Biomolecules in Disease and Disease Treatment

University of South Florida Scholar Commons

Graduate Theses and Dissertations Graduate School

11-15-2019

The Activity and Structure of Cu2+-Biomolecules in Disease and Disease Treatment

Darrell Cole Cerrato University of South Florida

Follow this and additional works at: https://scholarcommons.usf.edu/etd

Part of the Chemistry Commons

Scholar Commons Citation Cerrato, Darrell Cole, "The Activity and Structure of Cu2+-Biomolecules in Disease and Disease Treatment" (2019). Graduate Theses and Dissertations. https://scholarcommons.usf.edu/etd/8628

This Dissertation is brought to you for free and open access by the Graduate School at Scholar Commons. It has been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of Scholar Commons. For more information, please contact [email protected]. The Activity and Structure of Cu2+-Biomolecules in Disease and Disease Treatment

Darrell Cole Cerrato

A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Chemistry College of Arts and Sciences University of South Florida

Major Professor: Li-June Ming, Ph.D. Randy Larsen, Ph.D. Jianfeng Cai, Ph.D. Yu Chen, Ph.D.

Date of Approval: November 4th, 2019

Keywords: Alzheimer’s, beets, bacitracin, oxidative stress, thiabendzole, betanin, phytochemicals, salicylic acid, amyloid

DEDICATION

I dedicate this to you, Reader. If you are reading this, it is likely that you are family, friends, colleagues, researchers, or someone who is curious about the world we live in. The foundation of studying science is that we rely on each other. We use peer review; we train those that don’t have experience; we advance science based on what was discovered before us. For those of you reading this that are not chemists or professional scientists, this is for you as well. Science is not for a certain sect of society.

It is for everyone, and we all benefit from the curiosity of others. Your enthusiasm for learning more about the world around you is the same reason I remained motivated to complete this degree. If you have need of the information within this dissertation, I hope it helps you. We live in a world we must share, and I would like to share this contribution with you.

ACKNOWLEDGEMENTS

If I have seen further it is by standing on the shoulders of Giants.

- Isaac Newton, 1675

First and foremost, I must acknowledge my best friend Erin Cerrato. Your dedication and belief in me carried me more than I ever knew and will probably ever completely comprehend. Without you I do not think I would have ever applied to the

University of South Florida, much less have ever worked for a Ph.D. in chemistry. You worked just as hard for this as I did, seeing more in me than I often saw in myself. I would also like to acknowledge Max, Andi, and Alex. I realize you cannot read, because, you know, you’re a couple of dogs and a cat, but you guys, ironically, maintained my humanity and compassion throughout some of the toughest times. I don’t deserve any of you, and love you all so much more for it.

I must also acknowledge my academic family and friends, Shahed, Christie,

Yassin, Anne-Claire, Briana, Brian, Andrew, Anne-Marie, and Arthur, and so many more. You were all there for me both academically, emotionally, spiritually, scientifically, grammatically, etc. The mid-afternoon coffees, post-lab drinks, and dragging me from my apartment after marathon writing sessions maintained my motivation and sanity.

I would also like to recognize my mentors. Dr. Randy Larsen, you taught me that chemistry is more than facts, but it’s a lens through which we view and explore nature.

You have taught me how to communicate science and inspired me to keep looking at the big picture. Finally, I must acknowledge my major professor and advisor Dr. Li-

June Ming. In the moment I might have lamented how you would not give answers but made us figure it out instead. I cannot thank you enough for that now. I learned that a portion of earning this degree isn’t only learning the material and doing the research, but also learning how to think. Regardless of what life throws my way, I think I can handle it with the tools you have let me learn to use.

I am a better person for knowing all of you.

TABLE OF CONTENTS

List of Tables ...... iii

List of Figures ...... iv

List of Abbreviations and Units ...... xiii

Abstract...... xv

Chapter One: Life is So Metal: A Survey of Copper in Life, Disease, and Disease Treatment from Humans to Agriculture ...... 1 A Very Brief Historic Overview of Metals in Biology ...... 1 Copper in Biology ...... 7 Role of Copper in Neurodegenerative Disease ...... 22 Metallo-Antibiotics ...... 28 Copper and Secondary Metabolite Interactions ...... 32 Conclusion ...... 35 References ...... 36

Chapter Two: “Beeting” Alzheimer’s Disease: The Structure-Activity Role of Betanin Towards the Redox Active CuAβ Implicated in Alzheimer’s Disease ...... 55 Background ...... 55 Results and Discussion Antioxidation of CuAβ by Bn ...... 65 Betanin-Copper Binding ...... 86 Beet Extract Redox Inhibition and Metal Binding ...... 99 Concluding Remarks ...... 103 Materials and Methods...... 106 References ...... 108

Chapter Three: Inhibition of Copper(II)-Bacitracin Mediated Oxidation By Salicylic Acid and Thiabendazole ...... 120 Background ...... 120 Results and Discussion Salicylic acid as an inhibitor against Cu-bactracin ...... 134

Thiabendazole as an inhibitor against Cu-bacitracin ...... 159 Concluding Remarks ...... 173 Materials and Methods ...... 175 References ...... 176

Chapter Four: Three-Dimensional Structure Elucidation of a Novel Peptide “Foldamer” and NMR Evaluation of a Metallo-Peptoid ...... 187 Background ...... 187 Results and Discussion D-sulfono-γ-AApeptide/L-amino acid structure and conformation elucidation ...... 192 Ac-FHFH, γ-Peptoid, and Peptide metal binding ...... 203 Concluding Remarks ...... 211 Materials and Methods ...... 212 References ...... 214

Appendix A ...... 221 Permissions ...... 221

LIST OF TABLES

Table 2.1 – Michaelis-Menten parameters of betanin inhibition of aerobic oxidation determined by fitting each line in Figure 2.6 to the Michaelis- Menten hyperbolic equation. (a) based on dimeric CuAβ concentration of 2.5 μM ...... 72

Table 2.2 - MM parameters of betanin inhibition of peroxide mediated oxidation determined by fitting each line in Figure 2.12 to the MM hyperbolic equation. (a) based on monomeric concentration of CuAβ...... 79

Table 2.3 - MM parameters of betanin inhibition of hydrogen peroxide mediated oxidation with respect to H2O2 determined by fitting each line in Figure 2.14 to the MM hyperbolic equation. (a) based on dimeric concentration of CuAβ ...... 82

Table 3.1 - MM parameters determined from fitting Figure 3.11 as a function of DTBC titration with 100 μM CuBc ...... 141

Table 3.2 - MM parameters for SA inhibition of 2 mM DTBC turnover as H2O2 was titrated in the presence of 10 μM CuBc ...... 148

Table 3.3 - MM parameters determined by fitting the non-linearized data from Figure 3.32, where kcat and catalytic efficiency were determined using 100 μM CuBc ...... 162

Table 3.4 - MM parameters of hydrogen peroxide and 10 μM CuBc mediated turnover in the presence of Tbz ...... 167

Table 4.1 - complete signal assignment for oligomer 5 of D-sulfono-γ- AApeptide/L-amino acid ...... 199

iii

LIST OF FIGURES

Figure 1.1 - Comparison of chemical distribution of the most common inorganic elements in earth’s crust, oceans, and of an average 80.7 kg human body ...... 3

Figure 1.2 - Plastocyanin (PDB: 2PCF) enzyme of Spinacia oleracea with focus on the copper active site and its ligands on the right ...... 11

Figure 1.3 - PDB ID: 3MND. Oxidized form of Cu,Zn superoxide dismutase enzyme (left) and the active site of copper (orange) and zinc (right) ...... 14

Figure 1.4 - PDB ID: 2EIE galactose oxidase (left) with emphasis on the copper active site (right) ...... 16

Figure 1.5 – Oxidation of D-galactoside catalyzed by galactose oxidase and O2 ...... 16

Figure 1.6 - a proposed simplified reaction mechanism for electron transfer at the copper center of galactose oxidase ...... 17

Figure 1.7 – oxidation of DTBC to form o-quinone product ...... 20

Figure 1.8 - PDB ID: 2P3X Catechol oxidase with a focus on the di-copper active site, with a ~45° horizontal rotation in the final frame for better visualization of the binding pocket histidine residues ...... 20

Figure 1.9 – abbreviated catechol oxidase mechanism, showing key mechanistic structures ...... 21

Figure 1.10 – depiction of “Healthy Brain and Severe AD Brain” by National Institutes of Health ...... 25

Figure 1.11 - line structure of bleomycin (bottom) and cobalt-bound structure of the metal binding portion of BM (top), adapted from PDB ID 1JIF ...... 31

Figure 1.12 – general flavonoid structure with number assignments (left), myricetin (middle), and quercetin (right)...... 34

Figure 2.1 – Aqueous solution NMR structure of apo-Aβ adapted from Vivekanandan, et al., highlighting the metal-binding sidechains, His-6, His-13, and His-14, for reference, relaxed-eye view. PDB ID: 2LFM ...... 57

Figure 2.2 - a proposed structure of cobalt(II) bound to Aβ1-16 ...... 59

Figure 2.3 - line representation of betanin and its predominant source, beetroot. Photo by Cole Cerrato ...... 61

Figure 2.4 – derivatives and degradation products of betanin ...... 63

Figure 2.5 – chemical structure of catechols including neutotransitters ...... 66

Figure 2.6 - Adapted reaction mechanism for catechol (adapted for DTBC) with CuAβ proposed by da Silva, et al ...... 67

Figure 2.7 – inhibition of 3.2 mM DTBC oxidation by 3.5 μM CuAβ as a function of Bn in a pH 6.0 buffered solvent of 1:1 MES:MeOH ...... 68

Figure 2.8 - simplified reaction pathway for CuAβ oxidation of DTBC and its quantification using Michaelis-Menten equation (Equation 2.1) and its linearized form (Equation 2.2) ...... 69

Figure 2.9 - 5 μM CuAB oxidizing DTBC in the presence of Bn, in 1:1 MeOH:MES at pH 6.0 at 25 °C, Michaelis-Menten fitting (top) and Lineweaver-Burk fitting (bottom) ...... 71

Figure 2.10 - mixed inhibition pathway of Bn towards CuAβ oxidation of DTBC featuring competitive (binary EI complex) and uncompetitive (ternary ESI complex pathways) ...... 73

Figure 2.11 - MM fitting for Figure 2.7 as a function of Bn for the determination

of the Kic dissociation constant of Bn–CuA binary complex...... 74

Figure 2.12 - Plot of the data from Table 2.1 fitted to Equation 2.5 in determination of dissociation constant, Kiu, of Bn ternary binding of Bn- CuAβ-DTBC (R2 = 0.9) ...... 75

Figure 2.13 - Oxidation of 3.2 mM DTBC as a function of H2O2 titration to 5 μM (CuAβ)2 without inhibition in 1:1 MES:MeOH solvent buffered at pH 6.0, 25 °C, kcat of (9.3 ± 0.7) × 10—3 sec—1 ...... 77

Figure 2.14 - Double-reciprocal plot for DTBC titration in the presence of varying amounts of Bn with constant 30 mM H2O2 and 10 μM CuAβ ...... 78

Figure 2.15 - Replot for Kic dissociation constant determination of Bn binary complex in the presence of 30 mM hydrogen peroxide ...... 80

Figure 2.16 – Bn inhibition of H2O2 as a substrate for DTBC (1 mM) turnover by 5 μM CuAβ ...... 82

Figure 2.17 - simplified reaction pathway in describing Bn competitive and uncompetitive inhibition pathways towards H2O2 reactivity with CuAβ ...... 83

Figure 2.18 - Replot for Kic dissociation constant determination of Bn-CuAβ binary complex, inhibiting H2O2 binding ...... 84

Figure 2.19 - Replot for Kiu dissociation constant determination of Bn-CuAβ-H2O2 ternary complex turnover ...... 86

Figure 2.20 - normalized UV-vis spectrum for the titration of Cu2+ to Bn in MES pH 6.0 buffer ...... 87

Figure 2.21 - Job Plot (continuous variation) for Cu-Bn absorbance due to binding as a function of the mole fraction of Cu2+, χCu, in 100 mM pH 6.0 MES buffer...... 88

Figure 2.22 - Titration of Cu2+ to Bn at 9.1 μM, in 100 mM pH 6.0 MES buffer, where the change in absorbance at 535 nm is measured as a function of Cu2+ titrated...... 90

Figure 2.23 – paramagnetic NMR 1H spectrum of betanin with 0.25 (C), 0.5 (B), and 1.0 (A) stoichiometric equivalents of Co2+ bound to betanin at 1.1 mM in D2O ...... 92

Figure 2.24 - iminodiacietic acid compared to the proposed metal-binding region of betanin ...... 94

Figure 2.25 - apo-iminodiacetic acid (top) with 1 stoichiometric equivalent (2.0 mM) of Co2+ in D2O (bottom) ...... 95

Figure 2.26 - 1H NMR paramagnetic spectrum of 1 mM nitrolotriacetic acid with 1 stoichiometric equivalent of Co2+ in D2O ...... 96

Figure 2.27 - a tentative structure of Cu2+-betanin (glucose not included for minimization simplicity) minimized using Chem3D. Relaxed-eye view ...... 97

Figure 2.28 - Inhibition of free Cu2+ oxidation by Bn (1 μM Cu, 4 mM DTBC) in 1:1 100 mM pH 6.0 MES:MeOH solvents ...... 99

Figure 2.29 - inhibition of free Cu2+ oxidation by beet extract as a function of Bn concentration ...... 102

Figure 2.30 - 260 μM Bn in D2O as part of beet extract in the presence of 1 stoichiometric equivalent of Co2+ ...... 103

Figure 2.31 - general purification of Bn through homogenization, micropore filtration, and column chromatographies ...... 107

Figure 3.1 - relaxed-eye stereoscopic view of apo-bacitracin A1 minimized using ChemDraw3D software ...... 121

Figure 3.2 - Amino acid sequence for bacitracin with congener variations for R, X, and Y amino acids represented in Figure 3.3...... 122

Figure 3.3 - Side chain variants for R, X, and Y amino acids for the most common bacitracin congeners ...... 123

Figure 3.4 - Line representation for bacitracin A1 congener ...... 124

Figure 3.5 - Calculated structure of cobalt-bound bacitracin by NMR, relaxed-eye view, modeled based on NMR T1 data. Image used with permission by Elsevier ...... 126

Figure 3.6 - Aspirin bottle ca. 1899. Creative Commons 3.0 Image License ...... 129 vii

Figure 3.7 - line representation of salicylic acid ...... 129

Figure 3.8 - line representation of thiabendazole ...... 131

Figure 3.9 - Proposed catechol oxidation mechanism for CuBc ...... 135

Figure 3.10 – common polyphenol and catechol-containing secondary metabolites produced by plants ...... 136

Figure 3.11 - Titration of DTBC to 10 μM CuBc (1:1) under aerobic conditions ...... 137

Figure 3.12 - simplified aerobic catechol oxidase-like pathway for CuBc oxidation of catechol to quinone ...... 138

Figure 3.13 - titration of SA to 10 μM CuBc in 1:1 MeOH:HEPES (pH 7.0, 100 mM), 2 mM DTBC ...... 139

Figure 3.14 - Lineweaver-Burk plot for varying concentrations of salicylate as DTBC is titrated in the presence of 100 μM CuBc in 100 mM pH 7.0 HEPES:MeOH ...... 140

Figure 3.15 - competitive and uncompetitive SA inhibition pathways of CuBc oxidation of DTBC ...... 142

Figure 3.16 - Fitting of apparent Vmax and Km as a function of SA concentration to determine Kic of SA, fitted to Equation 2.4 ...... 143

Figure 3.17 - Fitting of apparent Vmax as a function of SA as fitted by Equation 2.5 (R2=0.92) ...... 145

Figure 3.18 - titration of hydrogen peroxide to 10 μM CuBc and constant 1 mM DTBC, normalized for fitting to determine Km,app ...... 146

Figure 3.19 - titration of SA to 30 μM CuBc and 30 mM H2O2 in 100 mM pH 7.0 1:1 HEPES:MeOH ...... 147

Figure 3.20 - LB plot of inhibition of DTBC oxidation rate by SA as a function of H2O2 catalyzed by 10 μM CuBc with constant 2 mM DTBC in 100 mM HEPES:MeOH (1:1) at pH 7.0 25.0 °C ...... 148 viii

Figure 3.21 - Vmax,app data fitted to Equation 2.5 as a function of inhibitor SA concentration in determination of uncompetitive dissociation constant, Kiu ...... 149

Figure 3.22 - 5 mM apo-salicylate in D2O (4.8 ppm) using a presaturation pulse sequence at the HDO frequency ...... 151

Figure 3.23 - 0.8 eq Co2+ salicylate in 10% D2O, with 1 stoichiometric equivalent of TEA to include phenolate binding ...... 152

Figure 3.24 - EXSY of 0.2 stoichiometric equivalents of Co2+ with salicylate in 10% D2O and signal assignments ...... 153

Figure 3.25 - 1H NMR of 4 mM apo-bacitracin in d6-DMSO ...... 154

Figure 3.26 - CoBc in d6-DMSO with 0.8 eq Co2+, using a presaturation pulse sequence, diagmagnetic signal omitted for presentation of paramagnetic signals ...... 154

Figure 3.27 - Paramagnetic 1H NMR spectra of Co2+-bacitracin as SA is titrated by the stoichiometric equivalent, where 1 equivalent is 10.2 mM and SA is 0.0 eq, 0.3 eq, 0.5 eq, 1.0 eq, and 4.0 eq (red) SA going down ...... 156

Figure 3.28 - diamagnetic region upon addition of SA to 10.2 mM CoBc in d6- DMSO. Apo-Bc (top), and CoBc with the addition of 0.0 eq (green), 0.3 eq, 0.5 eq, 1.0 eq, and 4.0 eq (pink) SA going down ...... 157

Figure 3.29 – Focused 1H spectrum of 0.8 eq CoBc 0 SA (top) and 4 eq SA (bottom) in d6-DMSO ...... 158

Figure 3.30 - inhibition of 2 mM DTBC oxidation as a condition of Tbz titration to 10 μM free Cu2+ in HEPES buffer (pH 7.0) with 10% DMSO for catechol and Tbz solubility ...... 160

Figure 3.31 - oxidation rate of 2 mM DTBC upon addition of Tbz to 10 μM CuBc in 1:1 DMSO:HEPES 100 mM buffer at pH 7.0 ...... 161

Figure 3.32 - linearized LB plot for DTBC titrated to 100 μM CuBc in 100 mM pH 7.0 1:1 HEPES:DMSO where Tbz varied up to 4.0 mM ...... 162

Figure 3.33 - fitting for determination of Kic of Tbz towards CuBc under aerobic conditions (R2 = 0.8) by fitting data from Table 3.3 ...... 163

Figure 3.34 - fitting for determination of Kiu of Tbz towards CuBc under aerobic conditions (R2 = 0.8) by fitting data from Table 3.3 ...... 164

Figure 3.35 - titration of Tbz in the presence of 10 μM CuBc, 30 mM H2O2, and 2 mM DTBC in 100 mM pH 7.0 1:1 HEPES:DMSO ...... 166

Figure 3.36 - LB plot for titration of hydrogen peroxide to 10 μM CuBc measuring oxidation rate 3 mM DTBC with varying amounts of Tbz up to 8 μM ...... 167

Figure 3.37 - Fitting of Vmax,app/Km,app as a function of [Tbz] according to Equation 2.5 for 2.0 mM DTBC oxidation by the CuBc complex in the presence of increasing amounts of H2O2 to afford the competitive dissociation constant, Kic ...... 168

Figure 3.38 - Fitting of Vmax,app as a function of [Tbz] according to Equation 2.5 for 2.0 mM DTBC oxidation by the CuBc complex in the presence of increasing amounts of H2O2 to afford the uncompetitive dissociation constant, Kiu ...... 169

Figure 3.39 - Paramagnetic spectrum of 0.9 molar equivalents Co2+ to Tbz in d4- MeOD (top) and d6-DMSO (bottom). Spectrum with DMSO zoomed 260k× relative to diamagnetic region, acquired using a Super-WEFT sequence ...... 171

Figure 3.40 - titration of Tbz to 4 mM CoBc in d6-DMSO, where Tbz increased going down: 0 Tbz (top, black), 0.5 eq Tbz (red), 1.0 eq (blue), 5.0 eq Tbz (orange), and 8.0 Tbz (bottom, greed). Best viewed in color for easy comparison ...... 172

Figure 4.1 - Hydrolytic reaction performed a by protease along peptide backbone ...... 190

Figure 4.2 - Variations in peptide backbones via addition of carbons (second and third) or inversion of stereo-center (last) compared to the most common natural peptide (first) ...... 191

Figure 4.3 – Structural motifs for the design of D-sulfono-γ-AApeptide/L-amino acid ...... 193

Figure 4.4 - line representation of D-sulfono-γ-AApeptide/L-amino acid oligomer 5 ...... 194

Figure 4.5 - full 1D 1H spectrum of D-sulfono-γ-AApeptide/L-amino acid oligomer 5 using WET solvent suppression pulse sequence using an Inova600 spectrometer, solvent signals at 3.3 and 4.8 ppm ...... 196

Figure 4.6 - 1D 1H spectrum of D-sulfono-γ-AApeptide/L-amino acid with focus on side chains (top), backbone CH groups (middle), and aromatic and backbone NH groups (bottom) ...... 197

Figure 4.7 - overlaid 2D spectra of D-sulfono-γ-AApeptide/L-amino acid in d6- DMSO, followed by focus on the NH/aromatic correlations (f1 spectrum) to the side chain correlations (f2 spectrum) (viewed best in color) ...... 198

Figure 4.8 – i+3 NOE signal for 6aHγ and the backbone NH of residue 9a (signal labeled) ...... 201

Figure 4.9 - i+3 NOE signal for 7Hγ’ and the backbone NH of residue 10 (signal labeled) ...... 201

Figure 4.10 - detected NOESY signals for i+1, i+2, and i+3 correlations. Best viewed in color ...... 202

Figure 4.11 - D-sulfono-γ-AApeptide/L-amino acid proposed minimized structure for oligomer 3 ...... 203

Figure 4.12 - peptide and peptoid structures designed to probe difference in metal binding ...... 204

Figure 4.13 - overlaid spectra for 4 mM Ac-FHFH (top) and γ-AA-FHFH (bottom) and 1 stoichiometric equivalent of Co2+, 50× zoom applied to γ-AA- FHFH far downfield region signal ...... 207

Figure 4.14 - drop of deuterated water added (top) to Ac-FHFH peptoid to confirm NH signal assignment ...... 208

Figure 4.15 – Line representation Ac-γ-FHFH binding Co2+ through N ε2 ...... 209

Figure 4.16 - 4 mM γ-AA-FHFH peptoid in d6-DMSO (top) and half d6-DMSO, half d3-MeOH (bottom) ...... 210

Figure 4.17 - 4 mM γ-AA-FHFH peptoid in d6-DMSO (top) and half d6-DMSO, half d3-MeOH (bottom) ...... 210

xii

LIST OF ABBREVIATIONS AND UNITS

δ Chemical shift ε Molar absorptivity λmax Lambda max Å Ångstrom, 10—10 meters Aβ Amyloid beta or beta amyloid AD Alzheimer’s disease ALS Amyloid lateral sclerosis ATP Adenosine triphosphate Bc Bacitracin BM Bleomycin Bn Betanin Cat Eff Catalytic Efficiency CD Circular dichroism CDC Centers for Disease Control cm Centimeter CO Catechol oxidase Da Daltons DMSO Dimethylsulfoxide DNA Deoxyribonucleic acid DTBC Di-tert-butylcatechol DTBQ Di-tert-butylquinone EDTA Ethylenediaminetetraacetic acid EPR Electron paramagnetic resonance EtOH Ethanol eV Electron volt EXAFS Extended X-ray absorption fine structure gCOSY Gradient correlated spectroscopy GO Galactose oxidase H Hydrogen or proton HEPES [4-(2-hydroxyethyl)piperazine-1-yl]ethanesulfonic acid kcat Turnover number Km Michaelis constant xiii

Km,app Apparent Michaelis constant Ki Inhibition constant LFSE Ligand field stabilization energy M Molar MeOH Methanol MES 2-morpholin-4-ylethanesulfonic acid MHz Megahertz MM Michaelis-Menten (kinetics) Mr Myricetin nm Nanometer NMR Nuclear magnetic resonance NOESY Nuclear Overhauser effect spectroscopy NTA Nitrilotriacetic acid RNA Ribonucleic acid ROS Reactive oxidative species RP-HPLC Reverse phase high performance liquid chromatography PDB Protein Databank Ppm Parts per million Qr Quercetin s seconds SA Salicylic acid SOD Superoxide dismutase Tbz Thiabendazole TOCSY Total correlated spectroscopy tRNA Transfer ribonucleic acid USF University of South Florida UV-vis Ultra violet-visible Vmax Maximum reaction rate/velocity Vmax,app Apparent Vmax V Volt WHO World Health Organization

xiv

ABSTRACT

Despite its relatively low concentration in the human body, copper is an essential transition metal ion. Under normal circumstances, it is a well-regulated trace element that is introduced via diet or environment depending on the organism. Its strong reactivity and binding to some biologically essential chemicals can be detrimental to the host organism. Herein the oxidative chemistry of copper and its possible prevention in humans and agricultural organisms is discussed, as well as the potential avenues as metal-targeting disease treatment via novel-peptide and peptoid products.

In the first study, copper is implicated in the progression and onset of

Alzheimer’s disease. In this study I show a chemical that is found in beetroot, betanin, can bind to the copper-amyloid beta complex, the complex implicated in AD, using paramagnetic NMR techniques. Betanin also shows a reduction in catechol oxidase-like activity of CuAβ towards DTBC under oxidative and peroxidative conditions with a Kic and Kiu in the presence of ambient of O2 of 1.11 ± 0.03 and 2.2 ± 0.4 μM, respectively.

Betanin also showed inhibition towards H2O2-mediated CuAβ oxidation with a Kic and

Kiu of 0.76 ± 0.04 and 1.28 ± 0.08 μM, respectively.

Secondly, the potential activity of copper from an agricultural perspective is presented where copper-bound bacitracin from the soil bacteria Bacillus subtilus is

xv inhibited by natural products that may occur in the environment. Natural and synthetic metal-binding oxidative inhibitors, salicylic acid and thiabendazole, respectively, were chosen based on the potential interaction with CuBc based on their place in medicine and/or the environment around plants. Paramagnetic NMR techniques show each compound can form the ternary complex with CuBc and the subsequent inhibition of

redox activity. Salicylic acid of the willow tree showed inhibition constants of Kic,O2 of

5.24 ± 0.03 mM, Kiu,O2 of 21 ± 3 mM, and a Kiu,H2O2 15 ± 3 μM. Thiabendazole revealed

inhibition constants of Kic,O2 of 1.1 ± 0.4 mM, Kiu,O2 of 1.5 ± mM, Kic,H2O2 6.3 ± 0.7 μM,

and Kic,H2O2 of 13.2 ± 2.6 μM.

In the last chapter a novel γ-AApeptide right-handed helical “foldamer” and the structure of a similar γ-AApeptide were examined. Peptoids are a type of peptidomimetic compound that is synthesized with the α-carbon sidechain substituted onto the nitrogen on the peptide backbone. One of the key advantages in using peptoids is their increase in bioavailability given the inability to be metabolized by proteases. The first half of this study uses NMR techniques to support the structural information about a novel γ-AApeptide. The examination of incorporating metals into these novel peptides in the second half of the study would help elucidate the role of metals as they are often under-utilized in drugs or drug targets. This study shows the differences in binding resulting in a difference in metal binding as well as a difference in reactivity.

xvi

CHAPTER 1: INTRODUCTION

Life Is So Metal: A Survey of Copper-Peptide Interactions in Life, Disease, and

Disease Treatment from Humans to Agriculture

Because the history of evolution is that life escapes all barriers. Life breaks free. Life expands to new territories. Painfully, perhaps even dangerously. But life finds a way.

- Jurassic Park by Michael Crichton

I. A Very Brief History and Introduction of Metals in Biology

Metals have an undeniable role in life and biology as we know it. The oxygen all aerobic organisms breathe was formed in the chloroplast of plants using copper in plastocyanin to transport electrons. Those electrons are bound for manganese clusters responsible for the formation of O2 from water and carbon dioxide. That dioxygen might then eventually be bound for the lungs of an aerobic organisms that uses iron in hemoglobin to bond to and transport the O2 to the proper muscle or organ. The example of using metals in respiration is only a very small fraction how biological systems uses transition metals for essential biological purposes. Their role in biology is complexly 1 interwoven with basic function of the organism, and their dysregulation are implicated in a number of diseases.

The history of metals in biology is intricately linked to the presence of atmospheric and dissolved oxygen concentrations. For billions of years early life developed under reductive conditions largely due to the low concentration of oxygen in the atmosphere (<0.001% of current levels) and oceans, making metals relatively inactive and insoluble in their reduced states.1 “Great oxygenation or oxidation events” that occurred around 2.5 billion years ago eventually raised the oxygen content to ~20% of the current atmospheric composition, which caused mass extinctions of the mainly anaerobic life.2 These early lifeforms did not have the molecular and physiological capacity to manage the reactivity of molecular oxygen and ambient oxidized compounds like transition metals; O2 was effectively a poison to most early life.3-5 The organisms that adapted in the harsh oxidative environment were those that could reduce the toxicity of O2 or developed the means to reduce it. Many of these organisms not only found a means to prevent toxicity, within limits, but developed means to incorporate O2 into biological functions that use it as an electron acceptor in the eventual production of ATP.6 Many of these systems that developed under aerobic conditions actually lead to a more efficient means of producing nearly 15 to 20 times more ATP! And many of them that evolved during this time period of an oxygen boom involved metals, the most prominent of which involved iron, given its several easily accessible oxidation states and high crust and oceanic concentrations compared to other

2 metals. The relative distribution of most common transition metals are constant between the soil, ocean, and higher organisms as shown in Figure 1.1. Many of the reactions discussed herein are between a copper ion, a reduced form of oxygen, and an

O2 molecule; and their associations in life are as old as life itself.

0 Mn Fe Co Ni Cu Zn -2

Log mass composition mass Log -4

-6 Earth's Crust (mg/kg) Earth's Oceans (mg/L) Human Body (mg/kg)

Figure 1.1 - Comparison of chemical distribution of the most common inorganic elements in earth’s crust, oceans, and of an average 80.7 kg human body. Data adapted with permission from CRC Handbook of Chemistry and Physics.7,8

Still, the role of metal in biology is diverse beyond its interaction with molecular oxygen. Copper(II) and iron(III) may have played an essential role in the formation of early amino acids by lowering the activation energy required to perform the condensation reaction to form peptide bonds.9,10 Transition metals can also aid in

3 maintaining protein structure, such as the zinc-finger motif, by acting as a binding site for dioxygen transport, like copper in the hemocyanin of mollusks, or by playing an active role in the chemistry of the enzyme, such as molybdenum and iron in nitrogenase for reducing N2 to ammonia in nitrogen-fixation bacteria, such as Azotobacteria.11-13

Despite their relatively low concentrations in the environment compared to organic compounds, metals are estimated to be associated with approximately one-third of all proteins. Iron, zinc, and copper comprise a majority of the transition metals in biology, however, many of the other transition metals, such as cobalt, nickel, manganese, and vanadium, can be found in trace or ultra-trace amounts in many biological systems.14 In fact, the deficiency of any of these metals can lead to health problems that include anemia, heart disease, and slowed growth of many organisms and microorganisms. In a recent story that made headlines, a “fussy” eating teenager that went blind and started to lose hearing was found to have dangerously low levels of iron, copper, selenium, and vitamin B12, a cobalt-bound cofactor!15

Dysregulation of homeostatic regulation of redox-active metals can lead to the formation of free-radicals and reactive oxidative species (ROS). Biologically, ROS are most commonly chemical byproducts of mitochondrial reactions with O2. The production of reactive oxidative species (ROS) is not inherently detrimental as some research has shown some ROS might also be used as messenger molecules for calcium regulation or in the chemical pathway triggering cell apoptosis.16,17 The need for strict regulation is essential, though, as dysregulation is implicated in a number of

4 diseases!18,19 Some of the most efficient enzymes in nature are those that deal with quenching the presence or production of reactive oxidative species (ROS), such as superoxide (O2—), peroxide (O22—), and hydroxyl radicals (•OH), given their strong and non-specific reactions with biomolecules. The sources of ROS can be endogenous, especially from the mitochondrial electron transport, and exogenous sources such as ionizing radiation from X-ray and ultraviolet radiations, and environmental toxins.20

The reduction of O2 to superoxide can occur biologically in mitochondrial complexes I and III (Reaction 1.1).21

O2 + e— → O2—• Reaction 1.1

Accumulation of superoxide can in turn generate another ROS molecule hydrogen peroxide via disproportionation (Reaction 1.2). This is also often mediated by enzyme superoxide dismutase in regulation of superoxide ions.

2 O2— + e— + 2H+ → H2O2 + O2 Reaction 1.2

Further generation of ROS hydroxyl radical, •OH, is produced via disproportionation of hydrogen peroxide (Reactions 1.3 and 1.4). Catalases and peroxidases can mediate the radical generation by converting hydrogen peroxide to water.

H2O2 + e— + H+ → H2O + •OH Reaction 1.3

•OH + e— + H+ → H2O Reaction 1.4

From the metal perspective reactive oxygen can be produced by a reduced form of metal, such as Fe2+ or Cu+ ions, overcoming the initial energy required to spontaneously oxidize:

O2 + Cu+ ↔ O2•— + Cu2+ Reaction 1.5

Additionally, Cu+ can further react in a Fenton-like reaction generating a hydroxyl radical (Reaction 1.6):

Cu+ + H2O2 ↔ Cu2+ + OH— + OH• Reaction 1.6

Metal reactions with O2•- can also form the highly reactive singlet oxygen, 1O2. .

While most biological systems contain enzymes that remove free-radicals to prevent immediate harm, such as the aforementioned superoxide dismutase and catalase, accumulation of small amounts of oxidation can lead to disease, discussed in more detail in subsequent sections.

Generally, ROS react with biological compounds without specificity. This can lead to a number complication in biological functions as they readily react with

6 proteins, DNA, lipids and more! Eventually, these reactions can lead to neurodegenerative diseases, cancer, aging and others.22 Metals in biology can play a role in both the generation of oxidative stress and quenching ROS activity. Enzymes like superoxide dismutase and glutathione peroxidase can protect against ROS activity using metal centers to catalyze their reduction.

The main transition metal explored in this dissertation is copper. Despite composing around 1/100,000 of a human body’s mass composition, it is an essential metal that performs all of the aforementioned roles metals can play in biology, and does so in all known forms of life to some extent, even with limited use in anaerobic bacteria.23-25 Herein, the investigation of a few metal-binding biomolecules are discussed for better understanding of copper in biology as well as its role in neurodegenerative diseases and how some organisms have adapted to the use of copper and prevent oxidative damage caused by copper.

II. Copper in Biology

The role of copper in biological systems, its role in disease, and some of the topics relevant to studies in subsequent chapters are subsequently discussed. The relatively low concentration of copper in biology should not understate its role in those systems. Copper is found in the active center of numerous essential enzymes in the body.

The preferred binding motifs for copper in biological systems tends to rely on the oxidation state of the ion and the electron charge distribution of the ligand. The most common oxidation states for copper are the +1 and +2 states, typically with a 3d9 paramagnetic electron valence. Copper(II) often forms octahedral geometries, which can sometimes elongate along the dz2 orbital through the Jahn-Teller effect. There are a few proposed examples of Cu3+, however, they are much rarer than their mono- and divalent counterparts, and have yet to be recorded in biological systems. The reduced state Cu1+ tends to coordinate to sulfur-based compounds like cysteine and methionine.

Loss of an electron due to oxidation of a substrate or a ligand forms Cu2+, which tends to bind more frequently to oxygen- or nitrogen-containing residues in biological systems. Changes in binding affinity based on charge state of the ion tends to change distances between the metal and the ligand, as well as altering the ligand field stabilization energy (LFSE).

2. Copper Proteins

The common forms of ionic copper, Cu1+/2+, can have highly reactive electron configurations. Biological systems evolved around these highly reactive ions, using them to perform many functions in biological systems, often modulating the reactivity by altering the active site accessibility or ligand interaction. The overall functionality of

8 copper is typically found within a few biologically well-conserved binding motifs or reaction profiles in proteins classified as types I, II, or III copper proteins.

2.1. Type I Copper Proteins

Type 1 copper proteins can be found in many different types of biological systems, from plants to bacteria. Some of the prototypical blue copper proteins can be found in electron transport chain in plant plastocyanin and the azurin proteins found in bacteria.

Type I copper proteins, also known as blue copper proteins or cupredoxins, are typically distinguished by a pseudo-tetrahedral coordination site consisting of two histidine residues bound at the imidazole Nδ atom, a cysteine sulfur, and a methionine residue within the protein. As their colloquial name suggests, these proteins have an intense blue color that absorbs light at around 600 nm with an absorptivity as high as

6000 M—1 cm—1 that is derived from the ligand-to-metal charge transfer transition from the S of the coordination cysteine residue to the Cu2+ ion with the cysteine residue.

2.1.1. Plastocyanin

The quintessential type I copper protein studied is plastocyanin. It functions, like other blue copper proteins, as an electron transfer protein. In plastocyanin, copper(II) is

9 reduced to copper(I) by cytochrome b6f, wherein the copper(I) is oxidized by cytochrome P700 in the photosystem I process.

Cytf• + PCCu2+ → Cytf + Cu1+ Reaction 1.7

PCCu1+ + P700+ → PCCu2+ + P700 Reaction 1.8

The copper ion resides in a distorted 4-coordinated tetrahedral geometry of a cysteine and two histidine residues, and typically a methionine residue. Similar type I copper the axially-coordinated methionine residue can vary based on the protein with glutamine, valine, leucine, or phenylalanine.26 The coordinating ligands cysteine and methionine have unusual bond distances with the oxidized copper ion of ~2.0 and 2.8

Å, respectively.27 The binding geometry around the copper active site is relatively rigid with only minor geometric changes during redox reactions. The rigid distorted tetrahedral geometry is not an ideal geometry for both Cu2+ and Cu1+, which is exemplified by the small ~0.7 eV reorganization energy of the relatively rigid coordination structure. Small reorganization energy allows for fast electron transfer.

Modulating the electron transfer reaction is a combination of protein-protein surface interactions based on hydrophobicity and hydrophilicity of the amino acid residues, pH-dependent gating, and small changes in the coordination environment around the metal.26 Cyrstallographic data revealed bond distance and geometric changes of His-44 and Gln-95 of umecyanin during redox cycling of the metal center. These residues

10 would be analogous His-44 and Met-92, respectively, in plastocyanin. The study found a more trigonal pyramidal geometry in the Cu1+ ion as bond distances increased for

His-44 and Gln-95. This is likely due to structural changes due to protein-protein interactions.28

Figure 1.2 - Plastocyanin (PDB: 2PCF) Spinacia oleracea with focus on the copper active site and its ligands on the right29

Like most things in life, there are always exceptions to the rule. Not all type I copper proteins are exactly the same, nor would they be based on differences in environmental pressures that lead to the protein as it exists in its respective system. In stellacyanin, the methionine residue is replaced by glutamine and shows a shorter Cu-

OGln bond distance at 2.2 Å than the Cu-Smet compared to 2.8 Å in plastocyanin, stabilized by a more tetrahedral geometry and solvent exposure.30

2.2. Type II Copper Proteins

Type II copper proteins, similar to type I copper proteins, are also mononuclear copper proteins, though, they tend to have a distorted 5-coordination planar geometry at the copper center. Unlike the electron-rich environment from the sulfur ligands, cysteine and methionine in Type I copper proteins, type II copper proteins have various coordination ligands that are typically a combination of histidine, aspartate/glutamate, and tyrosine. The functionality of type II copper proteins is varied, ranging from dioxygen transport and catalysis, to ROS quenching, and nitrogen assimilation with nitrite reductase in plants, to name a few.31 The overall function can be summed up for their ability to reduce their substrates. The following type II copper proteins are some of the quintessential type II proteins, which only represents some of the functions of all type II copper proteins.32

2.2.1. Cu,Zn Superoxide Dismutase (SOD1)

Cu,Zn-containing SOD (Cu,Zn-SOD or SOD1) is probably one of the most well- studied type II copper proteins, if not the most studied copper protein in general other

12 than potentially plastocyanin. Its primary role in biological systems is to quench superoxide, a byproduct produced during cycles involved in aerobic respiration and photosynthesis. The importance of SOD1 in biological systems cannot be understated as mutations or down-regulation of SOD1 transcription is implicated in many types of cancers, neurological diseases like ALS (discussed in more detail in section 3.2), and other age-related diseases.33,34 At the molecular scale, ROS, such as hydrogen peroxide

(H2O2) or superoxide (O2—), are indiscriminate in their reactions with DNA, carbohydrates, fatty acids, lipids, and many other molecules relevant to a living system, indicating the importance of Cu,ZnSOD in health and disease regulation.

The general reactions, in a ping-pong mechanism, that SOD1 can perform are

(1) O2- + Cu2+ZnSOD → O2 + Cu1+ZnSOD Reaction 1.9

(2) O2- + Cu1+ZnSOD +2H+ → H2O2 + Cu2+ZnSOD Reaction 1.10

These reactions occur with an efficiency of 2-6 × 109 M—1 sec—1, a diffusion limited reaction!35,36

There are four different kinds of SOD with an active site that can contain iron, manganese, nickel, or zinc and copper. The most common type in eukaryotes is the

Cu,ZnSOD.37 The coordination sphere of SOD proteins is comprised of three or four histidine residues (dependent on oxidation state). The 4 histidine residues typically form a distorted square planar geometry around the copper(II) active site. Unlike most

13 of the other enzymes discussed, SOD1 is a dimetallic enzyme as shown in Figure 1.3.

One of the histidine amino acids that is coordinated to the copper active site is also coordinated to a zinc ion, confirmed by a number of crystallographic and NMR studies.

The bridging histidine is currently known to be unique to Cu,ZnSOD. Reduction of the bridging imidazolate and copper center leads breaking the coordination between the two metal cations, thus forming a distorted trigonal planar geometry.38 The coordination of oxidized form of copper in the active site when bound to all four histidine residues yields a primary absorption band at 680 nm with an absorptivity of

155 M—1 cm—1.

Figure 1.3 - PDB ID: 3MND. Oxidized form of Cu,Zn superoxide dismutase enzyme

(left) and the active site of copper (orange) and zinc (right)39

The bridging His-63 and the stabilizing zinc ion regulate the electron density around the copper ion as it undergoes oxidation and reduction. EPR studies confirm this geometry with g-values corresponding to a dx2-y2 orbital as the partially occupied magnetic orbital (g|| > g⊥ > 2.0023), allowing for molecular oxygen binding to the dz2 orbital at the axial position.40

Interestingly, the role of zinc seems to be more structural than acting in any of the chemistry. In fact, replacing zinc with other divalent metal ions has little effect on the overall chemistry of SOD.41

2.2.2. Galactose Oxidase

Another example of a prototypical type-2 copper protein, to show the diversity of this type of copper protein, is galactose oxidase. Discovered in 1959 galactose oxidase is a ~68 kDa metallo-enzyme that is produced by Fusarium fungi.42

Figure 1.4 – PDB ID: 2EIE galactose oxidase (left) with emphasis on the copper active site (right).43

As part of the oxidoreductase family of enzymes, galactose oxidase catalyzes oxidation of the C6 hydroxyl on D-galactose and galactoside, and other primary alcohols. The generalized reaction is the oxidation of alcohol in galactose or galactoside to form an aldehyde in the presence of O2 in a two electron transfer process as part of a likely ping-pong mechanism:

D-Galactoside

Figure 1.5 – Oxidation of D-galactoside catalyzed by galactose oxidase and O2

The coordination of galactose oxidase is dependent on the oxidation state of the metal and the geometry of the binding pocket. The oxidized copper complex consists of two tyrosine and two histidine residues, along with a solvent molecule to form a distorted square pyramidal geometry. Upon reduction Cu+ yields a trigonal planar, three-coordinated copper center by dissociating from Tyr495 and the solvent, confirmed by X-ray absorbance techniques.44

One of the most interesting parts of the enzyme that attracts many researchers is the fact that the copper center is attached to a stable radical in the tyrosine.45 The radical tyrosine residue is stabilized in part by the post-translational cross-linkage of its ring to the adjacent cysteine residue, forming a thioether bond (Figure 1.4, right panel).46

Moreover, it is proposed that the free-radical tyrosine plays a role in the two electron transfer process given that Cu2+ can only facilitate a one-electron transfer reaction on its own.

Figure 1.6 – a proposed simplified reaction mechanism for the redox reaction at the copper center of galactose oxidase. 17

In the above mechanism, initial electron transfer to form the aldehyde from the alcohol takes place at the copper center, reducing Cu2+ to Cu1+. The fast nature of the radical single electron transfer and the nearly diffusion limited turnover of molecular oxygen make some of the smaller steps more difficult to propose for a more detailed mechanistic analysis. Many of the mechanistic approaches used in research currently uses models and computation to propose or support certain mechanisms of electron and proton transfer.47,48 The D-galactose turnover efficiency has been shown to be 0.93 ×

104 M—1 s—1 and the molecular oxygen turnover efficiency at a near diffusion limitation of 0.98 × 106 M—1 s—1 at pH 7.0. The galactoside methyl-α-D-galactopyranoside shows a more favorable reaction with a catalytic efficiency of 1.53 × 104 M—1 s—1 and a subsequent oxygen catalytic efficiency of 7.99 × 106 M—1 s—1.49,50 While these values vary between species, these studies have shown the efficiency for O2 and galactose/galactoside turnover by GO.

2.3. Type III Copper Proteins

The third type of copper protein discussed is the type 3 copper protein. These proteins can be found in all domains of life on Earth!51 The key differentiation between type 3 and copper protein types 1 and 2 is the di-copper center active site in type 3 proteins. Each copper ion in the active center is typically bound to the protein via three histidine residues. The most common and well-studied of the type 3 copper proteins are

18 likely tyrosinase, catechol oxidase and hemocyanin (the blue molluscan functional analogue of hemoglobin). The relation among the three most common type 3 copper proteins goes all the way into the DNA, showing high conservation of binding motifs on one of the coppers centers. It appears that the divergence of the second copper’s binding pocket leads to a difference in substrate specificity, where the changes in the binding pocket on the second copper cation modulate substrate specificity.55 They are generally involved in activation or transportation of O2, or the hydroxylation of monophenols for tyrosinase, often forming a side-on dioxygen bridge during the redox reaction.

2.3.1. Catechol Oxidase

Of the many type 3 copper proteins, catechol oxidase is briefly reviewed here for its relevance towards the activity observed in the subsequent chapters. Despite its presence in a variety of life, plant polyphenol oxidase, a type of catechol oxidase, is the predominant source of the protein in research. The size of the protein varies from species to species, ranging ~38-42 kDa based on submitted PDB structures.52 As their name suggests, catechol oxidase oxidizes catechols to form quinones. Most people are aware of its effects after leaving an apple or banana on the counter too long. Fruits and plants have many catechols and polyphenols that can be oxidized by catechol oxidase and the analogous polyphenol oxidase. The accumulation of yellow ortho-quinone

19 polymerizes into the brown-color polymers associated with an apple left out for too long as catechols and polyphenols are oxidized by the enzyme with ambient O2.53

Catechol o-Quinone

Figure 1.7 – oxidation of DTBC to form o-quinone product

The binding pocket of catechol oxidase is conserved in the typical type 3 binding center: two coppers bound by three histidine residues on each metal as seen below.

Figure 1.8 – PDB ID: 2P3X Catechol oxidase with a focus on the di-copper active site, with a ~45° horizontal rotation in the final frame for better visualization of the binding pocket histidine residues54

In the mechanistic cycle for catechol oxidase, the Cu-Cu bond distance is modulated by the ligand bound to one or both, though, they can vary based on species as the surrounding scaffold can affect metal-metal distance. The two most studied forms of type 3 proteins are the met and oxy forms, seen below, where hydroxide or molecular oxygen are bridges between the two coppers, respectively.

met form deoxy (reduced) form oxy form

Figure 1.9 – abbreviated catechol oxidase mechanism, showing key mechanistic structures

In one of the well-studied catechol oxidases from Ipomoea batata (sweet potatoes) the Cu-

Cu distances are 2.9 Å in the met form and 4.4 Å in the oxy form.55 EXAFS methods revealed the side-on μ-η2-η2 bonding of peroxide bridging the two copper(II) ions. The distance between the two copper cations was found to be essential in model studies; kinetic data indicates it likely loses it di-copper activity becoming two separate mono- copper active sites as the distances increase to >5 Å.56

The fascinating di-copper and bridging dioxygen has led to some in-depth studies on the electronic transitions as well. Electronic transitions at ~345 nm and ~580 nm showed relatively strong absorbances with absorptivities of 20,000 M-1 cm-1 and 1,000 21

M-1 cm-1, respectively. The strong charge transfer at ~345 nm is due to charge transfer from the πσ* of O22— to dx2-y2 of Cu2+. The less intense band at ~580 nm is due to the transition from the πυ* in O22— to dx2-y2 of Cu2+.57

III. Role of Copper in Neurodegenerative Disease

Despite 2 billion years to adapt to the toxicity of redox activity of oxygen and various biologically relevant metals, their dysregulation is still the source of toxicity, exemplified in numerous diseases caused by oxidative stress. The strong reactivity of most transition metals provide intense environmental pressures on biological systems on the evolutionary scale, making their integration into a larger protein structure a necessity for protection from these redox potentially harmful reactions. Dysregulation of metals in in vivo can lead to myriad health issues, such as several implications in cancer development, Creutzfeld-Jacob syndrome, Alzheimer’s disease, and amyotrophic lateral sclerosis. Research in neurodegenerative disease, often shows the presence of or mis-regulation of copper, among other metals, such as Fe+2/+3 and Zn2+.

Research in neurodegenerative diseases still leaves much to be concluded. Many research articles that discuss any role of metal in disease will often use modifiers like

“might”, “could”, implicated”, “probable”, or “likely” because the pathway of disease development is often not completely understood. Unlike diseases caused by viruses and bacteria where diagnostics can be performed with some reliability and with well-

22 understood etiology, neurodegenerative diseases are more subtle. These diseases often take years to develop where low levels of dysregulation accumulate over the course of decades! As chemical and clinical studies continue to develop, the picture fills in little- by-little giving us more insights into how the diseases function and how they develop in hopes that they can one day be reliably prevented treated.

3.1. Alzheimer’s Disease

The marvel of modern medicine has allowed humans to live longer, healthier lives than they ever have in history. Survey data from the Centers for Disease Control reveals the current average life expectancy to be approaching 80 years old, and an upward trend in age is anticipated to continue!58 Studies in the last decade show the most common cause of death, heart disease, has been in decline for nearly 50 years, while age-related neurodegenerative disorders continue to rise.59-61 One of the key challenges that an aging population creates is the growth of age-related disease incidences, such as Alzheimer’s disease and Parkinson’s disease.62,63 Currently, the most common neurodegenerative disorder is Alzheimer’s disease, affecting 10% of the

American population over the age of 65 and placing a $300 billion burden on both suffers and taxpayers.58

Like most age-related diseases, the generalized causes are varied, with implications towards environmental and genetic factors. The environmental factors that

23 can impact the propensity to develop the disease are the some of the usual suspects: air- quality, water quality, sedentary lifestyles, excessive alcohol consumption, etc.64-66 A high-impact study by Saunders, et al. (cited nearly 4000 times at this writing!) found a genetic link between late-onset AD and the ApoE ε4 allele, opposed to the protecting ε2 allele. Most people, though, have the ε3 allele which does not seem to play much of a role in the disease.67 The current hypothesis toward the role of ApoE ε4 influence is that it potentially upregulates γ-secretase, an enzyme involved in producing a peptide implicated in AD, amyloid β (Aβ), from its precursor protein.68 In a related idea, one of the leading ideas on copper’s involvement in Alzheimer’s disease is based on causing aggregation of the 40 to 42 amino acid long Aβ peptide on brain tissue. Along with the tau protein, aggregates of the amyloid beta peptide have been found on the brain tissue of deceased AD patients in post-mortem observations. Moreover, these plaques were found to have higher than normal amounts of iron, zinc, and copper.69 It is thought that the aggregation of the peptide on brain tissue induces inflammation of brain tissue which can trigger apoptosis. In a similar but alternative proposal, there is also research indicating that apoptosis might also be triggered by Aβ interaction with a 17- hydroxysteroid dehydrogenase 10, 17β-HSD10, an alcohol dehydrogenase in mitochondria, creating toxic conditions that induce cell death.70 Loss of brain tissue pockets then consequently leads to loss in memory and other mental faculties.

Figure 1.10 – “Healthy Brain and Severe AD Brain” by National Institutes of Health.

CC-BY-NC.2.0 2016 by Copyright Holder.

More in-depth associations of copper in the disease, such as peptide-copper structure and neurotoxic chemistry are covered in chapter 2. The propensity for the disease is also not understood well as it seems genetics and lifestyle might both play a role. Dysregulation of the amyloid protein or copper transport proteins, such as ceruloplasmin, due to genetic disorders could both play a role in the disease.71

Similarly, poor diet or excess intake of biological metals could slowly trigger the aggregate cascade.72,73

3.2. Amyotrophic Lateral Sclerosis

Dysregulation of the aforementioned copper protein SOD1 is also implicated in amyotrophic lateral sclerosis (ALS), another neurological disorder, also known as Lou

Gehrig’s disease. The disease presents in the slow loss of motor control as brain tissue associated with motor neurons undergo apoptosis. The loss of motor control eventually leads to paralysis and death due to organ failure for organs reliant on muscle function, like the respiratory system. ALS is considered a rare disease with around 15,000 people in the U.S. diagnosed, typically diagnosed after the age of 55 years.74 Though symptoms had been observed prior to its official discovery; the first published recognition of ALS occurred by Jean-Martin Charcot in 1874.75 Despite its rarity, it has been brought to public awareness by celebrities like Lou Gehrig and Dr. Stephen Hawking having the disease, as well as the viral Facebook “ice bucket challenge” that raised a substantial amount of money towards its research.

The cause(s) of ALS are still not well-understood. Only 10% of those that suffer from the disease have a family history of ALS, though, more current research is beginning to recognize a link between ALS and other familial instances of neurodegenerative diseases.76,77 A 2015 review found over 25 genes and hundreds of single-point mutations implicated in the familial form ALS. Some of the genes show mutation for

SOD1 and mitochondrial myopathy resulting in an increase in ROS activity and subsequent oxidative stress. Unfortunately, the genes implicated in familial ALS will correlate to <5% of all ALS occurrences, implying both the complexity of the genetics behind ALS and the role sporadic factors might play in its development.78

The genes for SOD1 regulation are implicated in up to ~20% of familial (non- sporadic) ALS cases. 78-80 Similar to Alzheimer’s disease, ALS has shown protein

26 aggregation on neuronal tissue, often a mutant of SOD1, for familial-type ALS. Research has revealed more than hundred mutations of SOD1 implicated in ALS, supporting its importance in neuronal health! Therein the most common mutation of the SOD1 allele is the D90A mutation, where the 90th codon for SOD1 is mutated from aspartic acid to alanine.81 The loss of hydrogen-bonding from the aspartic acid causes structural instability due to the change in polarity from the hydrophilic residue being replaced the hydrophobic mutant residue.

Several other mutations for the SOD1 gene can cause subsequent amino acid changes upon translation that can also affect SOD1 functionality. In U.S. populations

Ala4Val (alanine to valine mutation at codon 4) mutations of SOD1 is the most common and often has less than a two year survivability on average.82 Studies observed lower

Km values (comparatively larger association of the H2O2-SOD1 complex) for H2O2 product relative to the wild-type protein, thus producing more free radical activity.83 In the Gly93Ala mutant of SOD1 can lead to oxidative damage of the protein, oligomerization, and eventual aggregation.84,85 As stated previously, hundreds of genetic point mutation can lead to many SOD1 mutations that are implicated in ALS.

The role of all the mutations is still not fully understood, but case studies on some specific mutations associated with aggregates have revealed loss of proper SOD function and mis-folding around the metal center.86

Unfortunately, the role environmental factors play is not well-understood and seems to be contingent on the health of some protective genes and proteins. Research has

27 examined the effects of heavy metals, viruses, and factors that can induce oxidative stress, all potentially showing some correlation with certain types of sporadic cases.87-89

Cases like SOD1 gene mutations provide at least some clues that might lead towards treatment in the future.

IV. Metallo-Antibiotics

Metallo-drugs, as whole, are an incredibly diverse and effective method for treating disease. Some of the most effective and enduring metallo-drugs on the market are:

1. Cisplatin – platinum-centered antitumor drug

2. HPA-23 - antimonium tungstate, tungsten-based HIV antiviral

3. Pepto-Bismol – bismuth subsalicylate, treats stomach discomfort

4. Lithium carbonate – mood stabilizer for bipolar disorder90

5. Metallo-antibiotics:

A. Streptonigrin – Cu- or Fe-bound antibiotic and antitumorogenic91

B. Tetracycline – a broad spectrum antibiotic that can bind to various transition

metals92

C. Bleomycin – antibiotic and antitumor drug that binds Fe, Zn, or Cu, discussed

in Section 4.1 of this chapter

D. Bacitracin – a metallo-antibiotic that prevents cell wall synthesis, discussed in

detail in Chapter 3

As their name suggests, metallo-antibiotics are antibiotics that rely on a metal ion for their structure and/or antibiotic function. The metals involved range from the first row of transition elements on the periodic table to the third row! From the evolutionary perspective, the incorporation of metals into peptides as a means of preservation of the host was inevitable given the availability of metals in the environment, especially the soil and ocean, and the potential harm caused by metals without regulation. Culturally, the human species have been administering, metals, metallo-antibiotics and other metal-activated medicines for several thousand years, sometimes to the point of adding arsenic to their drinking sources!93

Compared to their non-metal counterparts, metallo-antibiotics are have been drastically under-represented in research and in application. Despite the lack of representative interest in metallo-antibiotics their current uses are varied and tend to be highly effective. The modes-of-action can range from DNA intercalators that prevent

DNA replication and transcription, to cell membrane disruption, to inhibiting cell wall synthesis. Another example of a quintessential metallo-antibiotic, bacitracin, is explored in more depth in Chapter 3.

4.1. Bleomycin

Bleomycin (BM, Figure 1.8) is a multifunctional drug that can act as an antibiotic and antitumor drug that is naturally produced by Streptomyces verticillus and several other strains of Streptomyces. The exact mechanism-of-action for bleomycin is not currently understood completely, though, there is evidence for it binding to and cleaving genetic material, including DNA, RNA, and tRNA.94,95 Research on the metal presence has shown their importance in performing either single or double stranded cleavage of DNA in the presence of Fe or Cu and DNA binding in the presence of

Zn.96,97 Bleomycin has been shown to have biological interactions against tumors as well, due to these biomolecular interactions.98 Predominant focus has been on the iron- bound form, owing to its high antineoplastic activity, although the copper form was the originally isolated antibiotic. Complexation of the copper ion with bleomycin has been evidenced by several methodologies. Electronic d-d transitions have been observed

~600 nm with ε600 of ~110 M—1 cm—1, indicating a likely pseudo-type 2 copper center. A crystallographic study of the cobalt(II) complex showed metal binding at the histidine imidazole, β-aminoalanine, pyrimidine, the β-hydroxyhistidine amide nitrogen, and the

β-aminoalanine amine as shown in Figure 1.8.102,99 The nitrogen binding motifs were also confirmed by electron spin-echo envelope spectroscopy and EPR.100,101

Figure 1.11 – line structure of bleomycin (bottom) and cobalt-bound structure of the metal binding portion of BM (top), adapted from PDB ID 1JIF102

Metallo-bleomycin’s mechanism-of-action as an antibiotic is also what gives it its versatility in treating cancer. The metallo- form of the drug can bind the minor groove of DNA, likely with stabilization by secondary binding of DNA with the bithiazoline rings.103 The redox-active metal-center in bleomycin can activate O2 or H2O2 to produce superoxides or free-radicals, in a manner discussed in previous sections, which in turn react with the deoxyribose component of DNA and, subsequently cleave DNA. A study on the Fe-bleomycin complex by Burger, et al. found metal-mediated deoxyribose

31 cleavage of the C3’-to-C4’, and the C1’-to-ring-oxygen.104 Earlier studies on the copper(II) complex with 1,10-phenanthroline showed similar activity, laying the groundwork for elucidating the iron-bleomycin complex.105

Metallo-antibiotics are likely a staple of life for as nearly as long as life, as we know it, has existed. The ubiquity of environmental metals, like copper and iron, have made antibiotics versatile and efficacious towards a number of diseases listed in the opening of this section. Only in the last hundred years have humans developed the means to understand, isolate, synthesize, and prescribe antibiotics. Metallo-antibiotics have yielded some of the long-lasting antibiotics, like bacitracin discussed in Chapter 3.

Biological systems have been using them for many, many millennia. As bacteria and fungi are the primary sources of natural antibiotics, they can influence disease in agricultural settings as well, affecting plant and animal life, growth, and soil conditions.106,107

V. Copper and Secondary Metabolite Interactions

An intersection of food, disease, drug design, and evolution is briefly described as it relates to effects observed in the following chapters. They are often recognized as antibiotics, defensive compounds, chromophores, or compounds that can protect against oxidative stress; often playing multiple roles in the biological system.

Humans have been interested in secondary metabolites indirectly for centuries. The prominence of these compounds in research has been steadily growing over the decades as well. They can often be the source for future drug design, such as salicylic acid being used as to form the effective aspirin known so well today. In furthering our understanding of human health and its relationship with diet, secondary metabolites have gained favor among the public as well, who are often seeking means of healthy living without the use of in vitro synthesized molecules. Unfortunately, the extent to which secondary metabolites can be beneficial has not had the same scrutiny as the synthetically derived molecules until more recently. Later chapters explore some of the relationships that specific secondary metabolites, such as betanin and salicylic acid, might have in the presence of copper.

5.1. Flavonoids (Quercetin and Myricetin)

With >10,000 unique compounds, flavonoids are a huge chemical class of secondary metabolites produced by plants, especially fruits and vegetables. Some of the highest concentrations in foods can be found in dried herbs like oregano and parsley (~4.5 g/kg), capers (0.49 g/kg), and some teas.108,109 They generally consist of three six- membered rings with varying substituents based on alcohol and ketone groups, with the exception of the chalcone class that is opened on the central 3 carbon ring. For the sake of brevity, I will only be discussing the a few classes and examples that are

33 involved in metal binding, but there are a few good reviews that go into detail on the qualities of all the classes and subclasses.110-112

Figure 1.12 – general flavonoid structure with number assignments (left), myricetin

(middle), and quercetin (right)

Structurally, many flavonoids are “ripe” for the capacity to bind metals. On the flavonol subclass the keto-enol moiety on carbons C3-C4 and the keto-phenol moiety at

C4-C5, along with some interaction with the catechol and polyphenol moieties, are the most likely candidates for metal-binding. In one comprehensive study, paramagnetic

NMR using Co2+ as a paramagnetic metal binding probe, have confirmed this interaction with quercetin and myricetin.113 Displacement of the C3-OH proton forming the negatively charged α-ketoenolate, long catechol T1 values, and shorter C3’ and C6’

T1 values indicate enolate binding at C3-O—; wherein T1 is proportional to the distance from the paramagnetic center to the sixth factor (T1 α rM-H6). Moreover, spectrophotometric studies show clear electronic transitions for flavonoids ~360-380 nm representing the A and C rings, and an absorbance band ~270 nm depending on the conjugation and electron density of side-groups. Binding of metals, such as copper, 34 calcium, and ytterbium, show an apparent red shift in electronic transitions. This allowed for facile determination of metal binding to show 1:1 M-Qr and M-Mr in

DMSO.113

Based on the 1:1 binding of metal-flavonoid interaction, this could yield inhibitory effects in the dysregulation of metals, allowing for ternary complex formations with metal centers in the metallo-biomolecules and prevention of ROS activation. This might give some insight towards versatility of flavonoids as antibiotic, anticancer, and antioxidant compounds. Flavonoids have been shown to potentially inhibit oxidation chemistry involved in AD by binding to CuAβ, and inhibiting cancer growth where metals are required in higher concentrations in growth factors.113,114

VI. Conclusion

As explored in this small introduction of metallo-biomolecules, metals have an undeniable role in life on Earth. With origins of metal involvement in aerobic life, metals are essential. They are even incorporated into biomolecules that reduce invasive biotic activity, as seen with bleomycin, and ROS activity, as discussed with SOD1. Their dysregulation, unfortunately, often has adverse effects on the host organism in diseases like Alzheimer’s and ALS. Subsequent chapters explore some of the environmental interfaces between metals and biomolecules. Human disease, plant metabolites, soil bacteria, and future drugs can all have a role in binding to metal active centers like

35 copper. Herein, these interactions are explored in order to better understand the role of copper in disease, disease treatment, and the chemistry that is occurring in the environment.

VII. References

1 Pavlov, A.A. and Kasting, J.F. Mass-independent fractionation of sulfur isotopes in

Archaen sediments: strong evidence for an anoxic Archaen atmosphere.

Astrobiology. 2004. 2(1)27-41. Doi: 10.1089/153110702753621321

2 Dupont, C.L., Butcher, A., Valas, R.E., Bourne, P.E., and Caetano-Annolles, G. History

of biological metal utilization inferred through phylogenomic analysis of protein

structures. Proceedings of the National Academy of Sciences. 2010. 107(23)10567-

10572. Doi: 10.1073/pnas.0912491107

3 Imlay, J.A. The molecular mechanisms and physiological consequences of oxidative

stress: lessons from a model bacterium. Nature Reviews: Microbiology. 2013.

11:443-454. Doi: 10.1038/nrmicro3032

4 Canfield, R.M. The early history of atmospheric oxygen: Homage to Robert M. Garrels.

Annual Review of Earth and Planetary Sciences. 2005. 33:1-36. Doi:

10.1146.annurev.earth.33.092203.122711

5 Cloud, Jr., P.E. Atmospheric and hydrospheric evolution on the primitive earth.

Science. 1968. 160(3829)729-736. Doi: 10.1126/science.160.3829.729

6 Donald, K.W. Oxygen poisoning in man. British Medical Journal. 1947. 1(4506)667-672.

Doi:10.1136/bmj.1.4507.712

7 John R. Rumble, ed., CRC Handbook of Chemistry and Physics, 99th Edition (Internet

Version 2018), CRC Press/Taylor & Francis, Boca Raton, FL

8 Walpole, S.C., Prieto-Merino, D., Edwards, P., Cleland, J., Stevens, G., and Roberts, I.

The weight of nations: an estimation of adult human biomass. Biomed Central.

2012. 12:439. Doi: 10.1186/1471-2458-12-439l

9 Schwendinger, M.G., Tauler, R., Saetia, S., Liedl, K.R., Kroemer, R.T., and Rode, B.M.

Salt induced peptide formation: on the selectivity of the copper induced

formation under possible prebiotic conditions. Inorganic Chimica Acta. 1995.

228(2)207-214. Doi: 10.1016/0020-1693(94)04186-Y

10 Iqubal, M.A., Sharma, R., Jheeta, S., and Kamaluddin. Thermal condensation of

glycine and alanine on metal ferrite surface: Primitive peptide bond formation

scenario. Life. 2017. 7(15)1-20. Doi: 10.3390/life7020015

11 Miller, J., McLachlan, A.D., and Klug, A. Repetitive zinc-binding domains in the

protein transcription factor IIIA from Xenopus oocytes. The EMBO Journal. 1985.

4(6)1609-1614. Doi: 10.1002/j.1460-2075.1985.tb03825.x

12 Bonaventura, J., and Bonaventura, C. Hemocyanins: Relationships in their structure,

function and assembly. American Zoologist. 1980. 20(1)7-17. Doi:

10.1093/icb/20.1.7

13 Goucher, C.R., and Kocholaty, W. Effect of various ions on the respiration of

Azotobacter. Journal of Biological Chemistry. 1954. 211(2)613-620.

14 Andreini, C., Bertini, I., Cavallaro, G., Holliday, G.M., and Thornton, J.M. Metal ions

in biological catalysis: from enzyme databases to general principles. Journal of

Biological Inorganic Chemistry. 2008. 13(8)1205-1218. Doi: 10.1007/s07755-008-0404-

15 Harrison, R., Warburton, V., Lux, A., and Atan, D. Blindness caused by junk food

diet. Annals of Internal Medicine. 2019. Doi: 10.7326/L19-0361.

16 Mazars, C., Patrice, T., Oliver, L., and Stephane, B. Cross-talk between ROS and

calcium in regulation of nuclear activities. Molecular Plant. 2010. 3(4)706-718. Doi:

10.1093/mp/ssq024

17 D’autreaux, B., and Toldedano, M.B. ROS as signaling molecules: mechanisms that

generate specificity in ROS homeostasis. Nature Reviews Chemical Cell Biology.

2007. 8:813-824.Doi: 10.1038/nrm2256

18 Lowenstein, C.J., Dinerman, J.L., and Snyder, S.H. Nitric oxide: A physiologic

messenger. Annals of Internal Medicine. 1994. 120(3)227-237. Doi: 10.7326/0003-

4819-120-3-199402010-00009

19 Rhee, S.G. Redox Signaling: hydrogen peroxide as intracellular messenger.

Experimental & Molecular Medicine. 1999. 31(2)53-59. Doi: 10.1038/emm.1999.9

20 Han, D., Antunes, F., Canali, R., Rettori, D., and Cadenas, E. Voltage-dependent anion

channels control the release of the superoxide anion form mitchochondria to

cytosol. Journal of Biological Chemistry. 2003. 278(8)5557-5563. Doi:

10.1074.jbc.M210269200

21 Grivennikova, V.G., and Vinogradov, A.D. Generation of superoxide by the

mitochondrial Complex I. Biochimica et Biophyica Acta (BBA) – Bioenergetics. 2006.

5-6:553-561. Doi: 10.1016/j.bbabio.2006.03.013

22 Ray, P.D., Huang, B.W., and Tsuji, Y. Reactive oxygen species (ROS) homeostasis and

redox regulation in cellular signaling. Cellular Signaling. 2012. 24(5)981-990. Doi:

10.1016/j.cellsig.2012.01.008

23 Andreini, C., Bertini, I., and Rosato, A. A hint to search for metalloproteins in gene

banks. Bioinformatics. 2004. 20(9)1373-1380. Doi: 10.1093/bioinformatics/bth095

24 Maret, W. The metals in biological periodic systems of the elements: Concepts and

conjectures. International Journal of Molecular Science. 2016. 17(1)66. Doi:

10.3390.ijms17010066.

25 Festa, R.A., and Thiele, D.J. Copper: an essential metal in biology. Current Biology.

2011. 21(21)R877-R833. Doi: 10.1016/j.cub.2011.09.040

26 Dennison, C. Investigating the structure and function of cupredoxins. Coordination

Chemistry Reviews. 2005. 249(24)3025-3054. Doi: 10.1016/j.ccr.2005.04.021

27 Shibata, N., Inoue, T., Nagano, C., Nishio, N., Kohzuma, T., Onodera, K., Yoshizaki,

F., Sugimura, Y., and Kai, Y. Novel insight into the copper-ligand geometry in

the crystal structure of Ulva pertusa plastocyanin at 1.6 Å resolution: structural

basis for regulation of the coper site by residue 88. Journal of Biological Chemistry.

1999. 274:4225-4230. Doi: 10.1074/jbc.274.7.4225

28 Koch, M., Velarde, M., Harrison, M.D., Echt, S., Fischer, M., Messerschmidt, A., and

Dennison, C. Crystal structures of oxidized and reduced stellacyanin from

horseradish roots. Journal of the American Chemical Society. 2005. 127(1)158-166.

Doi: 10.1021/ja046184p

29 PDB 2PCF Doi: 10.1016/S0969-2126(98)00035-5, modeled with Schrodinger

30 Hart, P.J., Nersissian, A.M., Herrmann, R.G., Nalbandyan, R.M., Valentine, J.S., and

Eisenberg, D. A missing link in cupredoxins: Crystal structure of cucumber

stellacyanin at 1.6 Å resolution. Protein Science. 1996. 5(11)2175-2183. Doi:

10.1002/pro.5560051104.

31 MacPherson, I.S., and Murphy, M.E. Type-2 copper-containing enzymes. Cellular and

Molecular Life Sciences. 64:2887-2889. Doi: 10.1007/s00018-007-7310-9

32 MacPherson, S. and Murphy, M.E.P. Type-2 copper-containing enzymes. Cellular and

Molecular Life Sciences. 2007. 64:2887-2899. Doi: 10.1007./s00018-007-7310-9

33 Griess, B., Tom, E., Domann, F., and Teoh-Fitzgerald, M. Extracellular superoxide

dismutase and its role in cancer. Free Radical Biology and Medicine. 2017. 112: 464-

479. Doi: 10.1016/j.freeradbiomed.2017.08.013.

34 Rosen, D.R., Siddique, T., Patterson, D., Figlewicz, D.A., Sapp, P., Hentati, A.,

Donaldson, D., Goto, J., O’Regan, J.P., Deng, H.-X., Ragmnai, Z., Krizus, A.,

McKenna-Yaskek, D., Cayabya, A., Gaston, S.M., Berger, R., Tanzi, R.E.,

Halperin, J.J., Herzfeldt, B., Van den Bergh, R., Hung, W.-Y., Bird, T., Deng, G.,

Mulder, D.W., Smyth, C., Laing, N.G., Soriano, E., Pericak-Vance, M.A., Haines,

J., Rouleau, G.A., Gusella, J.S., Horvitz, H.R., and Brown Jr., R.H. Mutations in

Cu/Zn superoxide dismutase gene are associated with familial amyotrophic

lateral sclerosis. Nature. 362:59-62. Doi: 10.1038/362059a0

35 Fee, J.A. and Bull, C. Steady-state kinetic studies of superoxide dismutase: saturative

behavior of the copper- and zinc-containing protein. The Journal of Biological

Chemistry. 1986. 261(28)13000-13005.

36 Gray, B. and Carmichael, A.J. Kinetics of superoxide scavenging by dismutase

enzymes and manganese mimics determined by electron spin resonance. The

Biochemical Journal. 1992. 281:795-802. Doi:10.1042/bj2810795

37 Tainer, J.A., Getzoff, E.D., Richardson, J.S., and Richardson, D.C. Structure and

mechanism of copper, zinc superoxide dismutase. Nature. 1983. 306:284-287. Doi:

10.1038/306284a0

38 Hart, P.J., Balbirnie, M.M., Ogihara, N.L., Nersissian, A.M., Weiss, M.S., Valentine,

J.S., and Eisenberg, D.. A structure-based mechanism for copper-zinc superoxide

dismutase. Biochemistry. 1999. 38(7)2167-2178. Doi: 10.1021/bi982284u.

39 Hernandez-Santoyo, A., Landa, A., Gonzalez-Mondragon, E., Pedraza-Escalona, M.,

Parra-Unda, R., and Rodriguez-Romero, A. Crystal structure of Cu/Zn

superoxide dismutase from Taenia solium reveals metal-mediated self-assembly.

The FEBS Journal. 2011. 278(18)3308-3318. Doi: 10.1111/j.1742-4658.2011.08247.x

40 McGuirl, M.A., and Dooley, D.M. Copper proteins with type 2 sites. Encyclopedia of

Inorganic and Bioinorganic Chemistry. 2011.

41 Lippard, S.J., Burger, A.R., Ugurbil, K., Pantoliano, M.W., and Valentine, J.S. Nuclear

magnetic resonance and chemical modification studies of bovine erythrocyte

superoxide dismutase: Evidence for zinc-promoted organization of the active site

structure. Biochemistry. 1977. 16(6)1136-1141. Doi: 10.1021/bi00625a017.

42 Copper, J.A.D., Smith, W., Bacila M., and Medina, H. Galactose oxidase from

Polyporus circinatus. Journal of Biological Chemistry. 1956. 234:445-448.

43 Rogers, M.S., Tyler, E.M., Akyumnai, N., Jurtis, C.R., Spooner, R.K., Deacon, S.E.,

Tamber, S., Firbank, S.J., Mahmoud, K., Knowles, P.F., Phillips, S.E., McPherson,

M.J., and Dooley, D.M. The stacking tryptophan of galactose oxidase: a second-

coordination sphere residue has profound effects on tyrosyl radical behavior and

enzyme catalysis. Biochemistry. 2007. 46:4606-4618. Doi: 10.1021/bio62139d

44 Clark, K., and Penner-Hahn, J.E. Structural characterization of the copper site in

galactose oxidase using x-ray absorption spectroscopy. Biochemistry. 1994.

33(42)1253-12557. Doi: 10.1021/bi00208a004

45 Whittaker, M.W. and Whittaker, J.W. The active site of galactose oxidase. Journal of

Biological Chemistry. 1988. 263(13)6074-6080.

46 Ito, N., Phillips, S.E.V., Stevens, C., Ogel, Z.B., McPherson, M.J., Keen, J.N., Yadav,

K.D.S., and Knowles, P.F. Novel thioether bond revealed by a 1.7 Å crystal

structure of galactose oxidase. Nature. 1991. 350(6313)87-90. Doi:

10.1038.350087a0

47 Gamez, P., Koval, I.A., and Reedijk, J. Bio-mimicking galactose oxidase and

hemocyanin, two dioxygen-processing copper proteins. Dalton Transactions.

2004. 24:4079-4088. Doi: 10.1039/B413535K

48 Himo, F., Eriksson, L.A., Maseras, F., and Siegbahn, P.E.M. Catalytic mechanism of

galactose oxidase: A theoretical study. Journal of the American Chemical Society.

2000. 122(33)8031-8036. Doi: 10.1021/ja994527r

49 Borman, C.D., Saysell, C.G., and Sykes, A.G. Kinetic studies on the reactions of the

Fusarium galactose oxidase with five different substrates in the presence of

dioxygen. Journal of Biological Inorganic Chemistry. 1997. 2(4)480-487. Doi:

10.1007/s007750050159

50 Whittaker, M.W., Ballou, D.P., and Whittaker, J.W. Kinetic isotope effects as probes of

the mechanism of galactose oxidase. Biochemistry. 1998. 37(23)8426-8436. Doi:

10.1021/bi980328t

51 Clause, H., and Decker, H. Bacterial tyrosinases. Systematic and Applied Microbiology.

2009. 29(1)3-14. Doi: 10.1016/j.syapm.2005.07.012.

52 Kaintz, C., Mauracher, S.G., and Rompel, A. Type-3 copper proteins: recent advances

on polyphenol oxidases. Advances in Protein Chemistry and Structural Biology.

2014. 97:1-35. Doi: 10.1016/bs.apcsb.2014.07.001

53 Harel, E., Mayer, A.M., and Shain, Y. Catechol oxidases, endogenous substrates and

browning in developing apples. Journal of the Science of Food and Agriculture. 1966.

17(9)389-392. Doi: 10.1002/jsfa.2740170901.

54 Virador, V.M., Reyes Grajeda, J.P., Blanco-Labra, A., Mendiola-Olaya, E., Smith,

G.M., Moreno, A., Whitaker, J.R. Cloning, sequencing, purification, and crystal

structure of Grenache (Vitis vinifera) polyphenol oxidase. Journal of Agriculture

and Food Chemistry. 2010. 58(2)1189-1201. Doi: 10.1021/jf902939q

55 Gerdemann, C., Eicken, C, and Krebs, B. The crystal structure of catechol oxidase:

New insight into the function of the type-3 copper proteins. Accounts of Chemical

Research. 2001. 35(3)138-191. Doi: 10.1021/ar990019a

56 Zheng, B., Liu, H., and Zhang, J. Effect of coordination sphere of the copper center

and Cu-Cu distance on catechol oxidase and nuclease activities of the copper

complexes. Applied Organometallic Chemistry. 2014. 28:372-378. Doi:

10.1002/aoc3138

57 Solomon, E.I., Baldwin, M.J., and Lowery, M.D. Electronic structures of active sites in

copper proteins: contributions to reactivity. Chemical Reviews. 1992. 92(4)521-542.

Doi: 10.1021/cr00012a003.

58 Kochanek, K.D., Murphy, S.L., Xu, J., and Tejada-Vera, B. Deaths: Final Data for 2014.

National Vital Statistics Reports. 2016. 65(4)

59 Jones, D.S., Greene, J.A. The decline and rise of coronary heart disease:

Understanding public health catastrophism. American Journal of Public Health.

2013. 103(7):1207-1218

60 Ford, E.S., Ajani, U.A., Critchley, J.A., Labarthe, D.R., Kottke, T.E., Giles, W.H., and

Capewell, S. Explaining the decrease in U.S. deaths from coronary disease, 1980-

2000. New England Journal of Medicine. 2007. 356:2388-2398

61 Hebert, L.E., Weuve, J., Scherr, P.A., and Evans, D.A. Alzheimer disease in the United

States (2010-2050) estimated using the 2010 census. Journal of Neurology. 2013.

80(19):1778-1783.

62 Nunomura, A., Chiba, S., Kosaka, K., Takeda, A., Castellani, R.J., Smith, M.A., and

Perry, G. Neuronal RNA oxidation is a prominent feature of dementia with Lewy

bodies. Neuroreport. 2002. 13(16)2035-2039

63 Dalfó, E., Portero-Otín, M., Ayala, V., Martinez, A., Pamplona, R., Ferrer, I. Evidence

of oxidative stress in the neocortex in incidental Lewy body disease. Journal

Neuropathology and Experimental Neurology. 2005. 64(9):816-830

64 McKay, J.A. and Mathers, J.C. Diet induced epigenetic changes and their implications

for health. Acta Physiologica. 2011. 202(2)103-18.

65 Wheeler, M.J., Dempsey, P.C., Grace, M.S., Ellis, K.A., Gardiner, P.A., Green, D.J., and

Dunstan, D.W. Sedentary behavior as a risk factor for cognitive decline? A focus

on the influence of glycemic control of brain health. Alzheimer’s & Dementia:

Translational Research & Clinical Interventions. 3(3)291-300. Doi:

10.1016/j.trci.2017.04.001.

66 Huang, W-J., Zhang, X., and Chen, W-W. Association between alchol and Alzheimer’s

disease: Review. Experimental and Therapeutic Medicine. 12(3)1247-1250. Doi:

10.3892/etm.2016.3455.

67 Saunders, A.M., Strittmatter, W.J., Schmechel, D., George-Hyslop, P.H., Pericak-

Vance, M.A., Joo, S.H., Rosi, B.L., Gusella, J.F., Crapper-Machlachlan, D.R.,

Alberts, M.J., Hulette, C., Crain, B., Goldgaber, D., and Roses, A.D. Association of

apolipoprotein E allele ε4 with late-onset familiar and sporadic Alzheimer’s

disease. Neurology. 1993. 43(8)1467-1472. Doi: 10.1212/wnl.43.8.1467.

68 Fuentealba, R.A., Barria, M.I., Lee, J., Cam, J., Araya, C., Escudero, C.A., Inestrosa,

N.C., Bronfman, F.C., Bu, G., and Marzolo, M.P ApoER2 expression increases Aβ

production while decreasing Amyloid Precursor Protein (APP) endocytosis:

Possible role in the partitioning of APP into lipid rafts and in the regulation of γ-

secretase activity. Molecular Neurodegeneration. 2007. 2(14)1-19. Doi: 10.1186/1750-

1326-2-14.

69 Lovell, M.A., Robertson, J.D., Tessdale, W.J., Campbel, J.L., and Markesbery, W.R.

Copper, iron and zinc in Alzheimer’s disease senile plaques. Journal of

Neurological Sciences. 1998. 158(11)47-52. Doi: 10.1016/S0022-510X(98)00092-6.

70 Lustbader, J.W., Cirilli, M., Lin, C., Xu, H.W., Takuma, K., Wang, N., Caspersen, C.,

Chen, X., Pollak, S., Chaney, M., Trinchese, F., Liu, S., Gunn-Moore, F., Lue, L.F.,

Walker, D.G., Kupposamy, P., Zewier, Z.L., Arancio, O., Stern, D., Yan, S.S., and

Wu, H. ABAD directly links Aβ to mitochondrial toxicity in Alzheimer’s disease.

Science. 2004. 304(5669)448-452. Doi: 10.1126.science.1091230.

71 Brewer, G.J., Kanzer, S.H., Zimmerman, E.A., Celmins, D.F., Heckman, S.M., and

Dick, R. Copper and ceruloplasmin abnormalities in Alzheimer’s disease.

American Journal of Alzheimer’s Disease & Other Dementias®. 2010. 25(6)490-497.

Doi: 10.1177/1533317510375083

72 Squitti, R., Siotto, M., and Polimanti, R. Low-copper diet as a preventative strategy for

Alzheimer’s disease. Neurobiology of Aging. 2014. 35:S40-S50. Doi:

10.1016/j.neurobiolaging.2014.02.031.

73 Brewer, G.J. Alzheimer’s disease causation by copper toxicity and treatment with

zinc. Frontiers in Aging Neuroscience. 2014. 6(92)1-5. Doi:

10.3389/fnagi.2014.00092.

74 Amyotrophic Lateral Sclerosis. U.S. Department of Health and Human Services,

National Institute of Health:Bethesda, MD, 2017.

75 Rowland, L.P. How amyotrophic lateral sclerosis got its name: The clinical-pathologic

genius of Jean-Martin Charcot. Archives of Neurology. 2001. 58(3)512-515. Doi:

10.1001/archneur.58.3.512

76 Renton, A.E., Chiò, A., and Traynor, B.J. State of play in amyotrophic lateral sclerosis

genetics. Nature Neuroscience. 2014. 17:17-23. Doi: 10.1038/nn.3584

77 Bannwarth, S., Ait-El-Mkadem, S., Chaussenot, A., Genin, E.C., Lacas-Gervais, S.,

Fragaki, K., Berg-Alonso, L., Kageyama, Y., Serre, V., Moore, D.G., Verschueren.

A,, Rouzier, C., Le Ber, I., Augé, G., Cochaud, C., Lespinasse, F., N'Guyen, K., de

Septenville, A., Brice, A., Yu-Wai-Man, P., Sesaki, H., Pouget, J., Paquis-

Flucklinger, V. A mitochondrial origin for frontotemporal dementia and

amyotrophic lateral sclerosis through CHCHD10 involvement. Brain. 137(8)2329-

2345. Doi: 10.1093/brain/awu138

78 Marangi, G., and Traynor, B.J. Genetic causes of amyotrophic lateral sclerosis: New

genetic analysis methodologies entailing new opportunities and challenges. Brain

Research. 2015. 1607:75-93. Doi: 10.1016/j.brainres.2014.10.009.

79 Cozzolino, M., Ferri, A., Valle, C., Carrì, M.T. Mitochondria and ALS: Implication

from novel genes and pathways. Molecular and Cellular Neuroscience. 2012. 55:44-

49. Doi: 10.1016/j.mcn.2012.06.001.

80 Chiò, A., Traynor, B.J., Lombardo, F., Fimorgnari, M., Calvo, A., Ghiglione, P.,

Mutani, R., and Restagno, G. Prevalence of SOD1 mutations in the Italian ALS

population. Neurology. 2008. 70(7)533-537. Doi:

10.1212/01.wnl.0000299187.90432.3f

81 Pansarasa, P., Bordoni, M., Diamanti, L., Sporviero, D., Gagliardi, S., and Cereda, C.

SOD1 in amyotrophic lateral sclerosis: “Ambivalent” behavior connected to the

disease. International Journal of Molecular Science. 2018. 19(5)1345-1358. Doi:

10.3390.ijms19051345.

82 Saeed., M., Yang, Y., Dend, H.-X., Hung, W.Y, Siddique, N., Delleface, L., Gellera, C.,

Andersen, P.M., and Siddique, T. Age and found effect of SOD1 A4V mutation

causing ALS. Neurology. 2009. 72:1634-1639.

83 Yim, H.S., Kang, J.H., Chock, P.B., Stadtman, E.R., and Yim, M.B. A familial

amyotrophic lateral sclerosis-associated A4V Cu, Zn-superoxide dismutase

mutant has a lower Km for hydrogen peroxide. Correlation between clinical

severity and the Km value. Journal of Biological Chemistry. 1997. 272(14)8861-

8863. Doi: 10.1074/jbc.272.14.8861

84 Valentine, J.S., and Hart, P.J. Misfolded CuZnSOD and amyotrophic lateral sclerosis.

Proceedings of the National Academy of Sciences. 100(7)3617-3622. Doi:

10.1073/pnas.0730423100

85 Andrus, P.K., Fleck, T.J., Gurney, M.E., and Hall, E.D. Protein oxidative damange in a

transgenic mouse model of familial amyotrophic lateral sclerosis. Journal of

Neurochemistry. 2002. 71(5)2041-2048. Doi: 10.1046/j.1471-4159.1998.71052041.x

86 Banci, L., Bertini, I., Boca, M., Girotto, S., Martinelli, M., Valentine, J.S., and Vieru, M.

SOD1 and amyotrophic lateral sclerosis: mutations and oligermizations. PLoS

One. 2008. 3(2)1-13. Doi: 10.1371/journal.pone.0001677.

87 Gordon, P.H. Amyotrophic lateral sclerosis: Pathophysiology, diagnosis, and

management. CNS Drugs. 2011. 25(1)1-15. Doi: 10.2165/11586000- 000000000-

00000

88 Celeste, D.B. and Miller, M.S. Reviewing the evidence for viruses as environmental

risk factors for ALS: A new perspective. Cytokine. 2018. 108:173-178. Doi:

10.1016/j.cyto.2018.04.010

89 Nani, F., Cifra, A., and Nistri, A. Trasient oxidative stress evokes early changes in the

functional properties of neonatal rat hypoflossal motoneursons in vitro. European

Journal of Neuroscience. 2010. 31:951-966. Doi: 10.1111/j.1460-9568.2010.07108.x

90 Mjos, K.D. and Orvig, C. Metallodrugs in medicinal inorganic chemistry. Chemical

Reviews. 2014. 114:4540-4563. Doi:10.1021/cr400460s

91 Long, G.V., Harding, M.M., Fan, J.Y., and Denny, W.A. Interaction of the antitumour

antibiotic streptonigrin with DNA and oligonucleotides. Anticancer Drug Design.

1997. 12(6)453-472.

92 Zakeri, B., and Wright, G.D. Chemical biology of tetracycline antibiotics. Biochemistry

and Cell Biology. 2008. 86(2)124-136. Doi: 10.1139/O08-002

93 Sadler, P.J. Inorganic chemistry and drug design. Advances in Inorganic Chemistry.

1991. 36:1-48. Doi: 10.1016/S0898-8838(08)60035-5.

94 Abraham, A.T., Jin, J-J., Newton, D.L., Rybak, S., and Hecht, S.M. RNA cleavage and

inhibition of protein synthesis by bleomycin. Chemistry and Biology. 2003. 10(1)45-

52. Doi: 10.1016/S1074-5521(02)00306-X

95 Hüttenhofer, A., Hudson, S., Noller, H.F., and Maschacark, P.K. Cleavage of tRNA by

Fe(II)-bleomycin. Journal of Biological Chemistry. 1992. 267(34)24471-24475.

96 Follet, S.E., Murray, S.A., Ingersoll, A.D., Reilly, T.M., and Lehmann, T.E. Structural

changes of Zn(II)bleomycin complexes when bound to DNA hairpains

containing the 5’-GT-3’ and 5’-GC-3’ binding sites studied through NMR

spectroscopy. Magnetochemistry. 2018. 4(1)1-26. Doi:

10.3390/magnetochemistry4010004

97 Chen, J., Ghorai, M.K., Kenney, G., and Stubbe, J. Mechanistic studies on bleomycin-

mediated DNA damage: multiple binding modes can result in double-stranded

DNA cleavage. Nucleic Acids Research. 2008. 36(11)3781-3790. Doi:

10.1093/nar/gkn302.

98 Dabrowiak, J.C., Greenaway, F.T., and Grulich, R. Transition-metal binding site of

bleomycin A2. A carbon-13 nuclear magnetic resonance study of the zinc(II) and

copper(II) derivatives. Biochemistry. 1978. 17(19)4090-4096. Doi:

10.1021/bj00612a034.

99 Sugiyama, M., Kumagi, T., Hayashida, M., Maruyama, M., and Matoba, Y. The 1.6-Å

crystal structure of the copper(II)-bound bleomycin complexed with the

bleomycin-binding protein from bleomycin-producing Streptomyces verticillus.

Journal of Biological Chemistry. 2001. 277:2311-2320. Doi: 10.1074/jbc.M103278200

100 Burger, R.M., Adler, A.D., Horwitz, S.B., Mims, W.B., and Peisach, J. Demonstration

of nitrogen coordination in metal-bleomycin complexes by electron spin-echo

envelope spectroscopy. Biochemistry. 1981. 20(6)1701-1704. Doi:

10.1021/bi00509a045

101 Ming, L.-J. Structure and function of “Metalloantibiotics”. Medicinal Research Reviews.

2003. 23(6)697-762. Doi: 10.1002/med.10052

102 Sugiyama, M., Kumagai, T., Hyashida, M., Maruyama, M., and Matoba, Y. The 1.6

Å crystal structure of the copper(II)-bound bleomycin complexed with the

bleomycin-binding protein from bleomycin-producing Streptomyces verticillus.

The Journal of Biological Chemistry. 2002. 277(3)2311-2320. Doi:

10.1074/jbc.M103278200

103 Manderville, R.A., Ellena, J.F., and Hecht, S.M. Solution structure of a Zn(II)-

bleomycin A5-d(CGCTAGCGC)2 complex. Journal of the American Chemical

Society. 1994. 116(23)10851-10852. Doi: 10.1021/ja00102a087.

104 Burger, R.M., Drlica, K., and Birdsall, B. The DNA cleavage pathway of iron

bleomycin: Strand scission deoxyribose 3-phosphate bond cleavage. The Journal of

Biological Chemistry. 1994. 269(42)25978-25985.

105 Que, B.G., Downey, K.M., and So, A.G. Degradation of deoxyribonucleic acid by a

1,10-phenanthroline-copper complex: The role of hydroxyl radicals. Biochemistry.

1980. 19:5987-5991.

106 Chang, Q., Wang, W., Regev-Yochay, G., Lipsitch, M., and Hanage, W.P. Antibiotics

in agriculture and the risk to human health: how worried should we be?

Evolutionary Applications. 2015. 8(3)240-247. Doi: 10.1111/eva.12185

107 Cycon, M., Mrozik, A., and Piotrowska-Seget, Z. Antibiotics in the soil environment –

degradation and their impact on microbial activity and diversity. Frontiers in

Microbiology. 2019. 10:1-45. Doi: 10.3389/fmicb.2019.00338

108 Kozlowska, A., and Szostak-Węgierek, D. Flavonoids – food sources and health

benefits. Bioactive Molecules in Food. 2014. 65(2)79-85. Doi: 10.1007/978-3-319-

54528-8_54-1

109 USDA Database for the Flavonoid Content of Selected Foods, Release 3.1; United

States Department of Agriculture, Agricultural Research Service. Beltsville, MD,

2014.

110 Panche, A.N., Diwan, A.D., and Chandra, S.R. Flavonoids: an overview. Journal of

Nutritional Science. 2016. 5(e47)1-92:10.1017/jns.2016.41

111 Spencer, J.P.E.; Crozier, A. Flavonoids and Related Compounds: Bioavailability and

Function. CRC Press: Boca Raton, 2012.

112 Selvaraj, S., Krishnaswamy, S., Devashya, V., Sethuraman, S., and Krishnan, U.M.

Flavonoid-metal complexes: A novel class of therapeutic agents. Medicinal

Research Reviews. 2013. 34(4)677.702.

113 Tay, W.M., da Silva, G.F.Z., and Ming, L.-J. Metal binding of flavonoids and their

distinct inhibition mechanisms toward the oxidation activity of Cu2+-β-amyloid:

Not just serving as suicide antioxidants! Inorganic Chemistry. 2013. 52(2)679-690.

Doi: 10.1021/ic301832p

114 Cherrak, S.A., Mokhtari-Soulimane, N., Berroukeche, F., Bensenane, B., Cherbonnel,

A., Merzouk, H., and Elhabiri, M. In vitro antioxidant versus metal ion chelating

properties of flavonoids: A structure-activity investigation. PLoS One. 2016. Doi:

10.1371/journal.pone.0165575

CHAPTER 2

“Beeting” Alzheimer’s: Prevention of Cu2+-Amyloid Beta Mediated Oxidationa,1,2

It is a physical disease, not a mystic curse; therefore it will fall to a physical cure. There’s time to kill the demon before it grows.

- Sir Terry Pratchett on his battle with Alzheimer’s disease before passing away in 20153

I. Background

1.1. Alzheimer’s Disease

The molecular cause of AD via neuronal degeneration is thought to be due, in part, to the aggregation of misfolded peptides called amyloid-β (Aβ) on neuronal

a Presented in part at the Spring 2018 National Meeting & Expo in poster and featured by an ACS press release in a press panel. The press release on this research has been shared by Newsweek, Phys.org, and social media like Youtube.1,2

55 tissue.4 While the broader causes of the disease through genetics and environmental factors were discussed in the first chapter, this chapter focuses on the role of CuAβ in the amyloidogenic pathway of the disease, especially the copper active center.

Physiologically the disease is currently thought to be based on neuronal cell death due to inflammation. Several ideas have been put forth to explain the induction of cell inflammation, including the presence of Tau proteins and/or accumulation of misfolded Aβ peptide. In the case of β-amyloid, inflammation is thought to occur as Aβ accumulates on brain tissue. The aggregation is thought to be induced in a cascade of misfolding Aβ peptides. The misfolded form of Aβ, having a relatively small size (40-42 amino acids long), can pass through the blood-brain-barrier. More recent studies have elucidated that both the polar and non-polar regions contribute towards aggregation in different ways. Early studies found beta-sheet and amorphous protein interaction could lead to aggregation in non-polar regions of the peptide.5,6 While some studies posit an inhibitory role of metals inhibiting aggregations, others show differences in the charge of the more polar regions upon binding to metals can also drive aggregation.7

The overall development of the disease from the amyloidogenic perspective begins with the enzyme β- and γ-secretases cleaving the tail-end of the amyloid precursor protein (APP).8,9 The amino acid sequence for the cleaved peptide sequence is typically 40-42 amino acids long with a sequence of

DAEFR5HDSGY10EVHHQ15KLVFF20AEDVG25SNKGA30IIGLM35VGGVV40IA42 (the predominant metal-binding amino acids, histidine, are underlined).

The fate of the Aβ peptide when properly regulated is not fully understood, but it is believed to be degraded into its amino acid components under normal conditions by the enzymes neprilysin and insulin-degrading enzymes. Some research, though, suggests that Aβ plays a role in protecting neuronal tissue in the hippocampus.10-12

There is currently no known method to break up the Aβ aggregates under physiological conditions, and thus no known cure for AD, though, therapies are being researched to ameliorate the progression of AD.13,14 By targeting the mis-folded Aβ peptide, the potential negative consequences that would be associated with targeting the healthy Aβ form can be avoided.

Figure 2.1 - Aqueous solution NMR structure of apo-Aβ adapted from Vivekanandan, et al., highlighting the metal-binding sidechains, His-6, His-13, and His-14, for reference, relaxed-eye view. PDB ID: 2LFM15

Misfolding of the peptide can occur under oxidative conditions in the presence of singlet oxygen and peroxides, and in the presence of unregulated metals, such as Cu2+ and Fe3+. Several studies have confirmed this interaction using a variety of methods, such as potentiometry, Raman spectroscopy, and NMR techniques.16-18 Structurally, metals Co2+ and Cu2+ were found to bind to His-6, His-13, and His-14 as shown in

Figures 2.1 & 2.2 from the JBC study by da Silva, et al. They were also able to conclude this conformation with relative minimized energies of each shape adopted for the Cu2+ and Cu+ ions as the metal went through a catalytic redox cycle. The Cu2+ ion forming an octahedron minimized to −385 kcal/mol and Cu+ a trigonal shape at −318 kcal/mol.19

Early studies in AD and its Aβ origins found that Al3+ could also play a role in the disease as certain places with high Al3+ concentrations in drinking water also had high incidences of AD. The correlation of Al3+ and AD, though, have come under more scrutiny in follow-up studies based on the earlier studies’ methodologies.20

To exacerbate AD and other diseases influenced by oxidative stress, the metal- bound form of Aβ has been shown to promote ROS activity. Some studies suggest this mechanism of metal binding could be related to biological sequestration of metals in cerebrospinal fluid (CSF), which would reduce potential effects of further ROS production in a short-term protective measure.21 Regardless of cause or effect in AD, metal-bound Aβ has been shown in several studies and confirmed in this study, to perform oxidative reactions which could lead to further oxidative stress implicated in

AD, neuronal inflammation, and other ROS-related diseases.

Figure 2.2 – a proposed structure of cobalt(II) bound to Aβ1-16. Adapted with permission. Used with permission of Journal of Biological Chemistry, from Catechol oxidase-like oxidation chemistry of the 1-20 and 1-16 fragments of Alzheimer’s disease- related β-amyloid peptide, Giordano F.Z. da Silva, Ph.D., William Tay, Ph.D., and Li-

June Ming, Ph.D. 280, 17, 2005.

Focus in recent years has recognized many potential chemicals that might inhibit metal-mediated ROS activity. By preventing the metal center from reacting with neurotransmitters and tissues in the brain, the inflammation and subsequent neuron apoptosis could be abated. Chelation therapies have shown some ability to reduce ROS activity and even reduce aggregation in vitro, though, studies have largely been unsuccessful in chelation therapies in vivo.22 The cultural popularity of dietary sources

59 of antioxidants has, in part, led to many studies examining phytochemicals like carotenoids, polyphenols, vitamins, and many others for the potential ability to ameliorate AD-related ROS activity.23-25 The requirements for this study for a novel inhibitor towards CuAβ were its ability to bind metal, potential stability in vivo, and can be found in a dietary source.

1.2 Betanin

Many a shirt, pants, and counter have been ruined by beets.b When staring at the new purple stain on your hands, shirt, or counter, you can now rest assured in the knowledge that, at its “root”, the red color from the beet is caused by betanin, or more formally betanidin-5-O-β-glucoside. Betanin is a phytochemical found predominantly in beetroots (Beta vulgaris), though, it can also be sourced in smaller amounts in the Swiss chard cultivar of beetroots and the prickly pear cactus (Opuntis ficus), and even some fungi in an interesting occurrence of convergent evolution.26 The secondary metabolite was “unearthed” for the first time in 1918 by Schudel, a fellow graduate student at that time, with its first recorded isolation in his dissertation.27 It is one of several betacyanic pigments of the betalain class of secondary metabolites. In beets, betanin is produced at

~30-300 mg per kilogram of beet.28 It is biochemically produced by spontaneous

b Sorry Erin and Dr. Ming.

60 condensation after two tyrosine amino acids are modified in the shikimate pathway using enzymes cytochrome P450 and DOPA 4,5 dioxygenase, respectively.29

Figure 2.3 – line representation of betanin and its predominant source, beetroot. Photo by Darrell Cole Cerrato

As synthetic food colorings have come under increased scrutiny for long-term safety, natural products have in turn become a more popular means of coloring for foods and drinks. Betanin is listed under E162 in the European E number coding, denoting its red color and ability to be consumed safely. It is often found in frozen foods given its sensitivity to heat and oxygen.56 The secondary metabolite has an

61 incredibly strong color with a molar absorptivity of 65,000 M—1 cm—1 at λmax of 535 nm.

The strong absorptivity is partly why it stains so well; a little goes a long way.

Like many secondary metabolites, the benefits of their production are not part of the necessary pathways for life. Instead, their production is likely to deal with environmental stressors, such as ROS.30 Many people have started looking at secondary metabolites like betanin as a source to reduce free-radical or oxidative activity given their availability in dietary supplements. Research has, in fact, shown that betanin could be a source of antioxidation, much like the “gold standard”, ascorbic acid (vitamin

C).31,56 One study found betanin could inhibit cytochrome c oxidation in mediated by peroxide of linoleate-tocopherol with an IC50 of 0.4 μM, stronger than the less conjugated betaxanthin and betanadin derivatives!32 Another extensive study found betanin performed free radical scavenging activity, in particular, which was further elucidated upon in earlier mouse-model studies that revealed that betanin might also play a role in tumor suppression.33-35

Unfortunately for researchers, and fortunately for your t-shirt, betanin’s degradation is relatively simple. It is easily photo-oxidizable and degrades under pH values >7.0. Depending on the conditions it will degrade into one of several de- carboxylated or hydrolytically cleaved products, L-DOPA and betalamic acid, to yield one of the yellow betaxanthin derivatives.66

Betaxanthin Betalamic acid Betacyanin

Figure 2.4 – derivatives and degradation products of betanin

Most people recognize betanin in their diet usually when their body cannot process it.c This condition, known as beeturia, affects about 1 in 10 people and is recognized by a distinct and often alarming red pigment in their urine.36 The carboxylates on betanin make its stability reliant on more acidic conditions, pH 3.0 to

c An older Russian gentleman was very excited to talk to me after I described my research to a group. He reminisced with me as he described a time when he was a little boy growing up in what was U.S.S.R. His mother frequently cooked borscht, a soup made up of mostly beets. As he thought back on it, he was laughing as he described how he started yelling for his mother when the color of his urine was an alarming red color!

7.0.37,38 The low pH stability makes it more stable in stomach acids and potentially other more acidic environments. The high concentration of the secondary metabolite, low renal absorption, and the inherently bright color can sometimes leave it intact and still visible through the digestive system even in small concentrations. In humans, beeturia has been shown to be higher in people with iron anemia, making the incidence of beeturia rise from around 13% to 80% for those suffering from untreated anemia.39

Interestingly, betanin production also seems to be higher in plants when grown in environments with higher than normal concentrations of several divalent metal cations, such as Cu2+, Fe3+, and Co2+. The reason for this is still not fully-understood, but is thought to likely be a response to increased amounts of cofactors required for betanin production or up-regulation in response to the environmental stressors.40

Several studies have also suggested a chelating role for betanin owing to the iminodiacetic-acid-like moiety of its structure, wherein it can act as an antioxidant by binding to and sequestering free-metals, especially copper and iron. In one study the stability of betanin was improved in the presence of citric acid, a strong antioxidant and also a chelator.41 Another study looked more directly at the chelating capabilities of betanin by examining an increase in betanin stability in the presence of the quintessential chelator ethylenediaminetetraacetic acid (EDTA).42 Missing from these reports, though, are direct measurements of metal-binding beyond observing secondary effects due to metal presence and the potential effects of preventing metal-mediating oxidation.

II. Results and Discussion

2.1 Antioxidation of CuAβ by Bn

In combining what was known about CuAβ and the potential metal binding capabilities of betanin, this study examined their interactions from the activity and structural perspective. For the purposes of examining the metal-binding capabilities of

Bn, Aβ1-16 was used for all experiments in order to maintain solubility in solution. From the metal-binding and reactivity perspective there is not a large difference between the short peptide and the large peptide. Moreover the shorter peptide contains more hydrophilic amino acids, maintaining solubility throughout the experiment.19

Previous studies by da Silva, et al have shown that CuAβ oxidizes catechols in the presence of aerobic O2 and H2O2.19 Testing catechols against CuAβ has significance given the necessity for catecholamine neurotransmitters, such as dopamine and serotonin, for proper brain function.19,47 Simple catechols can be used as models for early experiments to show catechol oxidase-like activity, which could give some indication of the potentially detrimental chemistry occurring in the brain. Catechol oxidation was measured using 3,5 DTBC in the following experiments, a standard chemical used in measuring catechol oxidase-like activity.43,44 Simple catechols often require forming an adduct with 3-methyl-2-benzothiazolinone hydrazine (MBTH) to stabilize and accurately visualize the quinone product, however the adduct band would overlap with the intense Bn band at 500-560 nm.

Catechol 3,5 di-tert-butylcatechol Norepinephrine Epinephrine Dopamine

Figure 2.5 – chemical structure of catechols including neutotransitters

Based on the da Silva, et al study on catechol oxidase-like activity, a type-3 di- copper reaction center was proposed, seen in the pathway below. The mechanism in

Figure 2.6 requires a two-electron transfer which can occur in vivo using O2 and/or

H2O2. As a type-3 copper, the di-copper center the active site forms bridged intermediates with substrates in a μ-η2-η2 fashion. Respiration and oxygen transport molecules make O2 present in a vast majority of the human body. Heightened production of ROS within mitochondria due to a number of mutations or dysregulation of the peroxide regulating protein superoxide dismutase could be a potential source for induced oxidative stress, making peroxide an “opportunistic” ROS. In combination with cationic metals, hydrogen peroxide can more efficiently form the highly reactive superoxide ion, responsible for the increased reactivity.45 In either the O2 or H2O2

66 pathways, an inhibitor binding to the metal in lieu of the oxygen, peroxide, or the catechol substrates could prevent turnover of the oxidized quinone.

Figure 2.6 - Adapted reaction mechanism for catechol (adapted for DTBC) with CuAβ proposed by da Silva, et al.19 Used with permission of Journal of Biological Chemistry, from Catechol oxidase-like oxidation chemistry of the 1-20 and 1-16 fragments of

Alzheimer’s disease-related β-amyloid peptide, Giordano F.Z. da Silva, Ph.D., William

Tay, Ph.D., and Li-June Ming, Ph.D. 280, 17, 2005.

An initial titration of up to 7 μM Bn to CuAβ shows an apparent decrease in oxidation of DTBC. The rate decreasing towards zero as betanin increases indicates reduction of DTBC turnover in the presence of Bn.

Rate (nM/sec) 2

0 0 2 4 6 8

[Betanin] ( M)

Figure 2.7 - inhibition of 3.2 mM DTBC oxidation by 3.5 μM (CuAβ)2 as a function of Bn in a pH 6.0 buffered solvent of 1:1 MES:MeOH

Experiments to determine inhibition were designed based on pre-equilibrium kinetics observed for CuAβ.19,47 For the purposes of applying Michaelis-Menten (MM) kinetics, the following simplified pathway can be used, which allows the standard MM pre-equilibrium kinetics equation to be applied.

Figure 2.8 – simplified reaction pathway for CuAβ oxidation of DTBC and its quantification using Michaelis-Menten equation (Equation 2.1) and its linearized form

(Equation 2.2)

푉푚푎푥×[퐷푇퐵퐶] 푣0 = Equation 2.1 퐾푚+[퐷푇퐵퐶]

1 퐾 1 1 = 푚 × + Equation 2.2 푣0 푉푚푎푥 [퐷푇퐵퐶] 푉푚푎푥

푘−1+ 푘푐푎푡 where 퐾푚 = Equation 2.3 푘1

Unfortunately, due to the high absorptivity of Bn, UV-vis kinetic measurements, while convenient for isolation and determining its concentration, are problematic during kinetics where even at small concentrations the light receiver can be quickly

69 reach the limitations of light detection between 500 and 600 nm. In the pre-equilibrium reaction kinetics, DTBC was measured at varying concentrations while holding CuAβ and Bn constant. Follow-up titrations would repeat the experiment with a different amount of Bn to determine the inhibition pattern based on the linearized plot of MM, the Lineweaver-Burk plot (Equation 2.2).

The data collected in Figure 2.9 was fitted to the Michaelis-Menten equation

(top), Equation 2.2. To determine the type of inhibition as betanin was increased the

MM fitting was converted to a double-reciprocal, Lineweaver-Burk plot and by evaluating the Vmax,app and Km,app derived from the MM fitting . The MM parameters from Figure 2.7 upon being fitted with the Lineweaver-Burk equation determined the

Vmax,app and the Km,app. Constants kcat and the catalytic efficiency were determined using this data, summarized in Table 2.1.

120

0 M Bn 100 1.50 MBn 2.50 M Bn 3.75 M Bn 80

Rate (nM/sec) 40

0 0 2 4 6 8 10 12 14

[DTBC] (mM)

3e+8 3.75 M Bn 3e+8 M Bn 1.50 M Bn 0 M Bn 2e+8

sec) 2e+8

-1

(M -1 1e+8

Rate 5e+7

0 -200 0 200 400 600 800 1000 1200 1400

-5e+7 -1 -1 [DTBC] (M )

Figure 2.9 - 5 μM CuAB oxidizing DTBC in the presence of Bn, in 1:1 MeOH:MES at pH

6.0 at 25 °C, Michaelis-Menten fitting (top) and Lineweaver-Burk fitting (bottom).

Table 2.1 - MM parameters of betanin inhibition of aerobic oxidation determined by fitting each line in Figure 2.6 to the MM hyperbolic equation. (a) based on dimeric

CuAβ concentration of 2.5 μM

Vmax Km,app kcata Catalytic Efficiencya (nM/sec) (mM) (sec—1) (M—1 sec—1) 0.0 μM 159.0 ± 9.6 6.6 ± 0.8 (63.6 ± 4.8) × 10—3 9.7 ± 0.1 1.25 μM 85.5 ± 5.6 7.5 ± 1.0 (34.2 ± 2.8) × 10—3 4.6 ± 0.2 2.50 μM 76.8 ± 7.7 10.0 ± 1.8 (30.7 ± 3.9) × 10—3 3.1 ± 0.2 3.75 μM 73.6 ± 6.4 13.9 ± 1.9 (29.4 ± 3.2) × 10—3 2.1 ± 0.2

Convergence of all the fits in Figure 2.9 reveals apparent mixed-type inhibition.

Titration of DTBC in the presence of varying amounts of Bn revealed an apparent mixed-type inhibition based on both an increase in Km,app and a decrease in Vmax,app.

Further evaluation for inhibitor constants Kic and Kiu were determined based on the following pathway in Figure 2.10:

Figure 2.10 – mixed inhibition pathway of Bn towards CuAβ oxidation of DTBC featuring competitive (binary EI complex) and uncompetitive (ternary ESI complex pathways)

The stability of the binary Bn-CuAβ complex can be described by the dissociation constant, Kic, where constants Km and Vmax are determined by fitting the MM equation to the data obtained under no inhibition by Bn. Kic was determined by fitting a modified

MM equation where Vmax,app/Km,app are a function of Bn, Figure 2.11.

푉푚푎푥,푎푝푝 푉푚푎푥/퐾푚 = Equation 2.4 퐾푚,푎푝푝 1+[퐵푛]/퐾푖푐

m,app

/K 20

max,app

0 0 1 2 3 4 5 6

[Betanin] ( M)

Figure 2.11 –MM fitting for Figure 2.7 as a function of Bn for the determination of the

Kic dissociation constant of Bn–CuA binary (EI) complex

Fitting Vmax,app/Km,app as a function of Bn in Table 2.1 to Equation 2.4 reveals a

Kic constant of 1.11 ± 0.03 μM, which has some interesting implications. The incubated concentration of CuAβ was 5 μM, indicating a sub-stoichiometric competitive inhibition of Bn towards mononuclear CuAβ. Taking into account the previous study by da Silva, et al that proposed a dimeric active site, this data fits well with 1:1 binding of the EI

(betanin-CuAβ or enzyme-inhibitor) complex. Under the condition that 5 μM CuAβ

74 would yield 2.5 μM active sites, tight binding of Bn towards a di-copper center under these conditions would reveal a Kic near 1.25 μM, which is close to what is observed.

Similarly, the dissociation constant for the Bn ternary complex of Bn-CuAβ-

DTBC, Kiu, in the uncompetitive pathway can be determined by:

푉푚푎푥 푉푚푎푥,푎푝푝 = Equation 2.5 1+[퐵푛]/퐾푖푢

Figure 2.12 - Plot of the data from Table 2.1 fitted to Equation 2.5 in determination of dissociation constant, Kiu, of Bn ternary binding of Bn-CuAβ-DTBC (R2 = 0.9)

The fitted data in Figure 2.12 reveals a relatively low dissociation constant of Kiu,

2.2 ± 0.4 μM. Unlike the clear 1:1 competitive nature revealed by the Kic, the Kiu reveals a more ambiguous value. This could be due in part to a more substrate favored equilibrium once catechol is bound to the active center. Alternatively, this could also be indicative of monomeric reactivity not bound to Bn under a different, but less predominate, reaction pathway.

Unfortunately, the nature of this type of study is not able to evaluate O2 as a substrate, where inhibition of O2 binding would prevent oxidation by preventing electron transfer and subsequent formation of H2O2. One of the byproducts of concern for metallo-protein oxidation is the production of H2O2 which can increase oxidation rates through a redox cycle with a copper ion.46,47 For example, prior to normalization of

Figure 2.13, the overall rate of DTBC oxidation rose five-fold compared to ambient O2.

Given the presence of H2O2 occurring naturally and as a by-product of O2 electron transfer, inhibition for H2O2 turnover is essential for any inhibitor to remain effective in preventing ROS activity in AD.

76 Figure 2.13 – Oxidation of 3.2 mM DTBC as a function of H2O2 titration to 5 μM

(CuAβ)2 without inhibition in 1:1 MES:MeOH solvent buffered at pH 6.0, 25 °C, kcat of

(9.3 ± 0.7) × 10—3 sec—1

Peroxide was incubated with varying amounts of Bn at 30 mM, a close approximation to the Km determined upon fitting Figure 2.13 to yield 34.8 ± 6.4 mM, closely resembling Km values of previous studies.19

77 1.2e+7 5.0 M Bn 1.0e+7 2.5 M Bn 1.0 M Bn 0 M Bn 8.0e+6

sec)

-1 6.0e+6

-1

Rate 4.0e+6

2.0e+6

0.0 -1000 -500 0 500 1000 1500

-1 -1 [DTBC] (M )

Figure 2.14 – Double-reciprocal plot for DTBC titration in the presence of varying amounts of Bn with constant 30 mM H2O2 and 10 μM CuAβ

Dissimilar to the O2 example, Bn shows more competitive character in the presence of H2O2 from Figure 2.14, though, could also be showing some mixed type inhibition as well. The ambiguity of the nature of inhibition is based on the proximity of the convergence point for all the fitted lines for each concentration of Bn. Pure competitive inhibition would converge on the y-axis, indicating similar turnover numbers and apparent maximum rates for all catalysts without Bn bound. Instead pure

78 competitive inhibition would differ in Km,app alone, revealing an equilibrium pushing away from the ES complex and more towards the binary EI complex.

Table 2.2 - MM parameters of betanin inhibition of peroxide mediated oxidation determined by fitting each line in Figure 2.14 to the MM hyperbolic equation. (a) based on monomeric concentration of CuAβ

Catalytic Vmax kcata Efficiencya Bn (μM) (nM/sec) Km (mM) (sec—1) (M—1 sec—1) 0.0 μM 597 ± 54 1.2 ± 0.2 (29.9 ± 2.7) × 10—3 24.7 ± 0.2 1.0 μM 569 ± 37 1.8 ± 0.2 (28.4 ± 1.9) × 10—3 17.2 ± 0.1 2.5 μM 502 ± 27 2.1 ± 0.2 (25.1 ± 1.4) × 10—3 11.4 ± 0.1 5.0 μM 502 ± 28 2.7 ± 0.1 (25.1 ± 1.4) × 10—3 9.2 ± 0.1

In a similar analysis compared to the inhibition of CuAβ by Bn with aerobic O2, the stability of the ES complex CuAβ-DTBC decreases as the apparent Km increases.

Investigating the competitive nature of Bn towards the metal center in the presence of

H2O2 reveals a Kic of 2.3 ± 0.3 μM, where the concentration of monomer CuAβ is 10 μM, fitted in Figure 2.15. Similar to the O2 evaluation, the sub-stoichiometric constant likely indicates inhibition of a dimeric catalyst, making the effective concentration of CuAβ at

5.0 μM. The Kic for DTBC in the presence of H2O2 is comparable to that of the Kic in the presence of O2 with reference to the Km of(CuAβ)2. This relative stability of the EI complex against DTBC binding is consistent despite the electron source for turnover, O2 versus H2O2.

Figure 2.15 - Replot for Kic dissociation constant determination of Bn binary complex in the presence of 30 mM hydrogen peroxide

As explored in detail in Chapter 1, excessive generation of H2O2 can further exacerbate redox activity when unregulated. Disproportionation or catalytic turnover into the highly redox active superoxide, O2•—. A strong inhibitor bound to the metal active could prove to be a potent redox inhibitor. Competitive inhibition would show a favor in equilibrium towards the EI complex, or an inhibitor as large as betanin could alternatively prevent H2O2 access to the binding.48,32

80 Inhibition towards CuAβ with reference to H2O2 as a substrate was performed in the presence of varying amounts of Bn, Figure 2.16. For each peroxide titration, betanin,

DTBC, and CuAβ were held constant, whereby Bn was varied in subsequent experiments to ascertain Vmax,app and Km,app as a function of Bn. Based on the data profile the type of inhibition was evaluated for its competitive, uncompetitive, or mixed inhibition pattern. Given the non-zero value for rate without peroxide (푦0), the data were fitted to a modified MM equation:

푉푚푎푥×[퐻2푂2] 푣0 = 푦0 + Equation 2.6 퐾푚+[퐻2푂2]

The profile for hydrogen peroxide titration, upon fitting the data to the double- reciprocal Lineweaver-Burk plot, Equation 2.6, revealed a mixed inhibition pattern.

This is further verified by analyzing Km (competitive) and kcat (uncompetitive) shown in

Table 2.3. The increase in apparent Km values from the Km without Bn reveals competitive inhibition via Bn binding directly to the active site. Moreover, an evaluation of Vmax shows uncompetitive type inhibition. These pathways can be tentatively applied to Figure 2.17, though, the uncompetitive pathway might require further evaluation to elucidate the nature of this type of inhibition towards H2O2 binding.

81 5e+8 0.00 mM Bn 0.625 mM Bn 4e+8 1.25 mM Bn 2.50 mM Bn 3e+8 sec) 1 2e+8

1/Rate (M 1e+8

0 0 500 1000 1500 2000 2500

-1e+8 1 1/[H2O2] (mM )

Figure 2.16 - Bn inhibition of H2O2 as a substrate for DTBC (1 mM) turnover by 5 μM

CuAβ

Table 2.3 – MM parameters of betanin inhibition of hydrogen peroxide mediated oxidation with respect to H2O2 determined by fitting each line in Figure 2.14 to the MM hyperbolic equation. (a) based on dimeric concentration of CuAβ

Catalytic Vmax kcata Efficiencya Bn (μM) (nM/sec) Km (mM) (sec—1) (M—1 sec—1) 0.0 μM 103 ± 11 3.5 ± 0.1 (41.2 ± 4.2) × 10—3 11.8 ± 0.2 0.625 μM 69 ± 13 4.1 ± 0.2 (27.5 ± 5.2) × 10—3 6.8 ± 0.4 1.25 μM 50 ± 15 4.6 ± 0.3 (18.8 ± 6.0) × 10—3 4.1 ± 0.6 2.50 μM 38 ± 3 6.0 ± 0.8 (15.3 ± 1.2) × 10—3 2.6 ± 0.2

82 Figure 2.17 – simplified reaction pathway in describing Bn competitive and uncompetitive inhibition pathways towards H2O2 reactivity with CuAβ.

In evaluating the competitive dissociation constant, Kic, for the the Bn-CuAβ complex, Vmax,app/Km,app were plotted as a function of Bn. Fitting this data to Equation

2.4 revealed a Kic toward H2O2 turnover of 0.76 ± 0.04 μM. This value is curiously low, given that 50% inhibition would occur at 1.25 μM betanin with dimeric CuAβ. This value could be indicative of error propagation in measuring slow reactions. Chemically, this value could also partially indicate decomposition of H2O2 through H2O2-Bn interactions. Studies have shown several pathways by which Bn can by oxidized by peroxidases in the presence of hydrogen peroxide. Moreover, the cited study found some of the antioxidant electron pathways occurring in the non-metal-binding motif.49

Secondary interactions likely would not account for a Kic twice as low as expected for most strong inhibitors and would require more exploration in the binding of Bn and

83 CuAβ. Moreover, the presence of peroxide might also induce small amounts of oxidation on the peptide that could trigger small amounts of aggregation of the peptide not observed in previous studies, thus sequestering the active metal center.

Figure 2.18 – Replot for Kic dissociation constant determination of Bn-CuAβ binary complex, inhibiting H2O2 binding

On top of the competitive inhibition exhibited by betanin towards hydrogen peroxide binding, betanin also showed uncompetitive inhibition. In this pathway the dissociation constant, Kiu, for the Bn-CuA-O22— ternary complex. Fitting the Vmax,app

84 data recorded in Table 2.3 with respect to the Bn concentration yields Figure 2.19. Upon fitting the apparent Vmax values as a function of betanin to Equation 2.5 reveals a Kiu for the ternary complex of 1.28 ± 0.08 μM. Uncompetitive inhibition towards the ESI

(betanin-CuAβ-DTBC or enzyme-substrate-inhibitor) complex indicates a lack of turnover upon binding of the ternary complex. This implies that Bn could play a role in the electron transfer process, making superoxide oxidation unfavorable. A study by

Kumorkiewicz, et al on Cu2+ and Cu1+ turnover in the presence of betanin, using the proposed binding motif explored in the subsequent section herein, have explored mechanisms by which betanin can interact in the redox process.50 The high conjugation of betanin makes one of several betacyanin derivatives stably possible.

Figure 2.19 - Replot for Kiu dissociation constant determination of Bn-CuAβ-H2O2 ternary complex turnover

2.2 Betanin-Copper Binding

In understanding the inhibitory effects of Bn towards CuAβ, several experiments were performed to examine the nature of the Cu2+ ion to Bn. Studies have examined the degradative qualities metals have in solution towards Bn that have also been observed in this lab when betanin and metal interact under unbuffered conditions.51 Fortunately, relatively “fast” experiments (10-20 minute incubation periods) showed relative

86 stability under buffered conditions at pH 6.0. Qualitatively, a titration of Cu2+ to Bn reveals a decrease in the Bn signal at 535 nm and an apparent new signal forming at 505 nm, shown in Figure 2.20. Apparent Cu2+ saturation appears at ~50 μM, indicating the equilibrium of the metal-ligand complex.

Figure 2.20 - normalized UV-vis spectrum for the titration of Cu2+ to Bn in MES pH 6.0 buffer.

Prior to determining the association constant for the Cu2+-Bn complex, the stoichiometric relationship of the two were determined. Job plots, or methods of continuous variation, can reveal how a physical attribute, such as light absorbance or chemical shift, can change based on the equilibrium of a complex. As the common name suggests, both compounds are simultaneously varied for each point to maintain a total concentration. Copper-ligand binding can often be observed by taking advantage of d-d transitions or charge transfer transitions that can be measured in a UV-vis spectrum.

Figure 2.21 shows a Job plot where the absolute value for the change in absorbance was

87 measured at 400 nm as a function the molar ratio of copper where the Cu:Bn concentrations were held at a total 9.1 μM throughout the experiment. The absorbance at 400 nm was chosen to reduce the influence unbound Bn would have on the measurement. Without fitting, the mole fraction ratio (χCu:χBn) is approximately 0.3:0.7, suggesting that a 1:2 Cu2+:Bn complexation is a predominate species under the given conditions.

0.0010

0.0008

0.0006

400nm

A 0.0004

0.0002

0.0000 0.0 0.2 0.4 0.6 0.8 1.0 1.2

Cu [Cu]/([Bn]+[Cu])

Figure 2.21 – Job Plot (continuous variation) for Cu-Bn absorbance due to binding as a function of the mole fraction of Cu2+, χCu, in 100 mM pH 6.0 MES buffer

This yields a complexation equilibrium of:

[퐶푢퐵푛] Where 퐾 = 푎1 [퐶푢][퐵푛]

[퐶푢퐵푛 ] and 퐾 = 2 Equation 2.6 푎2 [퐶푢퐵푐][퐵푛]

Which can be expanded to for binding constant terms, Ka1 and Ka2 (Equation 2.7):

3 2 0 = [퐶푢] 퐾푎1퐾푎2 + [퐶푢] {퐾푎1(2퐾푎2[퐵푛]0 − 퐾푎2[퐶푢]0 + 1))} + [퐶푢]{퐾푎1([퐵푛]0 − [퐶푢]0) + 1} − [퐶푢]0

While 1:1 complexes can still retain a relatively more trivial analysis, it still contains a quadratic function. In the 2:1 binding case, equilibrium constants determinations would require solving a cubic function! Many current software have some capabilities for fitting these complex equations with multiple parameters. The Nelder-Mead method of fitting optimization was employed through a Matlab platform, BindFit.52,53 Fitting of

Figure 2.22 for a copper titration to Bn using BindFit software revealed Ka1 of (6.4 ± 2.5)

×104 M—1 for the first equilibrium of 1:1 and (2.2 ± 0.3) ×104 M—1 for the 2:1 Ka2 equilibrium.

0.4

0.3

535 0.2

A Δ

0.1

0 0 50 100 150 200 [Cu2+] (μM)

Figure 2.22 – Titration of Cu2+ to Bn at 9.1 μM, in 100 mM pH 6.0 MES buffer, where the change in absorbance at 535 nm is measured as a function of Cu2+ titrated

Assuming the 2:1 constant is valid based on the Job plot these constant show strong binding towards Cu2+ in aqueous environments. While this is not quite as strong as most chelators exhibiting equilibrium constants at 109-1017 M—1, this could still be an advantage for betanin as a dietary source of preventing metal-mediated oxidation under environments where copper dysregulation is prevalent. Chelators with relatively large Ka constants could pose more risk than benefits as these types of chelators can leave people anemic, ill from dysfunctional proteins deficient of necessary metal resources required for metalloproteins and cofactors.

Most research has either assumed binding or examined binding by observing secondary effects. Direct evidence of metal binding, such as NMR or crystallography, have never been published prior to this study. The challenge of the latter technique was personally experienced as only potentially amorphous crystals could be isolated. Based on the structure of betanin, it can be assumed that the iminodiacetic acid portion of the molecule is likely responsible for metal binding given the close proximity of the electronegative carboxylates and amine nitrogen. The iminodiacetic acid moiety is also chosen over the carboxylate adjacent to the Schiff base nitrogen given the proximity of the repulsive positive charge the Schiff base would yield towards metals, yielding a local neutral charge.

Paramagnetic NMR can be an incredibly useful, though often under-used tool, to examine metal-ligand binding. Paramagnetic metals can cause isotropic shifts in the proton NMR signal, resulting in far shifted signals compared to the diamagnetic region.

While the technique is less sensitive than diamagnetic NMR, the pulse sequences can be optimized for the fast relaxation time protons near a paramagnetic metal center experience. The use for models in paramagnetic NMR is also helpful. Some paramagnetic metals, like Cu2+, can yield signals so broad and far up- or downfield from the transmitter pulse that NMR signals cannot be well-resolved. A common metal used in lieu of Cu2+ is Co2+, given their similar binding properties and bond distances.54

Moreover the signals for nuclei proximal to Co2+ tend to be relatively sharper and

91 within a reasonable spectral window for excitation.55 Figure 2.23 reveals metal binding of Bn to up to 1 molar equivalent of Co2+ in D2O.

Figure 2.23 – paramagnetic NMR 1H spectrum of betanin with 0.25 (C), 0.5 (B), and 1.0

(A) stoichiometric equivalents of Co2+ bound to betanin at 1.1 mM in D2O.

Using water as a reference for chemical shift and intensity in each spectrum, it is clear that several signals appear at 101.7, 38.7, 25.0 ppm. Smaller signals at 78.6, 24.4, and 16.3 ppm can likely be attributed to a small amount of the Bn isomer, iso-betanin, or one of several products betanin can form when degraded. Solvent, pH and presence of metals have shown form one of several oxidized forms of betanin, many of which maintain a metal-binding motif, such as neobetanin, a more conjugated oxidized form of betanin.49-51 After numerous rounds of purification, as discussed in the methods section, the most likely contaminant would be iso-betanin, a structurally identical molecule with the exception of a single stereocenteric carboxylate, which partially co-

92 elutes from RP-HPLC with betanin.56 Signals at 101.7, 38.7, and 25.0 ppm are likely the primary betanin signals based on their strong intensity. The most likely protons being observed are the α (CH) and β (CH2) protons adjacent to the stereocentric carboxylate and potentially the β proton adjacent to the conjugated carboxylate.

Ideally, an EXSY pulse sequence can be employed to correlate isotropically shifted signals to their diamagnetic counterparts by taking advantage of the fast exchange of the binding equilibrium. The typical paramagnetic EXSY pulse sequence is a modified NOESY using short recycle times and typical mixing times of ~1-20 ms.57,58

Based on the lack of EXSY correlations observed at varying amounts of Co2+, exchange could not be measured with the mixing times of 1, 5, 10, or 20 ms. Instead a tentative signal assignment and structural information was explored using the model systems discussed below.

The initial model system used was iminodiacetic acid based on the proposed metal-binding motif of betanin. The NMR spectra of iminodiacetic acid (Figure 2.25) showed very simple apo- form diamagnetic and metal-bound paramagnetic spectra. In the apo- form spectrum a single signal is apparent at 3.2 ppm, assigned to both sets of

CH2 α protons. The single signal reveals the molecular symmetry and magnetic equivalence of all four α-CH2 protons. In the paramagnetic spectrum, the signals at 84.5 and 12.3 ppm indicate two types of interactions α-CH2 protons with the electron spin- system from the paramagnetic Co2+ ion. Given the similar peak heights and widths for

93 both paramagnetic signals and a molar equivalent of both compounds, a single 1:1

Co:IDA is likely revealed.

Figure 2.24 – iminodiacietic acid compared to the proposed metal-binding region of betanin

This can imply electronically non-equivalent geminal and vicinal protons due to non- planar binding of the metal with respect to the relatively planar six rings of Bn. With only one somewhat flexible, non-conjugated CH2, the six-membered ring as well as the tryptophan moiety maintain a mostly planar shape. Despite the general planarity of the molecule, the non-conjugated carboxylate may force out-of-plane binding with respect

94 to the rest of the molecule. Differences in geminal and vicinal protons in relation to a metal center have been observed in previous studies, often showing a more downfield shifted signal for the equatorial proton compared to the axially positioned proton.59,60

The large difference in chemical shifts also indicates a more distinct equatorial-axial proton positions with respect to the Co2+-carboxylate plane.

Figure 2.25 – apo-iminodiacetic acid (top) with 1 molar equivalent (2.0 mM) of Co2+ in

D2O (bottom)

95 In determining the proposed geometric positions of the geminal protons with respect to the Co2+ center on both IDA and Bn, nitrilotriacetic acid (NTA) was mixed with an equivalent of Co2+. The NTA model would yield tridentate binding with the additional carboxylate. This type of binding would force all protons out of the plane, becoming equivalent on average to one another with reference to the metal center.

Figure 2.26 shows a single broad signal as opposed to the two signals found in IDA, confirming the magnetically equivalent geminal α-CH2 protons. The broad signal at 64 ppm indicates an averaging between both protons in a geminal position when compared to the two signals of IDA that have a 70 ppm difference.

Figure 2.26 – 1H NMR paramagnetic spectrum of 1 mM nitrolotriacetic acid with 1 molar equivalent of Co2+ in D2O

Based on the IDA data, the signal at 101.7 ppm in the CoBn spectrum can be tentatively assigned to the α-proton for the stereocentric carboxylate. Moreover, it suggests the equatorial geometry of the α-CH2 proton with respect to the metal center and out-of-plane binding of Co2+ with respect to the six-membered ring. This is represented by an approximation of the structure in the Figure 2.27.

Figure 2.27 – a tentative structure of Cu2+-betanin (glucose not included for minimization simplicity) minimized using Chem3D.61 Relaxed-eye view

A final study on the inhibitory effects of Bn towards free copper showed similar inhibitory effects as the CuAβ studies. In this experiment, though, incomplete inhibition is observed (Figure 2.28). The two methods by which this type of seemingly non-

97 competitive inhibition would occur would be through binding the substrate, which is unlikely given previous experiments, or by incomplete competition with the substrate towards the metal binding site. As observed in the Job plot, Bn could bind at most in a

2:1 Bn:Co fashion, which could leave d-orbitals exposed for quaternary binding of substrates O2, H2O2, and catechol. Apparent inhibition in the presence of free Cu2+ show apparent differences when compared to the Job plot, where Bn contribution towards inhibition appears to plateau after a single equivalent. This could be due in part to the competition of the substrate at 4000 times the concentration of Cu2+. Moreover, this study required organic solvent to maintain solubility of the quinone product, which was not necessary in the binding study.

2.0

1.8

1.6

1.4

1.2

1.0

0.8

Rate (nM/sec) 0.6

0.4

0.2

0.0 0 1 2 3 4 5

[Bn] ( M)

Figure 2.28 – Inhibition of free Cu2+ oxidation by Bn (1 μM Cu, 4 mM DTBC) in 1:1 100 mM pH 6.0 MES:MeOH solvents

2.3 Beet Extract Redox Inhibition and Metal Binding

In the above studies I have measured the effects of betanin alone to determine the type of chemistry it can employ towards copper-mediated oxidative stress. In the dietary sense betanin will never be isolated during consumption. For those interested in seeking to apply more beets in their diet to potentially take advantage of its anti-ROS effects, betanin, is

99 fortunately among the more concentrated chemicals in beets in addition to other ROS-inhibiting polyphenols and vitamins in beets.

The following studies are more qualitative in nature just show how beet extract is either inhibiting catechol oxidase-like activity or a cursory view of the metal-binding capabilities of products in the extract. A full quantitative analysis of all the products individually in beets would be far beyond the scope of this dissertation, given the numerous beneficial compounds currently known to be in beets would take years to catalogue and quantify.

In the first experiment I measured the inhibition of Cu2+ redox activity towards DTBC in the presence of beet extract. The beet extract used in this experiment was obtained from beets bought at a local grocery store. The extract was acquired after homogenization and simple filtration through a general, large pore filter to remove cellulose and large debris from the extract. Betanin was used as a reference for relative concentrations given its facile determination using its strong absorptivity in the UV-vis spectrum and due to its high concentration in relation to most other phytochemicals, especially antioxidants. Encouragingly, beet extract shows a rapid decrease in oxidative activity towards catechols in the presence of free Cu2+, reaching an equilibrium at ~10 μM Bn reference. The inhibition of Cu2+-mediated oxidative activity was potentially reduced by ~50% compared to the oxidation rate measured without any beet extract. This value might not be entirely representative of oxidation of catechol. Many polyphenols and several derivatives of betanin can turn yellow when exposed to metal ions, due to redox activity of the respective molecule, consistent with previous research on the effects of Cu2+ towards Bn in an extract.62 At this stage of the purification, polyphenols, ascorbic acid, and hundreds of other compounds still exist in the reaction vial. This could also indicate an equilibrium reached with catechol oxidizing enzymes present, such as catechol oxidase. In this

100 qualitative study, this at least shows the apparent ability for chemicals in beets to potentially bind to metal and show how beets as a whole might have a “nutraceutical” effect towards ROS activity.

This cursory examination of the overall antioxidant activity of beet extract could also be misleading in the favor of more antioxidation. Observing catechol oxidation is based on examination of a yellow color forming as the quinone is produced. Betanin has been shown to be more stable upon isolation, where in extract form it is more apt to decompose into one of several yellow derivatives.66 The halt in production of yellow color might reveal an equilibrium of chemical decomposition as well as oxidation of catechol. Therefore any reduction in ROS activity is occurring proportionally with product degradation with respect to molar absorptivities of all compounds.

101

1e-6

8e-7

6e-7

4e-7

Rate (mM/sec)

2e-7

0 0 10 20 30 40 50 [Bn] (uM)

Figure 2.29 – inhibition of free Cu2+ oxidation by beet extract as a function of Bn concentration

In a final examination, qualitative metal binding of beet extract towards Co2+ was observed with 260 μM Bn (71.8 μg/500 uL) as reference for the extract. Figure 2.30 shows a several strong signals. While signals at ~100 ppm and 25 ppm show potential betanin signals, several signals show across a 200 ppm window that were not observed

Figure 2.23. Using peak heights and widths as a reference for each compound, it can be surmised that a few predominant compounds exist in solution that have metal binding capabilities!

102

Figure 2.30 – 260 μM Bn in D2O as part of beet extract in the presence of 1 molar equivalent of Co2+

III. Concluding Remarks

From the perspective of betanin as a food source and not a medicine, the small

Kic and Kiu values for inhibiting CuAβ under aerobic O2 and H2O2 conditions are encouraging. This study shows that Bn can inhibit catechol oxidase-like activity in vitro caused by CuAβ. As more is understood about the contribution of CuAβ towards AD,

Bn provide a means of reducing the oxidative stress involved with the disease and potentially other metal-based neurological disorders. The mixed inhibition for Bn implies its versatility as an inhibitor towards Cu2+-mediated oxidative activity towards catechols. Active site competition between Bn and substrates expectedly shows tight

103 binding with low dissociation constants, making substrate binding difficult, especially for catechols. Interestingly, the low Kiu values could be showing the antioxidant versatility proposed by other previous studies.50,49 The highly conjugated structure of

Bn allows for a number of pathways that could all contribute towards the antioxidation by stabilizing electron transfer into its complex.

This is also the first study to acquire primary data that could propose structural information on Bn-Cu2+ binding. As previously mentioned, other studies have implied binding by examining the degradation of Bn. This is the first study to acquire data that shows metal binding to Bn itself. Unfortunately, the challenge of performing 2D experiments on paramagnetic molecules due to either lack of exchange or undetermined sequence optimization prevented the complete signal assignment, though, it does at least conclude metal binding.

As encouraging as these studies are in potentially using betanin or molecules inspired by betanin, there is still more to be observed. Part of the challenge in working with betanin is that it is often not soluble or stable in many solvents making more in- depth experiments, such as a binary or ternary EXSY, incredibly challenging for a complete signal assignment!

Outside of my lab’s expertise there is still more work to evaluate in betanin’s bioavailability. It is essential that betanin is able to pass through a number of membranes between intestinal linings to the blood-brain-barrier to affect Aβ.63,65 Many of the studies at the time of writing this leave room for improvement. Many secondary

104 metabolites and flavonoids are not well-absorbed, though, the absorbance has been shown to increase upon the addition of glucosides, similar to betanin. A similar metal- binding antioxidant, quercetin, was shown to have nearly 20 times the bioavailability at

10 times the gut absorption rate when bound to glucoside!64,65

This study could also be expanded upon, especially in the understanding of how the other molecules in the betalaine family can bind to metal. Betanin often degrades by decarboxylation or cleavage of the cyclo-DOPA moiety via cleavage of the aldimine bond that would form one of several yellow betaxanthin derivatives. Many of these could also have antioxidant properties given the often partially or wholly preserved iminodiacetic acid metal-binding motif.66,67 This could make a case for investigating betanin further against CuAβ redox chemistry since its degradation might not be the end of the story!

105

IV. Materials and Methods

4.1 Natural Product Treatment

All beet extract was removed directly from beetroot, Beta vulgaris, purchased in a local supermarket in Land O Lakes, Florida.d Extract was procured by peeling and grating the root, followed by homogenization in a food processor. The collected samples were then frozen at –5 °C overnight to further lyse cells. Cellular debris was removed by centrifugation at 3500 rpm for 10 min. The lysate was then purified by 0.2

μm pore-sized Whatman® filters to remove any remaining cellular debris and then by

Sephadex G-25 column filtration to isolate betanin from the extract. Quality and quantification of betanin was determined using UV-Vis spectrophotometer (ε535 = 65,000

M—1 cm—1). Further quality was measured by NMR spectroscopy (Varian Direct Drive

500 MHz spectrometer) and processed by VNMRJ.

d Betanin can be purchased by some companies. Frustratingly, some of the companies use techniques for “purification” do nothing to remove sugar from the compound yielding about 1000:1 w/w sugar NMR to betanin! The 5g bottle I received turned out to be about 4.995 g sugar. Thus, began the journey of learning natural product extraction.

106

Figure 2.31 - general purification of Bn through homogenization, micropore filtration, and column chromatographies

4.2 Antioxidation Kinetics

Reaction rates for oxidation and peroxidation were performed using 5 uM for ambient O2 studies and 10 uM Cu(II)Aβ for H2O2 studies titrating 0.25-12.8 mM 3,5-di- tert-butylcatechol (DTBC), in 1:1 (v/v) MES:MeOH. 30 mM H2O2 were used for peroxidative studies. The pH at 6.0 was held constant by 100 mM MES at 6.0 to maintain Bn stability. Varian Cary 50 spectrophotometer was used to measure reaction rate of DTBC forming oxidized o-quinone product at A410 (1900 M–1 cm–1). β-amyloid peptide 16-mer concentration was determined at A280, where ε280 was 1490 M—1 cm—1.

107

4.3 Paramagnetic NMR

1 mM betanin was dissolved in D2O or H2O with up to a stoichiometric equivalent of CoCl2. Paramagnetic isotropic chemical shifts for 1H nuclei were detected using the super-weft pulse sequence (180-τ-90) consisting of 10,000 scans with a recycle time of 100 ms. Chemical shifts were recorded using Agilent’s Varian Inova 600 MHz spectrometer and processed with VNMRJ software. Paramagnetic 1H spectra were processed with up to 20 Hz line broadening window function.

V. References

1 Covey, D. Beetroot could prevent progression of Alzheimer’s disease, study shows.

Newsweek. 2018. [Online] https://www.newsweek.com/betanin-alzheimers-

disease-drug-development-beets-855232.

2 Vegetable compound could have a key role in ‘beeting’ Alzheimer’s disease. ACS News

Release. [Online]

https://www.acs.org/content/acs/en/pressroom/newsreleases/2018/march/v

egetable-compound-could-have-a-key-role-in-beeting-alzheimers-disease.html

3 Pratchett, T. ‘A butt of my own jokes’: Terry Pratchett on the disease that finally

claimed him. The Guardian. 2015.

4 Matthews, K.A., Xu, W., Gaglioti, A.H., Holt, J.B., Mack, D., and McGuire, L.C. Racial

and ethnic estimates of Alzheimer’s disease and related dementias in the United

108

States (2015-2060) in adults aged ≥ 65 years. Alzheimer’s & Dementia. 2018.

15(1)17-24. Doi: 10.1016/j.jalz.2018.06.3063.

5 Roychaudhuri, R., Yang, M., Hoshi, M.M., and Teplow, D.B. Amyloid β-protein

assembly and Alzheimer disease. Journal of Biological Chemistry. 2008. 284(8):

4749-4753. Doi: 10.1074/jbc.R800036200.

6 Lyubchenko, Y.L. Amyloid misfolding, aggregation, and the early onset of protein

deposition diseases: insights from AFM experiments and computational

analyses. AIMS Molecular Science. 2018. 2(3)190-210. Doi:

10.3934/molsci.2015.3.190.

7 Jiang, D., Rauda, I., Han, S., Chen, S, and Zhou, F. Aggregation pathways of the

amyloid β(1-42) peptide depend on its colloidal stability and ordered β-sheet

stacking. Langmuir. 2012. 28(35)12711-12721. Doi: 10.1021/la30211436.

8 De Strooper, B., Saftig, P., Craessaerts, K., Vanderstichele, H., Guhde, G., Annaert, W.,

von Figura, K., and van Leuven, F. Deficiency of presenilin-1 inhibits the normal

cleavage of amyloid precursor protein. Nature. 1998. 397(6665)387-390. Doi:

10.1038/34910.

9 Wolfe, M.S., Xia, W., Ostaszewski, B.L., Diehl, T.S., Kimberly, W.T., and Selkoe, D.J.

Two transmembrane aspartates in presenilin-1 required for presenilin

endoproteolysis and gamma-secretase activity. Nature. 1999. 398(6727)513-517.

Doi: 10.1038/19077.

109

10 Iwata, N., Tsubuki, S., Takaki, Y., Shirotani, K., Lu, B., Gerard, N.P., Hama, E., Lee,

H.J., and Saido, T.C. Metabolic regulation of brain Aβ by neprelysin. Science.

2001. 292(5521)1550-1552. Doi: 10.1126/science.1059946.

11 Miller, B.C., Eckman, E.A., Sambamurti, K., Dobbs, N., Chow, K.M., Eckman, C.B.,

Hersh, L.B., and Thiele, D.L. Amyloid-β peptide levels in brain are inversely

correlated with insulysin activity levels in vivo. Proceedings of the National

Academy of Sciences. 2003. 100(10)6221-6226. Doi: 10.1073/pnas.1031520100.

12 Puzzo, D., Privitera, L., Leznik, E., Fa, M., Staniszewski, A., Palmeri, A., and Arancio,

O. Picomolar amyloid-β positively modulates synaptic plasticity and memory in

hippocampus. Journal of Neuroscience. 2008. 28(53). 14537-14545.

13 Ferrado, E., Arancibia, V., Norambuena, E., Olea-Azar, C., and Huidobro-Toro, J.P.

Stoichoimetry and conditional stability constants of Cu(II) or Zn(II) clioquinol

complexes; implications for Alzheimer’s and Huntington’s disease therapy.

Neurotoxicology. 2007. 28(3)445-449. Doi: 10.1016/j.neuro.2007.02.004.

14 Matsuo, K., Okamoto, H., Kawai, Y., Quan, Y.S., Kamiyama, F., Hirobe, S., Okada, N.,

and Nakagawa, S. Vaccine efficacy of transcutaneous immunization with

amyloid β using a dissolving microneedle array in a mouse model of Alzheimer’s

disease. Journal of Neuroimmunology. 2014. 266(1-2)1-11). Doi:

10.1016/j.jneuroim.2013.11.002

15 Vivekanandan, S., Brender, J.R., Lee, S.Y., and Ramamoorthy, A. A partially folded

structure of amyloid-beta (1-40) in an aqueous environment. Biochem, Biophys.

Res. Commun. 2011. (411)312-316. Doi: 10.1016/j.bbrc.2011.06.133 110

16 Damante, C.A., Osz, K., Nagy, Z., Pappalardo, G., Grasso, G., Impellizerri, G.,

Rizzarelli, E., and Sovago, I. The metal loading ability of β-amyloid N-terminus:

A combined potentiometric and spectroscopic study of copper(II) complexes

with β-amyloid (1-16), its short mutated peptide fragments, and it polyethylene

glycol (PED)-ylated analogue. Inorganic Chemistry. 2008. 47(20)9669-9683. Doi:

10.1021/ic8006052.

17 Dong, J., Atwood, C.S., Anderson, V.E., Siedlak, S.L., Smith, M.A., Perry, G., and

Carey, P.R. Metal binding and oxidation of amyloid-β within isolated senile

plaque cores: Raman microscopic evidence. Biochemistry. 2003. 42(10)2768-2773.

Doi: 10.1021/bio0272151.

18 Hou, L., and Zagorski, M.G. NMR reveals anomalous copper(II) binding to the

amyloid Aβ peptide of Alzheimer’s disease. Journal of the American Chemical

Society. 2006. 128(29)9260-9261. Doi: 10.1021/ja046032u.

19 Da Silva, G.F.Z., Tay, W.M., and Ming, L.-J. Catechol oxidase-like oxidation chemistry

of the 1-20 and 1-16 fragments of Alzheimer’s disease-related β-amyloid peptide:

Their structure-activity correlation and the fate of hydrogen peroxide. Journal of

Biological Chemistry. 2005. 280(17)16601-16609. Doi: 10.1047/jbc.M411533200

20 Martyn, C.N., Barker, D.J., Osmond, C., Harris, E.C., Edwardson, J.A., Lacey, R.F.

Geographical relation between Alzheimer’s disease and aluminum in drinking

water. Lancet. 1989. 1(8629)59-62. Doi: 10.1016/S0140-6736(89)91425-6

21 Lee, H-G., Castellani, R.J., Zhu, X., Perry, G., and Smith, M.A. Amyloid-β in

Alzheimer’s disease: the horse or the cart? Pathogenic or protective? Internation 111

Journal of Experimental Pathology. 2005. 86(3)133-138. Doi:10.111/j.0959-

9673.2005.00429.x.

22 Sampson, E.L., Jenagaratnam, L., and McShane, R. Metal protein attenuating

compounds for the treatment of Alzheimer’s dementia. Cochrane Systematic

Review. 2014. Doi: 10.1002/14651858.CD005380.pub5

23 Hu, B., Shen, Y., Adamcik, J., Fisher, P., Schneider, M., Loessner, M.J., and Mezzenga,

R. Polyphenol-binding amyloid fibrils self-assemble into reversible hydrogels

with antibacterial activity. ACS Nano. 2018. 12(4)3385-3396. Doi:

10.1021/acsnano.7b0869.

24 Ndam Ngoungoure, V.L., Schluesener, J., Moundipa, P.F., and Schluesener, H.

Natural polyphenols binding to amyloid: A broad class of compounds to treat

difference human amyloid diseases. Molecular Nutrition & Food Research. 2014.

59(1)8-20. Doi: 10.1002/mnfr.201400290.

25 Grimm, M.O.W., Thiel, A., Lauer, A.A., Winkler, J., Lehmann, J., Regner, L., Nelke, C.,

Janitschke, D., Benoist, C., Streidenberger, O., Stotzel, H., Endres, K., Herr, C.,

Beisswenger, C., Grimm, H.S., Bals, R., Lammert, F., and Hartmann, T. Vitamin D

and its analogies decrease amyloid-β (Aβ) formation and increase Aβ

degradation. International Journal of Molecular Science. 2017. 18(12)2764-2785. Doi:

10.3390/ijms18122764.

26 Gill, M., and Steglich, W. Pigments of fungi (Macromycetes). Progress in the

Chemistry of Organic Natural Products. Vol. 51. Springer, Vienna.

112 27 Schudel, G. The anthocyanins of Beta vulgaris L. and Raphans sativus L. Ph.D.

Dissertation. Edigenoss. Tech. 1918.

28 Clifford, T., Howatson, G., West, D.J., and Stevenson, E.J. The potential benefits of red

beetroot supplementation in health and disease. Nutrients. 2015. 7(4)2801-2822.

Doi: 10.3390/nu7042801.

29 Polturak, G., and Aharoni, A. “La Vie en Rose”: Biosynthesis, sources, applications of

betalain pigments. Molecular Plant. 2018. 1(8)7-22. Doi:

10.1016/j.molp.2017.10.008.

30 Singh, B., Bhat, T.K., and Singh, B. Potential therapeutic applications of some

antinutritional plant secondary metabolites. Journal of Agriculture and Food

Chemistry. 2003. 51(19)5579-5597. Doi: 10.1021/jf021150r

31 Sumaya-Martinez, M.T., Cruz-Jaime, S., Madrigal-Santillan, E., Garcia-Paredes, J.D.,

Carino-Cortes, R., Cruz-Cansino, N., Valadez-Vega, C., Martinez-Cardenas, L.,

and Alanis-Garcia, E.. Betalain, acid ascorbic, phenolic contents and antioxidant

properties of purple, red, yellow, and white cactus pears. International Journal of

Molecular Science. 12(10)6542-6468. Doi: 10.3390/ijms12106452.

32 Kanner, J., Harel, S., and Granit, R. Betalains—a new class of dietary cationized

antioxidants. Journal of Agriculture and Food Chemistry. 2001. 49(11)5178-5185. Doi:

10.1021/jf010456f

33 Esatbeyoglu, T., Wagner, A.E., Motafakkerazad, R., Nakajima, Y., Matsugo, S., and

Rimbach, G. Free radical scavenging and antioxidant activity of betanin: electron

113

spin resonance spectroscopy studies and studies in cultured cells. Food and

Chemical Toxicology. 2014. 73:119-126. Doi: 10.1016/j.fct.2014.08.007.

34 Zhang, Q., Pan, J., Wang, Y., Lubet, R., and You, M Beetrood red (betanin) inhibits

vinyl carbamate- and benzo(1)pyrene-induced lung tumorigenesis through

apoptosis. Molecular Carcinogenesis. 2012. 52(9)686-696. Doi: 10.1002/mc.21907

35 Clifford, T., Howatson, G., West, D.J., and Stevenson, E.J. The potential benefits of red

beetroot supplementation in health and disease. Nutrients. 2015. 7(4)2801-2822.

Doi: 10.3390/nu7042801.

36 Watts, A.R., Lennard, M.S., Mason, S.L., Tucker, G.T., and Woods, H.F. Beeturia and

the biological fate of the beetroot pigments. Pharmacogenetics. 1993. 3(6)302-311.

Doi: 10.1097/00008571-199312000-00004

37 Adams, J.P., and von Elbe, J.H. Betanine separation and quantification by

chromatography on gels. Journal of Food Science. 1977. 42(2)410-414.

38 Jackman, R.L. and Smith, J.L. Anthocyanins and betalains. In Natural food colorants;

Hendry, G.F and Houghton, J.D., Ed.; Blackie Academic & Professinal: London,

United Kingdom. Pp 244-309.

39 Sotos, J.G. Beeturia and iron absorption. The Lancet. 1999. 354(9183)1032. Doi:

10.1016/S0140-6736(05)76638-1

40 Trejo-Tapia, G., Jimenez-Aparicio, A., Rodriguez-Monroy, M., De Jesus-Sanchez, A,

and Gutierrez-Lopez, G. Influence of cobalt and other microelements on the

production of betalains and the growth of suspension cultures of Beta vulgaris.

114

Plant Cell, Tissue, and Organ Culture. 2001. 67:19-23. Doi:

10.1023/A:1011684619614

41 Sponsinska, A., Szot, D., and Wybraniec, S. The effect of citric acid and matrix of B.

vulgaris L .juice on thermal stability of betalains. PhD Interdisciplanary Journal.

2015. 1:193-200.

42 Attoe, E.L., and von Elbe, J.H. Oxygen involvement in betanin degradation: oxygen

uptake and influence of metal ions. Zeitschrift fur Lebensmittel-Untersuchung und

Forschung. 1984. 179(3)232-236. Doi: 10.1007.BF01041900.

43 Banu, K.S., Chattopadhyay, T., Banerjee, A., Bhattacharya, S., Suresh, E., Nethaji, M.,

Zangrando, E., and Das, D. Catehol oxidase activity of a series of new dinuclear

copper(II) complexes with 3,5-DTBC and TCC as substrates: syntheses, x-ray

crystal structures, and spectroscopic characterization of the adducts and kinetic

studies. Inorganic Chemistry. 2008. 47(16)7083-7093. Doi: 10.1021/ic701332w

44 Magherusan, A.M., Nelis, D.N., Twamley, B., and McDonald, A.R. Catehol oxidase

activity of comparable dimanganese dicopper complexes. Dalton Transactions.

2018. 48: 15555-15564. Doi: 10.1039/C8DT01378K

45 Bao, L., Avshalumov, M.V., Patel, J.C., Lee, C.R., Miller, E.W., Chang, C.J, and Rice,

M.E. Mitchondria are the source of hydrogen peroxide for dynamic brain-cell

signaling. The Journal of Neuroscience. 2009. 29(28)9002-9010. Doi:

10.1523/jneurosci.1706-09.2009.

46 Hangasky, J.A., Iavarone, A.T., and Marletta, M.A. Reactivity of O2 versys H2O2 with

polysaccharide monooxygenases. Proceeedings of the National Academy of Sciences 115

of the United States of America. 2018. 115(19)4915-4920. Doi:

10.1073/pnas.1801153115

47 da Silva, G.F.Z., and Ming, L.-J. Alzheimer’s disease related copper(II)-β-amyloid

peptide exhibits phenol monooxygenase and catechol oxidase activities.

Bioinorganic Chemistry. 2005. 44:5501-5504. Doi: 10.1002/anie.200501013.

48 Wasserman, B.P., Eiberger, L.L., and Guilfoy, M.P. Effect of horseradish peroxide and

phenolic compounds on horseradish peroxidase-catalyzed decolorization of

betalain pigments. Journal of Food and Science. 1984. 49(2)536-538. Doi:

10.1111/j.1365-2621.1984.tb12461.x.

49 Wybraniec, S., and Michalowski, T. New pathways of betanidin and betanin

enzymatic oxidation. Journal of Agriculture and Food Chemistry. 2011. 59(17)9612-

9622. Doi: 10.1021/jf2020107

50 Kumorkiewicz, A., Szmyr, N., Popenda, L., Pietrzkowski, Z., and Wybraniec, S.

Alternative mechanisms of betacyanin oxidation by complexation and radical

generation. Journal of Agricultural and Food Chemistry. 2018. 67(26)7455-7465. Doi:

10.1021/acs.jafc.9b01168.

51 Wybraniec, S., Starzak, K., Skopinsa, A., Szaleniec, M., Slupski, J., Mitka, K., Kowalski,

P., and Michalowski, T. Effect of metal cations on betanin stability in aqeous-

organic solutions. Food Science and Biotechnology. 2013. 22(2)353-363. Doi:

10.1007/s10068-013-008-7

116

52 Thordarson, P. Determining association constants from titration experiments in

supramolecular chemistry. Chemical Society Reviews. 2011. 40:1305-1323. Doi:

0.1039/C0CS00062K.

53 Hibbert, D.B., and Thordarson, P. The death of the Job plot, transparency, open

science and online tools, uncertainty estimation methods and other

developments in supramolecular chemistry data analysis. Chemical

Communications. 52:12792-12805. Doi: 10.1039/C6CC03888C

54 Bertini, I., Luchinat, C., and Parigi, G. Solution NMR of Paramagnetic Molecules:

Applications to Metallobiomolecules and Models. Current Methods in Inorganic

Chemistry, Vol 2. Elsevier Science, B.V. Amsterdam, The Netherlands. 2001.

55 Vila, A.J., Ramirez, B.E., Di Bilio, A.J., Mizoguchi, T.J., Richards, J.H., and Gray, H.B.

Paramagnetic NMR spectroscopy of cobalt(II) and copper(II) derivatives of

Pseudomonas aeruginosa His46Asp azurin. Inorganic Chemistry. 1997. 36(20)4567-

4570. Doi: 10.1021/ic9703282

56 da Silva, D.V.T., Baiao, D.d.S., de Oliveira Silva, F., Alves, G., Perrone, D., Del Aguila,

E.M., and Flosi Paschoalin, V.M. Betanin, a natural food additive: Stability,

bioavailability, antioxidant, and preservative ability assessments. Molecules. 2019.

24:458-471. Doi: 10.3390/molecules24030458.

57 Bertini, I., Luchinat, C., and Parigi, G. Solution NMR of Paramagnetic Molecules:

Applications to metallobiomolecules and models. Elsevier Science Ltd., New

York. 2001.

117

58 Goncalves, L.C.P., de Souza Trassi, M.A., Barbosa Lopez, N., Augusto Dorr, F., dos

Santos, M.T., Baader, W.J., Oliveira Jr., V.X., Bastos, E.L. A comparative study of

the purification of betanin. Food Chemistry. 2011. 1(1)231-238. Doi:

10.1016/j.foodchem.2011.08.067.

59 Ming, L.-J., and Que, Jr., L. NMR studies of the active site of isopenicillin N synthase,

a non-heme iron(II) enzyme. Biochemistry. 1991. 30(50)11635-11659. Doi:

10.1021/bi00114a007.

60 Epperson, J.D., Ming, L.-J., Baker, G.R., and Newkome, G.R. Paramagnetic cobalt(II)

as an NMR probe of dendrimer structure: mobility and cooperativity of dendritic

arms. Journal of the American Chemical Society. 2001. 123(35)8583-8592. Doi:

10.1021/ja15856y.

61 Schrodinger, L.L.C., Release 2019-3. New York, NY, 2019.

62 Ciriminna, R., Fidalgo, A., Danzi, C., Timpanaro, G., Ilharco, L.M., and Pagliaro, M.

Betanin: A bioeconomy insight into a valued betacyanin. Sustainable Chemistry

Engineering. 2018. 6:2860-2865. Doi: 10.1021/acssuschemeng.7b04163

63 McClements, D.J., Li, F., and Xiao, H. The nutraceutical bioavailability classification

scheme: Classifying nutraceuticals according to their oral bioavailability. The

Annual Review of Food Science and Technology. 2015. 6:299-327. Doi:

10.1146/annurev-food-032814-014043.

64 Hollman, P.C.H., Bijsman, M.N. van Gameren, Y., Cnossen, E.P., de Vries, J.H., and

Katan, M.B. The sugar moiety is a major determinant of the absorption of dietary

118

flavonoid glycoside in man. Free Radical Research. 1999. 6:569-573. Doi:

10.1080/1071576900301141.

65 Zhang, R. and McClements D.J. Enhancing nutraceutical bioavailability by controlling

the composition and structure of gastrointestinal contents: emulsion-based

delivery and excipient systems. Food Structure. 2016. 10:21-36. Doi:

10.1016/j.foostr.2016.07.006.

66 Azeredo, H.M.C. Betalains: properties, sources, applications, and stability – a review.

International Journal of Food Science and Technology. 2009. 44:2365-2376. Doi:

10.1111/j.1365-2621.2007.01668.x

67 Herback, K.M., Stintzing, F.C., and Carle, R. Impact of thermal treatment on color and

pigment pattern of red beet (Beta vulgaris L.) preparations. Journal of Food Science.

2004. 69(6)469-498. Doi: 10.1111/j.1365-2621.2004.tb10994.x.

119

CHAPTER 3

Inhibition of Copper(II)-Bacitracin Mediated Oxidation By Salicylic Acid and

Thiabendazole

We know more about the movement of celestial bodies than about the soil underfoot.

- Leonardo da Vinci

I. Background

1.1 Bacitracin

Bacitracin (Bc) is a topical antibiotic that was first discovered in 1945 by Johnson, et al of Columbia University1. Its discovery by this group was precipitated from a bacterial isolation from seven year old Margaret Treacy in Manhattan, New York.

Bacitracin was originally derived from Bacillus licheniformis and is now commercially developed from the Tracy I strain of Bacillus subtilus.2,3 It was named from a

120 portmanteau of its Bacillus bacterial source (bacitracin) and after a misspelling of

Treacy’s last name (bacitracin).

Figure 3.1 – relaxed-eye stereoscopic view of apo-bacitracin A1 minimized using

ChemDraw3D software4

Bacitracin has been found to be particularly useful as a narrow spectrum antibiotic for several strains of Staphylococcus and Streptococcus, and other Gram-positive bacteria. It is primarily used as a topical antibiotic, found in Neosporin® and Polysporin® among others.5 It is also commonly found in livestock feed for prevention of Gram-positive bacterial infections and subsequent growth promoter. Given its poor gastrointestinal absorption, it is generally safe for oral use; however, it is nephrotoxic in large quantities or if delivered in other non-oral methods.5 In Europe it maintains a “prudence” category for livestock infection therapy by the European Medicines Agency, though, it is banned a growth promoter.6,7 Despite the scale to which Bc is distributed, resistance is still relatively low, most likely owing to its structure and unique mechanism-of-action.6

The general amino acid sequence for Bc was proposed over the course of a few studies in the 1940s and 1950s. Separations techniques eventually resolved several

121 congeners of the cyclic antibiotic, subsequently labeled A1, A2, B1, B2, B3, D1, D2, D3, and

F.8-10 Revisiting these studies with more modern equipment revealed upwards of 50 potential congeners.11 Structural determination has also revealed the amino acid sequences and the variation among the congeners, summarized in Figures 3.2 and 3.3.12

The most biologically active forms have been found to be the A1 and B1 congeners that differ by a –CH3 on the N-terminal amino acid side chain, while F is the most biologically inactive form with a thiazole (oxidized thiazoline) at the second residue.

Figure 3.2 - Amino acid sequence for bacitracin with congener variations for R, X, and Y amino acids represented in Figure 3.3

122

Congener X5 Y8 R

A1 L-Ile L-Ile R1

A2 L-Ile L-Ile R2

B1 L-Ile L-Ile R3

B2 L-Val L-Ile R1

B3 L-Ile L-Val R1

D1 L-Val L-Ile R3

D2 L-Ile L-Val R3

D3 L-Val L-Val R1

F L-Ile L-Ile R4

Figure 3.3 – Side chain variants for R, X, and Y amino acids for the most common bacitracin congeners

123

Figure 3.4 – Line representation for bacitracin A1 congener

Determining the mechanism-of-action for bacitracin was initially difficult for researchers and was not fully understood for nearly two decades after its discovery.

Smith and Weinberg of Indiana University eventually determined that the zinc(II)- bound form of bacitracin inhibited bacterial growth ~10 times more than the apo- form of the antibiotic. They even hypothesized that some initial studies may have observed antibiotic activity by bacitracin scavenging metals from the glassware or media.13

Currently the most common commercial form of bacitracin is the zinc(II)-bound form.

The Bc-metal binding affinity has been established where

Cu2+>Ni2+>Co2+~Zn2+>Mn2+.14

124

The discovery of bacitracin’s metal-binding capability also eventually led to ascertaining the mechanism-of-action that could not be determined in early research using the apo- form. A 1971 study proposed inhibition of bacterial growth via the divalent metal in bacitracin binding to C55-isoprenyl pyrophosphate.15 The C55-isoprenyl pyrophosphate is responsible for the transport of disaccharide lipids across the cell membrane. Upon incorporation into the peptidoglycan layer, the disaccharide is released upon dephosphorylation of the lipid carrier into its monophosphate form by a membrane pyrophosphatase.16 By bacitracin binding to the pyrophosphate, dephosphorylation by pyrophosphatase is inhibited. Thus, the disaccharide turnover is inevitably prevented and the cell wall synthesis is compromised.17 Later studies confirmed the role of the divalent metal of metallo-Bc and pyrophosphate complexation via transmission electron microscopy and NMR.18,19

Among the earliest studies, His-10 and the Cys-2 thiazoline ring were proposed to be the main sources of metal-binding by Bc.20 This was later confirmed by studies in the 1990s using NMR to reveal the tail region of bacitracin bent towards the ring.21,22

While there was some early debate on the metal binding role of Asp-11 and the Ile-1 amino group, later studies concluded they were not involved in direct binding to metals.23, It was eventually concluded and confirmed the following residues were involved in metal-binding via EPR and NMR: the Cys-2 thiazoline nitrogen, the Glu-4

γ-carboxylate and the Nε of His-10.24,25 In conjunction with diamagnetic signal assignments, paramagnetically shifted signals in the presence of Co2+ were correlated to

125 their diamagnetic counterparts using EXSY techniques, resolving a more complete picture of the metal-bound structure.25

Figure 3.5 – Calculated structure of cobalt-bound bacitracin by NMR, relaxed-eye view, modeled based on NMR T1 data. Image used with permission by Elsevier.12

Studies in the last few decades have also focused on the versatile activity that metallo-bacitracin can perform. Part of the reason metals tend to be avoided in drug development processes is the lack of specificity that metals can have towards chemicals other than the drug target. Conversely, its versatility by binding to several targets might also hold the key towards bacitracin’s longevity. A few metal-bound forms of bacitracin have shown activity beyond cell wall synthesis inhibition. Cu2+-bacitracin has shown very strong reactivity towards bacterial DNA, performing both single and double- stranded DNA cleavage along the phosphate backbone.19 In fact, the Zn2+-bound form

126 showed some of the least amount of DNA cleavage compared to more redox active metals like Cu2+.

Additionally, the metallo-Bc has also exhibited potential redox bioactivity in performing catechol oxidase-like reactions. A study by Tay, et al. found the turnover of catechol to o-quinone of 7.0 × 10—3 s—1 and 0.38 s—1 in the presence of aerobic O2 and 32 mM H2O2, respectively.19 Medicinally some contact could occur with catechols and polyphenols as active ingredients in topical skin moisturizers.26 Interaction between the two could render the antibiotic and antioxidant activity for both, respectively, moot from outside influences. From the perspective of some of the inhibitors explored in this dissertation, catechol oxidase-like activity is of particular importance in nature and soils. Catechol-containing compounds and polyphenols are often produced by many plants and are often many of the chemical components in vegetables and fruits. As the natural source of bacitracin is derived from soil dwelling bacteria, the incidences of bacitracin and catechol moieties would occur most naturally in the presence of or by inhabiting plants. In fact, the presence of Bacillus subtilis can be used as a barometer for the health of the soil for plant growth.27 B. subtilis has been shown to promote root and leaf biomass, solubilize phosphates, and stimulate growth hormone production in a variety of plants!28,29 Additionally, measuring the redox activity against catechol moieties allows for facile, replicable, and comparable measurements of its turnover.

The premise of this study was to examine the effects of two environmentally available compounds that could interact with CuBc and potentially inhibit its redox

127 activity. Two products were chosen for their structural simplicity, longevity of use, and for their potential to bind metals. The two molecules studied to potentially interact with

CuBc were salicylic acid, a common secondary metabolite produced by many plants, particularly willow trees, and thiabendzole, a synthetic small molecule anti-helminthic drug frequently used in agriculture. By characterizing the redox activity of bacitracin in the presence of these metal-binding compounds, will give a sense of the chemistry occurring in the soil and during topical medicinal uses.

1.2 Salicylic Acid

Salicylic acid (SA), or 2-hydrobenzoic acid, is among the oldest and well-known medicines in the world. Produced by and named after the Salix alba, white willow tree, salicylic acid has been used medicinally from its tree bark source as far back as 1500

B.C. in Egypt. Reports have shown that willow bark was espoused even by Hippocrates for its pain killer properties!30,31 Its popularity is predominantly as a precursor to the painkiller aspirin (acetylsalicylic acid). The acetylated product was trademarked by

Bayer® in 1899, ~50 years after its initial synthesis by Frederic Gerhardt who developed a method to simply acetylate the hydroxyl group of the phenol ring.

128

Figure 3.6 – Aspirin bottle ca. 1899. Creative Commons 3.0 Image License

The sources of salicylic acid are actually varied among many different plants, including the rice, tobacco (Nicotiana tabacum), and even potatoes (Solanum tubersum) in small quantities.32,33 The ubiquity of the natural production has shown its importance for the plants where it has a variety of roles including antibiotic, a growth hormone, and an oxidative stress-reliever.34 A more recent study found that SA might play a role in antibiotic activity against E. coli. by decreasing extracellular biofilms by ~3-fold.35

Despite more than 3000 years of medicinal use and many millennia of natural production, the variety of roles SA can perform are still being elucidated.

Figure 3.7 – line representation of salicylic acid

129 The medicinal properties of salicylic acid are generally as an anti-inflammatory and pain reliever; it has also shown some efficacy as an antibiotic as well.36 The mechanism of anti-inflammation is derived from its interaction with cyclooxygenase-2

(COX-2). The COX-1 and COX-2 enzymes are responsible for producing inflammation- inducing prostaglandin hormones.37 Unlike the serine acetylation performed by aspirin on the COX-1 enzyme, SA shows no interaction with COX-1 enzymes in vitro.38 One study proposes that SA may inhibit expression of the COX-2 among other enzymes involved in inflammation responses. A study by Mitchell, et al found that salicylic acid can bind to the transcription factor responsible for several genes involved in prompting inflammation. As the genes, proteins, and protaglindin hormones are down-regulated, continued inflammation is not triggered.39 COX-2 has also been implicated in some cancers where it can play a role in triggering a number of tumorigenic responses, such as angiogenesis. Thus, the down-regulation of these enzymes for diseases like colon cancer by salicylic acid reduce tumor growth.40

The interaction of salicylic acid with metals has long been established. A 1958 study was initially able to show the metal binding capability of SA, which, however, did not quantify the binding capacity of the complexes in aqueous environments.

Qualitatively, the SA-metal complexes followed the Irving-Williams series of metal- ligand stability, where Mn2+ < Fe2+ < Co2+ < Ni2+ < Cu2+ > Zn2+.41 Potentiometric studies

130 eventually confirmed the earlier study and elucidating stability and thermodynamic constants associated with salicylic acid chelates towards several divalent metals.42

In conjunction with SA’s general antioxidant activity, it has also shown efficacy towards antioxidation mediated by redox active metals.43,44 The studies in the first part of this chapter explores the relationship between CuBc and SA in terms of redox functionality and structure.

1.3 Thiabendazole

Unlike salicylic acid and bacitracin, thiabendazole (Tbz) is a synthetic compound.

Tbz, or 2-(4’-thiazolyl)benzimidazole, was originally developed in the 1960s and was patented by Merck® in 1969.45 It is mainly used as an antihelminthic and antibiotic on food products. Its uses today are still much the same and is sold commercially under a variety of names, such as Arbotect, and Mintezol.

Figure 3.8 – line representation of thiabendazole

131 Thiabendazole is generally regarded to be an inhibitor towards fumarate reductase, an enzyme involved in electron transport in the eventual production of

ATP.46 Other studies have shown the mechanism-of-action might vary or play multiple roles in its antibiotic activity, including the inhibition of non-polymeric tubulin synthesis, ubiquinone, or even activation of myeloperoxidase to cause toxicity in mice.47-49

U.S. Environmental Protection Agency regulates Tbz as a crop antibiotic, where they allow it to be applied to most food crops. Additionally, it has found use as an antibiotic for paints and carpets to prevent molding. Moreover, it is also used in the treatment post-harvest as a preservative, especially apples, bananas, and citrus.50

Current EPA crop tolerances vary by food. EPA ranging from as high as 30-40 ppm or more for foods with waxy peels, while most allow for ~1 ppb. Environmental exposure naturally varies since its presence depends on those that apply it, local weather, and crop drainage. Though it was banned in the European Union as a crop preservative for a number of years, legislation in 2015 reinstated it as an accepted for commercial agriculture use.51

As previously discussed with betanin, synthetic chemicals in food have come under more scrutiny from the public regarding public health. The oral LD50 of Tbz for rats is ~3.5 g/kg, which is a relatively high dose for most drugs to induce death.

Conversely, fish seem to be far more sensitive towards the antihelminthic.50 More concern and research has currently examined low dose exposure for potential

132 cumulative toxic effects. While it is classified as potentially carcinogenic, early research in the 1970s up to more recent studies have shown Tbz to be antitumorogenic when application is more controlled in terms of application locale and concentration.52,53

Despite higher than normal concentration of Cu2+ under tumorigenic conditions, Tbz showed little antitumorogenic activity. Conversely, the study showed Tbz exhibited more chemotherapeutic activity when coordinated to copper prior to incubation.54

The capacity for Tbz to bind to metals has been known for nearly as long as the chemical has been in use as an antibiotic. Tbz:metal complexes of 1:1 and 2:1 were originally observed for Cu2+, Zn2+, Co2+, and Ni2+, and was eventually expanded to include even heavy metals like mercury.55,56

While many research aspects on Tbz have examined its antibiotic properties, antitumoregenesis, metal-binding capabilities, or bioremediation, this study herein sought to examine the chemistry that might be occurring in the soil. As some of the bacteria that produce bacitracin are essential for soil health, the interactions between

Tbz and metallobacitracin could yield some indication about how Tbz can affect the crop environment other than its intent as a preservative. CuBc is chosen as a prototype in this study owing to easy monitoring of its metal-mediated redox activity that can faithfully reflect its interactions with the phytochemical salicylic acid and synthetic Tbz.

The study is expected to provide further insights into the multiple equilibria microbrial metallomolecule (metallo-Bc), a metal-binding phytochemical (salicylic acid), and a

133 metal-binding synthetic compound Tbz, which would also “dig up more dirt” on the microecological balance within the soil at the molecular levels.

II. Results and Discussions

2. Salicylic acid as an inhibitor against Cu-bacitracin

2.1.1 Aerobic Oxidation

Previously, Cu2+-Bc (CuBc) was studies for its ability to bind pyrophosphate, cleave the phosphate backbone of DNA, and oxidative activity towards catechols.Error! Bookmark not defined. The metal center in all these cases played an essential role for the various activities. The previously observed catechol oxidase-like activity was a means to measure oxidative activity. In this study the catechol-containing substrate can also emphasize the role catechol-containing compounds CuBc might encounter in medicinal or agricultural roles. The mechanism of catechol oxidation by CuBc was determined in the aforementioned study by Tay, et al.19 It was proposed to follow a mechanism based on mono-copper center.19 To oxidize catechol into quinone, a two electron transfer process is required. The oxidation of the substrates was investigated under aerobic O2 and H2O2 conditions. The generation of peroxide under reductive conditions have been

134 extensively presented in literature which is also an essential mechanism to explore as it has shown strong reactivity in creating oxidative stress.

Figure 3.9 – Proposed catechol oxidation mechanism for CuBc. Adapted with permission from “1H NMR, Mechanism, and Mononuclear Oxidative Activity of the

Antibiotic Metallopeptide Bacitracin: The Role of d-Glu-4, Interaction with

American Chemical Society. A proposed mechanism for aerobic and hydrogen peroxide assisted oxidation of o-catechol to form o-quinone”.19

In the following studies DTBC was used as a substrate towards oxidation, similar to the previous studies. Herein DTBC acted as a model substrate for the investigation of 135 the oxidation of neurotransmitters discussed in Chapter 2 since DTBC can be representative of the numerous catechols found in nature. In the soil, catechols are often distributed by plants and bacteria. Testing against a model catechol like DTBC allows for comparison of oxidative activities occurring in the soil. The following are but a few catechols produced by most plants:

Catechin Gallate 4-ethylcatechol Quercetin

Figure 3.10 – common polyphenol and catechol-containing secondary metabolites produced by plants

Catechol-containing compounds are ubiquitous among plant-life. They are considered secondary metabolites given that they do not play a direct role in the metabolic activities required for the plant to live. They have shown the ability to reduced ROS activity, thus preventing genetic or metabolic damage, as well as some antibiotic activity.57,58 Potential antibiotic interactions are not limited to plant life either.

Several flavonoid catechols can also be found in skin-care products for their protection

UV-B light radiation and potential carcinogens.59,60

136 Oxidation of DTBC by CuBc under aerobic conditions revealed a similar substrate kinetic pattern to studies published by Tay, et al, showing a saturation in activity of the substrate catechol in Figure 3.10.19 This type of reaction kinetics can be described in the aforementioned Michaelis-Menten pre-equilibrium kinetics using the reaction pathway shown in detail in Figure 3.9, and simplified in Figure 3.12.

DTBC Oxidation

10 Rate (nM/sec)

0 0 5 10 15 20 25 30 [DTBC] (mM)

Figure 3.11 – Oxidation of various concentrations of DTBC by 10 μM CuBc (1:1) under aerobic conditions in 1:1 MeOH:HEPES (pH 7.0, 100 mM) at 25.0 °C

137 Figure 3.12 – simplified aerobic catechol oxidase-like pathway for CuBc oxidation of catechol to quinone by CuBc

A concerted effort was made to ensure comparability with previous studies, though, the SA-CuBc complex and DTBC required a mixed solvent to maintain solubility. The kinetic parameters, determined by fitting the data in Figure 3.11 to the

Michaelis-Menten, Equation 2.1, revealed a Km of 1.79 ± 0.3 mM and kcat of (1.74 ± 0.09)

× 10—3 sec—1, comparable to previous studies. Slight variations could occur due to general solvent environment, which could strengthen or weaken the binding of the ES complex (CuBc-DTBC), or change the diffusion of the substrate or products.

Initial titrations of salicylate took relatively large quantities of SA to perform significant inhibition towards constant amounts of CuBc and DTBC, shown in Figure

3.13. At 10 μM CuBc this shows relatively weak inhibition given oxidation is still occurring after 64 equivalents of SA were added.

138 16

Rate (nM/sec) 6

0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 [SA] (mM)

Figure 3.13 – titration of SA to 10 μM CuBc in 1:1 MeOH:HEPES (pH 7.0, 100 mM), 2 mM DTBC

The inhibition pattern for salicylic acid was determined by titrating DTBC in the presence of 100 μM CuBc and a fixed concentration of SA. Subsequent experiments varied the amount of SA to determine what manner SA inhibits DTBC turnover. CuBc was increased in these experiments given how slow some of the inhibited reactions could occur at high concentrations of SA and low concentrations of DTBC. Using 100

139 μM CuBc allowed for facile determination of absorbance change without increasing measurement error by approaching instrumentation limits.

0.08

10.0 mM SA 5.0 mM SA 0.06 2.5 mM SA 0 mM SA

sec) 0.04 -1

0.02 1/Rate (nM

0.00 0.0 0.2 0.4 0.6 0.8 1.0 1.2

1/[DTBC] (mM-1) -0.02

Figure 3.14 – Lineweaver-Burk plot for varying concentrations of salicylate as an inhibitor against oxidation of various concentrations of DTBC by 100 μM CuBc in 100 mM pH 7.0 1:1 HEPES:MeOH

140 Table 3.1 – MM parameters determined from fitting Figure 3.14 as a function of [DTBC] by 100 μM CuBc with different amounts of SA

Vmax kcat Catalytic Efficiency SA (mM) (nM/sec) Km (mM) (sec—1) (M—1 sec—1) 0.0 475.7 ± 28.5 9.3 ± 1.2 (4.76 ± 0.29) × 10—4 (5.11 ± 0.73) × 10—2 2.5 447.3 ± 14.2 12.9 ± 0.8 (4.47 ± 0.14) × 10—4 (3.48 ± 0.24) × 10—2 5.0 405.7 ± 27.1 15.6 ± 1.9 (4.06 ± 0.27) × 10—4 (2.60 ± 0.36) × 10—2 10.0 301.1 ± 15.2 17.1 ± 1.6 (3.01 ± 0.15) × 10—4 (1.76 ± 0.19) × 10—2

The inhibition pattern shown in Figure 3.14 shows apparent mixed inhibition with competitive and uncompetitive inhibition. The parameters for fitting the

Lineweaver-Burk plot were determined using Equations 2.1 and 2.2 (Chapter 2). The data shown in Table 3.1 supports this fit as apparent Km increases and apparent turnover (kcat) decreases. An increase in Km would indicate formation of the binary

CuBc-SA (EI) complex, yielding the term Kic, the dissociation constant for the EI complex. Alternatively, the uncompetitive pathway would indicate competition towards the ES complex, forming the ESI complex and preventing turnover, yielding the Kiu term as the dissociation for ESI complex.

141 Figure 3.15 – competitive and uncompetitive SA inhibition pathways of DTBC oxidation by CuBc

The competitive inhibition constant, Kic, was determined using Equation 2.4, where the inhibitor concentration was replaced by the concentration of salicylic acid.

The ratio of apparent Vmax and Km as a function of SA was plotted in Figure 3.16. The competitive inhibition constant for SA-CuBc was found to be 5.24 ± 0.03 mM. As previously observed in Figure 3.14, the inhibition of SA against CuBc is a relatively large value, 50 times higher than the concentration of CuBc. Under these conditions it is likely that SA is binding via its deprotonated carboxylate, which is comparable to catechol Michaelis constant, Km, of 9.3 mM. Given its high dissociation constant, SA would not likely be considered as a drug towards preventing oxidation of catechols moieties. SA could provide an alternative binding pathway against CuBc in soils especially in highly concentrated regions adjacent to plant materials. This could also be 142 indicative of the mutually beneficial relationship B. subtilis has with plants, where inhibition of CuBc or other metallo-Bc compounds would be counterproductive.

Figure 3.16 - Fitting of apparent Vmax and Km as a function of SA concentration to determine Kic of SA, fitted to Equation 2.4.

The uncompetitive pathway was explored by plotting Vmax,app as a function of the inhibitor concentration, Figure 3.17 This figure describes the uncompetitive binding of the ternary (ESI) complex between SA-CuBc-DTBC. The Kiu for this pathway was determined by fitting the data to Equation 2.5, where SA was substituted as the

143 inhibitor of interest. The poor fitting for this data likely indicates that the uncompetitive pathway is still within the linear region of fitting at the concentrations used for SA, especially given the Kiu of 21 ± 3 mM. This value is more than double the highest SA concentrations, indicating the SA concentrations used to investigate are relatively too low for the fit of this pathway. This value also indicates a relatively unstable ternary complex in favor of the CuBc-DTBC complex. Based on the more favorable equilibrium for the competitive pathway. This data could imply the space required for ternary binding is simply not favorable between the Bc ligands and DTBC. Solvent polarity might also play a role where the binding pocket might alter active site exposure based on hydrophobic/hydrophilic interactions.

144 Figure 3.17 – Fitting of apparent Vmax as a function of SA as fitted by Equation 2.5

(R2=0.92)

2.1.2 Oxidation in the Presence of H2O2

As part of the pathway shown in Figure 3.9 hydrogen peroxide can turnover the copper center in an alternative pathway. In aerobic biological systems and in the environment, peroxide can be present due to reaction byproducts and can be found in quantities of up to ~40 μM in rainwater.61,62 Addition of H2O2 to CuBc exhibited strong reactivity towards DTBC. At 10 μM CuBc and constant 1 mM DTBC, H2O2 titration

145 revealed showed a 50 ± 6 mM Km,app, 0.48 ± 0.02 sec—1 H2O2 turnover, and an efficiency of 9.71 ± 0.13 M—1 sec—1.

4000

Measured Rate Normalized Rate 3000

2000 Rate (nM/sec)

1000

0 0 20 40 60 80 100 120

[H2O2] (mM)

Figure 3.18 – titration of hydrogen peroxide to 10 μM CuBc and constant 1 mM DTBC, normalized for fitting to determine Km,app

CuBc inhibition of hydrogen peroxide turnover by SA was initially evaluated as a function of salicylic acid, Figure 3.18. Interestingly, much smaller concentrations of SA relative to CuBc were required to show inhibition of peroxide turnover.

146 100

40 Rate (nM/sec)

0 0 20 40 60 80 100 120 140

[SA] ( M)

Figure 3.19 - titration of SA to 30 μM CuBc and 30 mM H2O2 in 100 mM pH 7.0 1:1

HEPES:MeOH

Michaelis-Menten parameters, shown in Table 3.2, were derived by fitting the data from Figure 3.20. The overall data converging near the x-axis indicates more noncompetitive inhibition character. Moreover, fitting Vmax,app/Km,app data as a function of SA would not converge at the concentrations of SA being observed. Km,app determination remained relatively constant through each titration of SA, each within

147 the margins of error. This might indicate the competitive pathway would require a much higher concentrations of SA as observed under aerobic conditions.

5e+7 0 M SA 0.5 M SA 4e+7 2.0 M SA 4.0 M SA 3e+7 16.0 M SA sec) -1 2e+7

1/Rate(nM 1e+7

0 0.00.20.40.60.81.0

-1e+7 -1 1/[H2O2] (mM )

Figure 3.20 – LB plot of inhibition of DTBC oxidation rate by SA as a function of H2O2 catalyzed by 10 μM CuBc with constant 2 mM DTBC in 100 mM HEPES:MeOH (1:1) at pH 7.0 25.0 °C

Table 3.2 - MM parameters for SA inhibition against H2O2 activation for the oxidation of 2 mM DTBC was titrated in the presence of 10 μM CuBc

Vmax kcat Catalytic Efficiency SA (μM) (nM/sec) Km (mM) (sec—1) (M—1 sec—1) 0.0 489 ± 43 23 ± 6 (4.9 ± 0.2) × 10—2 2.1 ± 0.6 0.5 424 ± 102 26 ± 16 (4.2 ± 1.0) × 10—2 1.6 ± 1.0 148 Table 3.2 cont. 2.0 408 ± 85 27 ± 14 (4.1 ± 0.8) × 10—2 1.5 ± 0.9 4.0 359 ± 40 29 ± 8 (3.6 ± 0.4) × 10—2 1.3 ± 0.4 16.0 270 ± 35 21 ± 8 (2.7 ± 0.3) × 10—2 1.3 ± 0.5

The uncompetitive dissociation constant for SA in the ternary, ESI, complex was determined by fitting Vmax,app as a function of SA, Figure 3.21. Kiu for SA in the ternary complex of SA-CuBc-O22— was determined to be 15 ± 3 μM. In this pathway SA exhibited much stronger inhibition against H2O2 binding to the CuBc-DTBC complex than its inhibition against DTBC binding to CuBc (Kiu = 5.24 mM).

Figure 3.21 – Vmax,app data fitted to Equation 2.5 as a function of inhibitor SA concentration in determination of uncompetitive dissociation constant, Kiu

149 2.1.3 Paramagnetic NMR

Co2+ has been extensively used as a paramagnetic probe for the investigation of metal binding site in biomolecules by the use of NMR techniques given its similar binding capabilities as Cu2+. To encourage binding in initial studies, the base trimethylamine (TEA) was added at 1 equivalent to sodium salicylate to produce the phenolate binding site to push equilibrium towards bidentate binding of SA. Studies of the apo-SA were performed using a standard one 90° pulse followed by acquisition and recycle delay using an 8000 Hz spectral window at 600 MHz. Studies including Co2+ used a presaturation pulse to reduce HDO intensity, allowing for more gain to be applied to further resolve paramagnetic signals. Recycle times were much shorter, 250 ms, taking advantage of the short relaxation delays inherent to isotropically shifted signals with a spectral window of 120 kHz.

Simple structure elucidation could be performed on the deprotonated salicylate based on splitting and chemical shifts, which is in agreeance with previous studies of

SA signal assignments, Figure 3.22.

150 D A C D B C A B

Figure 3.22– 5 mM apo-salicylate in D2O (4.8 ppm) using a presaturation pulse sequence at the HDO frequency

Upon addition of 0.8 equivalents of Co2+, the paramagnetically shifted signals at

57.2, 50.0, 31.7, and 11.1 ppm all indicative of salicylate bound to the Co2+ paramagnetic metal, shown in Figure 3.23. Moreover the signals indicate binding through a bidentate

151 mode, including the phenolate moiety for easy delocalization of the unpaired electrons on the Co2+.

Figure 3.23 – 0.8 eq Co2+ salicylate in 10% D2O, with 1 stoichiometric equivalent of TEA to include phenolate binding

Further elucidation using EXSY pulse sequence allowed for facile correlation of paramagnetic signals with their diamagnetic counterparts, Figure 3.24. The EXSY measures chemical exchange between isotropic signals with their diamagnetic counterpart as the ligand undergoes association and dissociation with the metal center.

Using a modified NOESY pulse sequence, EXSY allows for correlation between two signals are undergoing exchange between two states; in this case it is the bound and unbound forms of the ligand.

The EXSY clearly reveals correlations for all four aromatic signals with the salicylate ring. The paramagnetic signals along the F2 dimension clearly show correlation to the diamagnetic counterparts along the F1 dimension. Expectedly, the most downfield shifted signals, labeled D and C, are those that are closest to the

152 deprotonated phenolate moiety, which would be closer to the paramagnetic Co2+ center without the added carbon on the carboxylate binding site.

D C A B

1H Diamagnetic Para- label (ppm) magnetic(ppm) A 7.68 (d) 31.7 B 6.79 (t) 11.1 C 7.30 (t) 50.0 D 6.81 (d) 57.2

Figure 3.24 – EXSY of salicylate with the addition of 0.2 equivalents of Co2+ with salicylate in 10% D2O and signal assignments

Unfortunately, the ternary complex is not observable in D2O for the same reasons a pure aqueous solution could not be used for kinetic studies: insolubility of the ternary complex at a high concentration in the millimolar range for NMR studies. Binding of SA to CuBc apparently makes the ternary complex as a whole much less polar and driving insolubility of the ternary complex in aqueous solutions.

Challenges were still encountered with other solvents as well with solubility and solvent interaction with the metal center was considered on choosing them. Solvents like d3-MeOH proved even less soluble for concentrations of Bc >4 mM. d6-DMSO was 153 ultimately used for subsequent investigations towards the ternary complex.

Unfortunately, a previous study had already determined that chemical exchange is immeasurable between Co2+ and Bc in DMSO. Thus, EXSY NMR signal assignment could not be properly performed using DMSO. Fortunately, this study was able to show ternary molecular interactions through the secondary effects of signal disappearance upon titration of SA.

In establishing a baseline for comparison, a control, the spectra in Figures 3.25 &

3.26 are comparable to a previous diamagnetic and paramagnetic studies.

Figure 3.25 – 1H NMR of 4 mM apo-bacitracin in d6-DMSO

Figure 3.26 – CoBc in d6-DMSO with 0.8 eq Co2+, using a presaturation pulse sequence, diagmagnetic signal omitted for presentation of paramagnetic signals

154 Upon addition of up to 4 equivalents of SA to Bc, most of the Bc signals appeared unchanged other than small changes in chemical shift Figure 3.27. Signals at 16.1 and

13.2 ppm for the two Lys-6δ protons maintain intensity as SA increases, indicating the stability of the nearby binding of the Glu-4 carboxylate. The thiazoline binding center of

Cys-2 appears to continue binding as SA was titrated. Based on previous signal assignments the Cys-2β proton at 33.5 ppm is retained as are the adjacent Ile-1β and

Ile-1α protons at 46.5 and ~52.2 ppm, respectively. Their continued presence as SA was titrated indicate the continued binding of the Cys-2 thiazoline ring. This likely indicates the ‘tail’ portion of bacitracin binding while the ternary with SA complex is formed.

The disappearance of some signals imply that SA disrupts bonding of the His-10 residue. As signals at 13 and 36.5 ppm appeared unchanged, in comparison with solvent signal, throughout the titration and were used as signal intensity references. As

Co2+ is bound via His-10ε nitrogen, the strong signal at 36.5 ppm was comparably assigned to His-10δ proton from the aforementioned studies. Figure 3.27 shows an apparent decrease in signal intensity for these signals potentially indicating interruption of the Co2+-His-10 interaction after 1 equivalent of salicylate was added. The signal in intensity is more apparent after the addition of the fourth equivalent. This is in agreeance with the disappearance of the upfield shifted signal at —3.5 ppm which are also assigned to the His-10β protons based on a TOCSY assignment of the region.25

155 Figure 3.27 – Paramagnetic 1H NMR spectra of Co2+-bacitracin as SA is titrated by the stoichiometric equivalent in d6-DMSO, where 1 equivalent is 10.2 mM and SA is 0.0 eq,

0.3 eq, 0.5 eq, 1.0 eq, and 4.0 eq (red) going down

156 Figure 3.28 – diamagnetic region upon addition of SA to 10.2 mM CoBc in d6-DMSO.

Apo-Bc (top), and CoBc with the addition of 0.0 eq (green), 0.3 eq, 0.5 eq, 1.0 eq, and 4.0 eq (pink) SA going down

In support of salicylate interaction with the Co2+ center are two signals that show an increase in intensity at 10.8 and 11.2 ppm, shown in Figure 3.27. A new signal at 10.8 ppm could indicate SA binding or formation of the ternary complex. Control studies of

SA in DMSO showed an approximate signal at 11.1 ppm for carboxylate γ proton on the ring of SA. The increase in intensity of the signal at 16 ppm, shown in more detail in

Figure 3.29, might also indicate SA interaction. However, the assignment remains ambiguous due to the overlapping Bc signal. Assuming the increase in intensity in sharpening of the signal, this signal might be indicative of the β proton adjacent to the carboxylate of SA. Confident assignment would require EXSY in a medium that provides both exchange and solubility.

157 Figure 3.29 – Focused 1H spectrum of 0.8 eq CoBc 0 SA (top) and 4 eq SA (bottom) in d6-

DMSO

Based on the addition of SA in DMSO to CoBc, some differences were observed in the isotropically shifted signals. The sustaining paramagnetic chemical shifts of CoBc with up to four equivalents of SA added indicate no chelation at these concentrations of

SA. Based on a comparison to prior CoBc spectra, it is likely that SA might be disrupting the His-10 interaction with the metal center when SA is greater than one equivalent. Assuming the binding pocket from the ring portion of Bc is interrupted by

SA, this would still leave space on the metal center for reactivity. Moreover, ligand replacement rather than blocking of the binding site could explain why the Kic for SA is so high under aerobic oxidative conditions.

158 2.2 Thiabendazole Inhibition of CuBc Redox Activity

2.2.1 Kinetics

The history of thiabendazole and metal binding was established early in its development.55 Tbz has shown up to 3-to-1 ligand-to-metal binding complexes revealed in crystallographic studies.63 Cu2+ binding did not show any significant spectrophotometric effects upon binding. It did, however, exhibit reduction in oxidative activity of Cu2+ toward catechol oxidation as shown in Figure 3.30, suggesting Tbz can block catechol from binding to and being oxidized by the metal center. Residual oxidation after addition of one equivalent of Tbz to Cu2+ reveals open binding sites for

DTBC turnover. Subsequent titration of Tbz would likely show an oxidation rate approaching zero, though, DTBC continues oxidizing at the concentrations of Tbz observed in this experiment.

159 120

100

Rate (nM/sec) 40

0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 [Tbz] (mM)

Figure 3.30 – Tbz inhibition of 2 mM DTBC oxidation by 10 μM free Cu2+ in HEPES buffer (pH 7.0) with 10% DMSO for catechol and Tbz solubility

Addition of Tbz to CuBc reveals a slightly different effect, shown in Figure 3.31.

Under aerobic conditions the rate of DTBC oxidation reduces as a Tbz increases. Larger concentrations of Tbz were required to affect oxidation drastically, still showing oxidation activity of CuBc in the presence of 100 equivalents of Tbz.

160 250

200

150

100 Rate (nM/sec) Rate

0 0 2000 4000 6000 8000 10000 12000

[Tbz] ( M)

Figure 3.31– oxidation rate of 2 mM DTBC upon addition of Tbz to 10 μM CuBc in 1:1

DMSO:HEPES 100 mM buffer at pH 7.0

In determining the type of inhibition towards DTBC turnover and the appropriate dissociation constant(s) for Tbz towards CuBc, oxidation of DTBC measured under varying amounts of Tbz, shown in Figure 3.32. Tbz showed more mixed inhibition in the linearized plot, where convergence occurs at a point other than the y- or x-axes. The data in Table 3.3 supports the conclusion of mixed-type inhibition revealing a slight increase in Km,app and decreased turnover as a function of increasing

161 amounts of Tbz. Based on the larger relative increase in kcat,app the Tbz inhibition shows more uncompetitive character.

0.08 4.0 mM Tbz 2.0 mM Tbz 0.06 1.0 mM Tbz 0.1 mM Tbz

sec) 0.04 -1

0.02 1/Rate (nM

0.00 0.0 0.5 1.0 1.5 2.0 2.5

1/[DTBC] (mM-1) -0.02

Figure 3.32 – linearized LB plot for DTBC titrated to 100 μM CuBc in 100 mM pH 7.0 1:1

HEPES:DMSO where Tbz varied up to 4.0 mM

Table 3.3 – MM parameters determined by fitting the non-linearized data from Figure

3.32, where kcat and catalytic efficiency were determined using 100 μM CuBc

Vmax kcat Catalytic Efficiency Tbz (mM) (nM/sec) Km (mM) (sec—1) (M—1 sec—1) 0.0 717.7 ± 19.0 4.46 ± 0.27 (7.2 ± 0.2) × 10—3 1.61 ± 0.1 0.1 552.4 ± 59.9 4.96 ± 1.17 (5.5 ± 0.6) × 10—3 1.11 ± 0.3 1.0 404.3 ± 22.5 4.77 ± 0.59 (4.0 ± 0.2) × 10—3 0.85 ± 0.1 4.0 269.7 ± 18.0 5.71 ± 0.79 (2.7 ± 0.2) × 10—3 0.47 ± 0.1

162 Based on a likely mixed inhibition patter surmised from the Lineweaver-Burk plot in Figure 3.32, dissociation constants Kic and Kiu were evaluated. Vmax,app/Km,app from the MM parameters, Table 3.3, were evaluated as a function of SA concentration,

Equation 2.1. Kic was determined by fitting data from Table 3.3 to Equation 2.4, and was found to be 1.1 ± 0.4 mM. Much like the SA inhibition of CuBc, Tbz required relatively large amounts to inhibit oxidative turnover of DTBC in air.

Figure 3.33 – fitting for determination of Kic of Tbz towards CuBc under aerobic conditions (R2 = 0.8) by fitting data from Table 3.3 163 As Tbz is an inhibitor that exhibits strong binding in the presence of free-metal, the relatively poor inhibition under aerobic conditions could be indicative of the physical inaccessibility to form a stable ternary complexes with CuBc.

The inhibition constant for the for the ternary complex of Tbz-CuBc-DTBC (ESI) was determined by fitting the Vmax,app as a function of Tbz concentration (Equation 2.5), shown in Figure 3.34. The fitting yielded a constant Kiu of 1.5 ± 0.5 mM Tbz. The relatively high dissociation constant could further indicate the lack of access for ligands and substrates to bind around the Bc metal-binding pocket.

Figure 3.34 – fitting for determination of Kiu of Tbz towards CuBc under aerobic conditions (R2 = 0.8) by fitting data from Table 3.3

164 Much like examining the inhibition of SA towards peroxide turnover, similarly

Tbz was evaluated. Interactions between the three compounds may be instigated by multi-antibiotic strategies used to control potential crop infections. Like Tbz, H2O2 can be topically used on crops in concentrations as high 35% v/v, as advised by the USDA and EPA in the United States; it is likely these concentrations are much lower upon application after dilution.64,65 Upon titration of Tbz to CuBc in the presence of H2O2, the rates of DTBC oxidation follow a complex pattern, shown in Figure 3.35. Sub- stoichiometric titration of Tbz to CuBc in the presence of constant H2O2 and DTBC showed a decrease in quinone production. When more than approximately one equivalent of Tbz is added, an apparent increase in rate is observed until reaching an equilibrium in rate. This pattern shows an obvious biphasic pathway for Tbz inhibition in the presence of H2O2. One probable pathway is the chelation of copper from the Bc active site. Given this was not observed in the aerobic case, it seems the hydrogen peroxide in conjunction with Tbz might play a role in destabilizing the metal-Bc interactions. In an effort to measure the Tbz inhibition against H2O2 toward CuBc- mediated oxidation of DTBC, sub-stoichiometric concentrations of Tbz were used.

The rates of 2 mM DTBC oxidation by CuBc in the presence of increasing amounts of H2O2 were measured at various amounts of Tbz, which allowed for determination the inhibition pattern for Tbz against peroxide turnover, Figure 3.36.

Unlike previous titrations, this inhibition pattern could contain an equilibrium of a few types of inhibition, potentially including chelation based on the previous figure. At the

165 concentrations used, though, of Tbz it was still clear that Tbz was inhibiting DTBC turnover via competitive and uncompetitive pathways based on the inhibition pattern in Figure 3.36. This type of mixed inhibition pattern is evidenced by a lowering Vmax,app and an increasing Km,app with an increasing amount of Tbz, as shown in Table 3.4.

200

150

100 Rate (nM/sec)

0 0.00 0.01 0.02 0.03 0.04 0.05 0.06 [Tbz] (mM)

Figure 3.35 - titration of Tbz in the presence of 10 μM CuBc, 30 mM H2O2, and 2 mM

DTBC in 100 mM pH 7.0 1:1 HEPES:DMSO.

166 0.030 8.0 M Tbz 0.025 4.0 M Tbz 2.0 M Tbz 0.020 1.0 M Tbz 0.0 M Tbz sec)

-1 0.015

0.010 1/Rate (nM 0.005

0.000 -0.050.000.050.100.150.200.25

-0.005 -1 1/[H2O2] (mM )

Figure 3.36 – LB plot for titration of hydrogen peroxide to 10 μM CuBc measuring oxidation rate 3 mM DTBC with varying amounts of Tbz up to 8 μM.

Table 3.4 – MM parameters of hydrogen peroxide and 10 μM CuBc mediated turnover in the presence of Tbz

Vmax kcat Catalytic Efficiency Tbz (μM) (nM/sec) Km (mM) (sec—1) (M—1 sec—1) 0.0 497 ± 42 28 ± 7 (5.0 ± 0.4) × 10—2 1.8 ± 0.3 1.0 445 ± 48 29 ± 9 (4.4 ± 0.5) × 10—2 1.53 ± 0.3 2.0 383 ± 37 31 ± 9 (3.8 ± 0.4) × 10—2 1.24 ± 0.3 4.0 371 ± 37 37 ± 10 (3.7 ± 0.4) × 10—2 1.00 ± 0.3 8.0 345 ± 34 39 ± 10 (3.4 ± 0.3) × 10—2 0.88 ± 0.3

167 Figure 3.37 – Fitting of Vmax,app/Km,app as a function of [Tbz] according to Equation 2.5 for 2.0 mM DTBC oxidation by the CuBc complex in the presence of increasing amounts of H2O2 to afford the competitive dissociation constant, Kic.

The inhibition of hydrogen peroxide binding to the CuBc metal center was found to have a Kic of 6.3 ± 0.7 μM for Tbz upon fitting of Vmax,app/Km,app as a function of Tbz concentration, shown in Figure 3.37, to Equation 2.4. The Kic indicates a relatively strong bonding towards the metal center, where affinity 1/Kic = 1.6 × 105 M—1.

Given the mixed inhibition exhibited in Figure 3.36, the apparent Vmax for each concentration of Tbz was fitted to Equation 2.5, shown in Figure 3.38.

168 Figure 3.38 - Fitting of Vmax,app as a function of [Tbz] according to Equation 2.5 for 2.0 mM DTBC oxidation by the CuBc complex in the presence of increasing amounts of

H2O2 to afford the uncompetitive dissociation constant, Kiu.

The fitting in Figure 3.38 revealed a Kiu 13.2 ± 2.6 μM for Tbz inhibition against H2O2.

Inhibition of the ternary complex exhibited a higher dissociation constant for ternary complex. While much lower than the Kiu for SA, this could still be indicative of difficulty for even small molecules to have the space to access the metal binding site or different binding modes in Bc.

169 2.2.2 Paramagnetic NMR

Paramagnetic NMR is able to determine structural information for the metal-Bc complex in the presence of Tbz by measuring differences in the isotropically shifted signals as Tbz was introduced to CoBc. Isotropically shifted signals of CoBc were directly comparable to signal assignments performed by Tay, et al in d6-DMSO.19 As in previous experiments, Co2+ was used as a paramagnetic probe in order to produce observable paramagnetically shifted signals.

While signal assignment for Tbz binding was partially possible in d4-MeOD, with a mix of broad and sharp paramagnetic signals based on proximity of paramagnetic influence of proton relaxation. 2D exchange correlation in DMSO was not possible, due in part to incredibly broad weak signals for Tbz, shown in Figure 3.39 (lower spectrum) and/or lack of exchange. Bacitracin, Co2+, and Tbz required the more polar solvent,

DMSO, to maintain solubility, thus limiting 2D exchange data as observed by Tay, et al and in these studies.19

170 Figure 3.39 – Paramagnetic spectrum of Co-Tbz (0.9 molar equivalents Co2+) in d4-

MeOD (top) and d6-DMSO (bottom). Spectrum with DMSO zoomed 50k× relative to diamagnetic region, acquired using a Super-WEFT sequence

Initial titrations of Tbz to CoBc revealed no new paramagnetic signals with up to

8 equivalents of Tbz. The lack of Tbz signals in ternary binding is likely due to the broad weak signals observed in Tbz-Co2+ binding observed in Figure 3.39. Figure 3.40

171 reveals a general weakening of most signals which likely indicates chelation of the metal from bacitracin.

Figure 3.40 – titration of Tbz to 4 mM CoBc in d6-DMSO, where Tbz increased going down: 0 Tbz (top, black), 0.5 eq Tbz (red), 1.0 eq (blue), 5.0 eq Tbz (orange), and 8.0 Tbz

(bottom, green). Best viewed in color for easy comparison

172 In comparison with the kinetic data that revealed binding and inhibition, the ternary complex is likely being formed at sub-stoichiometric concentrations, while higher concentrations pushes the equilibrium towards the removal of the metal and formation of the binary CoTbz complex. Alternatively, it might be performing stepwise ligand removal of bacitracin that is not observed using the concentrations in this study.

III. Concluding Remarks

The story of bacitracin is multifaceted given its presence in over-the-counter antibiotics and in the soil as it is produced by a few Bacillus species. From the medicinal perspective metallo-Bc and SA are both topical antibiotics. Based on the kinetic and

NMR data, the SA interaction might deem either compound less effective. The poor binding by SA might make it less susceptible to inhibiting CuBc, or potentially ZnBc as many of the over-the-counter form under aerobic conditions. SA seemed much more effective at inhibiting DTBC oxidation in the presence of H2O2. The lower Kic and Kiu for hydrogen peroxide indicates SA might effectively inhibit hydrogen peroxide turnover.

From the micro-ecological perspective, the inhibition of SA against CuBc might hold some clues towards the chemistry in the soil and biofilms around plants like the willow tree. SA in the soil could help regulate CuBc reactivity in the soil, helping maintain the redox activity around the host organism. Extrapolating from the metal-binding and

173 inhibition studies, SA could play a role in regulating other metallo-antibiotics and metallo-biomolecules.

The study of Tbz also shows how metal-binding synthetic drugs can have an effect on ambient metallo-antibiotics in the soil. Much like SA, Tbz is a topical antibiotic, although, the organism to which it is applied are fruits and vegetables rather than the scrapes and smalls skin wounds of people. Based on the data acquired in this dissertation, Tbz might act a double-edged sword. On the one hand it protects plants from fungal activity; and on the other hand, it might be at the sacrifice of the antibiotic activity of bacitracin from ambient Bacillus. Moreover it might fine-tune the soil microecological balance, and potentially may also sequester ambient metals in the soil.

Based on the strong metal-binding shown for Tbz likely plays a role in chelating environmental and biological metals like Cu2+.

Whether the chemistry is occurring in the soil or on the skin, understanding the chemistry of metal-based antibiotics ensure these interactions are taken into account when they are present together. By observing the binding of SA and Tbz towards an antibiotic metal center and inhibition of the metal center activity, this study helps provide more information behind the chemistry occurring under medicinal and microecological circumstances.

174 IV. Materials and Methods

4.1 Kinetics

All chemicals were purchased from Sigma Aldrich. Kinetics experiments were carried out using a Cary 50 UV-Vis spectrophotometer. Oxidation studies were performed on 3,5-di-tert-butylcatechol, where the quinone formed was measured at 410 nm (ε410 = 1900 M—1 cm—1).

Kinetic experiments revealing SA inhibition were performed in a buffered mixed solvent system to maintain pH and solubility of both the reactants and the products.

Ternary complexes were not soluble in 100% aqueous systems. 100 mM HEPES buffer at pH 7.0 was mixed with an equal volume (1:1) of methanol.

Tbz inhibition studies were performed in a mixed solvent system as well, where

1:1 mixture of 100 mM HEPES at pH 7.0 and DMSO was used to maintain solubility and pH control.

4.2 Nuclear Magnetic Resonance

All NMR studies were performed on an Agilent Inova and a Direct Drive spectrometers at a reference proton resonance of 600 MHz. All samples were prepared at ~5-10 mM in D2O, d6-DMSO, or d4-MeOD, as noted in the text. CoCl2 salt was used as

175 a paramagnetic metal source to investigate metal binding. Its metal binding properties reliably mimic Cu2+ binding in terms of binding geometry and motifs, while maintaining smaller spectral windows <120 kHz. 1D 1H spectra were obtained by the use of presaturation to reduce solvent intensity to better resolve the inherently weaker paramagnetic 1H signals. Alternatively, “Super-WEFT” was also employed by taking advantage of the fast relaxation of paramagnetic protons, wherein the delay times in the pulse sequence (D1-180°-τ-90°) are optimized to reduce the diamagnetic signals in order to comparatively improve the paramagnetic signals.66

NOESY and EXSY were the primary two dimensional NMR techniques employed to determine the binding capabilities of Co2+ and the ternary complex including Co2+, Bc, and SA. EXSY pulse sequences were modified based on the NOESY pulse sequence, using short mixing time, rather than longer mixing times, to reduce NOE buildup allowing only fast exchanging (1-20 ms) paramagnetic-diamagnetic protons to accrue signal. Spectra were acquired with 120 kHz spectral windows, 64 scans, 512 increments,

10-20 Hz line broadening, and 1K×1K points in both dimension for processing.

V. References

1 Jonhson, B.A., Anker, H., Meleney, F.L. Bacitracin: A new antibiotic produced by a

member of the B. subtilis group. Science. 1945, 102, 376-377. Doi:

10.1126/science.102.2650.376 176 2 Moore, M., and Wooldridge, W.E. Antifungal properties of various extracts of Bacillus

subtilis (Tracy strain obtained in the bacitracin recovery process. Journal of

Investigative Dermatology. 1950, 265-281. Doi: 10.1038/jid.1950.34

3 Bacitracin. Pharmaceutical Manufacturing Encyclopedia, 3rd ed.; William Andrew

Publishing: New York, 2007; Vol. 4, pp 509-512.

4 ChemDraw: Cambridgesoft, 100 CambridgePark Drive, Cambridge, MA 02140mercial:

http://www.chemdraw.com

5 Arky, R. Physician’ Desk Reference for Nonpresciption Drugs, 20th Ed. Medical

Economics Company: Montvale, NJ, 1999.

6 Phillips, I. The use of bacitracin as a growth promoter in animals produces no risk to

human health. Journal of Antimicrobial Chemotherapy. 1999. 44(6)725-728. Doi:

10.1093/jac/44.6.725.

7 Answer to the Request from the European Commision for Updating the Scientific

Advice on the Impact on Public Health and Animal Health of the Use of

Antibiotics in Animals – Categorisation of Antimicrobials. European Medicines

Agency, United Kingdom. Science Medicines Health: London, United Kingdom.

2019.

8 Barry, G.T., Gregoly, J.D., and Craig, L.C. The nature of bacitracin. Journal of Biological

Chemistry. 1948, 175, 485-486.

9 Craig, L.C., Weisiger, W., and Hausmann, E.J. The separation and characterization of

bacitracin polypeptides. Journal of Biological Chemistry. 1952, 259-266. 177 10 Newton, G.G.F., and Abraham, E.P. Some properties of bacitracin polypeptides.

Biochemical Journal. 1953, 4, 597-604. Doi: 10.1042/bj0530597

11 Kang, J-w, Reymaeker, G.D., Schepdael, A.V., Roets, E., and Hoogmartens, J. Analysis

of bacitracin by micellar electrokinetic capillary chromatography with mixed

micelle in acidic solution. Electrophoresis. 2001, 7, 1356-1362. Doi: 10.1002/1522-

2683

12 Ming, L.-J., and Epperson, J.D. Metal binding and structure-activity relationship of

the metalloantibiotic peptide bacitracin. Journal of Inorganic Chemistry. 2002.

91(1)46-58. Doi: 10.1016/S0162-134(02)00464-6.

13 Smith, J.L. and Weinberg, E.D. Mechanisms of antibacterial action of bacitracin. J. gen.

Microbiol. 1962. 559-569. Doi: 10.1099/00221287-28-3-559

14 Garbutt, J.T., Morehouse, A.L., and Hanson, A.M. Metal binding properties of

bacitracin. Journal of Agriculture and Food Chemistry. 1961, 9, 285-289. Doi:

10.1021/jf60116a013

15 Stone, K.J. and Strominger, J.L. Mechanism of action of bacitracin: complexation with

metal ion and C55-isoprenyl pyrophosphate. PNAS. 1971, 68, 3223-3227. Doi:

10.1073/pnas.68.12.3223.

16 Higashi, Y., Strominger, J.K., and Sweeley, C.C. Biosynthesis of the peptidoglycan of

bacterial cell walls. The Journal of Biological Chemistry. 1970. 245(14)3697-3702.

178

17 Storm, D.R., and Strominger, J.L. Complex formation between bacitracin peptides and

isoprenyl pyrophosphates. The Journal of Biological Chemistry. 1978. 248(11)3940-

3945.

18 Qi, Z.D., Lin, Y., Zhou, B., Ren, X.-D., Pang, D.-W., and Liu, Y. Characterization of the

mechanism of the Staphylococcus aureus cell envelope by bacitracin and

bacitracin-metal ions. Journal of Membrane Biology. 2008. 225: 27. Doi:

10.1007/s00232-008-9130-8

19 Tay, W.M., Epperson, J.D., da Silva, G.F.Z, and Ming, L.-J. 1H NMR, Mechanism, and

Mononuclear Oxidative Activity of the Antibiotic Metallopeptide Bacitracin: The

Role of d-Glu-4, Interaction with Pyrophosphate Moiety, DNA Binding and

Cleavage, and Bioactivity. Journal of the American Chemical Society. 2010.

132(16)5652-5661. Doi: 10.1021/ja910504t.

20 Craig, L.C., Phillips, W.F., and Burachik, M. Bacitracin A. isolation by counter double-

current distribution and characterization. Biochemistry. 1969. 8(6)2348-2356. Doi:

10.1021/bi00834a015.

21 Pons, M., Feliz, M., Molins, M.A., and Giralt, E. Conformational analysis of bacitracin

A1, a naturally occurring lariat. Biopolymers. 1991. 31(6)605-612. Doi:

10.1002/bip.360310604

22 Kobayashi, N., Takenouchi, T., Endo, S., and Munekata, E. 1H NMR study on the

conformation of bacitracin A in aqueous study. Federation of European Biochemical

Societies. 1992. 305(2)105-109. Doi: 10.1016/0014-5793(92)80874-G. 179

23 Wasylishen, R.E., and Graham, M.R. A nuclear magnetic resonance study of the metal

binding sites in bacitracin. Canadian Journal of Biochemistry. 1975. 53(12)1250-1254.

24 Seebauer, E.G., Duliba, E.P., Scogin, D.A., Gennis, R.B., Belford, R.L. EPR evidence of

the copper(II) bacitracin A complex. Journal of the American Chemical Society. 1983.

105(15)4926-4929. Doi: 10.1021/ja00353a015

25 Epperson, J.D., and Ming, L.-J. Proton NMR studies of Co(II) complexes of the peptide

antibiotic bacitracin and analogues: Insight into structure-activity relationship.

Biochemistry. 2000. 39(14)4037-4045. Doi: 10.1021/bi991997p

26 Zillich, O.V., Scheiggert-Weisz, I., Eisner, P., and Kerscher, M. Polyphenols as active

ingredients for cosmetic products. Journal of Cosmetic Science. 2015. 37:455-464.

Doi: 10.1111/ics.12218.

27 Radhakrishnan, R., Hashem, A., and Abd_Allah, E.F. Bacillus: A biological tool for

crop improvement through bio-molecular changes in adverse environments.

Frontiers in Physiology. 2017. 8(667)1-14. Doi: 10.3389/fphys.2017.00667

28 Barnawal, D., Maji, D., Bharti, N., Chanotiya, C.S., and Kalra, A. ACC Deaminase-

containing Bacillis subtilis reduces stress ethylene-induced damage and improves

mycorrhizal colonization and rhizolbial nodulation in Trigonella foenum-graecum

under drought stress. Journal of Plant Growth Regulation. 2013. 32:809-822. Doi:

10.1007/s00344-013-9347-3

180

29 Hashem, A., Tabassum, B., and Abd_Allah, E.F. Bacillus subtilis: A plant growth

promoting rhizobacterium that also impacts biotic stress. Saudi Journal of

Biological Sciences. 2019. 1291-1297. Doi: 10.1016/j.sjbs.2019.05.004.

30 Desborogough, M.J.R. and Keeling, D.M. The aspirin story – from willow to wonder

drug. British Journal of Haematology. 2017. 177(5)674-683. Doi: 10.1111/bjh.14520

31 Bryan, C.P. Ancient Egyptian Medicine: The Papyrus Eber; Ares Publishers, Inc: London,

1930.

32 Rashkin, I. Role of salicylic acid in plants. Annual Review of Plant Physiology and Plant

Molecular Biology. 1992. 43:439-463. Doi: 10.1146/annurev.pp.43.060192.002255

33 Vicente, M. R-S., Plasencia, J. Salicylic acid beyond defence: its role in plant growth

and development. Journal of Experimental Botany. 2011. 62(10)3321-3338. Doi:

10.1093/jxb/err031.

34 Slaymaker, D.H., Navarre, D.A., Clark, D., del Pozo, Martin, G.B. and Klessig, D.F.

The tobacco salicylic acid-binding protein 3 (SABP3) is the chloroplast carbonic

anhydrase, which exhibits antioxidant activity and plays a role in the

hypersensitive defense response. PNAS. 2002. 99(18)11640-11645. Doi

10.1073/pnas.18427699.

35 Catto, C., Grazioso, G., Dell’Orto, S., Gelain, A., Villa, S., Marzano, V., Vitali, A., Villa,

F., Cappitelli, F., and Forlani, F. The response of Escheria coli biofilm to salicylic

acid. The Journal of Bioadhesion and Biofilm Research. 2017. 33(3)235-251. Doi:

10.1080/08927014.2017.1286649 181

36 Wu, H.-S., Raza, W., Fan, J.Q., Sun, S.-G., Bao, W., Liu, D.-Y., Huang, Q.-W., Mao, Z.-

s., Shen, Q.-R., and Miao, W.-G. Antibiotic effect of exogenously applied salicylic

acid on in vitro soilborne pathogen, Fusarium oxysporum f.sp.niveum. Chemosphere.

2008. 74(1)45-50. Doi: 10.1016/j.chemosphere.2008.09.027

37 Miyamoto, T., Ogino, N., Yamamoto, S., and Hayaishi, O. Purification of

prostaglandin endoperoxide synthetase from bovine vesicular gland

microsomes. Journal of Biological Chemistry. 1976. 251(9)2629-2639.

38 Vane, J.R. Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-

like drugs. Nature New Biology. 1971. 231:232-235.

39 Mitchell, J.A., Saunders, M., Barnes, P.J., Newton, R., and Belvisi, M.G. Sodium

salicylate inhibits cylco-oxygenase-2 activity independently of transcription

factor (nuclear factor κB) activation: Role of arachidonic acid. Molecular

Pharmacology. 1997. 51(6)907-912. Doi: 10.1124/mol.51.6.907

40 Dannhardt, G., and Kiefer, W. Cyclooxygenase inhibitors – current status and future

prospects. European Journal of Medicinal Chemistry. 2001. 36(2)109-126. Doi:

10.1016/S0223-5234(01)01197-7

41 Perrin, D.D. Stability of metal complexes with salicylic acid and related substances.

Nature. 1958. 182. 741-742. Doi: 10.1038/182741a0.

42 Dhat, C.R. and Jahagirdar, D.V. Transition metal ion chelates of some substituted

salicylic acids: Part II-A thermodynamic study. Indian Journal of Chemistry. 1983.

22A:131-133. 182

43 Shi, Q., and Zhu, Z. Effects of exogenous salicylic acid on manganese toxicity, element

contents and antioxidative system in cucumber. Environmental and Experimental

Botany. 2007. 63:317-326.

44 Gaubert, S., Bouchaut, M., Brumas, V., and Berton, G. Copper-ligand interactions and

physiological free radical processes. Part 3. Influence of histidine, salicylic acid,

and anthranilic acid on copper-driven Fenton chemistry in vitro. Free Radical

Research. 2000. 32(5)451-461. Doi: 10.1080/10715760000300451

45 Brown, H.D., Matzuk, A.R., Ilves, I.R., Peterson, L.H., Harris, S.A., Sarett, L.H.,

Egerton, J.R., Yakstis, J.J., Campbell, W.C., Cuckler, A.C. Antiparasitic drugs. IV.

2-4(4’-thiazolyl)-benzimidazole, a new antihelmenthic. Journal of American

Chemical Society. 1961. 83(7)1764-1765. Doi: 10.1021/ja1468a052

46 Prichard, R.K. The fumarate reductase reaction of Haemonchus contortus and the mode

of action of some anthelmintics. International Journal of Parasitology. 1973. 3(3)409-

417. Doi: 10.1016/0020-7519(73)90121-5.

47 Cabanas, R., Castella, G., Abarca, M.L., Bragulat, M.R., and Cabanes, F.J.

Thiabendazole resistance and mutations of the β-tubulin gene of Penicillium

expansum strains isolated from apples and pears with blue mold decay. FEMS

Microbiology Letters. 297(2)189-195. Doi: 10.1111/j.1574-6969.2009.01670.x

48 Zhou, Y., Xu, J., Zhu, Y., Duan, Y., and Zhou, M. Mechanism of action of

benzimidazole fungicide on Fusarium graminearum: interfering with

polymerization of monomeric tubulin but not polymerized microtubule. Disease 183

Control and Pest Management. 2016. 106(8)807-813. Doi: 10.1049/PHYTO-08-15-

0186-R

49 Jamieson, J.D., Smith, E.B., Dalvie, D.K., Stevens, G.J., Yanochko, G.M.

Myeloperoxidase-mediated bioactivation of 5-hydroxythiabendazole: A possible

mechanism of thiabendazole toxicity. Toxicology In Vitro. 25(5)1061-1066. Doi:

10.1016/j.tiv.2011.04.007.

50 R.E.D. Facts 2002; U.S. Environmental Protection Agency.

51 Review report for the active substance thiabendazole. Food and Feed Safety,

Innovation, Pesticides, and Biocides, European Commission Directorate-General

for Health and Safety. 2015.

52 Lundy, J., Lovett, E.J. Conran, P., Goldblatt, P.J. Thiabendazole: a potential adjuvant

in cancer therapy. Surgery. 1976. 80(5)636-640.

53 Fu, X.B., Zhang, J.J., Liu, D.D., Gan, Q., Gao, H.W., Mao, Z.W., Le, X.Y. Cu(II)-

dipeptide complexes of 2-(4’-thiazolyl)benzimidazole: synthesis, DNA oxidative

damage, antioxidant and in vitro antitumor activity. Journal of Inorganic

Biochemistry. 2015. 143:77-87. Doi: 10.1016/j.jinorgbio.2014.12.006.

54 Devereux, M., O’Shea, D., Kellett, A., McCann, M., Walsh, M., Egan, D., Deegan, C.,

Kediora, K., Rosair, G., and Muller-Bunz, H. Synthesis, X-ray crystal structures

and biomimetic and anticancer activities of novel copper(II)benzoate complexes

incorporating 2-(4′-thiazolyl)benzimidazole (thiabendazole), 2-(2-

pyridyl)benzimidazole and 1,10-phenanthroline as chelating nitrogen donor 184

ligands. Journal of Inorganic Biochemistry. 2007. 101(6)881-892. Doi:

10.1016/j.jinorgbio.2007.02.002

55 Kowala, C., Murray, K.S., Swan, J.M., and West, B.O. Transition metal complexes of

thiabendazole. Australian Journal of Chemistry. 1971. 24:1369-1375. Doi:

10.1071/CH9711369.

56 Miller, V.L., Gould, C.J., Csonka, E., and Jensen, R.L. Journal of Agriculture and Food

Chemistry. 1973. 21(5)931-932. Doi: 10.1021/jf60189a042

57 Zhang, Y., de Stefano, R., Robine, M., Butelli, E., Bulling, K., Hill, L., Rejzek, M.,

Martin, C., and Schoonbeek, H.J. Different reactive oxygen species scavenging

properties of flavonoids determine their abilities to extend the shelf life of

tomato. Plant Physiology. 2015. 169:1568-1583. Doi: 10.1104/pp.15.00346.

58 Farhadi, F., Khameneh, B., Iranshahi, M., and Iranshahy, M. Antibacterial activity of

flavonoids and their structure-activity relationship: An update review.

Phytotherapy Research. 33:13-40. Doi: 10.1002/prt.6208.

59 Petrova, A., Davids, L.M., Rautenbach, F., and Marnewick, J.L. Photoprotection by

honeybush extracts, hesperidin, and mangiferin against UVB-induced skin

damage in SKH-1 mice. Journal of Photochemistry and Photobiology. 2011. 103(2)126-

139. Doi: 10.1016/j.jphotobiol.2011.02.020.

60 Das, M., Mukhtar, H., Bik, D.P., and Bickers, D.R. Inhibition of epidermal xenobiotic

metabolism in SENCAR mic by naturally occurring plant phenols. Cancer

Research. 1987. 41(3)760-766. 185 61 Boveris, A., Oshino, N., and Chance, B. The cellular production of hydrogen peroxide.

Biochemical Journal. 1972. 128(3)617-630. Doi: 10.1042/bj1280617

62 Hwang, H., and Dasgupta, P.K. Themodynamics of the hydrogen peroxide-water

system. Environmental Science and Technology. 1985. 19:255-256.

63 Miller, V.L., Gould, C.J., Csonka, E., and Jensen, R.L. Metal coordination compounds

of thiabendazole. Journal of Agricultural and Food Chemistry. 1973. 21(5)931-932.

Doi: 10.1021/jf60189a042

64 R.E.D. Facts: Peroxy Compounds. U.S. Environmental Protection Agency, U.S.

Government, Washington, D.C. [Online]

https://www3.epa.gov/pesticides/chem_search/reg_actions/reregistration/fs_

G-67_1-Dec-93.pdf

65 Hydrogen peroxide: crops. U.S. Department of Agriculture, U.S. Government

Printing Office: Washington, D.C., 2015. [Online]

https://www.ams.usda.gov/sites/default/files/media/Hydrogen%20Peroxide

%203%20TR%202015.pdf

66 Inubushi, T., and Becker, E.D. Efficient detection of paramagnetically shifted NMR

resonances by optimizing the WEFT pulse sequence. Journal of Magnetic

Resonance. 1983. 51(1)128-133. Doi: 10.1016/0022-2364(83)90109-9/

186

CHAPTER 4

Three-Dimensional Structure Elucidation of a Novel Peptide “Foldamer” and NMR

Evaluation of a Metallo-Peptoid

Adapted in part with permission from a Journal of the American Chemical Society

publication1

If you wish to make an apple pie from scratch, you must first invent the universe.

- Carl Sagan

I. Background

The importance of the development of non-resistant drugs cannot be understated! A

2013 Centers for Disease Control (CDC) study found approximately 20 million people are infected by antibiotic resistance diseases, and 23,000 people die from them each year! For individual patients this can be incredibly challenging as antimicrobial resistance will invariably require more costly drugs, longer hospital stays, and loss of

187

income. The overall cost of drugs, wage loss, hospitalization, and other peripheral costs- of-care are estimated to be $35 billion in the U.S. alone each year.2

Antibiotic resistance is as old as antibiotics themselves. Natural antibiotics are thought to have evolved as a means of self-preservation in a type of “arms race” for microbes.3,4 This, of course, took place over billions of years of evolution and natural selection for microbes to develop the antibiotics we observe and use today.

The ubiquity of bacteria makes human exposure to antibiotics inevitable. For example, there is evidence of trace tetracyclines in nearly 2000 year old skeletal remains from diets with high concentrations of the antibiotic.5 Most people, however, associate antibiotic use with the last ~80 years. The introduction of antibiotics worldwide in the

1940s has inarguably saved millions of lives. Unfortunately, wide-spread, inconsistent, and irresponsible use of antibiotics has led to many first generation drugs being no longer effective towards disease. Many of the early and current generations of antibiotics have suffered in efficacy due to mis-prescribing, such as prescribing an antibiotic for viral infections, and patients not completing a cycle of a prescription allowing for the more resistance bacteria to propagate resistance.2

Several approaches over the last several decades have been developed to curb antibiotic resistance. The CDC and the World Health Organization (WHO) have advocated with healthcare professionals and patients for responsible prescription of antibiotics and to take antibiotics as prescribed, respectively.2,6 Carefully researched studies have shown that some non-interacting drug combinations, or cocktails, can

188

prove more effective and less prone towards developing resistance, even towards drug resistant bacteria.7

Previous chapters have hopefully shown the structure of drugs and proteins can aid in explaining their functionality. Many current methods similarly approach disease control by designing structures that might induce a desired outcome. Some of the primary considerations in drug design are specificity and efficacy.8 For example, if a highly efficient drug is not specific towards its target, it could lead to unwanted side- effects, injury, or death as the drug interacts with essential biological systems of a host organism. Conversely, if the drug is highly specific towards a single, desired target, but does not effectively perform or bind to the target, it will not be a desirable drug since it would require high quantities to affect the target.

The development of new drugs can take several avenues of discovery, including target-specific design, natural product isolation or synthesis, high-throughput screening, and many others.8 The last couple of decades have had an influx of interest in proteins and protein-mimics to treat disease. The predominant issue with protein or peptides treatment for microbial disease is the relatively short bioavailability of most exogenous peptides. Peptides introduced into biological systems often encounter proteases that hydrolyze the foreign peptides, Figure 4.1.9 One of the common methods to increase longevity of drugs is to alter the backbones to prevent protease recognition and subsequent hydrolysis.

189

Figure 4.1– Hydrolytic reaction performed by a protease along peptide backbone

Altering or designing peptides with non-standard backbones, other than L-α configurations, have gained popularity for their increased in vivo longevity and for their ability to mimic protein secondary and tertiary structures. Addition of one or two carbons along the backbone to form beta- and gamma-peptides, respectively, have been shown to have longer bioavailability.10-12 More complex combinations and backbone structures, including the addition of bulky side groups and sulfono-peptides, are discussed in the subsequent section.

Unfortunately, synthesis of mimetic peptides are not directly analogous to their non-mimetic counterpart.13 Increased backbone length, allowing for more rotation around a peptoid α-carbon, can change the structure and, thus, change the function of the peptide.14 Drug design of peptido-mimetic structures can use strategies previously developed by researchers studying non-mimetic counterparts, including target-specific design and libraries of peptide variations for antibiotic screenings.15 Some studies, like the one discussed herein, focus on developing three-dimensional structural motifs that fold into specific patterns in order to maintain stability and potentially interact with specific targets.1,16 190

L-α-alanine L-β-leucine γ-aminobutyric acid D-α-alanine

Figure 4.2 - Variations in peptide backbones via addition of carbons (second and third) or inversion of stereo-center (last) compared to the most common natural peptide (first).

Many peptido-mimetic compounds have shown promise. Peptido-mimetic compounds Lytixar™ and brilacidin are both currently in phase 2 trials for treating

Gram-positive bacterial infections and cancer, respectively.17,18 In another example, specific folds were designed into the peptide, deemed foldamers, in order to bind to cell membranes for eventual membrane depolarization. The cited small foldamer was even shown to be efficacious towards multidrug resistant Staphylococcus aureus and

Escherichia coli.19

Two models for peptido-mimetic structure were explored in this dissertation. As structure is indicative of function, as discussed in previous chapters, structural information was ascertained using NMR. In the first study, a structural elucidation of a novel D-sulfono-γ-AApeptide that can form a specific tertiary structure was performed.

A second study using a peptido-mimetic γ-peptoid was evaluated for its metal-binding

191

capabilities compared to its non-mimetic counterpart was structurally explored using

NMR.

II. Results and Discussion

2.1 D-sulfono-γ-AApeptide/L-amino acid Structure and Conformation

Elucidation

In this study a novel peptide was synthesized using a variety of peptidomimetics motifs by Teng, et al. This particular peptide was designed to induce a specific three dimensional folding pattern, making this peptide a “foldamer”. D-sulfono-γ-

AApeptide/L-amino acid peptide was synthesized for its potential to fold into a helical structure, based on previous literature.25

As opposed to the earlier studies discussed herein, this peptide does not contain a metal center. In this case, it was included for its novelty and potential in drug design.

As discussed in chapter 1, many proteins like SOD1 require a flexibility in the scaffold to modulate reactivity at the metal center. In understanding the design of three- dimensional folding of peptide mimics, metallo-centers could be designed in future developments with higher bioavailability and specificity. As metallo-antibiotics and proteins have largely been disproportionately under-represented in research, by including this study herein, it may help begin bridging the gap between peptido-

192 mimetics and metallo-peptides, which only more recent studies have begun investigating.20-23

Published studies have previously shown the propensity for folding among peptides containing the γ-backbone motif which has been further expanded by the D- sulfono-γ-AApeptide backbones.24,25 Moreover the N-acetylated-N-aminoethyl amino acids (AApeptides) have shown potential for increased bioavailability by avoiding proteolysis due to the lack of recognition sites required for proteolytic enzymes. As study on these scaffolds have grown since their initial discovery, current designs are also beginning to approach the efficiency of natural peptides.26,27 Figure 4.4 shows a basic design scheme for these peptide motifs. These peptides use a combination N- substituted functional groups, like peptoids, lengthened backbones up to γ-carbons, and the addition of bulky sulfono side-chains. For the initial four structures synthesized, each peptide contained a repetition of L-α/D-sulfono-γ-AA amino acid, L- alanine, and a L-phenylalanine. Each sequence differed in combination of 3 or 4 repeating flexible amino acids, L-alanine and L-phenylalanine, and the presence of a terminal L-alanine.

Figure 4.3 – Structural motifs for the design of D-sulfono-γ-AApeptide/L-amino acid

193

For the purposes of using NMR as a technique to determine “foldamer” properties, the peptide was synthesized maintaining the general scheme and length as oligomer 3 in this publication. The peptide studied, oligomer 5, was synthesized using slightly different sulfono- side chains and several amino side-groups to facilitate accurate signal assignment, avoiding chemical shift overlap of repeating residues.

Oligomer 3 contained four repeating sulfono- side chains of para-chlorobenezene, leading which would reveal NMR signal ambiguity due to overlapping chemical shifts, especially for the internal amino acids. In the publication several similar oligomers of varying length were studied using circular dichroism (CD) and X-ray crystallography to elucidate three-dimensional folding in as much detail as possible. The synthesis of oligomer 5 yielded the structure as shown in Figure 4.4.

Figure 4.4 – line representation of D-sulfono-γ-AApeptide/L-amino acid oligomer 5

Initial 1H NMR spectra were obtained using the water elimination (WET) solvent suppression pulse sequence given the use of d3-MeOH solvent, which yields strong

194 signals at 4.8 and 3.3 ppm, shown in Figure 4.5. Despite the relative simplicity of the peptide, careful measurements of the gCOSY (gradient correlation spectroscopy) and zTOCSY (total correlation spectroscopy) spectra were required to perform a complete structural elucidation. In conjunction with the 1D 1H spectrum, the TOCSY allowed for assignment of each amino acid residue based the number protons per residue and the chemical shifts to make confident assignments. The TOCSY spectrum yields information via long-range homonuclear scalar coupling. Longer times in mixing or spinlock can achieve more through-bond coupling allowing for magnetization between protons to show correlation up to several bonds from the irradiated signal. For example, alanine would show three signals corresponding to one in the 7-9 ppm region, one signal in ~2.5-4.5 ppm for the backbone proton, and a strong signal ~1-2 ppm for the -

CH3 side group. In comparison, a valine, such as residue 14, would yield five to six signals in TOCSY corresponding to one proton in the NH region, one from the backbone proton, and 3 or 4 from the side chain depending on any loss in symmetry of the terminal CH3 groups due to any spin-coupling. Signal assignments for non-standard amino acids proved more of a challenge since coupling across heteroatoms is unlikely to occur, such as the sulfono- moiety. Measuring residues between heteroatoms loses

TOCSY correlation, relying on nuclear Overhauser effect spectroscopy (NOESY) to perform longer range signal assignment for the adjacent residues, i+1 NOE. The i+1 assignments were predominantly observed for adjacent NH amide protons with D- sulfono-γ-AApeptide backbone protons. Overlaying TOCSY signal assignment with

195 NOESY signal assignments aided in assigning signals for the D-sulfono-γ-AApeptide side chains, as did the chemical shifts by using differing aromatic functional groups.

The complete signal assignment for the peptide is shown in Table 4.1 based on the data gathered in Figure 4.5 and 4.6.

Figure 4.5 - full 1D 1H spectrum of 4.1 mM D-sulfono-γ-AApeptide/L-amino acid oligomer 5 using WET solvent suppression pulse sequence using an Inova600 spectrometer, solvent signals at 3.3 and 4.8 ppm

196 Figure 4.6– 1D 1H spectrum of D-sulfono-γ-AApeptide/L-amino acid with focus on side chains (top), backbone CH groups (middle), and aromatic and backbone NH groups (bottom)

197 Figure 4.7– overlaid 2D spectra of D-sulfono-γ-AApeptide/L-amino acid in d6-DMSO, followed by focus on the NH/aromatic correlations (f1 spectrum) to the side chain correlations (f2 spectrum) (viewed best in color)

198 Table 4.1 – complete signal assignment for oligomer 5 of D-sulfono-γ-AApeptide/L- amino acid

Residue NH γ β α Hβ' Hγ' Hδ' Hε' Hζ' 1 8.05 4.27 1.33 1.72 1.64, 1.71, NH : 2 8.19 4.35 2.93 2 1.86 1.46 1.64 7.73, 7.75 2.92 4.00 3a 8.02 4.26 1.18 3.39 3.97 CH : 3b 7.64 3 7.37 7.43 7.63 2.35 2.89 7.21, 4 8.74 4.75 3.33 7.26 5 8.47 4.31 1.43 3.12 3.91 6a 4.14 1.13 3.21 3.82 OCH : 6b 7.87 7.06 7.65 3 7.41 7.74 3.82 7 4.30 1.36 CH : 8 8.36 4.36 4.19 3 1.18 2.89 3.88 9a 7.97 4.15 1.18 3.30 3.84 9b 6.99 7.93 7.92 7.03 4.01 10 8.94 4.26 4.08 11 7.91 4.31 1.43 2.81 3.57 12a 7.67 4.23 1.16 3.38 3.92 CH : 12b 7.05 7.79 3 7.80 7.01 2.43 13 8.81 4.21 1.50 2CH3: 14 7.74 4.27 1.44 1.54 0.69, 0.77 C-terminus 7.74 7.45

199 Three-dimensional interactions and potential folding were determined using several techniques, supported by NOE measurements. NOE interactions are particularly useful in determining structural interactions as magnetization is transferred via spin-lattice (through-space) relaxation. NOESY is limited by distance where signal intensity is proportional to the inverse sixth power of the distance between the nuclei (I

α 1/r6); signal is limited by a 4-5 Å atomic distance through space depending on the size of the molecule.

In terms of folding, particular interest was paid toward NOE signals pertaining to nonadjacent residues. Seven i+2 (NOE signals two residues from NOE irradiated proton) and two i+3 (three residues from the NOE irradiated proton) NOE signals. Of particular interest are the two i+3 signals that are 17 bonds and 9 bonds away, Figure

4.8 and 4.9. The two i+3 signals reveal correlating interactions between the γ -proton of residue 6b with the α-proton of residue 9b and the terminal methyl group of L- threonine of residue 7 with the NH proton of residue 10. The relative weakness of the signals signifies the upper limits of NOE interactions as their distances approach ~5 Å in solution. These signals would not show correlation without organized folding of the interior structure, similarly shown in previous studies of similar foldameric compounds.25 The long time scales required for NMR obtains an average structure. The ability to observe i+3 and multiple i+2 NOE interactions shows a relative stability this structure at the timescales used for NMR.

200 Figure 4.8 - i+3 NOE signal for 6aHγ and the backbone NH of residue 9a (signal labeled).

Figure 4.9 – i+3 NOE signal for 7Hγ’ and the backbone NH of residue 10 (signal labeled).

All of the NOE interactions can be summarized shown in Figure 4.10.

201 Figure 4.10 – detected NOESY signals for i+1, i+2, and i+3 correlations. Best viewed in color

In conjunction with crystal structures, and circular dichroism (CD) studies of the similar oligomer 3, observed by Dr. Teng in this study, the i+2 and i+3 signal correlations confirm an alpha helix conformation in methanol, shown in Figure 4.1 and

4.10. Crystal structure torsional angles of flexible backbone and the 16-16-14 hydrogen bonding pattern also yield information on how NOE buildup could occur through space by folding into the right handed π-helix secondary structure. This is further shown as structural changes from the other 3 oligomers showed the same hydrogen bonding pattern during crystallographic measurements. The secondary structure of oligomer 5 would likely be akin to oligomer 3, as shown in Figure 4.9, proposed via molecular dynamics calculations by Dr. Teng.1

202 Figure 4.11 – D-sulfono-γ-AApeptide/L-amino acid proposed minimized structure for oligomer 3

These de novo structures of the D-sulfono-γ-AApeptide with mixed with L-amino acid residues reveals three-dimensional right-handed helical conformation. The successful formation of a helical foldamer for D-sulfono-γ-AApeptide could provide a synthetic scaffold for future oligomers. These motifs and structures could prove invaluable during the development of bioavailable antimicrobials. Development of sidechains designed for hydrophobic membrane interaction could yield antimicrobial activity via membrane depolarization as discussed in earlier studies while still resembling the structures synthesized herein.8,9

2.2 Ac-FHFH γ-Peptoid and Peptide metal binding

Among an older class of peptido-mimetic compounds are the peptoids, poly-N- substituted amino acids. These poly-N-substituted peptides have been deemed 203 ‘peptoids’ by their creators, Reyna Simon et al in 1992.28 Similar to the β-peptides, peptoids have the increased flexibility at the β-carbon as well as increased bioavailability, having no natural proteolytic enzyme that can recognize this type of backbone motif.29,30 Their uses are primarily in building peptide-like scaffolds using side-chains on the backbone nitrogen rather than the α-carbon. Besides their use to avoid proteolysis in drug design, they have also gained popularity to for their abilities to form complex secondary structures due to the free-rotating β-carbon.

Peptide Ac-FHFH Peptoid γ-AA-FHFH

Figure 4.12 – peptide and peptoid structures designed to probe difference in metal binding

Most of the focus for peptoids has been on their structures as potential protein mimics with only limited studies on using them as novel metal ligands. One of the challenges in designing protein-like compounds are differences in specificity due to structural changes. Peptoid-metal interaction studies are still relatively new with many

204 studies published in the last couple years.22,23 Structural and reactivity modulation has been shown for peptoid-based metallo-porphyrins, much like the native form!31,32 As these novel class of peptides and peptoids continue to develop, the interactions and chemistry they can perform in the presence of metals could yield much more stable, reliable, and novel drugs. Only more recent studies have even started considering them as potential targets for in metal-mediated neurodegenerative diseases, such as AD and

Parkinson’s disease.33 In understanding how peptoids might be inhibitory towards diseases like AD, the structure and reactivity of metal-bound peptides were examined in the following herein.

In this study a very simple peptide and peptoid model were synthesized using standard solid-phase synthesis techniques. The sequences produced were the acetylated

Ac-FHFH with potential bi-dentate binding capacity in each compound. 1H NMR was employed to probe metal bind capacities in the presence of paramagnetic Co2+ with observable hyperfine shifts of ±100 ppm. As was the case for previously discussed studies, protons closer to the metal binding sites experience isotropic shift due to influence of the strong magnetic nature of the unpaired electron from Co2+.

Upon titration of one molar equivalent of Co2+ shows the isotropically shifted signals for both the peptide and the γ-AA-peptoid, Figure 4.13. In the peptide experiment (top spectrum) four broad signals appear signals at 55.2, 57.3, 66.5, and 70.4 ppm. This spectrum is comparable to the Mn-to-cobalt and Cu-to-cobalt substituted

MnSOD and azurin, respectively.34,35 Typically, ortho protons with reference to the

205 coordination site on the imidazole ring would yield broader, less intense signals, if they are visible. The more intense signals between ~40-90 ppm are typically indicative of meta protons from the coordinated nitrogen. When Co2+ is bound to Nδ1, two meta protons at CδH and Nε2H would be apparent.36a For a bidentate binding through the delta nitrogen four signals would be apparent in this region, as shown in the Figure

4.13. Conversely for the peptoid, only two signals appear in this region at 65.8 and 44.6 ppm (bottom spectrum). The appearance of only two sharp signals likely indicates a different binding mode through the Nε2. In the Nε2 metal coordination only the meta Nδ1 can be observed. Potentially, the ortho signals for the peptoid might also be apparent at

95.8 ppm given its far downfield shift and its broadness.

a A caution to the reader: older texts could lead to confusion based on how some label histidine. The approach I use is a more current labeling scheme, shown in Figure 4.13, though, I do deviate from proper labeling of the peptoid with an additional carbon on the side-chain to allow for easier comparison for both myself and the reader. Papers by very well-respected chemists, like Ivano Bertini, Ph.D., are still very much correct in their assignment, though, the scheme has been updated.

206 Figure 4.13 – overlaid spectra for 4 mM Ac-FHFH (top) and γ-AA-FHFH (bottom) and 1 stoichiometric equivalent of Co2+, 50× zoom applied to γ-AA-FHFH far downfield region signal.

In support for the tentative assignment of the peptoid bound to the ε2-nitrogen, a drop of deuterated water was added to the NMR tube. Hydrogens on imidazole nitrogens are prone to solvent exchange. Exchange from N-H to N-D would remove any

N-H signals. Figure 4.14 shows a drastic decrease in any signal intensity for both of the apparent signals, confirming the Nε2-H coordination and signal assignment for the two

Nδ1 protons.

207 Figure 4.14 – drop of deurated water added (top) to Ac-FHFH peptoid to confirm NH signal assignment

The cause for the differences in the binding mode for the peptide and γ-

AApeptide histidine residues, likely lies in the difference in the number of rotational carbons for each residue. Each peptoid histidine residue contains three rotatable CH2 moieties over only 2 on the peptide. The extra rotation seems to play a role allowing for different stable coordination site between binding motifs.

In many cases, solvation can also play a role in the folding and binding around metals.

In fact, solvation is often essential for protein functionality.37,38

208 Figure 4.15 – Line representation Ac-γ-FHFH binding Co2+ through Nε2

To determine how solvation or solvent polarity might be driving the influence of the preferred binding motifs, d3-MeOH was added to each compound as shown in

Figures 4.16 and 4.17. In these experiments water was avoided due to the general hydrophobicity of the phenylalanine residues.

In a comparison of the peptide and the peptoid, both show differences upon addition of d3-MeOH. The peptide showed much more change in the chemical shifts.

Changes in chemical shift and intensity of two of the signals (using the diamagnetic region as an intensity reference) decrease in intensity. This could potentially indicate a change structural conformation or a potential weakening of the binding between the metal and Nε.

209 Figure 4.16 - 4 mM FHFH in d6-DMSO (top) and half d6-DMSO, half d3-MeOH (bottom)

Figure 4.17 – 4 mM γ-AA-FHFH peptoid in d6-DMSO (top) and half d6-DMSO, half d3-

MeOH (bottom)

In comparison, the peptoid chemical shifts show a general downfield shift while approximately maintaining chemical shift difference. The more downfield shifted 210 signals likely indicate a closer proximity to the metal center as the unpaired electron from Co2+ influences the Nδ-H chemical shift more. Solvation in this case may not be competing with imidazole binding or change in polarity of the solvent forces tighter binding.

As peptoid-metal reactivity is still a mostly untapped resource in medicine, the reactivity of the peptoid was also observed. In collaboration on this project with

Christian Tang, Ph.D., 1:1 Cu2+ and γ-AApeptoid were incubated with DTBC to determine the redox activity of the metal center. Much like many of the previous Cu2+ redox centers, Cu2+ mediated oxidation of DTBC via O2 and H2O2 pathways yielding intrinsic binding constants, Km,int, of 0.17 ± 0.02 mM and 23.5 mM towards DTBC and

H2O2, respectively. Secondary analysis using Hanes plots also revealed cooperativity between both DTBC and H2O2 substrates.39 In terms of medicinal properties, this type of reactivity would require tight regulation to avoid unnecessary and unwanted redox activity. Scaffolds, as shown in the first half of this study, could provide more regulated control of the active site and potentially a target.

III. Concluding Remarks

The progress and sophistication of the peptidomimetic field has had a massive growth over the last 30 years, though. The advent of many chemical techniques and solid-phase syntheses have made once drugs impossible to produce, possible. Despite

211

the rapid growth in the field, we are still learning and catching up to what nature had millions of years to develop. Complex structures and intricate reaction pathways can take years to develop. In developing novel methods for disease treatment, metals are often overlooked. As the methodologies for peptidomimetics develop, metals could play a role just as they do in nature: structural and/or reactive sites. The need for understanding the role metal might play in reactivity is essential as shown in the second half of the study. As the field develops more, three-dimensional folding motifs will be essential for driving and modulating reactivity.

IV. Materials and Methods

4.1 D-sulfono-γ-AApeptide/L-amino acid

Synthesis and purification of oligomers were performed by Teng, et al from the aforementioned publication.1 NMR studies of oligomer 5 was performed using 5 mM sample in d3-MeOH.

All NMR spectra were obtained on Varian 600 MHz magnet with an Inova600 console using VNMRJ to process all data using a WET pulse sequence to eliminate residual solvent signal. All spectra were acquired with a 7k Hz spectral window using a

90° pulse width of 16 μs. The 1D spectrum was collected using 24k points with a 1.5 sec delay for 32 scans. The f2×f1 dimensions were collected for 8k×8k points. The

212 gDQFCOSY experiment was performed for 500 increments, 16 scans per increment, using applying a 1.0 ms gradient pulse and 2.0 ms steady state pulses. zTOCSY was performed using a mixing time of 80 ms with a DIPSI2 pulse pattern, 2.0 ms steady state gradient pulses, and zero quantum filters calculated “on-the-fly” by VNMRJ at 30 and

40 ms. 32 scans were performed for 256 increments. To best visualize long-range NOE coupling, 32 scans were performed for 500 increments with a mixing time of 200 ms. A

30 ms ZQ filter was used with a 2.0 ms steady state gradient pulse.

4.2 Peptide/Peptoid Ac-FHFH

Paramagnetic NMR spectra were performed on a Varian 600 MHz magnet with an Inova console. All data were processed using VNMRJ 4.2. The superWEFT pulse sequence was employed to minimize diamagnetic contribution to the receiver, allowing for higher signal sensitivity of paramagnetic signals. The superWEFT pulse sequence is a two-pulse sequence where first pulse is optimized for a 180°-90° tilt angle pulse, followed by a short delay to allow paramagnetic proton relaxation, and finally a final

90° pulse (180°-τ-90°). Total sequence recycle times were 140 ms. A 130k Hz spectral window was acquired using 16k data acquisition points. All samples were solvated in d6-DMSO or d3-MeOH. Addition of water to determine N-H solvent exchangeability was performed using 18.0 MΩ deionized water.

213 V. References

1 Teng, P, Ma, N., Cerrato, D.C., She, F., Odom, T., Wang, X., Ming, L.-J., van der Vaart,

A., Wojtas, L., Xu, H., and Cai, J. Right-handed helical foldamers consisting of de

novo d-AApeptides. Journal of the American Chemical Society. 2017.

139(21)7363-7369. Doi: 10.1021/jacs.7b03007.

2 Antibiotic Resistance Threats in the United State, 2013; Centers for Disease Control.

U.S. Department of Health and Human Services. Atlanta, GA, 2013.

3 Bliven, K.A., and Maurelli, A.T. Evolution of bacterial pathogens within the human

host. Microbiology Spectrum. 2016. 4(1)1301-1313. Doi:

10.1128/microbiolspec.VMBF-0017-2015

4 Hede, K. Antibiotic resistance: An infectious arms race. Nature. 2014. 509:S2-S3. Doi:

10.1038/509S2a

5 Basset, E.J., Keith, M.S., Armelagos, G.J., Martin, D.L., and Villanueva, A.R.

Tetracycline-labeled human bone from ancient Sudanese Nubia (A.D. 350).

Science. 1980. 209(4464)1532-1534.

6 Antibiotic Resistance. World Health Organization. Geneva, Switzerland, 2018. [Online]

https://www.who.int/news-room/fact-sheets/detail/antimicrobial-resistance

7 Worthington, R.J., and Melander, C. Combination approaches to combat multi-drug

resistant bacteria. Trends in Biotechnology. 2013. 31(3)177-184. Doi:

10.1016/j.tibtech.2012.12.006. 214

8 Gajdacs, M. The concept of an ideal antibiotic: implications for drug design. Molecules.

2019. 24(5)892-908.

9 Mendez-Samperio, P. Peptidomimetics as a new generation of antimicrobial agents:

current progress. Infection and Drug Resistance. 2014. 7:229-237. Doi:

10.2147/IDR.S49229

10 Cheng, R.P., Gellman, S.H., and DeGrado, W.F. β-peptides: From structure to

function. JACS: Chemical Review. 2001. 101(10)3219-3232. Doi: 10.1021/cr000045i.

11 Chongsiriwatana, N.P, Patch, J.A., Czyzewski, A.M., Dohm, M.T., Ivankin, A.,

Gidalevitz, D., Zuckermann, R.N., and Barron, A.E. Peptoids that mimic

structure, function, and mechanism of helical antimicrobial peptides. Proceedings

of the National Academy of Sciences. 2008. 105(8)2794-2799. Doi:

10.1073/pnas.0708254105.

12 Hook, D.F., Bindschädler, P., Mahajan, Y.R., Ŝebesta, R., Kast, P., and Seebach, D. The

proteolytic stability of ‘designed’ β-peptides containing α-peptide-bond mimics

and of mixed α,β -peptides: application to the construction of MHC-binding

peptides. Chemistry and Biodiversity. 2005. 2(5)591-632. Doi:

10.1002/cbdv.200590039.

13 Kirshenbaum, K., Barron, A.E., Goldsmith, R.A., Armand, P., Bradley, E.K., Truong,

K.T.V., Dill, K.A., Cohen, F.E., and Zuckerman, R.N. Sequence-specific

polypeptoids: A diverse family of heteropolymers with stable secondary

structure. Proceedings of the National Academy of Sciences. 1998. 95(8)4303-4308. 215

Doi: 10.1073/pnas.95.8.4303

14 Kim, J.-K., Lee, S.-A., Shin, S., Lee, J.-Y., Jeong, K.-W., Nan, Y.H., Park, Y.S., Shin, S.Y.,

and Kim, Y. Structural flexiblility and the positive charges are the key factors in

bacterial cell selectivity and membrane penetration of peptoid-substituted analog

of Piscidin 1. Biochimica et Biophysica Acta (BBA) – Biomembranes. 2010.

1798(10)1913-1925. Doi: 10.1016/j.bbamem.2010.06.026

15 Niu, Y., Wu, H., Li, Y., Hu, Y., Padhee, S., Li, Q., Cao, C., Cai, J. AApeptides as a new

class of antimicrobial agents. Organic & Biomolecular Chemistry. 2013. 11:4283-

4290. Doi: 10.1039/C3OB40444G

16 Gadermann, K., Ernst, M., Hoyer, D., and Seebach, D. Synthesis and biological

evaluation of a cyclo-β-tetrapeptide as a somatostatin analogue. Angewandte

Chemie. 1999. 38(9)1223-1226. Doi: 10.1002/(SICI)1521-

3773(19990503)38:9<1223::AID-ANIE1223>3.0.CO;2-A

17 A study to evaluate safety, tolerability and efficacy of Lytixar™ (LTX-109) on uncomplicated

Gram-positive, skin infection. U.S. National Library of Medicine, U.S. National

Institutes of Health. Washington, D.C. 2019.

18 Study of the effects of brilacidin oral rinse on radiation-induced oral mucositis in patients with

head and neck cancer. U.S. National Library of Medicine, U.S. National Institutes of

Health. Washington, D.C. 2019.

19 Choi, S., Isaacs, A., Clements, D., Liu, D., Kim, H., Scott, R.W., Winkler, J.D., and

DeGrado, W.F. De novo design and in vivo activity of conformationally 216

restrained antimicrobial arylamide foldamers. Proceedings of the National Academy

of Sciences. 2009. 106(17)6968-6973. Doi: 10.1073/pnas.0811818106

20 Cohen, S.M. A Bioinorganic approach to fragment-based drug discovery targeting

metalloenzymes. Accounts of Chemical Research. 2017. 50(8)2007-2016. Doi:

10.1021/acs.accounts.7b00242

21 Mohan, A., Koh, A.H.M., Gate, G., Calkins, A.K., McComas, K.N., and Fuller, A.A.

Solid-phase synthesis of azole-comprising peptidomimetics and coordination of

a designed analog to Zn2+. Molecules. 2018. 23(5)1035-1047. Doi:

10.3390/molecules23051035

22 Baskin, M., Zhu, H., Qu, Z.-W., Chill, J.H., Grimme, S., and Maayan, G. Folding of

unstructured peptoids and formation of hetero-bimetrallic peptoid complexes

upon side-chain-to-metal coordination. Chemical Science. 2019. 10:620-632. Doi:

10.1039/C8SC03616K

23 Nguyen, A.I., Spencer, R.K., Anderson, C.L., and Zuckermann, R.N. A bio-inspired

approach to ligand design: folding single-chain peptoids to chelate a

multimetallic cluster. Chemical Science. 2018. 9:8806-8813. Doi:

10.1039/C8SC04240C

24 Dragulescu-Andrasi, A., Rapireddy, S., Frezza, B.M., Garathri, C., Gil, R.R., and Ly,

D.H. A simple γ-backbond modification preorganizes peptide nucleic acid into a

helical structure. Journal of the American Chemical Society. 128(31)10258-10267. Doi:

10.1021/ja0625576 217

25 Wu, H., Qiao, Q., Hu, Y., Teng, P., Gao, W., Zuo, X., Wojtas, L., Larsen, R.W., and Cai,

J. Sulfono-γ-AApeptides as a new class of nonnatural helical foldamer. Chemistry

– A European Journal. 2015. 21: 2501-2507. Doi: 10.1002/chem.201406112.

26 Hu, Y., Li, X., Sebti, S.M., Chen, J., and Cai, J. Design and synthesis of AApeptides: A

new class of peptide mimics. Bioorganic & Medicinal Chemistry Letters. 2011.

21:1469-1471. Doi: 10.1016.j.bmcl.2011.01.005

27 Shi, Y., Teng, P., Sang, P., She, F., Wei, L., and Cai, J. γ-AApeptides: Design, structure,

and application. Accounts of Chemical Research. 2016. 49: 428-411. Doi:

10.1021/acs.accounts.5b00492.

28 Simon, R.J., Kania, R.S., Zuckerman, R.N., Huebner, V.D., Jewell, D.A., Banville, S.,

Ng, S., Wang, L., Rosenberg, S., and Marlowe, C.K. Peptoids: a modular

approach to drug discovery. Proceedings of the National Academy of Sciences. 1992.

89(20)9367-9371. Doi: 10.1073.pnas.89.20.9367

29 Miller, S., Simon, R.J., Ng, S., Zuckerman, N., Kerr, J.M., and Moos, W.H. Comparison

of proteolytic susceptibilities of homologuous L-amino acid, D-amino acid, and

N-substituted glycine peptide and peptoid oligomers. Drug Development Research.

1995. 35:20-32.

30 Miller, S.M., Simon, R.J., Ng, S., Zuckermann, R.N., Kerr, J.M., and Moos, W.H.

Proteolytic studies of homologous peptide and N-substituted glycine peptoid

oligomers. Bioorganic & Medicinal Chemistry Letters. 1994. 4(22)2657-2662. Doi:

10.1016/S0960-894X(01)80691-0. 218

31 Yang, W., Kang, B., Voelz, V.A., and Seo, J. Control of porphyrin interactions via

structural changes of a peptoid scaffold. Organic & Biomolecular Chemistry. 2017.

15:9670-9679. Doi: 10.1039/C7Ob02398G.

32 Kang, B., Wang, W., Lee, S., Mukherjee, S., Forstater, J., Kim, H., Goh, B., Kim, T.-Y.,

Voelz, V.A., Pang, Y., and Seo, J. Precisely tuneable energy transfer system using

peptoid helix-based molecular scaffold. Scientific Reports. 2017. 7:4768. Doi:

10.1038/s41598-017-04727-0.

33 Wolf, L.M., Servoss, S.L., and Moss, M.A. Peptoids: emerging therapeutics for

neurodegeneration. Journal of Neurology Neuromedicine. 2017. 2(7)1-5.

34 Renault, J.P., Beaur-Verchere, C., Morgenstern-Badarau, I., and Piccili, M.

Paramagnetic NMR spectroscopy of native and cobalt substituted manganese

superoxide dismutase from Escherichia coli. FEBS Letters. 1997. 401(1)15-19. Doi:

10.1016/S0014-5793(96)31420-2.

35 Romero, C., Moratal, J.M., and Donaire, A. Metal coordination of azurin in the

unfolded state. FEBS Letters. 1998. 440(1-2):93-98. Doi: 10.1016/S0014-

5793(98)01438-0.

36 Banci, L., Bertini, I., Luchinat, C., and Viezzoli, M.S. A comment on the 1H NMR

spectra of Cobalt(II)-substituted superoxide dismutases with histidines

deuterated in the ε1-position. Inorganic Chemistry. 1990. 29(7)1438-1440. Doi:

10.1021/ic00332a031.

37 Schein, C.H. Solubility as a function of protein structure and solvent components. 219 Nature Biotechnology. 1990. 8:308-317.

38 Capdevila, D.A., Edmonds, K.A., Campanello, G.C., Wu, H., Gonzalez-Gutierrez, G.,

and Giedroc, D.P. Functional role of solvent entropy and conformational entropy

of metal binding in a dynamically driven allosteric system. Journal of the American

Chemical Society. 2018. 140(29)9108-9119. Doi: 10.1021/jacs.8b02129

39 Tang, C.T. Structure and Activity of Metallo-Peptides. Ph.D. Dissertation, University

of South Florida, Tampa, FL, 2018.

220 APPENDIX A: PERMISSIONS

Reprinted from Journal of Inorganic Biochemistry, Vol 91, Issue 1, Metal binding and structure–activity relationship of the metalloantibiotic peptide bacitracin, Pages 46-58, Copyright 2019, with permission from Elsevier

Adapted with permission from 1H NMR, Mechanism, and Mononuclear Oxidative Activity of the Antibiotic Metallopeptide Bacitracin: The Role of d-Glu-4, Interaction with Pyrophosphate Moiety, DNA Binding and Cleavage, and Bioactivity. Copyright 2019. American Chemical Society.

Used with permission of Journal of Biological Chemistry, from Catechol oxidase-like oxidation chemistry of the 1-20 and 1-16 fragments of Alzheimer’s disease- related β-amyloid peptide, Giordano F.Z. da Silva, Ph.D., William Tay, Ph.D., and Li-June Ming, Ph.D. 280, 17, 2005.

221