 1,4-dioxane (neat, 75%(v/v), 50%(v/v)). The purpose of this study was to evaluate the relationship between solvent polarity or hydrogen bonding strength and the fluorescence quenching properties of 2-naphthol-8-sulfonate and then to relate this to the behavior of the compound in the presence of albumin. In order to evaluate the hydrogen bonding in the ground state, the absorption spectrum of 2-naphthol-8-sulfonate in the presence of pyridine at pH 5.0 was also recorded. (2) 2.0 ml of the albumin solution of concentration 5 x 10'' M dissolved in either pH 5.0 or pH 9.0 phosphate buffer, placed in a 1 cm quartz cell, were injected with 2 |al of 1 x 10"^ M stock solution of 2-naphthol-8-sulfonate. The fluorescence intensity was monitored at the emission maximum of 2-naphthol-8-sulfonate-albumin complex. Then, above solution was titrated by successive additions of 5 [i\ of appropriate concentration stock solution of pyridine. After each addition of quencher, the fluorescence intensity was monitored. Same procedures were used for monitoring the fluorescence intensity of free 2-naphthoI-8-sulfonate in buffer solutions, and albumin solutions in the absence and presence of pyridine. For the fluorescence quenching experiments, the same excitation wavelength was used as in the Job study. After each addition of 2-naphthol-8-sulfonate solution, a Fisher Scientific Vortex Genie Mixer was used to mix the solution before reading. Fluorescence values were corrected for dilution resulting from the addition of the quencher stock solution. All fluorometric determinations were performed at least three times, and the reported results are the mean of the data.

PAGE 74

58 Circular Dichroism Experiments As discussed in the chapter 2 to this dissertation, studies of the CD spectra of Hgand-protein complex can (a) probe changes in the conformation of the macromolecule, and (b) probe macromolecule interaction with small molecules, especially achiral ones whose induced CD is due solely to their interaction with the macromolecule. Since the wavelength of the extremum of a circular dichroism curve almost coincides with the position of the maximum in the absorption spectroscopy. Proper choices of protein concentration, pathlength, and solvent are essential for obtaining good CD spectra and for avoiding artifacts. Since the CD instrument measures very small differences in the transmitted light, the total absorbance of the sample in the desired spectral region is of utmost importance. A good signal-to-noise ratio is achieved when the absorbance is around 1 .0. Therefore, the CD experiment was carried out as follows: (1) absorption spectra of the samples and buffer at pH values of 5.0 and 9.0 in the desired spectral region were recorded prior to the CD experiment to determine the protein concentration and the ligand to protein ratio, and to select the optimal conditions for the CD measurement at different spectral region. (2) the albumin, and albumin and 2-naphthol-8-sulfonate ( at a constant ratio and concentration, more specifically, albumin concentration is 4.0 x 10"' M, 2.0 x 10'^ M, and 7.5 X 10-' M in the wavelength region 200-250 nm, 250-300 nm, and 300-400/450 nm, respectively, and molar ratio of 2-naphthol-8-sulfonate to albumin is 1:1, and 3:1 for BSA and HSA in the region below 300 nm and 2.7:1 above 300nm, respectively) were brought to pH value either 5.0 or 9.0. Then, the baseline of the buffer and the CD spectra of

PAGE 75

59 albumin and the 2-naphthol-8-sulfonate-albumin complex were recorded successively under identical instrumental settings in the same CD cell (i. e.l cm circular quartz cell). Results are expressed as molar ellipticity calculated with reference to the albumin concentration, using a molecular weight of 66,500. Each CD spectrum reported is the average of three scans for short or long wavelength.

PAGE 76

CHAPTER 4 RESULTS Photophsicochemical Properties of 2-Naphthol-8-Sulfonate Absorption Spectra of 2-Naphthol-8-Sulfonate Figure 4-1 shows the absorption spectra of 2-naphthol-8-sulfonate in the Hammett acidity pH range -1.97 14.0. The respective maxima are 333 nm for monoanion (protonated form) and 360.4 nm for dianion (deprotonated form) in the wavelength range 300 nm 400 nm, and two isosbestic points in the same wavelength range were clearly shown at 309 imi and 337 nm, which indicated a ground state acid base reaction. Fluorescence Excitation Spectra of 2-Naphthol-8-Sulfonate The fluorescence excitation spectra of 2-naphthol-8-sulfonate are depicted in Figure 4-2. The excitation spectral properties are similar to those shown in the absorption spectra. Fluorescence Emission Spectra of 2-Naphthol-8-Sulfonate The recorded fluorescence emission spectra of 2-naphthol-8-sulfonate are shown in Figure 4-3. The respective maxima are 379 nm and 456 nm. An isoemissive point indicates an excited state acid base reaction. At Ph of about 5, the hydroxy 1 group in the ground state is fully protonated, but not so in the lowest excited singlet state (see below). The excited molecules dissociate and majority of the emission is at the wavelength of the 60

PAGE 77

61 excited 2-naphthol-8-sulfonate dianion (456 nm). The dissociation can be prevented if 2naphthol-8-sulfonate is dissolved in acid solution, pH < pK* , such as 2.88 M perchloric acid. Under such conditions we observe the emission of 2-naphthol-8-sulfonate monoanion with maximum at 379 nm. This reveals that 2-naphthol-8-sulfonate in aqueous solutions undergoes excited state proton transfer from the hydroxyl group to a water molecule, which is manifested by appearance of the fluorescence band of the protonated form accompanied by fluorescence emission band of the deprotonated form while only the protonated form is directly excited. Determination of Ground and Excited State Ionization Constants Titration experiments in the Hammett acidity pH range -1 .97 14.0 for 2naphthol-8-sulfonate were performed, and the absorbances and the relative fluorescence intensities of 2-naphthol-8-sulfonate as a function of pH are shown in Figures 4-4 and 4-5. The values of pATand pK* at ionic strength 0.1 were determined graphically to be 9.7 and 1.6, respectively. These titration results of 2-naphthol-8-sulfonate indicated that 2-naphthol-8-sulfonate monoanion is the predominant species present in the ground state and the dianion is the predominant species present in the excited state at pH < pK, say, 7.4. Steady-State Kinetics of Excited State Proton Transfer Reaction of 2-Naphthol-8Sulfonate The rate constants for excited state proton transfer reaction (1-1) were determined by two methods:

PAGE 78

62 ki2 ^ OOH* + H2O ^y.^^ 00-* + H3O+ k(r + nr) hv\ hv2 k'(r + nr) OOH + H2O ^ 00+ H3O+ (1-1) (a) from the spectrofluorimetric titration at low hydrogen ion concentration (3 < pH < 9), (b) from the spectrofluorimetric titration in acidic pH range. In method (a), the following equation A' (|)' (—) (") J. ' -' ^ J. ' const
PAGE 79

63 Using linear dependence of the ratio of relative fluorescence quantum yields ^Ij^ on F[ H3O" ], from the ordinate intercept the value of 1/ k,2To and from the slope /0 the value of k2,To' / kj^Xo were obtained. Then the excited state proton transfer rate constants k,2 and k2, were calculated using the lifetimes Tq and To', 7.1 x 10'' s and 13.3 x 10'' s, respectively, and the results are presented in Table 4-1. The value of pA:* of 2-naphthol-8-sulfonate was calculated from the ratio of excited state proton transfer rate constants k,2 and k2i and the result is presented along with the graphically determined pK* value in Table 4-1 . Table 4-1: Photophysicochemical properties of 2-naphthol-8-sulfonate. k|2 k2, pK* 2.7 X 10' ± 0.38 s-' Method (a) 9.2 X 10'° ±0.75 M-' s' 1.53 (calculated) 6.2 X 10" ±0.13 S-' Method (b) 2.1 X 10" ±0.75 M-' S-' 1.6 (graphically determined) Solvent Studies Binding of chromophores to protein affects their electronic transitions and hence generates difference absorption and fluorescence spectra. Since the chromophores are generally surrounded by a less polar environment upon binding to protein, similar effects are expected when the chromophores are transferred from a purely aqueous solvent to an organic solvent. Thus we first investigated solvent perturbation on absorption and fluorescence spectra of 2-naphthol-8-sulfonate.

PAGE 80

64 For the sake of comparison, the absorption spectra of 2-naphthol-8-sulfonate monoanion and its conjugate base dianion dissolved in aqueous and organic solvents are shown in Figures 4-6 and 4-7. The maximal wavelengths of 2-naphthol-8-sulfonate in various solvents are presented along with solvent dielectric constant values, where available, in Table 4-2. The peaks of 2-naphthol-8-sulfonate in organic solvents corresponded to red shifts of the original bands with respect to those in aqueous solution. The fluorescence emission spectra of 2-naphthol-8-sulfonate monoanion and its conjugate base dianion in aqueous and organic solvents are shown in Figures 4-8 and 4-9, respectively. The maximal wavelengths of 2-naphthol-8-sulfonate in various solvents are presented in Table 4-2. The fluorescence peaks of 2-naphthol-8-sulfonate, especially monoanion, in organic solvents shifted to shorter wavelengths compared with that in aqueous solution. As shown in Figure 4-10, the fluorescence emission spectra of 2-naphthol-8sulfonate in ethanol-water mixtures strongly depend on the concentration of water. A similar set of the emission spectra was also found for 2-naphthol-8-sulfonate in 1,4dioxane-water mixtures as depicted in Figure 4-1 1 . In pure organic solvents, only a single emission band which is attributed to the protonated form (monoanion) of the excited 2naphthol-8-sulfonate is observed at ~ 368 nm. However, upon addition of water at constant solute concentration, the fluorescence intensity of the protonated band decreases gradually and a new band appears at ~ 455 nm. The new band is attributed to the conjugate base of the excited 2-naphthol-8-sulfonate dianion. This behavior is accounted for by deprotonation of the excited 2-naphthol-8-sulfonate since the hydroxyl group of 2naphthol-8-sulfonate behaves like a stronger acid in the excited state than in the ground

PAGE 81

65 state. The shift of the isoemissive point to the longer wavelength region where water concentration is 50% and 25% by volume or higher in the 1 ,4-dioxanewater and ethanolTable 4-2: Solvent studies absorption and fluorescence data. Solvent Dielectric Constant Wavelength of Absorption Maxima (nm) Wavelength of Fluorescence Maxima (nm) Monoanion TA * * Dianion Monoanion Dianion Water no '5 o 333.0 N* 379 4j6 Perchloric Acid (2.88 M) NA** 1)1)11 N 379 XT Sodium Hydroxide (1 N) NA N 360.4 XT 4i6 Methanol 32.66 336.0 361.6 369.5 455 Ethanol 24.3 336.0 364.0 369.5 455 75% (VA^) Ethanol 32.8 <8 <43.4 ** 371 Broad A AC\ ~ 449 50% (VA^) Ethanol 43.4< e < 55.0 372 456 1,4-Dioxane 336.6 -> /r\ o 360.8 367 A A C 445 75% (VA^) 1 A TA * 1 ,4-Dioxane 12.1<8<27.5 370 N 50% (VA^) 1 ,4-Dioxane 27.5
PAGE 82

66 water mixtures, respectively, might be caused by the change in solvent structure in the organic-water mixtures. These results also reveal that no dissociation will take place unless water molecules are in the immediate vicinity to act as proton acceptors. The observations of such an isoemissive point in both aqueous solution and organic-water mixtures demonstrates the presence of only protonated (monoanion) and deprotonated (dianion) forms of 2-naphthol-8-sulfonate in the excited state. Binding Studies Qualitative Spectrophotometric Examination of 2-Naphthol-8-Sulfon ateAlbumin Complex Interaction Upon ligation to albumin, the spectroscopic properties of 2-naphthol-8-sulfonate were changed, which reflect the perturbation caused by changing the environment of 2naphthol-8-sulfonate from aqueous solution to protein. The absorption and fluorescence emission spectra for 2-naphthol-8-sulfonatealbumin complex in the pH range 3-10 at various ionic strengths were recorded and summarized in Figures 4-12 4-37. The absorption spectra of 2-naphthol-8-sulfonate bound to albumin were obtained as a difference spectra (2-naphthol-8-sulfonate-albumin versus albumin). The respective maxima for the albumin bound 2-naphthol-8-sulfonate complex are presented in Table 4-2. As can be seen, the absorption spectra of BSA-2naphthol-8-sulfonate complex show maxima at 336 nm and 363.5 nm, and the maximal absorption bands for HSA-2-naphthol-8-sulfonate complex show little change compared to 2-naphthol-8-sulfonate alone in buffer. A peak was observed at 334 rmi in the difference absorption spectra of 2-naphthol-8-sulfonate produced by HSA at molar ratio

PAGE 83

67 of 2-naphthol-8-sulfonate to HSA 1:5. This maximum was positive and corresponded to a small red shift of the original absorption peak. The fluorescence of BSAand HSAbound 2-naphthol-8-sulfonate still consists of two bands, typical for the dianion (445 nm and 453 nm, for BSAand HSAbound 2naphthol-8-sulfonate, respectively,) and monoanion (376 imi), but shifted to shorted wavelengths. BSA-2-naphthol-8-sulfonate complex are much more intensely fluorescent but varying with pH while HSA-2-naphthol-8-sulfonate complex shows less profound increase in intensity but constant with pH. The results show that the absorption bands and fluorescence bands of 2-naphthol8-sulfonate are perturbed to produce the difference spectral maxima upon its binding to albumins. Similarities between the albuminand organic solvent-generated difference spectra indicate that 2-naphthol-8-sulfonate binds to the binding site with less polar envirormient in the albumin molecule. In this system, the perturbant is the albumin molecule, or more precisely the less polar environment of the binding site. To ensure that the intensities at monitored wavelengths are independent of the 2naphthol-8-sulfonate / albumin ratio, the fluorescence emission spectra of three sets of solutions containing different molar ratio of 2-naphthol-8-sulfonate to albumin, i.e., 1:1, 1 :5 and 1:10, were recorded and are depicted in Figures 4-38 and 4-39 for BSAand HSA-bound 2-naphthol-8-sulfonate, respectively. The fluorescence spectra for the latter two sets consisting different molar ratios of 2-naphthol-8-sulfonate to albumin were found to be essentially identical and did not change with increased concentration of albumin.

PAGE 84

68 Two different lots of albumin from each species were examined by recording the fluorescence emission spectrum. And it was found that the emission peaks of albumin bound 2-naphthol-8-sulfonate are identical for the lots derived from the same species. These differences in the spectra of albumin bound 2-naphthol-8-sulfonate are due to species differences in the albumins tested in this study. Fluorometric Determination of 2-Naphthol-8-Sulfonate -Albumin Complex Stoichiometry The results of the Job study are presented in Table 4-3. Two of the Job plots for albumin bound 2-naphthol-8-sulfonate monoanion and dianion are shown in Figures 4-40 and 4-41. The rounded appearance of the maximum of the curve indicates that the complex formed is somewhat dissociated. Therefore the extrapolations of these Job curves can only be considered to approximate the complex stoichiometry. Extrapolation along the more linear portions of the curves monitored at wavelengths 445 nm and 453 nm gave a value of jc « 0.5 which corresponds to N = 1 for BSAand HSA-bound 2naphthol-8-sulfonate dianion, respectively. However, extrapolation along the more linear portions of the curves monitored at wavelength 376 nm which corresponds to albumin bound 2-naphthol-8-sulfonate monoanion gave different values of x for BSAand HSA2-naphthol-8-sulfonate under various pH and ionic strength conditions. Since all x values for HSA2-naphthol-8-sulfonate are around 0.75, which corresponds to a N value of 3, it is reasonably certain that each HSA molecule binds roughly three 2-naphthol-8-sulfonate monoanion molecules in the pH range 5-9 under the experimental conditions. For BSA-2naphthol-8-sulfonate, with increasing ionic strength from 0.001 to 0.3 in the pH range 5-9, the X values decrease from 0.67 to 0.5 corresponding to a N value from 2 to 1 . It is

PAGE 85

69 clear that the number of binding sites on BSA is smaller than that on HSA for 2-naphthol8-sulfonate. 2-Naphthol-8-Sulfonate-Albumin Titration Experiments-Determination of Binding Constants The fluorometric titrations of albumin with 2-naphthol-8-sulfonate are shown in Figures 4-42 and 4-43. For compounds like 2-naphthol-8-sulfonate, which show increased fluorescence intensity upon binding, the fraction of ligand bound, a, is usually calculated by using the following equation: a = (Fp F,)/(F, F,) (2-27) Table 4-3: Results of Job Study number of binding sites. Ionic pH5.0 pH7.4 pH 9.0 Strength Monoanion Dianion Monoanion Dianion Monoanion Dianion BSA 0.001 x = 0.66 x = 0.5 X = 0.69 x = 0.5 X = 0.69 x-0.5 N= 1.9 N1 N = 2.2 N= 1 N = 2.2 N= 1 0.01 X = 0.62 x = 0.5 x-0.6 x = 0.5 x-0.58 x = 0.5 N= 1.6 N= 1 N= 1.5 N= 1 N= 1.4 N= 1 0.1 x = 0.56 x = 0.5 x = 0.59 x = 0.5 X = 0.57 x = 0.5 N= 1.3 N= 1 N= 1.4 N= 1 N= 1.3 N= 1 0.3 x = 0.51 x = 0.5 x = 0.56 x = 0.5 x = 0.57 x = 0.5 N= 1 N= 1 N= 1.3 N= 1 N= 1.3 N= 1 HSA 0.001 X = 0.74 x = 0.5 x = 0.78 x = 0.5 x = 0.78 x = 0.5 N = 2.8 N1 N = 3.5 N= 1 N = 3.5 N= 1 0.1 X = 0.74 x = 0.5 X = 0.77 x = 0.5 X = 0.75 x-0.5 N = 2.8 N= 1 N = 3.3 N1 N = 3 N= 1 In order for the above equation to yield good values of a, the fluorescence intensity of the bound ligand must be linear function of its concentration. As mentioned in chapter 3, fluorescence intensity is related to absorbance in a power series:

PAGE 86

70 F = (t)Io[2.3A (2.3A)V2! + (2.3A)V3! (2.3A)V4! + -] (3-1) Only when the absorbance at the exciting wavelength is very low (< 0.02) does fluorescence intensity show a direct linear relationship to ligand concentration. When the absorbance lies between 0.02 and 0.15, the second term in the equation (3-1) becomes significant, and each point in the titration curve must be corrected for this value. At higher absorbances, the third term in the equation (3-1) may have to be considered. The molar absorptivity for the albumin-2-naphthol-8-suIfonate complex at the excitation wavelength of 337 nm was determined to be 2.2 x 10^ liter mole"' (2-naphthol8-sulfonate) cm"'. Therefore, a 2-naphthol-8-sulfonate concentration greater than 9x10"* M would result in an absorbance value greater than 0.02, and a deviation from linearity would result unless the second term in equation (3-1) is considered. Because the concentration of 2-naphthol-8-sulfonate in the albumin titration experiments exceeded this 9 X 10"* M limit, it was inappropriate to use equation (2-27) to calculate the concentration of bound 2-naphthol-8-sulfonate. Consequently, the observed fluorescence intensity at each point during the titrations with high protein concentration was corrected for the second term in equation (4-3). Curve (•) in Figures 4-42 and 4-43 was plotted after applying the corrected factor, a straight line plot was obtained. The fact that the corrected values yielded a straight line verified that the deviation from linearity was indeed due to the absorbance effect. The curve (A) in Figure 4-42 approached a plateau indicating the saturation of BSA binding sites by 2-naphthol-8-sulfonate.

PAGE 87

71 In the HSA titration experiment, the high HSA concentration titrations were initially performed using HSA concentrations of 1 x lO"" M and 1 x 10"' M with a 2naphthol-8-sulfonate stock solution concentration of 1x10"^ M. The results from these experiments showed the maximum intensity was obtained in the 1 x 10"' M HSA titration. Lower intensity was obtained in the higher HSA concentration experiment (1 x 10"* M) due to nonlinearity of HSA fluorescence at this concentration level. Titrations were then performed at HSA concentrations of 1 x 10"' M and 2.5 x 10"^ M. In these albumin titration experiments, the maximum volume of ligand solution which could be added to the 2.0 ml of albumin without encountering an error due to sample volume change was placed at 30 ^1. Thus, in the 2.5 x lO'*" M HSA titration, the titration curve had not completely reached a plateau after the addition of 30 |il of the 2-naphthol-8-sulfonate solution. A common method of treating binding data makes use of the Scatchard equation (29), which was discussed in the background to this dissertation. A basic assumption of the Scatchard treatment is that the binding sites are independent, noninteracting and have the same binding affinity for the ligand. Such assumptions are not necessary when N = 1, and in that case, a Scatchard plot of r /[L] versus r should yield a straight line. However, as the number of binding sites increases, both the likelihood of the existence of different types of binding sites and of electrostatic interaction between the sites, even if they are similar, increases and the Scatchard-type plots often deviate from linearity. Participation of multiple classes of binding sites is easily apparent from the Scatchard plot since there is a marked concave curvature of the plot. In such case, extrapolation along the linear portions of the curve should yield a set of slopes and intercepts, which correspond to

PAGE 88

72 different classes of binding sites. Scatchard plots of r/[L] versus r for the BSAand HSA2-naphthol-8-sulfonate titration with the protein concentration of 2.5 x IQ-* M at pH 7.4 and ^ 0.1 are shown in Figures 4-44 and 4-45, respectively. Since the slope of a Scatchard plot corresponds to the negative k value and y-intercept corresponds to Nk, the value of N may be determined simply by dividing the y-intercept value by slope of the line (times -1). Scatchard analysis of the fluorescence data showed a non-linear curve, suggesting the presence of at least two classes of sites for the binding of 2-naphthol-8sulfonate to albumins. The best fitting values for the binding parameters obtained by the fluorescence method are shown in Table 4-4. Table 4-4: Resuhs of Scatchard plot binding constants and number of binding sites. Protein Concentration (M) pH 7.4 BSA 2.5 X IQ-^ k, = 3.9x 10', N, = 1.6 (r' = 0.9947) k2 = 2.5 X 10', = 2.2 (r' = 0.9972) HSA 2.5 X IQ-^ k, =2.3 X 10^N, = 1.4 (r' = 0.5877) kj = 1.2 X 10', = 6.2 (r" = 0.9665) Medium Effects on the Affinity of the Site in Albumin for the 2-Naphthol-8-Sulfonate The fraction of the bound ligand was measured by the fluorescence intensity of excited 2-naphthol-8-sulfonate monoanion and dianion. The fluorescence of excited 2naphthol-8-sulfonate monoanion increases upon binding, therefore any dissociation of the ligand from the site will lower the emission at 376 nm and increase the emission at 445 nm and 453 nm for BSA and HSA bound 2-naphthol-8-sulfonate, respectively. Thus dissociation will lower the ratio of emission ImJIaa^ and I,?/,/!^,, for BSA and HSA bound

PAGE 89

73 2-naphthol-8-sulfonate, respectively. The effects of pH and ionic strength on the fluorescence emission ratio of 2-naphthol-8-sulfonate in the presence of albumin were depicted in Figures 4-46 and 4-47. As seen in Figures 4-46 and 4-47, the fluorescence emission ratio of 2-naphthol-8-sulfonate in the presence of BSA but not HSA decreases with rising pH in the pH range 5-9 at a constant ionic strength, and the fluorescence emission ratio of 2-naphthol-8-sulfonate in the presence of BSA but not HSA increases with rising ionic strength at a fixed pH in the pH range 5-9. Displacement of 2-Naphthol-8-Sulfonate from Albumin To determine the binding sites for 2-naphthol-8-sulfonate on HSA and BSA, displacement studies were carried out using the typical site I binding markers for human albumin, namely, phenylbutazone and salicylic acid, and site II binding markers, diazepam and ibuprofen. The displacement results are shown in Figures 4-48 and 4-49 for HSAand BSA-bound 2-naphthol-8-sulfonate, respectively. A decrease in the fluorescence intensity of the albumin-2-naphthol-8-sulfonate complex can be interpreted as a displacement of 2-naphthol-8-sulfonate from its binding site by the added specific site binding marker, probably through a competitive mechanism. As can be seen in Figure 4-48, the fluorescence intensity of 2-naphthol-8-sulfonate bound to HSA was remarkably decreased by one of the site II binding markers, namely, ibuprofen, but not by diazepam, a specific site II binding marker and site I binding markers. On the other hand, it is clear from the Figure 4-49 that the fluorescence intensity of 2-naphthoI-8-sulfonate bound to BSA was more extensively decreased by site II binding markers but not by site 1 binding markers.

PAGE 90

74 Fluorescence Quenching of Albumin-Bound 2-Naphthol-8-S ulfonate Many parameters can be obtained by fluorescence spectroscopic methods to provide insights into the environment, structure, and dynamics of a fluorescent probe that is either covalently bound or liganded to a biological molecule. One important, commonly used method is the addition of a quenching agent to reduce the fluorescence of the probe. By comparing the quenching efficiency, the environment of the probe, and thus a specific region of the biomolecule, can be characterized in terms of solvent accessibility. Therefore, quenching studies were carried out using pyridine. The absorption spectra of pyridine and pyridine-containing solution at pH 5.0 are shown in Figure 4-50. Compared with that of 2-naphthol-8-sulfonate in pyridine-free solution, the changes in the absorption spectra upon addition of pyridine are very minor. Unlike the absorption spectrum, the emission bands of unbound 2-naphthol-8sulfonate observed in pH 5.0 and 9.0 buffer solutions decrease in intensity upon pyridine titration as depicted in Figures 4-5 1 and 4-52, the position and shape of the band was not changed. These results are consistent with those observed by Mataga (94). The effects of pyridine on the fluorescence emission of albumin-2-naphthol-8sulfonate complex were shown in Figures 4-53 4-60. At low molar ratio of albumin to 2-naphthol-8-sulfonate, upon addition of pyridine at both pH values, the intensities for albumin bound 2-naphthol-8-sulfonate monoanion and dianion decreased but to a lesser degree as compared to the unbound 2-naphthol-8-sulfonate. However, at high molar ratio of albumin to 2-naphthol-8-sulfonate, the emission bands of HSA-bound 2-naphthol-8sulfonate shows little change at pH 5, while at pH 9, the intensity of HSA-bound 2-

PAGE 91

75 naphthol-8-sulfonate dianion decreases with increasing concentration of pyridine. In all cases, the reduction in intensity of the HSA-bound 2-naphthol-8-sulfonate by pyridine is considerably less than that found for free 2-naphthol-8-sulfonate. For BSA-bound 2naphthol-8-sulfonate, the intensity of monoanion decreases with increasing concentration of pyridine at both pH values concomitant with increasing in intensity of dianion at pH 5. In no case did the intensity increase for the long-wavelength emission match the intensity lost from the monoanionic emission. It is reported by several groups (95, 108) that the effects of alkylamine on the fluorescence emission of hydroxyaromatics such as naphthol in organic solvent were dramatic. Either an increase in emission intensity or a new peak at the long-wavelength which corresponds to the naphtholate anion accompanying with the quenching of the neutral form was observed upon addition of alkylamine. In order to investigate the abnormal quenching behavior of BSA-bound 2-naphthol-8-sulfonate, the quenching experiments of unbound 2-naphthol-8-sulfonate in organic solvents and organic-aqueous mixtures were conducted and the results are depicted in Figures 4-61 4-66. It is clear from these figures that in all cases the intensities of 2-naphthol-8-sulfonate monoanion and dianion decrease with increasing concentration of pyridine. Therefore, the increase in intensity of BSA-bound 2-naphthol-8-sulfonate dianion might be caused by the alteration of structure or affinity of BSA to 2-naphthol-8-sulfonate upon addition of pyridine. The emission intensities of albumin bound and unbound 2-naphthol-8-sulfonate in pyridine-containing solutions relative to the intensities in pyridine-free solutions had been determined at a number of pyridine concentrations. The emission intensities of unliganded albumins in the absence and presence of pyridine had been also determined.

PAGE 92

76 Since the emission intensities of unliganded albumins at the monitored wavelengths are very low, the quenching effect of pyridine on albumins can only be regarded as an estimate. All results are shown in Figures 4-67 and 4-68. As depicted in these figures, the fluorescence intensities of unliganded albumins are slightly affected by the addition of pyridine, especially for unliganded BSA. Circular Dichroism Experiments As mentioned in chapter 2, the CD spectrum is very useful to study the biological system because it can provide information about structural transitions and conformational changes of macromolecules induced by binding small molecules. Thus, the binding of 2naphthol-8-sulfonate to albumin was quantitatively monitored by circular dichroism spectra at pH 5 and 9 in the wavelength range 200 nm 450 nm. The CD spectra of HSA and BSA in the far-UV region are typical for a-helical proteins as shown in Figures 4-69 4-74. The CD spectra of the aIbumin-2-naphthol-8sulfonate complex were found to be practically identical to the spectra of unliganded albumin. Similarly, CD bands in the near-UV region of albumin-2-naphthol-8-sulfonate complex did not differ from those of unliganded albumin. These indicated that the CD spectra of albumin below 300 nm did not appear to be affected by the binding of 2naphthol-8-sulfonate. A notable positive Cotton effect at 332 imi in the CD spectra for the HSA-2-naphthoI-8-sulfonate complex was observed at both pH values as shown in Figure 4-74, which is attributed to the induced Cotton effect of 2-naphthoI-8-sulfonate upon binding to HSA. No significant changes in the same wavelength region of the CD spectra of BSA-2-naphthol-8-sulfonate were observed.

PAGE 93

77 5.000 210.0 355.0 500.0 wavelength (nm) Figure 4-1. Absorption spectra of 2-naphthol-8-sulfonate at different Hq or pH values: i, -1.97 7.7; 2, 8.6; 3, 9.6 10.28; 4, 1 1.8; 5, 12.9.

PAGE 94

78 1B989wavelength (nm) Figure 4-2. Fluorescence excitation spectra of 2-naphthol-8-sulfonate in different pH values: i, 4.08 6.92; 2, 8.02; 3, 8.99; 4, 9.54; 5, 10.07.

PAGE 95

79 12 wavelength (nm) Figure 4-3. Fluorescence emission spectra of 2-naphthol-8-sulfonate in different Hq or pH values: i, -1.97; 2, -1.25; 3, -0.63 ;4, 0.02; 5, 0.63; 6, 1.04; 7, 1.31; 8, 2.06; 9, 3.04; 10, 4.03 9.02; U, 9.99; 12, 1 1 .8 14.0.

PAGE 96

80 0.5 -4 -2 0 2 4 6 8 10 12 14 16 HjOrpH Figure 4-4. The variation of absorbances with Hq or pH of 2-naphthol-8-sulfonate monoanion (•) and its conjugate base dianion (O).

PAGE 97

81 or pH Figure 4-5. The variation of relative fluorescence intensities with Hq or pH of 2-naphthol8-sulfonate monoanion (•) and its conjugate base dianion (O).

PAGE 98

82 280 300 320 340 360 380 wavelength (nm) Figure 4-6. Absorption spectra of 2-naphthol-8-sulfonate monoanion in different solvents: 2.88 M perchloric acid; methanol; ethanol; • 1 ,4-dioxane; • DMSO; t-butanol; acetonitrile.

PAGE 99

83 3.800, — . — . — . — . — — . — . — . — . — , — • • — . — — — ' — . — — — I — — — — — ' ' — -—r 2S0.0 300.0 350.0 400.0 4S0.0 Wavttlangth (nm. ] ure 4-7. Absorption spectra of 2-naphthol-8-sulfonate dianion in different solvents: — IN sodium hydroxide; ethanol; • • methanol; • • 1,4-dioxane; • — glycerol; acetonitrile.

PAGE 100

84 2 wavelength (nm) Figure 4-8. Fluorescence emission spectra of 2-naphthol-8-sulfonate monoanion in different solvents: i, 2.88M perchloric acid; 2, DMSO; 3, 1,4-dioxane; 4, methanol; 5, ethanol; 6, acetonitrile; 7, glycerol.

PAGE 101

85 1 ^ r — " I I I 1 1 1 1 1 — I I ^^^^^^n En 3Sa 388 488 428 448 468 488 388 328 548 968 388 688 EX 337 wavelength (nm) Figure 4-9. Fluorescence emission spectra of 2-naphthol-8-sulfonate dianion in different solvents: i, IN sodium hydroxide; 2, glycerol; 3, ethanol; 4, methanol; 5, DMSO; 6, 1,4-dioxane; 7, acetonitrile.

PAGE 102

86 lee 8aI En 3£e 3sa 4ee 42a 44a 468 4ea sae sza S4e sea sea eaa EX 337 wavelength (nm) Figure 4-10. Relative fluorescence intensities of 2-naphthol-8-sulfonate as a function of ethanol concentration (volume %): i, 100%; 2, 75%; 3, 50%.

PAGE 103

87 Figure 4-11. Relative fluorescence intensities of 2-naphthol-8-sulfonate as a function of 1,4-dioxane concentration (volume %): i, 75%; 2, 100%; 3, 50%.

PAGE 104

88 250.0 300.0 350.0 400 Havelencfth < lui . ) Figure 4-12. Absorption spectra of free ( ), BSA-bound () and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 3.04.

PAGE 105

89 0.300, . . . . . . 1 ' ' • r 250.0 300.0 350.0 400.0 Wav«l«ngth (nm. ) Jure 4-13. Absorption spectra of free ( ), BSA-bound () and HSA-bound -) 2-naphthol-8-sulfonate at pH 4.03 and ^ 0. 1 .

PAGE 106

Figure 4-14. Absorption spectra of free ( ), BSA-bound (• ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 5.01 and n 0.1.

PAGE 107

91 Figure 4-15. Absorption spectra of free ( ), BSA-bound (-•-•) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 5.93 and ^ 0.1.

PAGE 108

Figure 4-16. Absorption spectra of free ( ), BSA-bound (( ) 2-naphthol-8-sulfonate at pH 6.93 and 0.1. •) and HSA-bound

PAGE 109

93 0.300| . . . . — I . . . , — — — — — r 0.200 b iOO.O 330.0 340.0 3C0.0 380.0 4M.0 WavalwMth rim.) Figure 4-17. Absorption spectra of free ( ), BSA-bound (• ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 8.09 and ^ 0.1.

PAGE 110

94 Figure 4-18. Absorption spectra of free ( ), BSA-bound (• ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 9.50 and |a 0.1.

PAGE 111

95 Figure 4-19. Fluorescence emission spectra of free ( ), BSA-bound ( ) HSA-bound ( ) 2-naphthol-8-sulfonate at pH 3.04.

PAGE 112

96 Figure 4-20. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 4.04 and ^ 0.001.

PAGE 113

97 Figure 4-21 . Fluorescence emission spectra of free ( ), BSA-bound ( — HSA-bound ( ) 2-naphthol-8-sulfonate at pH 5.07 and ^ 0.001. — ) and

PAGE 114

98 Figxire 4-22. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 6.07 and \i 0.001.

PAGE 115

99 En 3Sa 380 496 429 448 4e9 469 30« 529 34* f}i9 389 498 EX 3 37 wavelength (nm) Figure 4-23. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate as well as BSA (• • -) and HSA (• -) atpH 7.05 and ^ 0.001.

PAGE 116

100 Figxire 4-24. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 8.12 and ^ 0.001.

PAGE 117

101 Figure 4-25. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 8.94 and ^ 0.001.

PAGE 118

102 Figure 4-26. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 10.01 and n 0.001.

PAGE 119

103 t»»EX 337 wavelength (mn) Figure 4-27. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 4.03 and ^ 0.1.

PAGE 120

104 99 Figure 4-28. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 4.98 and ^ 0.1.

PAGE 121

105 Figure 4-29. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 5.93 and ^ 0.1.

PAGE 122

106 Figure 4-30. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate as well as BSA (• • -) and HSA (• -) atpH 6.93 and ^0.1.

PAGE 123

107 Figure 4-3 1 . Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 8.09 and ^ 0.1.

PAGE 124

108 Figure 4-32. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 9.02 and ^i 0.1.

PAGE 125

109 Figure 4-33. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 9.99 and |i 0.1.

PAGE 126

110 MEM 3Se 388 400 42B 44e 4«9 480 39B 520 346 360 380 SBO EX 337 wavelength (nm) Figure 4-34. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 4.03 and \x 1.0.

PAGE 127

Ill Figure 4-35. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 4.95 and \i 1.0.

PAGE 128

112 Figure 4-36. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate as well as BSA (• • -) and HSA (• -) atpH 6.93 and^ 1.0.

PAGE 129

113 89 EX 337 wavelength (nm) Figure 4-37. Fluorescence emission spectra of free ( ), BSA-bound ( ) and HSA-bound ( ) 2-naphthol-8-sulfonate at pH 8.92 and ^ 1.0.

PAGE 130

114 Figure 4-38. Fluorescence emission spectra of BSA-2-naphthol-8-sulfonate complex at different molar ratios of 2-naphthol-8-sulfonate to BSA: i, 1:1; 2, 1:5 and 1:10.

PAGE 131

115 EM 3€9 389 499 429 449 4€9 489 599 329 549 5£9 589 699 EX 337 w»velength (lun) Figure 4-39. Fluorescence emission spectra of HSA-2-naphthol-8-sulfonate compli at different molar ratios of 2-naphthol-8-sulfonate to HS A: i, 1 : 1 ; 2, 1 : 5 ; 3 , 1:10.

PAGE 132

116 160 140 -\ 120 100 80 60 H 40 20 -\ 0 -20 -40 -60 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Mole fraction of 2-naphthol-8-sulfonate Figure 4-40. Job's plot of the difference between the observed fluorescence intensity (k J, = 337 nm) and the corresponding value in the absence of albumin as a function of mole fraction of 2-naphthol-8-sulfonate monoanion at pH 7.4 and |i 0.1. The total concentration of BSA and 2-naphthol-8-sulfonate was kept constant at 10 ^M.

PAGE 133

117 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Mole fraction of 2-naphthol-8-suIfonate Figure 4-41 . Job's plot of the difference between the observed fluorescence intensity (X, „ = 337 run) and the corresponding value in the absence of albumin as a function of mole fraction of 2-naphthol-8-sulfonate dianion at pH 7.4 and |i 0.1. The total concentration of BSA and 2-naphthol-8-sulfonate was kept constant at 10 \xM.

PAGE 134

118 180 0 5 10 15 20 2-Naphthol-8-sulfonate concentration x 10* Figure 4-42. Plots of relative fluorescence intensity (measured at emission wavelength 445 nm with excitation at 337 nm) as a function of total 2-naphthol-8sulfonate concentration for the BSA-2-naphthol-8-sulfonate titrations with a constant amount of the protein (Pj) at pH 7.4 and ^ 0. 1 . Key: curve (O), P^ = 0 M; curve (A), P^ = 2.5 x 10"* M; curve (•), P^ = 1 x 10 ' M and is obtained after correcting the fluorescence intensities for the absorbance effect.

PAGE 135

119 Figure 4-43. Plots of relative fluorescence intensity (measured at emission wavelength 453 nm with excitation at 337 nm) as a function of total 2-naphthol-8sulfonate concentration for the HSA-2-naphthol-8-sulfonate titrations with a constant amount of the protein (P^) at pH 7.4 and |i 0.1 . Key: curve (O), P^ = 0 M; curve (A), Pt = 2.5 x 10^ M; curve (•), P^ = 1 x 10"' M and is obtained after correcting the fluorescence intensities for the absorbance effect.

PAGE 136

120 7 Figure 4-44. Scatchard plot of r/[L] versus r for the bovine serum albumin-2naphthol-8-sulfonate titration with the protein concentration of 2.5 x 10"* M (measured at emission wavelength 376 nm with excitation at 337 rmi).

PAGE 137

121 35 30 25 Figure 4-45. Scatchard plot of r/[L] versus r for the human serum albumin-2naphthol-8-sulfonate titration with the protein concentration of 2.5 x 10"* M (measured at emission wavelength 376 nm with excitation at 337 nm).

PAGE 138

122 0.1 4 5 6 7 8 9 10 PH Figure 4-46. The fluorescence emission ratio of BSA-2-naphthol-8-sulfonate complex as a function of pH at different ionic strengths. Curve (•), ^ = 1 .0; curve (O), ^ = 0. 1 ; curve (), ^ = 0.001 .

PAGE 139

123 pH Figure 4-47. The fluorescence emission ratio of HSA-2-naphthol-8-sulfonate complex as a function of pH at different ionic strengths. Curve (•), [i= 1.0; curve (O), 0.1; curve (), n = 0.001.

PAGE 140

124 Molar ratio of marker/albumin Figure 4-48. Marker-induced changes in fluorescence of 2-naphthol-8-sulfonate bound to HSA. Solution containing HSA (10 pM) and 2-naphthol-8-sulfonate (10 |iM) was titrated with ibuprofen (), diazepam (•), phenylbutazone (O) and salicylic acid (O), added at molar marker/albumin ratios varying from 0.025 to 5.

PAGE 141

125 Figure 4-49. Marker-induced changes in fluorescence of 2-naphthol-8-sulfonate bound to BSA. Solution containing BSA (10 ^M) and 2-naphthol-8-sulfonate (10 |aM) was titrated with ibuprofen (), diazepam (•), phenylbutazone (O) and salicylic acid (O), added at molar marker/albumin ratios varying from 0.025 to 5.

PAGE 142

126 -o.osot— 300.0 }40.0 160.0 Havel Angth { nm . ) «00.0 Figure 4-50. Absorption spectra of pyridine (• •), 2-naphthol-8-sulfonate in the presence of pyridine ( ) at pH 5.01 and |a 0.1.

PAGE 143

127 Ex 337 wavelength (nin) Figure 4-5 1 . The fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of pyridine at pH 5.0 and |i 0.1. The concentrations of pyridine are: 0 M; 2.0 X 10-" M; 2.2 x 10'^ M; 5.9 x 10"' M; 0.01 M and 0.014 M (from top to bottom) with constant amount of 2-naphthol-8-sulfonate (10 [iM).

PAGE 144

128 ee: En 368 EX 337 I 42a 1 — I — -tee — I — 4e« — I — 329 1 548 1 see SB* wavelength (nm) Figure 4-52. The fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of pyridine at pH 9.0 and |i 0.1. The concentrations of pyridine are: 0 M; 2.0 X 10^ M; 2.2 x 10"^ M; 5.9 x IQ-^ M; 0.01 M and 0.014 M (from top to bottom) with constant amount of 2-naphthol-8-sulfonate (10 |iM).

PAGE 145

129 Figure 4-53. The fluorescence emission spectra of HSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 5.0 and ^ 0.1 with constant amount of 2-naphthol-8sulfonate (10 ^M) and HSA (7 [iM). The concentrations of pyridine are: 0 M; 2.0 x lO"" M; 2.2 X 10"' M; 5.9 X 10"^ M; 0.01 M and 0.014 M (from top to bottom).

PAGE 146

130 lee "T I 1 1 1 1 1 1 1 1 1 1 1 EH 36* see 4ee ize -«'«• lee 4»b see 92e S4e see see see EX 337 wavelength (iim) Figure 4-54. The fluorescence emission spectra of HSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 9.0 and |i 0.1 with constant amount of 2-naphthol-8sulfonate (10 |iM) and HSA (7 nM). The concentrations of pyridine are: 0 M; 2.0 x lO"" M; 2.2 X 10"' M; 5.9 x 10 ' M; 0.01 M and 0.014 M (from top to bottom).

PAGE 147

131 8e E 4> I — I — 3ee — I — IBB 1 42B 1 44B 1 1 4BB 1 SB* 1 9S« -I wavelength (nm) Figure 4-55. The fluorescence emission spectra of HSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 5.0 and |i 0.1 with constant amount of 2-naphthol-8sulfonate (10 \iM) and HSA (50 ^M). The concentrations of pyridine are: 0 M ( ); 2.0 X 10"' M ( ); 5.9 x lO'^ M (• -); 9.8 x 10"' M (•• -); and 0.014 M ( ).

PAGE 148

132 s 8 o i 01 oe» EK 337 1 1 1 **9 450 488 wavelength (mn) — I — see — 1 — 32e — I — — I — see -r 38S I ••• Figure 4-56. The fluorescence emission spectra of HSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 9.0 and \i 0. 1 with constant amount of 2-naphthol-8sulfonate (10 \iM) and HSA (50 |aM). The concentrations of pyridine are: 0 M ( ); 2.0 X 10"' M ( ); 5.9 x 10'^ M (-); 9.8 x lO"^ M (•• -); and 0.014 M ( ).

PAGE 149

133 IMwavclength (nm) Figure 4-57. The fluorescence emission spectra of BSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 5.0 and 0.1 with constant amount of 2-naphthol-8sulfonate (10 |iM) and BSA (7 |iM). The concentrations of pyridine are: 0 M; 2.0 x 10-" M; 2.2 X 10 ' M; 5.9 x 10'' M; 0.01 M and 0.014 M (from top to bottom).

PAGE 150

134 1098B '-\ 1 1 1 1 1 n 1 1 1 1 I I en see 38b 488 4ae a*9 4se vs* saa sas see sea see see EX 337 wavelength (nm) Figure 4-58. The fluorescence emission spectra of BSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 9.0 and 0.1 with constant amount of 2-naphthol-8sulfonate (10 nM) and BSA (7 ^M). The concentrations of pyridine are: 0 M; 2.0 x 10"^ M; 2.2 X 10"' M; 5.9 X 10 ' M; 0.01 M and 0.014 M (from top to bottom).

PAGE 151

135 Figure 4-59. The fluorescence emission spectra of BSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 5.0 and |i 0.1 with constant amount of 2-naphthol-8sulfonate (10 \iM) and BSA (50 \iM). The concentrations of pyridine are: 0 M ( ); 2.0 X 10"^ M ( ); 5.9 x 10"^ M (• -); 9.8 x W M (•• -); and 0.014 M ( ).

PAGE 152

136 EH 3E8 EX 337 388 488 T 428 T 1 T 44B 468 488 wavelength (nm) 388 328 368 Figure 4-60. The fluorescence emission spectra of BSA-2-naphthol-8-sulfonate in the presence of pyridine at pH 9.0 and \i 0.1 with constant amount of 2-naphthol-8sulfonate (10 \iM) and BSA (50 ^M). The concentrations of pyridine are: 0 M ( ); 2.0 X 10-' M ( ); 5.9 x 10"^ M (• -); 9.8 x 10"' M (•• -); and 0.014 M ( )•

PAGE 153

137 — 1 1 1 1 1 1 1 — I I I I I 3se 42a i-ta 4ee 4ea sea sze 34a sea ssa saa wavelength (nm) Figure 4-61. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of pyridine in ethanol with constant amount of 2-naphthol-8-sulfonate (10 ^M). The concentrations of pyridine are: 0 M; 2.0 x 10"* M; 2.2 x IQ-^ M; 6.2 x lO^' M; 0.01 M; 0.014 M; and 0.022 M (from top to bottom). e EH 36S EX 337

PAGE 154

138 8 eFX 337 1 38a 4£e 48a wavelength (nm) — I — see — I — 32e — I — 34a — I — 36B I 688 Figure 4-62. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of pyridine in 75% (VfV) ethanol with constant amount of 2-naphthol-8sulfonate (10 |iM). The concentrations of pyridine are: 0 M; 2.0 x 10"^ M; 2.2 x 10' M; 6.2 X 10-^ M; 0.01 M; 0.014 M; and 0.022 M (from top to bottom).

PAGE 155

139 IMwavelcngth (nm) Figure 4-63. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of pyridine in 50% (V/V) ethanol with constant amount of 2-naphthol-8sulfonate (10 nM). The concentrations of pyridine are: 0 M; 2.0 x lO"* M; 2.2 x 10' M; 6.2 X 10-^ M; 0.01 M; 0.014 M; and 0.022 M (from top to bottom).

PAGE 156

140 T 1 — I 1 1 1 1 1 1 I I r EH 3ea 38« ^SS -t2« f4« -tCO -fSB 568 32» 34e 3«8 58« EX 337 wavelength (nm) Figure 4-64. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of pyridine in 1 ,4-dioxane with constant amount of 2-naphthol-8-sulfonate (10 nM). The concentrations of pyridine are: 0 M; 2.0 x 10"^ M; 2.2 x 10^^ M; 6.2 x 10"^ M; 0.01 M; 0.014 M; and 0.022 M (from top to bottom).

PAGE 157

141 Figure 4-65. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of pyridine in 75% (VA^) 1 ,4-dioxane with constant amount of 2-naphthol8-sulfonate (10 |iM). The concentrations of pyridine are: 0 M; 2.0 x 10"^ M; 2.2 x 10-^ M; 6.2 X 10"^ M; 0.01 M; and 0.022 M (from top to bottom).

PAGE 158

142 •EH 3C« EX 33? 1 420 1 52e — I — S4a see — I — wavelength (ran) Figure 4-66. Fluorescence emission spectra of 2-naphthol-8-sulfonate in the presence of pyridine in 50% (VA^) 1 ,4-dioxane with constant amount of 2-naphthol8-sulfonate (10 \iM). The concentrations of pyridine are: 0 M; 2.0 x 10"^ M; 2.2 x 10-^ M; 6.2 X 10"^ M; 0.01 M; and 0.022 M (from top to bottom).

PAGE 159

143 Figure 4-67. Plots for fluorescence quenching of 2-naphthol-8-sulfonate in the absence and presence of albumin as well as albumin with pyridine at pH 5 and \i 0.1. Curve (•), unbound; curve (A), HSA bound; curve (A), BSA bound 2naphthol-8-sulfonate; curve (T), HSA; and curve (V), BSA.

PAGE 160

144 2.0 -r 1.5 1.0 I c:0.5 -0.5 -10 H 1 1 1 1 1 1 0.000 0.005 0.010 0.015 0.020 0.025 0.030 0.035 [Q] (mol/1) Figure 4-68. Plots for fluorescence quenching of 2-naphthol-8-sulfonate in the absence and presence of albumin as well as albumin with pyridine at pH 9 and |a 0.1. Curve (•), unbound; curve (A), HSA bound; curve (A), BSA bound 2naphthol-8-sulfonate; curve (T), HSA; and curve (V), BSA.

PAGE 161

145 600 400 -1200 H 1 \ \ 1 1 200 210 220 230 240 250 wavelength (nm) Figure 4-69. CD spectra of HSA ( ), (• • -); HSA-2-naphthol-8-sulfonate ( ), ( ) at pH 9.0 and pH 5.0, respectively. Molar ellipticity is calculated with regard to the HSA concentration (0.4 |iM). Molar ratio of 2-naphthol-8sulfonate to HSA in complex is 3:1. The spectra were recorded in a 1 cm cell at room temperature.

PAGE 162

146 1 Figure 4-70. CD spectra of HSA ( ), (• • -); HSA-2-naphthol-8-sulfonate ( ), ( ) at pH 9.0 and pH 5.0, respectively. Molar ellipticity is calculated with regard to the HSA concentration (20 ^M). Molar ratio of 2-naphthol-8sulfonate to HSA in complex is 3: 1 . The spectra were recorded in a 1 cm cell at room temperature.

PAGE 163

147 2000 -8000 -\ 1 \ 1 \ [ 1 r300 320 340 360 380 400 420 440 wavelength (nm) Figure 4-71. CD spectra of HSA ( ), (• • -); HSA-2-naphthol-8-sulfonate ( ), ( ) at pH 9.0 and pH 5.0, respectively. Molar ellipticity is calculated with regard to the HSA concentration (75 ^M). Molar ratio of 2-naphthol-8sulfonate to HSA in complex is 2.7: 1 . The spectra were recorded in a 1 cm cell at room temperature.

PAGE 164

148 40 200 210 220 230 240 250 wavelength (nm) Figure 4-72. CD spectra of BAS ( ) and BSA-2-naphthol-8-sulfonate ( ) at pH 5.0 and room temperature. Molar ellipticity is calculated with regard to the BSA concentration (0.4 \iM). Molar ratio of 2-naphthol-8-sulfonate to BSA in complex is 1:1. The spectra were recorded in a 1 cm cell.

PAGE 165

149 1.0 Figure 4-73. CD spectra of BAS ( ) and BSA-2-naphthol-8-sulfonate ( ) at pH 5.0 and room temperature. Molar ellipticity is calculated with regard to the BSA concentration (20 ^M). Molar ratio of 2-naphthol-8-sulfonate to BSA in complex is 1:1. The spectra were recorded in a 1 cm cell.

PAGE 166

150 1000 "o B -a B o ab u -a B^ 5 -4000 o -1000 -2000 -3000 -5000 -6000 -7000 300 320 340 360 380 400 wavelength (nm) 420 440 Figure 4-74. CD spectra of BAS ( ) and BSA-2-naphthol-8-sulfonate ( ) at pH 5.0 and room temperature. Molar ellipticity is calculated with regard to the BSA concentration (75 ^M). Molar ratio of 2-naphthol-8-sulfonate to BSA in complex is 2.7:1. The spectra were recorded in a 1 cm cell.

PAGE 167

CHAPTER 5 DISCUSSION The Effect ofpH The absorption spectra of 2-naphthol-8-suIfonate in water as a function of Hf/pH are shown in Figure 4-1 . With an increase in pH, the absorption peaks move to longer wavelengths (red shift) and two isosbestic points were clearly shown at 337 nm and 309 nm. These indicated a ground state acid-base equilibrium for 2-naphthol-8-sulfonate as depicted in equation (1-1) bottom panel. In molecules like 2-naphthol-8-sulfonate containing an electron-donating group (hydroxyl group), the excitation of these molecules from the Sg to the S, state results in the movement of electronic charge to the aromatic ring. This causes the Sq state to be stabilized to a greater degree than is the S, state by protonation at the electron-donating group and increases the energy difference between Sq and S, state in the protonated molecule. Thus, the spectral band of the deprotonated form always appears at longer wavelength than that of protonated form. The fluorescence excitation spectra of 2-naphthol-8-sulfonate in water as a fianction of Ho/pH are shown in Figure 4-2. The properties of these excitation spectra are similar to those found in the absorption spectra. 151

PAGE 168

152 The fluorescence emission spectra of 2-naphthol-8-sulfonate were studied from Hq -1.97 to pH 14 and the fluorescence spectra of the two species (protonated and deprotonated forms) were observed as shown in Figure 4-3. The relative fluorescence quantum yields for the protonated {^/^o) and deprotonated i^'/^ol fonns were determined from fluorescence measurements. Examination of the fluorescence titration curves indicates that there are stretched sigmoidal curves, which reveal that the rate of proton transfer in S, is comparable with the rate of fluorescence. The values of the relative fluorescence quantum yields {^/^o and ^'Z^) are plotted as a function of Hf/pH in Figure 4-5. There is a plateau region in the pH range 4-9.5. The sum of ^/^o and ^'/^o' is approximately equal to unity except that at Hq around -1, which exceeds unity. This apparently high value of ^/^q + ^'/^o' might be a result of protonation of the sulfonate group (naphthalenesulfonic acid pA: = 0.57 cited from CRC Handbook of Chemistry and Physics 1993 1994) in the 2-naphthol-8-sulfonate molecule with increasing solution acidity. Further investigation is needed. Steady-State Kinetics of Excited State Proton Transfer Reaction of 2-Naphthol-8Sulfonate The rate constants for the excited state proton transfer of 2-naphthol-8-sulfonate are presented in Table 4-1. As can be seen the rate constants k|2 and obtained using different methods somehow do not agree with each other very well. The value of kji calculated using method (b) is one order larger than the rate constant calculated by the method (a). These apparently inconsistent values of kj, might be a result of protonation of the sulfonate group in the 2-naphthol-8-sulfonate molecule with increasing solution acidity.

PAGE 169

153 Solvent Studies The Effects of Solvent on the Spectral Properties of 2-NaDhthol-8-Sulfonate The results of the solvent studies (Table 4-1) show that 2-naphthol-8-sulfonate exhibited a red shift in the absorption spectrum and a blue shift in the fluorescence emission spectrum on going from water to the less polar solvents. As mentioned in chapter 2, the effect of solvents on electronic spectral bands depends upon the nature and relative strength of the interaction of the solvent molecule with the ground and excited states of the solute molecule. Figure 5-1 depicts a potential energy diagram of the radiative transition (absorption or emission). When the molecule is surrounded by a liquid solvent each state is stabilized (or destabilized) by an energy which is the solvation energy. As can be seen, in the process of light absorption, E^ = E° + E" (e* E') , in the emission process, Ej.=Ef-Ej+(E°-E"), where E^ is the relaxation energy. Stronger interacting (more polar) solvent makes E" E[ more negative in both fluorescence and absorption if the fluorophore is more polar in the S, than in the Sq, but stronger interaction between solute and solvent makes E" and E' bigger also. Since E" E] and E° are both subtracted from E°, Ef is always smaller or red shift. However, E3 can be red or blue shifted depending on the magnitude of E" E* and E* . Electronic excitation increases the degree of charge separation in the 2-naphthol8-solfonate molecule and thus, increases the dipole moment. In such a case, the energy of the excited state should be decreased to a greater extent than that of the ground state by

PAGE 170

154 S',ytMin solution) h nonaquilibrium ttate £a hf* lolution <4o (tn solution) •quilibrium ttste y4MisoUtad) >4Min solution) •quilibrium st«t« Si Ai>' solution \ (In solution) nonaquilibrium stat* •Si Figure 5-1. A simplified schematic representation of energy levels (A) attained by a molecule in solution in the course of light absorption and emission (B), according to a solute-solvent interaction model. Modified from reference 46.

PAGE 171

155 increasing the polarity of the solvent. However, the ground state molecule has two electrons in the nonbonding orbital of the O atom, the lone pair forms strong hydrogen bonds with protic solvents, and these bonds strongly stabilize the ground state. In the excited state an electron is transferred from the O lone pair to the aromatic system, so that there is a substantial decrease of electron density on the oxygen, consequently, the hydrogen bond is broken down and this will lead to the relaxed excited state in which there is no specific association (hydrogen bond) between the solute and the solvent. As a result, the energies of n -> ti* absorptions increase (the spectra shift to higher frequencies or shorter wavelengths) with increasing solvent hydrogen bond donor capacity. As mentioned in chapter 2, the magnitude of the spectral shifts resulting from specific solvent-solute interactions frequently exceeds that due to the general interactions. Spectral shifts to longer or shorter wavelengths resulting from dipole or hydrogen bonding interactions may be constructively or destructively additive. It appears that the observed blue shift in the absorption spectra of 2-naphthol-8-sulfonate on going from aprotic solvents to protic solvents is the result of strong hydrogen bond formation between the protic solvent and the hydroxyl group of the 2-naphthol-8-sulfonate molecule in the ground state. Solvent molecules (• ) in the immediate vicinity of a solute molecule (C3 ) are oriented in a configuration (— > or <— ) which is energetically favorable to the ground state (So) of the solute molecule. When such a solute molecule absorbs radiation, its charge distribution changes from that of its ground state to that of its excited state in a much shorter time than is required for molecular vibration or diffusion. Therefore, the excited

PAGE 172

156 solute molecule will be surrounded by solvent molecules in a configuration favorable to its ground state. Such a situation is referred to as the Franck-Condon excited state (S,'). In order to account for the fluorescence spectra observed in relaxing molecular systems it is necessary to draw potential energy surfaces along at least two coordinates. The vibrational coordinate, Q^, represents a specific vibration which is coupled with the electronic transition. The relaxation coordinate, Qp may represent an angle of twist, the distance between a proton and a specific atom in the molecule, a coordinate specifying the separation and orientation of two chromophores involved in excimer formation, or a generalized solute-solvent cage coordinate. Figure 5-2 shows how vertical excitation (process 1) of the equilibrium ground state configuration, Sq^^ ^j, leads to the FranckCondon excited state, S|(r in which both and did not change. This state undergoes rapid vibrational relaxation (process 2) leading to a state, 8,^^ where did not change but has the value of the equilibrium excited state configuration. Fluorescence may originate from S.^^ y ), where both Q^, and assume their values in the equilibrium excited state configuration, from state S^^ ^^-y or from an intermediate configuration depending on the relative magnitudes of the rate constants of the relaxation process k^ and fluorescence kf. Processes 6, 3, and 5 in Figure 5-2 represent those situations respectively. As mentioned before, the Ef is always smaller or red shift upon going from apolar to polar solvent for polar solute. An electronically excited molecule is a chemically distinct entity, thus, the tendency of a molecule to form hydrogen bonds with solvent

PAGE 173

157 molecules may be altered dramatically by excitation. In hydroxyaromatics such as 2naphthol-8-sulfonate, the excited state resulting from the transfer of an electron on the O Figure 5-2. Potential energy cross sections drawn along two coordinates Qy and for the ground and excited states. Taken from reference 109. lone pair to the aromatic system is so weakly hydrogen-bonded to the solvent that its energy (and that of the Franck-Condon ground state) is not very sensitive to changes in the hydrogen-bonding ability of the solvent. In such a case, the major factor which governs the fluorescence spectral properties is solute-solvent dipole-dipole interactions. At room temperature, the solvent molecules reorient themselves in about 10 " s to a configuration energetically favorable to the excited state of the solute molecule (equilibrium excited state (S,)). Reorientation has the effect of lowering the energy of the equilibrium excited state relative to that of the Franck-Condon excited state. The magnitude of this effect will depend to some extent upon the difference in charge

PAGE 174

158 distribution between the ground and excited states of the solute molecule, and the polar nature of the solvent, as well as the viscosity of the solvent. At room temperature, the lifetimes of fluorescent states, which are generally about 10'* s, are much longer than the time required for solvent reorientation, so emission is observed from the equilibrium excited state rather than from the Franck-Condon excited state to the ground state. The ground state initially formed when fluorescence occurs is a Franck-Condon ground state (So'), which then relaxes, again accompanied by solvent cage reorientation, to the equilibrium ground state (Sq). Therefore, the spectral shifts for a solute molecule with greater dipole moment in the excited state than the ground state show solvent-dependent emission spectra reflecting the mobility and polar character of the solvent. It is evidently that under our experimental conditions, the recorded fluorescence emission spectrum of 2-naphthol-8-sulfonate in glycerol, a viscous solvent, did not show a blue shift compared with those in other organic solvents as presented in Table 4-1. Thus, the observed fluorescence emission of 2-naphthol-8-suIfonate is not from the Franck-Condon excited state rather from the equilibrium excited state configuration or an intermediate excited state configuration as depicted in Figures 4-8 and 4-9. Also the hydrogen bonding donor capacity of solvent does not have profound effect on the spectral shifts. Since once the equilibrium excited state forms, the hydrogen bond at the lone pair electrons has been broken due to the charge transfer to the aromatic ring £md to solvent cage rearrangement. Thus the blue shift in the fluorescence emission spectrum on going from water to the less polar solvents may be contributed by the polar character of the solvent. In brief, it appears that the hydrogen bond donor capacity of the solvent dominates the absorption spectrum, while solvent polarity dominates the fluorescence emission spectrum. Furthermore, it can

PAGE 175

159 be stated that the observed spectral changes in 2-naphthol-8-sulfonate are due to interactions between the functional group, and its environment. The Effect of Solvent on Fluorescence Quantum Yields The variation of the fluorescence quantum yields with changes in solvent can be explained by an hypothesis which invokes several mechanisms of nonradiative deactivation of the excited state. The first of these is that if the lowest excited singlet interacts with the solvent, the probability of nonradiative transitions to the ground state will be increased. Solvents which exhibit minimal interaction with the excited molecule should allow the highest fluorescence quantum yields. A second means of nonradiative deactivation is afforded by intramolecular motion. Increasing solvent viscosity or immobilizing the solute molecule will increase the fluorescence quantum yields of the solute because intramolecular rotation can cause internal losses of energy. A third path of deactivation is afforded by the hydrogen bonding. Hydrogen bonding with the solvent brings about a substantial increase in the rate of internal conversion, at the expense of fluorescence. The Effects of Solvent on the Excited State Proton Transfer As mentioned in Chapter 2, the "abnormal" proton mobility is to be believed to result from a chain of water-water proton transfer reactions which is predominantly an incoherent process, driven by cleavage and formation of hydrogen-bonds in the second solvation shell of the H,0| cation. In other words, the proton moves faster in water having more labile hydrogen bonds.

PAGE 176

160 When the fluorescence emission spectra of 2-naphthol-8-sulfonate are compared between aqueous, aqueous-organic, and organic solvents, the different features can be seen. One is the disappearance of the long wavelength band attributed to the conjugate base of the excited 2-naphthol-8-sulfonate dianion in pure organic solvents while only the protonated (monoanion) form is directly excited. Upon addition of water at constant solute concentration as shown in Figures 4-10 and 4-11, the fluorescence intensity of the emission from the protonated form decreases gradually and a new band (dianion) appears at ~ 455 nm while water concentration is 50% and 25% by volume or higher in the 1,4dioxane-water and ethanol-water mixtures, respectively. These observations are accounted for by the change in solvent structure in the organic-water mixture (composition of solvents) and subsequent events, which affect the excited state proton transfer process. There are several factors involving solvent composition that affect the excited state proton transfer. First, it is known that the proton transfer rate is found to decrease in different organic-water mixtures. Many studies have been focused on the alcohol-water binary mixtures. It is found that the proton transfer rate of a hydroxyaromatic varies as a function of solvent composition in methanol-water mixtures. The proton is thought of as residing on a water molecule which is solvated by either water or methanol molecules since a hydroxyaromatic does not dissociate appreciably in methanol (1 10, and references cited therein. 111). Hopping of a proton from water to methanol is unfavorable by about 2 pK units (4.6 RT units, ca. 2.7 kcal/mol at room temperature). Therefore, proton conduction is limited to water molecule chains. In an organic-water mixture, the water

PAGE 177

161 pathways emanating from a HjO^ moiety could be blocked by the organic solvent molecule(s) in the first solvation shell, i.e., poor solvation of the proton in the organicwater mixtures. This lack of solvation by water may prevent the proton from leaving the excited protonated form, and eventually the proton transfer process would be diminished. Second, the anomalous behavior of thermodynamic and physical properties of alcohol-water mixtures has long been known. Evidence from these data strongly suggests that at low alcohol concentrations, water molecules tend to organize around the hydrophobic groups of the alcohol forming low entropy structures or "cages" of fairly regular and longer lived hydrogen bonds (112, and references cited therein). The enhancement of the structure of water is not only observed in alcohol-water mixtures, but also in other mixtures, such as those of water with dimethyl sulfoxide, acetonitrile or acetone. This means that at room temperature, addition of organic solvents into water enhances the strength of hydrogen bonding between water molecules compared with that in pure water (113-116). Molecular mobility for water near the organic solvent molecule is seen to be retarded, as evidenced in the slower translational diffusion and rotational reorientation (because the translational as well as the rotational dynamics of water molecules involve a breaking and formation of hydrogen bonds), e.g., solvation shell water molecules have correspondingly longer bond halflives compared to those of bulk water molecules, which could be caused by the rate differences between hydrogen-bond making and breaking. One such example is the DMSO-water mixture where the values for hydrogen-bond breaking constants in the DMSO-water mixture are much smaller than in pure water, while the values for hydrogen-bond making constants remain

PAGE 178

162 approximately the same (116). Therefore, the stronger hydrogen bond in an organic-water mixture would prolong the rotation time of water (i.e., cleavage and formation of hydrogen bond), and eventually lead to decreased proton transfer. In summary, excited state hydroxyaromatics require a water structure ( HjO)^ (n « 4 ± 1) to activate the proton dissociation, and furthermore the rate-limiting step is regulated by the rotational time of water, i.e., cleavage and formation of hydrogen-bonds in the second solvation shell of the H9O4 cation. Any changes of solvent composition caused by the addition of organic solvent into water would alter the rate of proton transfer. Binding Studies The retardation of the excited state proton dissociation as reflected by the enhancement of the fluorescence intensity for monoanion as well as the blue shift of the emission in the presence of albumin indicate binding of 2-naphthol-8-sulfonate to albumin. Moreover, the binding parameters correlate with this result. Thus the albumin-2naphthoI-8-sulfonate complex provides the basis for study of the role of solvent-water dynamics in the binding cavity. Identification and Characterization of the Binding Sites on Albumin On human serum albumin, there are specific binding sites called site I (warfarin site) and site II (diazepam site), which were characterized by Sudlow et al. (43). Recently, as demonstrated from the analysis of the X-ray crystallographic structure of human albumin, site I and site II are located in subdomain IIA and subdomain IIIA, respectively

PAGE 179

163 (38). Figure 5-2 shows the amino acid sequence of HSA in an arrangement reflecting the heart-shaped structure based on the design by Carter et al (38). Approximate helical regions are boxed and numbered at their starting points, numbered 1-10 for each domain. Ligand sites in subdomains IIA and IlIA are indicated by asterisks. The calculated net charge by quadrants for this configuration is as follows: upper left, -8; upper right, +2; lower left, -6; lower right, -3. The one-letter abbreviations for amino acids are A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, He; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gin; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; Y, Tyr. The three-dimensional structures of albumins appear to be roughly similar among the species, since equine albumin, is approximately 80% homologous with human albumin, possessing similar X-ray crystallographic structure to human albumin (93). This suggests that, with respect to binding sites, other mammalian albumins are analogous to human albumin, considering the structural similarities between the molecules. To identify the binding sites for 2-naphthol-8-sulfonate on HSA and BSA, displacement studies were performed using typical specific binding site markers: phenylbutazone and salicylic acid (site I markers), and ibuprofen and diazepam (site II markers). As can be seen in Figure 4-48, the fluorescence intensity of 2-naphthol-8sulfonate bound to HSA was remarkably decreased by ibuprofen, one of the site II binding markers, but not by diazepam, a specific site II binding marker and site I binding markers. On the other hand, it is clear from the Figure 4-49 that the fluorescence intensity of 2-naphthol-8-sulfonate bound to BSA was much more extensively decreased by site II binding markers but not by site I binding markers. Thus, it can be said that the primary

PAGE 180

164 binding site of 2-naphthol-8-sulfonate on HSA is located within the site II region, and 2naphthol-8-sulfonate binds to site II on BSA. ft <*| « '•), w v>. !* o o. jo *^>. .'f. «i -. .<* lA IB IIA IIB^ IIIA IIIB I I f Figure 5-3. Amino acid sequence of HSA in an arrangement reflecting the heart-shape structure. Taken from reference 117.

PAGE 181

165 The absorption and fluorescence emission spectra and binding parameters of 2naphthol-8-suIfonate to HSA and BSA were notably different. These results may reflect, to some extent, differences in the microenvironments of the site II ligand binding sites on these albumins. According to the X-ray study of crystalline HSA (38), the principle hydrophobic residues in IIIA are ^""Pro, ^^'Leu, '''Uc, ^''Phe, ""^eu, ""Val, "'"Val, ^"Leu, ''<*Val, "^"Leu, ""Val, ""Leu, "'"Val, ""Leu, """Leu, ""Val and ""^Phe. These residues form a focal point around ""Tyr and "'"Arg in HSA. The phenolic oxygen atom of ""Tyr interacts with "'"Arg and lays within 4 A of the carboxylate of "'"Glu. The carboxylate of triiodobenzoic acid, which is bound more or less equally in both subdomains IIA and IIIA, interacts primarily with "'"Arg and is within 4.0 A of the oxygen of ""Tyr, and the aromatic ring of triiodobenzoic acid interacts with the hydrophobic residues, e.g.. Leu, He, Pro, Val and Ala (38). However, "'"Arg and ""Tyr are conserved among all species and hydrophobic residues of these albumins are also well conserved except that one amino acid residue in HSA, "'^Val, related to the ligand binding to site II, is substituted with "'^Ile for BSA. Thus, it is difficult to imagine that the replacement of one hydrophobic residue by another (also hydrophobic) may cause the different binding behavior of 2-naphthol-8-sulfonate between HSA and BSA. It is accepted by many that site II on human serum albumin, originally described by Wanwimolruk et al. (1 18), is a hydrophobic cleft, approximately 16 A deep and 8 A wide, with a cationic group located near the surface. However, this is challenged by several research findings. Maruyama et al (1 19) studied the interaction between either benzodiazepine or other site II drugs (flufenamic acid and mefenamic acid) and bilirubin in binding to HSA. They found that the binding of benzodiazepine to HSA was inhibited by site II drugs but was not affected

PAGE 182

166 by bilirubin. When 1 or 2 mol of benzodiazepine per mol of HSA was added to bilirubinHSA complex (0.6:1), the amount of bound bilirubin increased and the extrinsic (induced ) Cotton effect of the bilirubin-HSA complex was also strengthened. On the other hand, 1 or 2 mol of site II drugs per mol of HSA did not affect the bilirubin-HSA complex. According to the observation they strongly suggest that the diazepam site is independent of the bilirubin site, and the benzodiazepine and site II drug binding sites may not be identical but rather overlapping as they demonstrated that ""Try in HSA involves both the diazepam and site II drug binding sites. More recently, Noctor et al (120) found that different displacement patterns of a series of benzodiazepine drugs by other site II drugs (ibuprofen, L-tryptophan, and medium chain (C^ to €,2) free fatty acids) to HSA could not be interpreted in terms of binding of the solute to a single site but rather could be better described by considering the attachment of the benzodiazepine drugs to several loci within the subdomain IIIA on HSA. It is also found that diazepam shows increased affinity for HSA up to pH 9 (121) while several other site II drugs show either no effect of the N B transition (122) or decreased binding constants with increasing pH (123). Irikura et al (124) studied the drug binding sites on HSA using a set of 7alkylaminocoumarin-4-acetic acids as fluorescent probes, and found that the hydrophobic cleft at site II is about 21 25 A in depth and the distance between the lone tryptophan residue in HSA and probes bound to site II is estimated to be 15 17 A. All of these findings pointed to the conclusion that site II on HSA consists of at least two binding regions. According to the displacement studies of HSA bound to 2-naphthol-8-sulfonate,

PAGE 183

167 it appears that the binding site for 2-naphthol-8-sulfonate is more likely located in the site II region which also involves binding of 2-arylpropionate NSAIDs such as ibuprofen. By comparing the drug binding sites on mammalian albumins, Panjehshahin et al (92) concluded that bovine, dog, horse and sheep albumins have binding sites for warfarin and dansylsarcrosine with similar properties but some quantitative differences from sites I and II on human albumin. There are many examples of enantioselective binding to serum albumins such as the binding of the chiral benzodiazepin oxazepam hemisuccinate to HSA and BSA. The ratios of the affinity constants of the two enantiomers when binding to HSA may be as high as 50, depending on the conditions of observation. BSA, on the other hand, exhibits only a very small enantioselectivity in comparison to HSA, the ratio of the binding constants of d-and 1benzodiazepin oxazepam hemisuccinate being only in the region of 0.8 (125). It is clear that despite apparent small differences in the primary structures of serum albumins from different species and gross similarities in the binding of racemates, the examination of the binding behavior of enantiomers reveals many dissimilarities. The observation of 2-naphthol-8sulfonate binding to HSA and BSA to different extents would imply that the molecular processes involved in the binding of small ligands to serum proteins are highly dependent on the structure and conformation of both participants in the interaction. More recently, Otagiri et al (126) found that binding affinities of site I drugs for bovine, rabbit and rat albumins were reasonably similar to those for human albumin. On the other hand, binding parameters of diazepam to bovine, rabbit and rat albumins were apparently different from those to human albumin while ibuprofen binding shows similarity of binding parameters between different albumins. Therefore, they suggest that these differences are best

PAGE 184

168 explained by microenvironmental changes in the binding sites resulting from change of size rather than a variation in amino acid residues as mentioned before. Matsushita et al. (91) indicated that a cleft (10 15 A deep) with properties similar to site II on HSA molecule exists on the BSA molecule. Thus, the interaction between 2-naphthol-8sulfonate and BSA might be the result of a tertiary structural change in cavity size rather than the loss of hydrophobic residues. According to the displacement pattern of BSA bound to 2-naphthol-8-sulfonate, it seems likely that the diazepam site is identical or very close to the ibuprofen site in the BSA molecule. These results suggest that while the overall arrangement of sites on BSA and HSA is similar, differences in the definition of individual sites occur. Therefore, it can be expected that there exist binding differences between HSA and BSA. Fluorescence Quenching Experiments The absorption spectra of 2-naphthol-8-sulfonate in pyridine-free and pyridinecontaining solutions at pH 5.0 are similar in nature, which is contradictory to what is expected. Namely, the formation of hydrogen bonding at ground state will shift the maximal absorption band towards longer wavelength. The reason might be that both 2naphthol-8-sulfonate and pyridine are strongly solvated by surrounding water molecules which suppress or weaken the formation of hydrogen bonding between 2-naphthol-8sulfonate and pyridine in the ground state. However, using excitation at an isosbestic point, the emission bands observed in pH 5.0 and 9.0 buffer solutions decrease in intensity upon pyridine titration of 2naphthol-8-sulfonate as depicted in Figures 4-51 and 4-52. It is well established that the

PAGE 185

169 H-bond donating ability of aromatic alcohols and amines is stronger in the lowest excited singlet state than in the ground state. The pK's of 2-naphthol-8-sulfonate in the ground state and the lowest excited singlet state are respectively 9.5 and 1 .3. It is not surprising that the strength of a hydrogen bond of a hydroxyaromatic compound with a given base increases with the acidity, and hydrogen bond formation occurs much more easily in the lowest excited singlet state than in the ground state. Mataga (94) has addressed the issue of H-bond formation in the excited state where the proton is capable of conjugating within a Ti-electron system, and stated that for proton donors such as 2-naphthol, the fluorescence yields decrease in the presence of nitrogen heterocycles due to the nonradiative degradation of the excited state by means of delocalization of n electrons through the H bond. Our observations are consistent with those of Mataga et al (94). The effects of pyridine on the fluorescence emission of 2-naphthol-8-sulfonate bound to albumins are shown in Figures 4-53 4-60. At higher molar ratios of albumin to 2-naphthol-8-sulfonate, upon addition of pyridine, the emission band for HSA-bound 2naphthol-8-sulfonate shows little change at pH 5.0, while at pH 9.0, the intensity decreases with increasing concentration of pyridine, which are similar to those observed for the unbound 2-naphthol-8-sulfonate. However, in all cases, the quenching of the HSA-bound 2-naphthol-8-sulfonate fluorescence by pyridine is considerably less than that found for free 2-naphthol-8-sulfonate. For BSA-bound 2-naphthol-8-sulfonate, the intensity of 2-naphthol-8-sulfonate monoanion emission decreases with increasing concentration of pyridine at both pH values concomitant with increase in intensity of the

PAGE 186

170 emission of dianion. In no case did the intensity increase for the long-wavelength emission match the intensity lost from the monoanion emission. As reported by several groups (95, 108), the effects of alkylamine on the fluorescence emission of hydroxyaromatics such as naphthol in organic solvent were dramatic. Either an increase in emission intensity or a new peak at the long-wavelength which corresponds to the naphtholate dianion, accompanying the quenching of the neutral form, was observed upon addition of alkylamine. In order to investigate the abnormal quenching behavior of BSA-bound 2-naphthol-8-sulfonate, the quenching of unbound 2naphthol-8-sulfonate fluorescence in various organic solvents and organic-aqueous mixtures were conducted and the results are depicted in Figures 4-61 4-66. It is clear from these figures that in all cases the intensities of 2-naphthol-8-sulfonate monoanion and dianion decrease with increasing concentration of pyridine. Moreover, the fluorescence quenching of albumin bound 2-naphthol-8-sulfonate by pyridine at low molar ratio of albumin to 2-naphthol-8-sulfonate were conducted. As shown in Figures 457 4-60, upon addition of pyridine at both pH values, the fluorescence intensities for albumin bound 2-naphthol-8-sulfonate monoanion and dianion decreased but to a lesser extent as compared to the unbound 2-naphthol-8-sulfonate. Therefore, combining the quenching results of 2-naphthol-8-sulfonate at low albumin concentration and in organic and aqueous-organic solvents it was postulated that the increase in intensity of emission from BSA-bound 2-naphthol-8-sulfonate dianion might be caused by the alteration of structure or affinity of BSA to 2-naphthol-8-sulfonate upon addition of pyridine. To support this postulation, the emission intensities of uncomplexed albumins in the absence and presence of pyridine have been determined. Also, the emission intensities of albumin

PAGE 187

171 bound and unbound 2-naphthol-8-sulfonate in pyridine-containing solutions, relative to the intensities in pyridine-free solutions, have been determined at a number of pyridine concentrations. The estimation of the quenching effect of pyridine on albumins is made difficult because the emission intensities of uncomplexed albumins at the monitored wavelengths are very low and not easy to measure precisely. The mechanism of quenching must ultimately be dependent upon encounters between quencher and fluorophore. The degree of quenching depends upon the accessibility of quencher to fluorophore. In the case of HSA-bound 2-naphthol-8sulfonate, the encounter between 2-naphthol-8-sulfonate and pyridine is hindered by the binding of 2-naphthol-8-sulfonate to HSA as reflected by a lesser degree of decrease in intensity. In other words, the quenching of the bound ligand is diminished by its relative inaccessibility to quencher. As mentioned before, the fluorescence of excited 2-naphthol8-sulfonate monoanion increases upon binding, therefore any dissociation of the ligand at the site will lower the emission at 376 nm and increase the emission at long-wavelength for albumin-bound 2-naphthol-8-sulfonate. Thus, dissociation will lower the ratio of emission I376/I445 for BSA bound 2-naphthol-8-sulfonate. In the case of fluorescence quenching of BSA-bound 2-naphthol-8-sulfonate, our observation suggests that addition of pyridine might inhibit the binding of 2-naphthol-8-sulfonate to BSA and to some degree dissociation of the ligand from the binding site could occur. However, since pyridine is a larger molecule than common quenchers such as oxygen, the amplitude of fluctuations of the protein would need to be larger to facilitate its inward diffusion. As a result, the penetration rate of pyridine into the binding site to collide with 2-naphthol-8sulfonate is considerably slower than that found for free 2-naphthol-8-sulfonate.

PAGE 188

172 Therefore, the decreased ratio of emission I376/I445 for BSA bound 2-naphthol-8-sulfonate might be caused by the alteration of structure or affinity of BSA upon addition of pyridine and the inabiHty or low ability of pyridine to penetrate into binding sites, which might explain the apparent abnormal quenching effect of BSA bound 2-naphthol-8sulfonate by pyridine. The results of the above experiments have shown that the binding of 2-naphthol8-sulfonate to albumins may result in both a decrease in the access to the ligand of quencher from the aqueous solvent and, immobilization of the ligand binding site, which hampers the diffusion of the quencher through the protein structure. The results also showed that 2-naphthol-8-sulfonate is extensively shielded from the solvent when bound to HSA. In other words, 2-naphthol-8-sulfonate is bound preferentially on the surface of BSA and is located inside HSA. Circular Dichroism Experiment The CD spectra of the albumin-2-naphthol-8-sulfonate complex below 300 nm did not appear to be affected by the binding of 2-naphthol-8-sulfonate. As depicted in the Figures 4-69 4-74, the spectra of the complexes in the far-UV region at pH 5 and 9 were found to be practically identical to the spectra of free albumin. Therefore, the CD analysis shows that interaction with 2-naphthol-8-sulfonate does not induce any notable changes in the secondary structure of both albumins. Induced Cotton effects are often observed when the electrons of a chromophore are perturbed by electrostatic forces associated with a nearby asymmetrical locus (97). The sign of such an induced Cotton effect is governed by the configuration of the

PAGE 189

173 asymmetrical center and its spatial relationship to the perturbed chromophore. Because induced Cotton effects result from an electrostatic interaction, their intensity is inversely proportional to the distance between the asymmetrical locus and the perturbed chromophore (97). For extrinsic Cotton effects, where the asymmetrical center and the perturbed chromophore are not part of the same molecule, the rigidity of the ligandmacromolecule complex is of paramount importance. A loose complex may allow the ligand sufficient freedom of movement so that the protein asymmetrical center moves into regions of positive and negative contribution to a Cotton effect. Under these conditions no optical activity would be observed. 2-Naphthol-8-sulfonate had an absorption maximum at 333 nm for the monoanion. The ellipticity band generated by binding to HSA was located at 332 nm. This observation suggested that it arises from the intramolecular charge-transfer transition of the hydroxyl group on the aromatic ring. Since a positive ellipticity band was observed when 2-naphthol-8-sulfonate was bound to HSA, it was obvious that the protein asymmetrical center was located in a region of 2-naphthoI-8-sulfonate which made a positive contribution to a Cotton effect, and that the HSA-2-naphthol-8-sulfonate complex was rigid enough to prevent the asymmetrical center from entering regions of negative contribution. Prominent changes in CD ellipcities due to the binding of 2naphthol-8-sulfonate to HSA as a function of pH were not observed. The results of the effect of pH on the binding of 2-naphthol-8-sulfonate to HSA, estimated by the fluorescence method at pH range 5.0-9.0, validated the above observation by showing that the fluorescence intensity ratio of 2-naphthol-8-sulfonate monoanion to dianion in HSA solution was not affected by a pH change.

PAGE 190

174 When the induced CD spectral results are compared between HSA and BSA complexes, different features can be seen. One is the appearance of a very weak positive CD band for 2-naphthol-8-sulfonate in BSA solution. As mentioned before, induced Cotton effects are often observed when the electrons of a chromophore are perturbed by electrostatic forces associated with a nearby asymmetrical locus. However, it would seem probable that 1 -point electrostatic interaction with albumin would leave a bound 2naphthol-8-sulfonate molecule fairly free to rotate. Under such conditions, the generation of an induced Cotton effect would appear unlikely. On the other hand, if the aromatic ring could form van der Waals bonds with a hydrophobic region of albumin, a fairly rigid complex might result. It appears from the CD spectra that the interactions between 2naphthoI-8-sulfonate and HSA are stronger and the complex is held fairly firm judging fi'om the notable induced Cotton effect of 2-naphthol-8-sulfonate in HSA. This result is consistent with the binding parameters obtained by the fluorescence method. The magnitude of induced CD bands generated by ligand binding to albumin depends on the distance and the spatial relationships between the ligand chromophore and the asymmetrical center in the binding site, respectively, as well as on the rigidity of the ligand-albumin complex formed (100). It seems likely from this viewpoint that the notable difference observed with 2-naphthol-8-sulfonate sensitively reflects subtle differences in the structure and environment of the binding site and in the bound state of 2-naphthoI-8-sulfonate molecule between HSA and BSA.

PAGE 191

175 Model for Binding of 2-Naphthol-8-Sulfonate to Albumin It has been accepted for a long time that the sulfonate group in molecules like 1anilino-8-naphthalenesulfonate ( ANS") is a solubilizing group for what would otherwise be a nearly water-insoluble anilinonaphthalene moiety. Assuming that the SO^ anion is merely a convenience, the sulfonate anion presumptively had minor bearing on, perhaps nothing to do with the thermodynamics or stoichiometry (numbers of binding sites) of ANS' -protein molecule interaction. It also was generally assumed that if ANS' became brilliantly fluorescent upon binding, host protein binding sites were nonpolar and hydrophobic. However, evidence about the direct participation of the sulfonate group in the interaction with proteins has been reported (127 131). Nowadays it is widely recognized that at least three types of interaction, namely ionic, hydrogen bonding, and hydrophobic, are involved in the ANS' -protein complex. With the results of displacement experiments and the knowledge of the suggested albumin binding area (site II) on one hand, and the number of different chemical groups in the 2-naphthol-8sulfonate molecule on the other hand, it should be assumed that more than one kind of chemical interaction takes part in the binding between 2-naphthoI-8-suIfonate and albumin. In addition, both the fluorescence quenching and CD experiment results indicate that the binding site for 2-naphthol-8-sulfonate in BSA is located on the surface while 2naphthol-8-sulfonate in HSA preferentially located inside the binding site. Therefore, we regard the binding mode of 2-naphthol-8-sulfonate to albumin to be as follows: the organic hydroxynaphthalene moiety of 2-naphthol-8-sulfonate is inserted into the hydrophobic cleft, while the sulfonate group interacts with a cationic sub-site located at

PAGE 192

176 or near the hydrophobic surface of albumin. It is likely that there exists hydrogen bonding between the hydroxyl group as well as the sulfonate group on the 2-naphthol-8-sulfonate molecule and the hydroxyl group of tyrosine (""Tyr) and the amino acid residues with polar side chain such as arginine, threonine, and lysine, which are in the immediate vicinity of ""Tyr. After the ligand is bound, the geometry of the complex is rather fixed (constrained), as is the configuration of the remaining water molecules. As mentioned before, there are multiple interactions between 2-naphthol-8-sulfonate and albumin. As a consequence of these multiple interactions, the motion of the bound 2-naphthol-8sulfonate will be reduced relative to its free state, with a consequent loss in translational and rotational entropy but a gain in enthalpy due to hydrophobic interactions. The entropic gain on displacement of ordered water molecules along with the enthalpic gain by the protein-Iigand interaction compensate the entropic loss by the constrained motion of ligand. This ensures the stability of the complex in solution. Albumin consists of what is essentially a preformed cavity in water. Although it is filled with water molecules, these are bound by the protein structure. This does not mean that there is no exchange with bulk water, or even that this exchange is necessarily slow with respect to self-diffusion rates in bulk water, but rather that the number of hydrogen bond connections from cavity water molecules that extended to large distances into the bulk-water phase is small, compared with the size of these networks in the bulk phase. There is an effective switching off of the water-water potential of mean force at these large distances, and hence a cavity, i.e., a place where 2-naphthol-8-sulfonate can go. The presence of this cavity, possessing a cation-attracting feature near or at the surface, means

PAGE 193

177 that it should be legitimate to continue to refer to albumin-2-naphthol-8-sulfonate binding as a predominantly (even though conditional) hydrophobic effect, despite the water molecules present in the cavity. The Effect of Binding on the Spectral Properties of 2-Naphthol-8-Sulfonate As can be seen from the absorption and fluorescence emission spectra depicted in Figures 4-12-4-37 and the spectral data given in Table 4-1, 2-naphthol-8-sulfonate in albumin solutions shows a red shift in the absorption and a blue shift, especially for the dianion, in the fluorescence emission spectra, relative to the same spectra in water. Absorption spectra of 2-naphthol-8-sulfonate undergo small changes after binding to albumin relative to the spectra in free state in the pH range 5-9. As postulated binding model, 2-naphthol-8-sulfonate binds albumin through multiple interactions which are most likely electrostatic interaction, hydrophobic interaction, and hydrogen bonding. After absorbing of light, among these interactions some may be weakened, some may be strengthened, this compensates the change of dipole moment of 2-naphthol-8-sulfonate upon excitation, i.e., the excited state is stabilized to the similar extent as the ground state, therefore, the energy gap between the ground and excited state is rather small compared that in free state. This leads to apparent red shift of albumin bound 2-naphthol8-sulfonate in absorption spectra compared to those in the free state in water. When the fluorescence emission spectra of 2-naphthol-8-sulfonate are compared between aqueous, organic solvents, organic-aqueous, and albumin solutions, different features can be seen. One is that the magnitude of the shifts for 2-naphthol-8-sulfonate on going from water to albumin solutions is smaller than those occurring on going from

PAGE 194

178 water to organic solvents, rather the properties of 2-naphthol-8-sulfonate bound to albumin are more resembling those in an organic-aqueous environment where water concentration is 50% by volume or higher. These comparative results suggest that the binding site for 2-naphthol-8-sulfonate is not necessarily hydrophobic in the sense that only apolar amino acid residues are in the vicinity. Judging from the extent of excited state proton transfer, the binding site cannot be considered wholly hydrophobic; certainly, it is not anhydrous in character. Although the presence of water in the binding site may affect the degree of blue-shift of the emission, relative to the fluorescence of 2-naphthol8-sulfonate in pure aqueous solutions, the enhancement of the fluorescence intensity of bound 2-naphthol-8-sulfonate appears to reside in other factors as well. It is known that the dipole-dipole interactions of the chromophore in the electronically excited state with the surrounding groups of atoms in the protein molecule or with solvent molecules give rise to shifts of the fluorescence spectra during the relaxation process. The observed spectral shift depends on both the properties of the chromophore itself (the vectorial difference between the dipole moments in the ground and the excited state, ^g ^i^) and also on the properties of the environment interacting with it. One important factor may have to do with the restriction of the geometry of solvents surrounded the 2-naphthol-8sulfonate molecule which bound to the protein, compared with the conformational freedom of those surrounded the 2-naphthol-8-sulfonate in pure aqueous solutions. As mentioned before, solvent molecules in the immediate vicinity of a solute molecule are oriented in a configuration which is energetically favorable to the ground state (Sq) of the solute molecule. When such a solute molecule absorbs radiation, its charge distribution changes from that of its ground state to that of its excited state in a much shorter time

PAGE 195

179 than is required for molecular vibration or diffusion. Therefore, the excited solute molecule will be surrounded by solvent molecules in a configuration favorable to its ground state. At room temperature, the lifetimes of fluorescent states, which are generally about 10'^ s, are much longer than the time required for solvent reorientation, so emission is observed from the equilibrium excited state rather than from the Franck-Condon excited state to the ground state. A protein molecule is an environment with special dielectric properties and contains in high concentrations fairly large electrical dipoles (the dipole moment of the peptide group is 3.6 D, which is twice as high as that of the water molecule), but their ability to reorient under the influence of an electric field is limited by steric effects. In addition, polar (and nonpolar) side groups may be arranged into clusters. Therefore, the orientational and translational movement of the chromophore surrounded by protein would be ultimately determined by the relaxation of protein. Thus, binding interactions could effectively freeze the groundand excitedstate solvent cages into coincidence, i.e., 2-naphthol-8-sulfonate molecule bound to albumin is in effect "locked into" the ground state solvent orientation which does not allow a sufficiently fast relaxation of the excited 2-naphthol-8-sulfonate molecules and fluorescence occurs from the Franck-Condon excited state. In such a case, the fluorescence of 2-naphthol-8sulfonate bound to albumin would occur at higher energy (shorter wavelength) than would be the case if complete solvent relaxation were able to occur prior to fluorescence as in free 2-naphthol-8-sulfonate in pure aqueous solution. Moreover, the complexity of the composition and structure of protein molecules explains why proteins such as albumin, besides their "main" binding site with high affinity for the ligand, may contain other ligand-binding sites with low affinity and

PAGE 196

180 specificity. These differences in the polarity of the environment surrounding the ligand and in the rate of the corresponding relaxation processes could make the observed spectra of the ligand more complicated. In brief, albumin bound 2-naphthol-8-sulfonate exhibits a red shift in absorption spectrum and a blue shift in fluorescence spectrum, probably because it is immobilized and protected from solvent reorientation, rather than the "hydrophobicity" per se of the cavity. The Effect of Binding on Excited State Proton Transfer As can be seen in Figures 4-19 4-37, the dissociation of a proton from excited 2naphthol-8-sulfonate in the presence of albumin is slowed down compared with that in the absence of albumin. This is the consequence of the protein-water interactions which retard the water dynamics. It is understandable that water molecules cover and interact with all exposed surfaces of the protein, undergoing Brownian motion, colliding with each other and with the protein surface. As a consequence of the slower motion of the protein, water molecules that are near the protein surface generally have a decreased diffusion coefficient. The water molecules that collide with apolar portions of the protein surface hit at a different position each time, so that a time average shows a fairly evenly distributed pattern of hydration. The motion of this type of water in association with a non-polar surface tends to be more strongly correlated with that of neighboring water molecules (reflecting the entropy loss seen in the hydrophobic effect) and thus the water diffuses still more slowly. Those water molecules that interact with polar surfaces tend to

PAGE 197

181 be steered into favored positions which complement the local electrostatic (hydrogenbonding) potential of the surface. The favored positions are more narrowly defined in clefts and crevices of the protein surface where there is less protein motion and less freedom of motion for water. The hydrogen-bonded interaction causes these water molecules to diffuse more slowly. These water molecules, having molecular orientations very dependent on the local character of the protein surface (either nonpolar, or positively or negatively charged), are usually within the range 3.2 4.5 A of the polar and nonpolar groups on the protein (16), respectively, and have residence times over 10"'" s. Additional layers of water molecules which extend continuously toward the bulk are, apart from slight residual orientational preferences of the dipole moments, in the first few A of these inner one when the closest protein atom is positively or negatively charged. These layers of water molecules have translational and rotational times larger than in pure water by 1 or 2 orders of magnitude and are thus, in the range 10 " 10''° s. In other words, the influence of the protein on the solvent structure is manifest within the ~ 4 A region, but on the solvent mobility it is manifest within the ~ 10 A region. Schematic representation of a protein molecule, with a layer or two of strongly associated water, suspended in aqueous solution is depicted in Figure 5-4. The hydration layer moves with the protein molecule during orientation, and beyond this layer the water molecules progressively adopt the normal tetrahedral geometry (shown in insert) and orientational relaxation behavior of bulk water. In brief, the protein-water interactions cause an overall reduction of the reorientational rate. This is similar to the situation of 2-naphthol-8-sulfonate in organic-water mixtures where the

PAGE 198

182 Figure 5-4. Schematic representation of a protein molecule, with a layer or two of strongly associated water, suspended in aqueous solution. Taken from reference 19. molecular mobility for water near the organic solvent molecule is seen to be retarded, as evidenced in the slower translational diffusion and rotational reorientation. Therefore, it is not surprising that retardation of excited state proton transfer of 2-naphthol-8-sulfonate occurs in the presence of albumin. This observation reveals that the cavity of the binding site on albumin is filled with water molecules, and the binding of 2-naphthol-8-sulfonate to albumin displaces one or more of these water molecules but there still are retained water molecules with some degree of ordering judging from the extent of excited state proton transfer of 2-naphthol-8-sulfonate in the presence of albumin. As illustrated by the fluorescence quenching experiment, in albumin-2-naphthol8-sulfonate complexes the 2-naphthol-8-sulfonate is shielded, to a large extent, from solvent by neighboring groups of the macromolecule. This implies a decrease in the steric

PAGE 199

183 accessibility for the proton transfer reaction to the solvent for 2-naphthol-8-sulfonate bound to albumin. If this reaction were diffusion controlled as those that take place in aqueous solution, a higher reaction rate would be expected for sites of better accessibility to the solvent, such as those in BSA-2-naphthol-8-sulfonate, than for those in HSA-2naphthol-8-sulfonate where the accessibility of 2-naphthol-8-sulfonate is much lower. The present results contradict the prediction. This phenomenon can be explained by the different mechanisms of proton transfer reaction. As reported by Zundel et al (132), the hydrogen bond between glutamic acid and glutamate residues shows large proton polarizability due to fluctuation of the proton within the bond. They performed ab initio calculations with formic acid water formate and formic acid water water formate systems and demonstrated that these systems show large proton polarizability which increases with increasing chain length. They also found if one water molecule is bound to a formate molecule, the proton potential is highly asymmetrical and this bond has no proton polarizability. Additionally, the hydrogen bond between two water molecules is completely asymmetrical. If these water molecules are present in the formic acid water water formate system, a symmetrical four-minima proton potential is present. The proton may fluctuate and the whole system shows large proton polarizability due to collective proton motion. These hydrogen bonded chains include not only homoconjugated B^H B B-H3, but also heteroconjugated AH-B A" H^B. These proton transfer processes in hydrogen-bonded chains with large proton polarizability are present in biological systems such as the proton conducting system in the bacteriorhodopsin molecule; the

PAGE 200

184 proton pathway in the Fq subunit of ATP synthase; and the proton pathway in the enzyme alcohol dehydrogenase. There are accumulating experimental evidences for hydrogenbonded networks in bacteriorhodopsin which contain amino acids such as Asp, Arg, Glu etc., and a few firmly bound water molecules (133). Kinetic isotope effects reveal an icelike intramolecular proton transfer mechanism in the proton release pathway in bacteriorhodopsin (134). Most of the papers that mention bacteriorhodopsin or hydrogenbonded chains did not address the fundamental rate constants for proton transfer. It is known that in order to participate in a direct proton transfer to a moiety, say a carboxylate group, the acceptor must be located about 0. 1 nm from the dissociating proton, otherwise the primary acceptor will be a water molecule. A carboxylate group located at such proximity is analogous to proton dissociation of an excited molecule in a concentrated buffer which serves as a proton acceptor. As shown by Weller (74) and Eigen et al (135), in the presence of such buffers the rate of proton dissociation is k^pp = k^ + kb[ B ] where kapp is the measured constant, ko is the rate constant of proton transfer to water, k^ is the rate constant of direct proton transfer to the buffer, and [ B ] is the buffer concentration. For the reaction of carboxylate with excited phenol, k^ « 2.9 x 10' M' s"'. Thus, participation of a carboxylate will increase the rate of proton dissociation. This is not the case in HSA bound 2-naphthol-8-sulfonate. Recently, the rate for proton transfer across a biomembrane through a single file of 10 water molecules stabilized by their interaction with the dipoles of carbonyl group of the amino acids (136) in Gramicidin (containing 15 amino acids), has been reported. The rate measured under low driving voltage and ~ 3 M HCl is ~ 2.5 X 10' protons s"'. A more realistic rate of proton transfer along the hydrogen-

PAGE 201

185 bonded chains in Gramicidin was measured to be 6.5 x 10'* s"'. As reported by the research group led by Gutman (136), the rate constants of proton binding to macromolecular structures (carried out with purified proteins, micelle systems, phospholipid membranes, and submitochondrial fragments) were all in the range of diffusion-controlled reactions: 2 6 x 10'" M'' s '. All of these results indicate that the proton transfer can take place via hydrogen-bonded chains containing side groups of some amino acids and/or the peptide backbone as well as a few firmly bound water molecules, and the rate is fast enough to compare with that in pure water. According to the relatively high rate of proton dissociation for 2-naphthol-8sulfonate bound to HSA, we can assume that the proton transfer in the HSA-2-naphthol8-sulfonate complex is most likely through hydrogen-bonded chains. Chains of hydrogen bonds could be formed by side groups of some amino acids in the binding site such as "'"Arg and ""Tyr and/or by the peptide backbone i.e. the carboxylate of "'"Glu. It is also very likely that these hydrogen-bonded chains contain a few firmly bound water molecules inside the protein macromolecule. In other words, the proton transfer in the BSA-2-naphthol-8-sulfonate complex takes place via the loosely associated surface water molecules while the proton transfer in the HSA-2-naphthol-8-sulfonate complex takes place via the strongly bound water molecules in the binding sites. Again, the different binding affinity and therefore the different environment of the binding sites in an albumin could complicate the excited state proton transfer process of albumin bound 2-naphthol-8-sulfonate.

PAGE 202

186 The Effects of pH and Ionic Strength on the Binding of 2-Nap hthol-8-Sulfonate to Albumin The conformation of albumin in the pH range of 5.0 to 9.0 is well documented, the N form occurring mainly below neutral pH and the B form at higher pH. These conformations arise from changes in the tertiary structure and not from the secondary conformation of albumin. The structural changes in the neutral-basic transition occur in domain II, with a contribution from domain 1(117). The change in domain III is small (117). It is well documented that on the human albumin molecule, the binding properties for site I and the diazepam binding site at site II are known to be influenced by the N -> B transition. With the rise of pH from 5.0 to 9.0, the observed ellipticity and fluorescence intensity ratio of monoanion to dianion were found to be unaffected for HSA. Therefore, it is not surprising that 2-naphthol-8-sulfonate shows no effect of the N -> B transition as the displacement experiments indicate that 2-naphthol-8-sulfonate binds primarily to the ibuprofen binding site at site II on human albumin. However, with the rise of pH from 5.0 to 9.0 at constant ionic strengths, the fluorescence intensity ratio of monoanion to dianion was found to be decreased for BSA. The possible explanation could be that the affinity of the binding site in BSA for 2-naphthol-8-sulfonate decreases with increasing pH, since the changes of the numbers of the binding sites is only small and may, within experimental error, be regarded as being independent of the pH as indicated by the data of Job's study. The lowered affinity might be caused either by the conformational transition, namely N B transition, or by the increased repulsion between the ligand bearing a negative charge and net negative charged BSA. The change in binding affinity is unlikely

PAGE 203

187 to be due to changes in ionization of 2-naphthol-8-sulfonate as its pK value is well outside this pH range rather due to the increasing net negative charge of BSA. The conformational changes of albumins in the range of 3.0-5.0 are different from those in the pH range 5.0 to 9.0. The conformational change in the acid range is attributed to the change in secondary and not tertiary structure, judging from the decrease in a-helix content. Around pH 4, the N ^ F transition which is observed even in distantly related species occurs. The F form of HSA becomes longer and increasingly asymmetrical. The N -> F transition involves the separation of the two halves of the molecule, domains I and IIA and domains IIB and III, from each other, and as a consequence the separation exposes the peptide cleavage site, ^"'Asp-^^^Phe in domain III, to solvent, whereas ^"^Phe is found to be buried in the native structure (38). As the pH of the albumin solution is lowered below 3.5, a further unfolding occurs. As a result, under acidic conditions the albumin bound 2-naphthol-8-sulfonate was considered to become exposed to water, to a greater extent as reflected in the fluorescence spectra of the albumin bound 2-naphthol-8sulfonate at low pH which were similar to those of free 2-naphthol-8-sulfonate; that is, the emission peaks shifted to wavelengths similar to those of free 2-naphthol-8-sulfonate, and the fluorescence intensities also resembled those of free 2-naphthol-8-sulfonate as depicted in Figure 4-19. The acid induced expansion is suppressed by increasing ionic strength, an indication that salt forces predominate. When all carboxyl groups have become protonated, the positive charges of lysine, arginine, and histidine residues would cause mutual repulsion between the domains and subdomains of the molecule. As illustrated in Figure 4-27, the fluorescence spectrum of BSA bound 2-naphthol-8-

PAGE 204

188 sulfonate at pH 4 and ionic strength 0. 1 is different to some degree from that at ionic strength 0.001, the former more resembles that of BSA bound 2-naphthol-8-sulfonate at pH5. With an increase of ionic strength from 0.001 to 1 .0 in the pH range 5-9, the observed fluorescence intensity ratios of monoanion to dianion at constant pH values were found to be increased for BSA but not for HSA. The retardation of excited state proton dissociation for BSA bound 2-naphthol-8-sulfonate could be caused by the inorganic salt-induced destruction of water clusters resulting from production of the hydrated ions. There are no changes on the fluorescence intensity ratio of monoanion to dianion for 2-naphthol-8-sulfonate bound to HSA under the same experimental conditions. This is understandable because proton transfer in HSA-2-naphthol-8-sulfonate complex takes place through the hydrogen bonds which involve no interaction with the surface water molecules as presented in the BSA-2-naphthol-8-sulfonate complex.

PAGE 205

CHAPTER 6 CONCLUSIONS In this study the capability of 2-naphthol-8-sulfonate as a photodissociative probe of water in the albumin binding sites has been tested and characterized. This study on the interaction between 2-naphthol-8-sulfonate and albumins has yielded data which indicate that (1) the kinetic studies of excited singlet state proton transfer can disclose the properties of the immediate environment of the proton emitter. This method can be utilized to probe the properties of water in a small defined volume as found in a specific binding site of a protein because the rate changes reflect directly the physical-chemical properties of the water molecule in the nearby hydration shell. (2) the reduced proton dissociation rate of the BSA (surface) bound 2-naphthol-8sulfonate is due to the protein-water interactions which cause an overall reduction of the reorientational rate of water. This is similar to the situation of 2-naphthol-8-sulfonate in an organic-water mixtures, where molecular mobility for water near the organic solvent molecule is seen to be retarded, as evidenced in the slower translational diffusion and rotational reorientation. (3) the relatively high proton dissociation rate for the HSA (internally) bound 2-naphthol8-sulfonate compared with surface-bound one is probably due to proton transfer via the hydrogen bond networks which contain side groups of some amino acids and/or the 189

PAGE 206

190 peptide backbone as well as a few firmly bound water molecules. (4) there exists ordered water within a binding site on albumin, since the 2.8 A resolution of the available crystal structure of the highly homologous human serum albumin (38) does not allow individual water molecules to be located, little is known about the role of water in the ligand-albumin interactions. (5) the primary factor involved in the 2-naphthol-8-sulfonate-albumin complex formation is the existence of the structured water hydration around the albumin molecule. The character in the vicinity of the binding site affects the binding indirectly through its influence on the structured water formation. (6) the primary binding site for 2-naphthol-8-sulfonate in both BSA and HSA is site II although there could be other binding sites for 2-naphthol-8-sulfonate with low affinity and specificity. (7) despite apparently small differences in the primary structure of serum albumins from different species and gross similarities in the binding of small ligands, the examination of the binding behavior of 2-naphthol-8-sulfonate reveals many dissimilarities between albumins from different species. Thus, 2-naphthol-8-sulfonate can be used to detect surface differences between proteins that perform the same functions in different species, more specifically, differences in the surface properties of proteins differing only in a few amino acids.

PAGE 207

REFERENCES 1. Jusko, W. J., Gretch, M., Drug Metab. Rev . 5, 43 (1976). 2. Chervenak, M. C, and Toone, E. J., J. Am. Chem. Soc . 116 . 10533 (1994). 3. Rand, R. P., Fuller, N. L., Butko, P., Francis, G., and Nicholls, P.. Biochemistry 32. 5925 (1993). 4. Rupley, J. A., Careri, G., Adv. Protein Chem . 41, 37 (1991). 5. Li, J., Vrielink, A., Brick, P., Blow, D. M., Biochemistry 32, 1 1507 (1993). 6. Cowan, S. W., Newcomer, M. E., Jones, T. A., Proteins 8, 44 (1990). 7. Lalondo, J. M., Levenson, J. A., Rose, J. J., Bemlohr, D. A., Banaszak, L. J., J. Biol. Chem. 269, 25339(1994). 8. Lalondo, J. M., Bemlohr, J. J., Banaszak, L. J., Biochemistry 33, 4885 (1994). 9. Quiocho, F. A., Wilson, D. K., Vyas, N. K., Nature 340 . 404 (1989). 10. Zacharias, M., Straatsma, T. P., McCammon, J. A., Biochemistry 32, 7428 (1993). 11. Taylor, T. C, Andersson, 1., Nature Struc. Biol . 3, 95 (1996). 12. Lam, P. Y. S., Jadhav, P. K., Eyermann, C. J., Hogge, C. N., Ru, Y., Bacheler, L. T., Meek, J. L., Otto, M. J., Rayner, M. M., Wong, Y. N., Science 263, 380 (1994). 13. Grzesiek, S., Bax, A., Nicholson, L. K., Yamazaki, Y., Wingfield, P., Stahl, S. J., Eyermann, C. J., Torchia, D. A., Hodge, C. N., Lam, P. Y. S., J. Am. Chem. Soc . il6, 1581 (1994). 14. Pethig, R., Ann. Rev. Phys. Chem . 43, 177 (1992). 15. Saenger, W., Ann. Rev. Biophvs. Bioohys. Chem . 16, 93 (1987). 16. Teeter, M. M., Ann. Rev. Biophvs. Biophys. Chem . 20, 577 (1991). 191

PAGE 208

192 17. Wuthrich, K., Billeter, M., Giintert, P., Luginbuhl, P., Riek, R., Wider, G., Faraday Discuss . 103 . 245 (1996). 18. Zhang, M. S., Matthews, B. W., Protein Sci . 3, 1031 (1994). 19. Pethig, R., In Protein Solvent Interactions : Gregory, R. B., Ed.; Marcel Dekker, Inc.: New York, 1995; Chapter 4, p265. 20. Otting, G., Wuthrich, K., J. Am. Chem. Soc . 111 . 1871 (1989). 21. Otting, G., Liepinsh, E., Farmer, B. T., Wuthrich, K., J. Biomol. NMR i, 209 (1991). 22. Clore, G. M., Bax, A., Wingfield, P. T., Gronenbom, A. M., Biochemistry 29, 5671 (1990). 23. Qian, Y. Q., Otting, G., Wuthrich, K., J. Am. Chem. Soc . 115, 1 189 (1993). 24. Sreenivasan, U., Axelsen, P. H.. Biochemistry 31. 12785 (1992). 25. Finer-Moore, J. S., Kossiakoff, A. A., Hurley, J. H., Earnest, T., Stroud, R. M., Proteins 12. 203 (1992). 26. Weller, A.. Prog. Reaction Kinetics 1. 189(1961). 27. Schulman, S. G., Fluorescence and Phosphorescence Spectroscopy Pergamon Press, Oxford, (1977). 28. Ireland, J. F., Wyatt, P. A. H., Ady. Phys. Org. Chem . 12, 131 (1976). 29. Bell, R. P., In The Proton in Chemistry p.250 (1973). 30. Schulman, S.G., Kelly, R. N., and Gonzalez, N. J., Pure & ApdI. Chem . 59, 655 (1987). 31. Schulman, S. G., Townsend, R. W., Baeyens, W. R. G., Anal. Chim. Acta. 303 . 25 (1995). 32. Huppert, D., & Kolodney, E., Chem. Phys . 63, 401 (1982). 33. Loken, M. R., Hayes, J. W., Gohlke, J. R., & Brand, L., Biochemistry ii, 4779 (1972). 34. Peters, Jr., T., Ady. Protein Chem . 37, 161 (1985). 35. Halliwell, B., Biochem. Pharmacol . 37, 569 (1988).

PAGE 209

193 36. Emerson, Jr., T. E., CRC Crit. Care Med . 17, 690 (1989). 37. Kakizoe, T., Sugimura, T., Jpn. J. Cancer Res . 79, 775 (1988). 38. He, X. M., & Carter, D. C, Nature 358 . 209 (1992). 39. Steinhardt, J., Krijn, J., Leidy, J.. Biochemistry 10. 4005 (1971). 40. Leonard, W. J., Vijai, K. K., & Foster, J. F., J. Biol. Chem . 238 . 1984 (1963). 41. Steiner, R. F., «& Edelhoch, H.. Nature 192 . 873 (1961). 42. Sudlow, G., Birkett, D. J., & Wade, D. N., Mol. Pharmacol . 12, 1052 (1976). 43. Wanwimolruk, S., Birkett, D. J., In Protein Binding and Drug Transport . Tillement, J. P., & Lindenlaub, E., Eds.; Schattauer Verlag. Stuttgart, 1986; pi 33. 44. Fehske, K. J., Muller, W. E., & Wollert, U., Biochem. Pharmacol . 30, 687 (1981). 45. Agmon, N., Chem. Phvs. Lett . 244 . 456 (1995). 46. Wehry, E. L., In Practical Fluorescence ; Guilbault, G. G., Ed.; Marcel Dekker, Inc.: New York, 1990; Chapter 4, pi 27. 47. Kelly, R. N., Schulman, S. G., In Molecular Luminescence Spectroscopy. Part 2 : Schulman, S. G., Ed.; John Wiley & Sons, Inc. 1985; Chapter 6, p461. 48. Narten, A. H., Levy, H. A., J. Chem. Phys . 55, 2263 (1971). 49. Davis, Jr., C. M., Litovitz, T. A., J. Chem. Phys . 42, 2563 (1965). 50. Walrafen, G. E., Fisher, M. R., Hokmabadi, M. S., Yang, W. H., J. Chem. Phvs . 85, 6970(1986). 51. Jorgensen, W. L., J. Chem. Phvs . 77, 4156 (1982). 52. Walrafen, G. E., Fisher, M. R., Hokmabadi, M. S., Yang, W.-H., J. Chem. Phvs . 85, 6970(1986). 53. Montrose, C. J., Bucaro, J. A., Marshall-Coakley, J., Litovitz, T. A., J. Chem. Phvs . 60, 5025 (1974). 54. Bellissent-funel, M.-C, Teixeria, J., J. Mol. Struct . 250, 213 (1991).

PAGE 210

194 55. Tuckerman, M., Laasonen, K., Sprik, M., J. Chem. Phys . 103, 150 (1995). 56. Tuckerman, M., Laasonen, K., Sprik, M., Parrinello, M., J. Phvs. Chem . 99, 5749 (1995). 57. Shi, Z., Ford, J. V., Wei, S., Castleman Jr., A. W., J. Chem. Phvs . 99, 8009 (1993). 58. Halle, B., Karlstrom, G.. .1. Chem. Soc. Faradav Trans. II 79. 1047 (1983). 59. Onsager, L., In Physics and Chemistry of Ice : Whalley, E., Jones, S. J., Gold, L. W., Eds; Royal Society of Canada; Ottawa; 1973; plO. 60. Kunst, M., Warman, J. M., J. Phys. Chem . 87, 4093 (1983). 61. Franck, E. U., Hartmann, D., Hensel, F., Discussion Faraday Soc . 39, 200 (1965). 62. Robinson, R. A., Stokes, R. H., Electrolyte Solutions 2nd Ed.; Butterworths, London; 1959. 63. Westheimer, F. H., Chem. Rev . 61, 265 (1960). 64. Agmon. N.. Int. J. Chem. Kinetics 13. 333 (1981). 65. Meiboom, S., J. Chem. Phvs . 34, 375 (1961). 66. Agmon, N., J. Chim. Phys . 93, 1717 (1996). 67. Mijoule, C, Latajka, Z., Borgis, D., Chem. Phys. Lett . 208. 364 (1993). 68. Zundel, G., Fritsch, J., In The Chemical Physics of Solvation Vol. 2 ; Dogonadze, R. R., Kalman, E., Kornyshev, A. A., Ulstrup, J., Eds.; Elsevier, Amsterdam; 1986; Ch.2. 69. Borgis, D., Tarjus, G., Azzouz, H., J. Chem. Phvs . 97, 1390 (1992). 70. Lobaugh, J., Voth, G. A., J. Chem. Phvs . 104 . 2056 (1996). 71. Inoue, H., Hida, M., Nakashima, N., Yoshihara, K., J. Phys. Chem . 86, 3184 (1982). 72. Waluk, J., Komorowski, S. J., Chem. Phys. Lett . 133 . 368 (1987). 73. Weller, A., Z. Phvs. Chem . 15, 438 (1958). 74. Weller, A., Z. Phvs. Chem . 17, 224 (1958).

PAGE 211

195 75. Brensted, J. N., Chem. Rev . 5, 231 (1928). 76. Stem, O., Volmer, M., Phvs. Z . 20, 183 (1919). 77. Meyer, M. C, Gutman, E. E., J. Pharm. Sci . 57, 895 (1968). 78. Solomon, H. M., Schrogie, J. J., Williams, D., Biochem. Pharmacol . 17, 143 (1968). 79. Pinkerton, T. C, Koeplinger, K. A., Anal. Chem . 62, 21 14 (1990). 80. Chignell, C. F., Fluorescence News 5, 1 (1970). 81. Jun, H. W., Mayer, R. T., Himel, C. M., Luzzi, L. A., J. Pharm. Sci . 60, 1821 (1971). 82. Jun, H. W., Luzzi, L. A., Hsu, P. L., J. Pharm. Sci . 61, 1835 (1972). 83. Sudlow, G., Birkett, D. J., & Wade, D. N., Mol. Pharmacol . U, 824 (1975). 84. Davison, C, In Fundamentals of Drug Metabolism and Drug Disposition : LaDu, B. N., Mandel, H. G., Way, E. L., Eds; Williams and Wilkins, Baltimore; 1971; ch.4. 85. Naik, D. V., Paul, W. L., Schulman, S. G., J. Pharm. Sci . 64, 1677 (1975). 86. Scatchard, G., Ann. N. Y. Acad. Sci . 5i, 660 (1949). 87. Rosenthal, A. E., Anal. Biochemistry 20, 525 (1967). 88. Klotz, I. M., Humston, D. L., Biochemistry 10, 3065 (1971). 89. Job, P., Ann. Chim . 9, 1 13 (1928). 90. Vosburgh, M. J., Cooper, G. R., J. Am. Chem. Soc . 63, 437 (1941). 91. Matsushita, Y., Takahashi, M., Moriguchi, I., Chem. Pharm. Bull . 34, 333 (1986). 92. Panjehshahin, M. R., Yates, M. S., Bowmer, C. J., Biochem. Pharmacol . 44, 973 (1992). 93. Ho, J. X., Holowachuk, E. W., Norton, E. J., Twigg, P. D., Carter, D. C, Eur. J. Biochem . 2 15 . 205 (1993). 94. Mataga, N., Bull. Chem. Soc. Jpn . 3i, 487 (1958). 95. Tolbert, L. M., Nesserlroth, S. M., J. Phvs. Chem . 95, 10331 (1991).

PAGE 212

196 96. Stryer, L. S., Blout, E. R., J. Am. Chem. Soc . 83, HI 1 (1961). 97. Schellman, J., Accounts Chem. Res . 1, 144 (1968). 98. Trynda-Lemiesz, L., Kozlowski, H., Bioorg. Med. Chem . 4, 1709 (1996). 99. Xu, L., Qu, C, Ma, Z.. Chung-Hua-Chung-Liu-Tsa-Chih 17. 115 (1995). 100. Chignell, C. F., Mol. Pharmacol . 5, 244 (1969). 101. Sjodin, T., Roosdorp, N., Sjoholm, I., Biochem. Pharmacol . 25, 2131 (1976). 102. Muller, W. E., Wollert, U., Biochem. Pharmacol . 25, 147 (1975). 103. Chignell, C. F., Mol. Pharmacol . 5, 455 (1969). 104. Fleury, F., Kudelina, I., Nabiev, I., FEBS Lett . 406. 151 (1997). 105. Klotz, I. M., Urquhart, J. M., J. Phvs. Coll. Chem . 53, 100 (1949). 106. Jorgenson, M. J., Hartter, D. R., J. Am. Chem. Soc . 85, 878 (1963). 107. Rosin, J., In Reagent Chemicals and Standards : D. Van Nostrand Company, Inc., Princeton, New Jersey; 1 967. 108. Hasselbacher, C. A., Waxman, E., Galati, L. T., Contino, P. B., Ross, J. B. A., Laws, W. R., J. Phvs. Chem . 95, 2995 (1991). 109. El-Bayoumi, M. A., J. Phvs. Chem . 80, 2259 (1976). 110. Agmon, N., Goldberg, S. Y., Huppert, D., J. Mol. Liq . 64, 161 (1995). 111. Htun, M. T., Suwaiyan, A., Klein, U. K. A., Chem. Phvs. Lett . 243 . 506 (1995). 1 12. Roses, M., Buhvestov, U., Rafols, C, Rived, F., Bosch, E., J. Chem. Soc. Perkin Trans . 2, 1341 (1997). 113. Soper, K. A., Faradav Discuss . 103 . 41 (1996). 114. Tanaka, H.. J. Chem. Phvs . 86, 1512 (1987). 115. Zichi, D. A., Rossky, P. J., J. Chem. Phvs . 84, 2814 (1986). 1 16. Luzar, A., Faradav Discuss . 103 . 29 (1996).

PAGE 213

197 117. Peters, Jr., T., All about albumin . Academic Press, Inc.; 1996. 118. Wanwimolruk, S., Birkett, D. J., Brooks, P. M., Mol. Pharmacol . 24, 458 (1983). 119. Maruyama, K., Nishigori, H., Iwatsuru, M., Chem. Pharm. Bull . 33, 5002 (1985). 120. Noctor, T. A. G., Pham, C. D., Kaliszan, R., Wainer, I. W., Mol. Pharmacol . 42, 506 (1992). 121. Wilting, J., 'T Hart, B. J., De Gier, J. J.. Biochim. Biophvs. Acta 626, 291 (1980). 122. Wanwimolruk, S., Birkett, D. J., Biochim. Biophvs. Acta 709 . 247 (1982). 123. Otagiri, M., Masuda, M., Imai, T., Imamura, Y., Yamasaki, M., Biochem. Pharmacol . 38, 1 (1989). 124. Irikura, M., Takadate, A., Goya, S., Otagiri, M., Chem. Pharm. Bull . 39, 724 (1991). 125. Muller, W. E., Wollert, U., Mol. Pharmacol . U, 52 (1975). 126. Kosa, T., Maruyama, T., Otagiri, M., Pharm. Res . 14, 1607 (1997). 127. Matulis, D., Lovrien, R., Biophvs. J . 74, 422 (1998). 128. Varela, A. S., Macho, M. I. S., Minones, J., J. Pharm. Sci . 81, 842 (1992). 129. Gerig, J. T., Katz, K. E., Reinheimer, J. P.. Biochim. Biophvs. Acta 534 . 196 (1978). 130. Ma, J. K. H., Hsu, P-L., Luzzi, L. A., J. Pharm. Sci . 63, 32 (1974). 131. Kirk, W. R., Kurian, E., Prendergast, F. G., Biophvs. J . 70, 69 (1996). 132. Brzezinski, B., Zundel, G., Faradav Discuss . 103, 363 (1996). 133. Rammelsberg, R., Huhn, G., Liibben, M., Gerwert, K., Biochemistry 37 . 5001 (1998). 134. Coutre, J., Gerwert, K., FEBS Lett . 398, 333 (1996). 135. Eigen, M., Hammes, G. G., Kustin, J., J. Am. Chem. Soc . 32, 3481 (1960). 136. Gutman, M., Nachliel, E., Ann. Rev. Phvs. Chem . 48, 329 (1997).

PAGE 214

BIOGRAPHICAL SKETCH Qiao-qing Di was bom in Shanghai, China. In August, 1983, she received a Bachelor of Science degree in medicinal chemistry from the School of Pharmacy, Shanghai Medical University, China. In 1993 she entered graduate school at the University of Florida College of Pharmacy. She received her Doctor of Philosophy degree in medicinal chemistry in 1998 under the guidance of Dr. Stephen G. Schulman. 198

PAGE 215

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in sc^pe and as a dissertation for the degree of Doctor of Philosophy. Stephen p. Schulman, Chair Professor of Meaicinal Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fiilly adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. ^ // JohtfWPerrin, Professor of Medicinal Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosoph)^ Kenneth B. Sloan Professor of Medicinal Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fiilly adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. les D. Winefordner faduate Research Pro/essor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Assistant Professor of Pharmaceutics

PAGE 216

This dissertation was submitted to the Graduate Faculty of the College of Pharmacy and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December, 1998 Dean, College of Pharjpacy Dean, Graduate School