<%BANNER%>

The relative signs of coupling constants in fluorinated cyclopropanes and the aggregation of amyloid-beta (25-35) peptide.

Permanent Link: http://ufdc.ufl.edu/UFE0041562/00001

Material Information

Title: The relative signs of coupling constants in fluorinated cyclopropanes and the aggregation of amyloid-beta (25-35) peptide.
Physical Description: 1 online resource (103 p.)
Language: english
Creator: Richardson, David
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: 19f, amyloid, cyclopropane, diffusion, nmr
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE RELATIVE SIGNS OF COUPLING CONSTANTS IN FLUORINATED CYCLOPROPANES AND THE AGGREGATION OF AMYLOID-BETA (25-35) PEPTIDE By David W. Richardson May 2010 Chair: Clifford R. Bowers Major: Chemistry Cyclopropanes present interesting molecules to study due to their unique bond angles and strain in the three-membered ring. Protonated cyclopropanes have been extensively studied using nuclear magnetic resonance (NMR), but little work has been done with their fluorinated analogues. Double-resonance, or spin-tickling, NMR was used to determine the signs of three-bond vicinal fluorine-fluorine coupling constants and their relationship with temperature. Certain molecules were shown to conflict with the previously published relationship between the sign of the coupling constants in cyclopropanes and their temperature dependence. Amyloid-beta peptide (A-b) is the main component of plaques found in the brains of patients with Alzheimer?s disease. A-b is a soluble 39 to 42 residue peptide resulting from the proteolytic cleavage of the amyloid precursor protein. As a result of conformational changes, A-b transforms from independent soluble monomers to insoluble, plaque-forming oligomers. This transformation is assumed to lead to the deleterious effects of Alzheimer?s disease (AD). Some debate remains as to whether the symptoms associated with AD are a direct result of the plaque formation, or if the aggregation of A-b is a secondary result. However, there is no question that the brains of patients who display signs of AD at time of death contain abnormally high amounts of A-b plaque. NMR was shown to be a useful analytical tool for the in vitro analysis of A-b. Pulsed-field-gradient NMR (pfg-NMR) is particularly useful as it allows for the determination of diffusion coefficients. By subjecting the sample to a gradient magnetic field, the nuclei of an analyte can be spatially marked along an axis. Any movement along that axis results in signal attenuation. The signal attenuation was plotted against an array of the gradient strength and diffusion coefficients were calculated. This work involved altering sample conditions and measuring the diffusion of A-b with the intent of finding a trigger for the aggregation of the peptide. The diffusion of Ab(25-35), an 11 residue fragment of the A-b peptide, was examined over a range of temperatures, pH, and in solutions with common biological metal ions.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by David Richardson.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Bowers, Clifford R.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041562:00001

Permanent Link: http://ufdc.ufl.edu/UFE0041562/00001

Material Information

Title: The relative signs of coupling constants in fluorinated cyclopropanes and the aggregation of amyloid-beta (25-35) peptide.
Physical Description: 1 online resource (103 p.)
Language: english
Creator: Richardson, David
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: 19f, amyloid, cyclopropane, diffusion, nmr
Chemistry -- Dissertations, Academic -- UF
Genre: Chemistry thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE RELATIVE SIGNS OF COUPLING CONSTANTS IN FLUORINATED CYCLOPROPANES AND THE AGGREGATION OF AMYLOID-BETA (25-35) PEPTIDE By David W. Richardson May 2010 Chair: Clifford R. Bowers Major: Chemistry Cyclopropanes present interesting molecules to study due to their unique bond angles and strain in the three-membered ring. Protonated cyclopropanes have been extensively studied using nuclear magnetic resonance (NMR), but little work has been done with their fluorinated analogues. Double-resonance, or spin-tickling, NMR was used to determine the signs of three-bond vicinal fluorine-fluorine coupling constants and their relationship with temperature. Certain molecules were shown to conflict with the previously published relationship between the sign of the coupling constants in cyclopropanes and their temperature dependence. Amyloid-beta peptide (A-b) is the main component of plaques found in the brains of patients with Alzheimer?s disease. A-b is a soluble 39 to 42 residue peptide resulting from the proteolytic cleavage of the amyloid precursor protein. As a result of conformational changes, A-b transforms from independent soluble monomers to insoluble, plaque-forming oligomers. This transformation is assumed to lead to the deleterious effects of Alzheimer?s disease (AD). Some debate remains as to whether the symptoms associated with AD are a direct result of the plaque formation, or if the aggregation of A-b is a secondary result. However, there is no question that the brains of patients who display signs of AD at time of death contain abnormally high amounts of A-b plaque. NMR was shown to be a useful analytical tool for the in vitro analysis of A-b. Pulsed-field-gradient NMR (pfg-NMR) is particularly useful as it allows for the determination of diffusion coefficients. By subjecting the sample to a gradient magnetic field, the nuclei of an analyte can be spatially marked along an axis. Any movement along that axis results in signal attenuation. The signal attenuation was plotted against an array of the gradient strength and diffusion coefficients were calculated. This work involved altering sample conditions and measuring the diffusion of A-b with the intent of finding a trigger for the aggregation of the peptide. The diffusion of Ab(25-35), an 11 residue fragment of the A-b peptide, was examined over a range of temperatures, pH, and in solutions with common biological metal ions.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by David Richardson.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Bowers, Clifford R.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041562:00001


This item has the following downloads:


Full Text

PAGE 1

1 THE RELATIVE SIGNS OF COUPLING CONSTANTS IN FLUORINATED CYCLOPROPANES AND THE AGGREGATION OF AMYLOID-BETA (25-35) PEPTIDE By DAVID W. RICHARDSON A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010

PAGE 2

2 2010 David Richardson

PAGE 3

3 To my loving wife and beautiful daughter

PAGE 4

4 ACKNOWLEDGMENTS I would like to express my utmost gratitude to my mentor, Dr. Wallace Brey, for his guidance and endless patience. Through my academic car eer he taught me NMR from the ground up. The innumerable conversation s about both the history and principles of NMR, and science and life in general, were inva luable to my growth as a scientist and person. I would like to thank my friends and co lleagues in the NMR facility, Dr. Ion Ghiviriga and Robert Harker. They furthered my knowledge of advanced NMR techniques and the function and maintenance of NMR spectrometers and groomed me for my future as a spectrosc opist. I would also like to t hank my lab mate Dr. Bethany Bechtel for often serving as a sounding board for my research thoughts and for providing ideas when I was at a loss. I would like to acknowledge the members of my committee, Dr. Richard Yost, Dr. David Powell, and Dr. Joanna Long, for their patience and encouragement and for answering my many questions. I especially want to thank Dr. C. R. Bowers for his assistance in the final stages of editing this document. I am thankful for my fellow graduate students. For t heir encouragement, assistance, and friendship. I especially want to thank Dr. Frank Kero, Dr. John Bowden, Dr. Jennifer Bryant, Dr. Joshua Caldwell, Daniel Magparangalan, and Mike Napolitano. I would like to thank all the members of my family, both in blood and through marriage, for their support. I especially want to thank my parents. Throughout my life they have blessed me with endless love and support and have sacrificed more for me than I could ever hope to express in words.

PAGE 5

5 I thank my daughter, Audrey, for always greeting me with a smile after a long day of work. I owe an eternal debt to my wife and best friend Samantha. Her companionship, understanding, and encouragement are immeasurable. Without her support this would not have been possible. Lastly, I would like to thank God for the many blessings.

PAGE 6

6 TABLE OF CONTENTS page ACKNOWLEDG MENTS .................................................................................................. 4 LIST OF TABLES............................................................................................................ 8 LIST OF FIGURES .......................................................................................................... 9 ABSTRACT................................................................................................................... 12 CHA PTER 1 INTRODUCTION.................................................................................................... 14 Overview of Nuclear Magnetic Resonance ............................................................. 14 Spin-Spin Coupling ........................................................................................... 15 Double-Res onance NMR.................................................................................. 16 Chemical Shift .................................................................................................. 17 Pulsed-Field-Gr adient NMR ............................................................................. 18 Cyclopro panes ........................................................................................................ 19 Amyloid................................................................................................................... 20 2 ANALYSIS OF THE RELATIVE SIGNS OF COUPLING CONSTANTS IN FLUORINATED CY CLOPRO PANES ..................................................................... 28 Introduc tion ............................................................................................................. 28 Experim ental ........................................................................................................... 30 Results and Discussion........................................................................................... 31 Temperature Dependence of Coupling Constants ........................................... 31 Signs of Coupling Const ants ............................................................................ 31 3 ANALYSIS OF THE AGGREGATION OF AMYLOID BET A (25-35)....................... 44 Introduc tion ............................................................................................................. 44 Experim ental ........................................................................................................... 44 Gradient Calibration ......................................................................................... 46 Temperature Regulat ion ................................................................................... 47 Chemical Shift Referenc ing .............................................................................. 48 Results and Discussion........................................................................................... 48 Chemical Shift Eval uation and As signm ent ...................................................... 48 Assignment of Dup licate Re sidues ................................................................... 50 Chemical Shi ft Results ..................................................................................... 52 Effect of TFE.............................................................................................. 53 Effect of metal ions.................................................................................... 55 Effect of te mperature ................................................................................. 57 Diffusion Re sults .............................................................................................. 58

PAGE 7

7 4 CONCLUSION........................................................................................................ 94 LIST OF RE FERENCES ............................................................................................... 96 BIOGRAPHICAL SKETCH .......................................................................................... 103

PAGE 8

8 LIST OF TABLES Table page 2-1 The change in magnitude of the vicinal fluorine-fl uorine coupling constants (Hz) with a change in te mperature from -20oC to 25oC....................................... 422-2 List of the connected transitions for CF2-CFCl-CF(OCCl3)................................. 432-3 The sign and magnitude of the vicinal fluorine-fluorine coupling constants (Hz) in fluori nated cyclopr opanes. ...................................................................... 433-1 Chemical shifts (ppm) of 6.71 mM A (25-35) in 90% H2O and 10% D2O at pH 2.53.................................................................................................................... 883-2 Chemical shifts (ppm) of 3.47 mM A (25-35) in 90% H2O and 10% D2O at pH 2.60.................................................................................................................... 883-3 Chemical shifts (ppm) of 0.69 mM A (25-35) in 80% H2O and 20% d3-TFE at pH 3. 23............................................................................................................... 893-4 Chemical shifts (ppm) of 0.69 mM A (25-35) in 80% H2O and 20% d3-TFE at pH 3.19 with 1:1 molar ratio of Zn2+.................................................................... 893-5 Chemical shifts (ppm) of 0.60 mM A (25-35) in 80% H2O and 20% d3-TFE at pH 3.24 with 3:1 molar ratio of Zn2+.................................................................... 903-6 Chemical shifts (ppm) of 0.69 mM A (25-35) in 80% H2O and 20% d3-TFE at pH 3.23 with 0.1:1 molar ratio of Cu2+................................................................ 903-7 Chemical shifts (ppm) of 0.69 mM A (25-35) in 80% H2O and 20% d3-TFE at pH 3.08 with 1:1 molar ratio of Cu2+................................................................... 913-8 Temperature dependence of the amide chemical shifts of 2.82 mM A (25-35) in 90% H2O and 10% D2O.................................................................................. 913-9 The diffusion coefficients of three concentrations of A (25-35) in 90% H2O and 10% D2O from 5-55oC................................................................................. 923-10 The diffusion coefficients of two pH values of 2.82 mM A (25-35) in 90% H2O and 10% D2O from 5-70oC................................................................................. 923-11 The diffusion coefficients 2.82 mM A (25-35) in 90% H2O and 10% D2O with and without zinc from 5-55oC.............................................................................. 93

PAGE 9

9 LIST OF FIGURES Figure page 1-1 Nuclear and electron spins of two spinnuclei. .............................................. 23 1-2 An example of the effects of a spin-tickling experiment. 23 1-3 The pfg-se pulse s equence. ............................................................................... 24 1-4 The pfg-ste pulse s equence. .............................................................................. 25 1-5 Representativ e cyclopr opanes. ........................................................................ 25 1-6 The amyloid precursor protein and the positions at which the three enzymes can cleave the protein to produce amyl oid bet a. ................................................ 26 1-7 The sequence of am yloid beta (25-35). .............................................................. 27 2-1 A theoretical energy-level diagra m and corresponding spectrum for the transitions of a two-spin firs t-o rder system wit h no coup ling............................... 34 2-2 A theoretical energy-level diagra m and corresponding spectrum for the transitions of a two-spin fi rst-o rder system with coupl ing.................................... 35 2-3 A theoretical energy-level di agram for a thr ee-spin system. ............................... 36 2-4 The effect of the sign of the coupling constants in a three-spin AMX system on the theoretical spectr um. ............................................................................. 37 2-5 The difference between progressively and regressively connected transitions and their effect on the spectrum o f a double-irradiat ion experiment................... 38 2-6 The numbering scheme assigned to t he substituent positions of the cyclopropa ne. ..................................................................................................... 39 2-7 The effect of a spin-t ickling experiment on the 19F spectrum of CF2-CFClCF(OCCl3).......................................................................................................... 40 2-8 The possible sets of coupling-cons tant signs and their corresponding energylevel d iagrams for CF2-CFCl-CF(OCCl3). ......................................................... 41 3-1 Amplitude of 10 mg/mL lysozyme signal for three spectral regions over a range of gradient strengths. ................................................................................ 66 3-2 Amplitude of 10 mg/mL BSA signal for th ree spectral regions over a range of gradient st rengths. .............................................................................................. 67

PAGE 10

10 3-3 Diagram of external variable-temperature unit used to regulate the sample temperat ur e........................................................................................................ 68 3-4 1H spectrum of 0.69 mM A (25-35) in 80% H2O and 20% d3-TFE at 10oC referenced to inte rnal D SS. .............................................................................. 69 3-5 TOCSY spectrum of 6.71 mM A (25-35) in 90% H2O and 10% D2O at 25oC. .. 70 3-6 Dependence of the homonuclear nuclear Overhauser enhancement ( ) on the spectrometer frequency ( o) and correlation time (c) for A) NOESY and B) RO ESY. ......................................................................................................... 71 3-7 Expansion of the ROE SY spectrum of 0.69 mM A (25-35) in 80% H2O and 20% d3-TFE........................................................................................................ 72 3-8 Expansion of the ROE SY spectrum of 6.71 mM A (25-35) in 90% H2O and 10% D2O. .......................................................................................................... 73 3-9 plot for the alpha protons of 6.71 mM A (25-35) in 90% H2O and 10% D2O at pH 2.53 compared to ran dom-coil chemic al shifts.................................. 74 3-10 plot for the alpha protons of 0.69 mM A (25-35) in 80% H2O and 20% d3TFE at pH 3.23 compared to ra ndom-coil chemic al shifts.................................. 75 3-11 Two visibly cloudy A (25-35) samples and one clear sa mple. ........................... 76 3-12 The equilibrium of A ( 25-35) in solution, and the effect of each transition. ...... 77 3-13 Two spectra showing the decrease in A (25-35) signal intensity of the samples with visi ble aggr egation. ....................................................................... 78 3-14 The chemical shifts of the amide protons of 1.88 mM A (25-35) in 90% H2O and 10% D2O over a range of temperatur es. .................................................... 79 3-15 Diffusion of lysozyme over a range of temperatures. ........................................ 80 3-16 Typical results from a pfg ex per iment. .............................................................. 81 3-17 Diffusion of 2.82 mM A (25-35) over a range of temperatures with linear regression applied over all temper atures (yellow) and from 25-70oC (red). ...... 82 3-18 The diffusion coefficient of A (25-35) for three concentrations in 90% H2O and 10% D2O from 5-55oC. .............................................................................. 83 3-19 The diffusion coefficient of 2.82 mM A (25-35) with a change in pH over a range of te mper atures. ..................................................................................... 84

PAGE 11

11 3-20 The diffusion coefficient of A (25-35) with the addition of 3:1 molar ratio Zn(II) over a range of temper atures. ................................................................... 85 3-21 The diffusion coefficient of A (25-35) at 25oC over a large range of concentrati ons. ................................................................................................. 86 3-22 The calculated line widths over a range of degrees of oligomerization for three geometri c model s. ..................................................................................... 87

PAGE 12

12 Abstract of Dissertation Pr esented to the Graduate School of the University of Florida in Partial Fulf illment of the Requirements for t he Degree of Doctor of Philosophy THE RELATIVE SIGNS OF COUPLING CONSTANTS IN FLUORINATED CYCLOPROPANES AND THE AGGREGATION OF AMYLOID-BETA (25-35) PEPTIDE By David W. Richardson May 2010 Chair: Clifford R. Bowers Major: Chemistry Cyclopropanes present interesting molecu les to study due to their unique bond angles and strain in the three-member ed ring. Protonated cyclopropanes have been extensively studied using nuc lear magnetic resonance (NMR), but little work has been done with their fluorinated analogues. Double-re sonance, or spin-tickling, NMR was used to determine the signs of three-bond vi cinal fluorine-fluorine coupling constants and their relationship with temper ature. Certain molecules were shown to conflict with the previously published relationship betwe en the sign of the c oupling constants in cyclopropanes and their temperature dependence. Amyloidpeptide (A ) is the main component of plaques found in the brains of patients with Alzheimers disease. A is a soluble 39 to 42 residue peptide resulting from the proteolytic cleavage of the amyloid precursor pr otein. As a result of conformational changes, A transforms from independent soluble monomers to insoluble, plaque-forming oligom ers. This transformation is assumed to lead to the deleterious effects of Alzheimers disease (AD). Some debate remains as to whether the symptoms associated with AD are a direct result of the plaque formation, or if the

PAGE 13

13 aggregation of A is a secondary result. However, t here is no question that the brains of patients who display signs of AD at time of death c ontain abnormally high amounts of A plaque. NMR was shown to be a useful analytical tool for the in vitro analysis of A Pulsed-field-gradient NMR (pfg-NMR) is parti cularly useful as it allows for the determination of diffusion coeffi cients. By subjecting the sa mple to a gradient magnetic field, the nuclei of an analyte can be spat ially marked along an axis. Any movement along that axis results in signal attenuation. The signal attenuation was plotted against an array of the gradient strengt h and diffusion coefficients were calculated. This work involved altering sample conditions and measuring the diffusion of A with the intent of finding a trigger for the aggregation of the peptide. The diffusion of A (25-35), an 11 residue fragment of the A peptide, was examined over a range of temperatures, pH, and in solutions with common biological metal ions.

PAGE 14

14 CHAPTER 1 INTRODUCTION Overview of Nuclear Magnetic Resonance Nuclei containing an odd number of prot ons or neutrons behave as though they are spinning and possess an angular momentum. The concept of nuclear spin angular momentum is rather abstract. It i s a property of quantum mechanics and has no macroscopic corollary. The existence of nuc lear spin was first suggested by Pauli in 19241 and later confirmed experimentally by Rabi in 19392. The math describing nuclear spin angular momentum is similar to that for rotational angul ar momentum, but it is incorrect to imagine nuclei spinning about their own axis3. Nuclear spin angular momentum, p is quantized in units of the reduced Planck constant, 2 Ih pI (1-1) The nuclear spin quantum number, I is nucleus-dependent and is either an integer or half-integer. Nuclei which have a non-ze ro nuclear spin quantum number possess a magnetic moment, p (1-2) The magnetogyric ratio, is a nucleus-dependent constant and is typically expressed in units of rads-1T-1. Nuclei with a nuclear spin of zero, notably 12C and 16O, are not observable by nuclear magnetic resonance (NMR) techniques as they have no angular momentum and thus no magnetic moment. A nucleus has 2 I + 1 orientations relative to an ar bitrary axis. These orientations are often referred to as spin states, m where: ,1,...,1, mIIII (1-3)

PAGE 15

15 Absent an external magnetic field, these spin states are of equal energy and an equal distribution of nuclei exists across all spin st ates. In the presence of a magnetic field, however, the spin states split in energy with: EmB (1-4) where B is the applied field, usually expressed in tesla (T). Following Eq. 1-3, a spinnucleus has two possible spin states, + and In terms of NMR, these spin states are often referred to as paralle l and anti-parallel, or spin-up and spin-down, relative to an applied magnetic field. Spin-Spin Coupling The presence of different spin states leads to a phenomenon known as spin-spin coupling. Ramsey and Purcell first suggest ed that information abo ut the spin of one nucleus is transferred to a neighbor ing nucleus via the bonding electrons4. Consider the case of two spinnuclei (Figure 1-1). The Pauli exclusion prin ciple dictates that the electrons must have opposite spins, so it follows that two st ates exist for a twonucleus system. In one, the nuclei are of opposite spins, leading to energetically favorable interactions (Figure 1-1A), and in the other the nuclei are of equal spins and one is the same spin as one of the elec trons (Figure 1-1B). When one nucleus undergoes resonance, or changes spin states the energy of the transition depends on the orientation of the other nucleus5. The difference in energy between these two different transitions is very sm all (of the order of hertz) but results in two spectral lines separated in frequency by what is known as the spin-spin coupling constant, J. It would seem logical that the first state where t he nuclei are anti-parallel to one another results in the lowest energy state; however, this is not always the case. The interaction of

PAGE 16

16 nuclear spins across one or more bonds can result in the opposite being true. These two cases are distinguished by the sign of the coupling constant; the first is said to have a positive coupling constant (J>0) and the second a negativ e coupling constant (J<0)5. The signs of coupling const ants are often ignored in NMR spectroscopy, as a change in sign results in no visible change in first-order spectra. However, t he sign of the coupling constant is useful when one examines the effect of different functional groups or structural changes on the magnitude of J. With no knowledge of the sign, it is impossible to know if an increase in magnitude is the result of a positive J becoming more positive or a negative J becoming more negative and vice versa. It is also necessary to know the signs of the coup ling constants when one tries to simulate nonfirst-order specta. Additionally, the sign of J is of use to theoretical chemists and physicists in terms of molecu lar modeling and dynamics. Double-Resonance NMR Double-resonance NMR utilizes a second weaker rf pulse over a much narrower range of frequencies compared to the initial ex citation pulse. The concept of using a spin-decoupling (or spin-tickling) pulse was first suggested by Bloch6. Maher and Evans were the first to apply this technique to the determination of the relative signs of coupling constants7. They used double-resonance (or double-irradiation) NMR to show that the Tl-H(CH2) and Tl-H(CH3) coupling constants in thallium diethyl cation were of opposite signs. The technique was furt her developed by Freeman and Evans to determine the signs of both proton and fluor ine coupling constants in three-spin systems8-11. Double-resonance NMR differs from conventional decoupling experiments in that it is usually a homonuc lear experiment and the second rf pulse is not strong enough to cause rapid changes in the orientation of the spins to a point that

PAGE 17

17 any coupled nuclei only sense an averaged, or decoupled, state. The pulse only slightly perturbs the spins, hence the term spin-tickli ng. This causes the resonance line of any transition connected to the tickled transition to be slightly split5 (Figure 1-2). A more thorough explanation of double -resonance NMR and the signs of coupling constants and their effect on the spectrum is included in chapter two. Chemical Shift Another important parameter in NMR is the chemical shift (). The magnetic field at a nucleus (Bnucleus) is not equal to the external magnetic field produced by the superconducting magnet (Bo). Since electrons are magnet ic particles, their motion is influenced by the external field as well. As stated by Lenzs law, the motion of a particle induced by a magnetic field is in a such a direction as to oppose the field5. The motion of the electrons and other nuclei in the vici nity of the observed nucleus cause local fields that shield the nucleus from the external field so: 000(1)nucleusBBBB (1-5) where is the shieldin g factor. When Bnucleus is substituted into the Larmor equation, 02B (1-6) it shows the resonance frequency (or c hemical shift expressed in hertz), of a particular nucleus is a function of t he local fields surrounding that nucleus: 0(1) 2B (1-7) To avoid the ambiguity of the chemical shift being dependent on the strength of the magnetic field of a particular instrument, it is commonly reported in the dimensionless

PAGE 18

18 unit of parts per million (ppm)5. The resonance frequency () is compared to that of a reference material (r) and: 610R R (1-8) Since the electron cloud, bond angles, and bond lengths can all be affected by things such as solvent, temperature, and confor mational changes, the chemical shift is a powerful tool to help analyze sa mples at the atomic level. Pulsed-Field-Gradient NMR Pulsed-field-gradient (pfg) NMR is a s pecialized NMR technique which can be used to measure self-diffusion coefficient s. The stimulatedecho pfg pulse sequence (pfg-se) was first described by Stejskal and Tanner12. With the use of a gradient pulse, the spins can be marked spatially, and upon refocusing by a second gradient pulse, the spin-echo can be observed (Figure 1-3). The time between the two gradient pulses, is the diffusion time. The second rf pulse, a (or 180o) pulse, simply inverts the spins in the xy plane. With the second gradient pulse of equal strength and duration to the first, the spins are unwound. Absent diffusion, t he spins are in the same plane along the z axis and they are refocused without any signal attenuation. However, when diffusion does occur, the spins move along the gradient in the z direction and no longer experience an equal gradient from the seco nd gradient pulse resulting in spin incoherence and a loss of signal (Figure 1-3). For macromolecules such as peptides, a slightly different pulse sequence called the stimulated-echo pfg (pfg-ste) is used (Fi gure 1-4). There are two types of nuclear relaxation which result in a return to equilibrium following the rf pulse. T1 relaxation, also known as spin-lattice or longitudinal relaxation, is the process by which localized

PAGE 19

19 and fluctuating fields cause the angle at which spins precess around the magnetic field to change until thermal equilibrium is restored3. T2 relaxation, also known as spin-spin or transverse relaxation, involves the lo ss of spin phase coher ence due to different nuclei experiencing different magnetic fields and precessing at slightly different frequencies3. In liquid NMR, T1 and T2 are typically of the same order of magnitude (seconds); however, for large molecules T2 can be very short (milliseconds). This short T2 would result in an extreme loss of signal over the diffusion time in the pfg-se sequence. The pfg-ste pulse sequence utilizes two /2 (or 90o) pulses in place of the pulse in the pfg-se sequence (Figure 1-4). This results in the nuclear magnetization being stored on the z axis for most of and being subjected to only T1 relaxation effects. The pfg experiment makes it possi ble to calculate the relative diffusion coefficient, D, by acquiring spectra over an array of gradient strengths. More details about pfg NMR and the measurement of D can be found in chapter three. Cyclopropanes Cyclopropane was first synthesized in 188113, and Henderson and Lucas discovered its anesthetic properties in 192914. From the 1960s to the 1980s the chemists at W.R. Grace & Co synthesiz ed dozens of fluori nated cyclopropanes (Figure 1-5) in an attempt to create an effective anes thetic to replace the potentially hazardous ethers and halogenated compounds used at the time. The sa mples were sent to the University of Florida NMR facility for char acterization. The project was later abandoned, but the samples were preserved. Fluor inated cyclopropanes present an interesting research opportunity to analyze via NMR usi ng modern high-field m agnets. The strain associated with the small bond angles c an have an effect on physical properties, and

PAGE 20

20 the effect of different substituents on the chemical shift and coupl ing constants of the fluorine can be examined. Also, the results may shed light on t he NMR properties of fluorine, which is widely used in pharmaceuticals due to the stability of carbon-fluorine bonds. Amyloid Alzheimers disease (AD) is the most common form of dementia and it is estimated that 4.5 million Americans have AD This includes more than 10% of the population over 65 and 50% of those over 8515;16. It is characterized by a loss of memory and changes in personality and behavior. There is some debate as to the root cause of AD, but the most widely accepted hypothesis pres umes that cleavage of the amyloid precursor protein ( APP) into neurotoxic amyloid(A) peptide results in neurfibrillary tangles and cell death17;18. The brain tissue of patients demonstrating symptoms of AD at the ti me of death contains abnor mal amounts of plaques, and the main component of these plaques is the A peptide. The full biological function of APP, a ty pe I transmembrane prot ein, is not yet understood; but the mechanism by which APP is cleaved into A has been well characterized. Three enzymes, -, -, and -secretase, cleave APP in different regions of the protein, with cleavage by and resulting in AFigure 1-6. -secretase cleaves APP within the A sequence creating a benign soluble peptide while secretase cleaves the protein in the extrac ellular region creating the N-terminus of Aandsecretase cleaves within the cell membrane to create the C-terminus19. A can range from 30-42 residues depending on where APP is cleaved; the plaques however are composed mainly of 39-42 residue peptides20;21. It has been demonstrated

PAGE 21

21 that in membrane-mimicking environments the peptide adopts a typical transmembrane helical conformation20;21; in aqueous solutions the pictur e is less clear. Initially the peptide adopts a soluble random-coil format ion, and upon environm ental changes forms a self aggregating -sheet structure. A is a normal metabolic product of APP processing and is found in all in dividuals. The trigger that causes the conformational changes and subsequent plaque formation and onset of AD is of great interest18. Extensive work has been performed with the amyloidpeptide and it has been shown that there are two regions of the peptide that form -sheets, residues 16-24 and 31-40. It has also been shown that the conformational changes which lead to the -sheets and aggregation can be brought about in vitro by changes in pH and temperature and by the addition of metal ions such as Al3+ and Zn2+ 20-23. The primary limitation of t he amyloid hypothesis is the lack of a strong correlation between the amount of plaques found in the brain and the degree of symptoms evident at time of death. Oligomers of A remain soluble up to dime r and trimer states before they continue to aggregate and bec ome insoluble plaques. Rec ent research suggests it is these soluble oligomers, not the plaques themselves, whic h elicit the toxic effect and lead to eventual cell death. Upon interacti on with a membrane the oligomers can create an ion channel which destabilizes the ionic homeostasis and can lead to cell death20;24;25. Unfortunately not enough is known about the structure of the peptides in the membrane and how that structure diffe rs from that in an aqueous environment. A (25-35) (Figure 1-6), contai ning the kink between the two -sheet regions and a portion of the second -sheet area, has been proposed to represent the biologically active region of the peptide; it represent s the shortest fragment that exhibits large -

PAGE 22

22 sheet aggregated structures and retains t he toxicity of the full length peptide21. The 11residue peptide has been demonstrated to undergo a conformational change from a soluble random-coil to an aggregated -sheet structure depending on environmental conditions21.

PAGE 23

23 Figure 1-1. Nuclear and electron spins of two spinnuclei. A) The instance where the nuclei are anti parallel. B) The instance where the nuclei are parallel. When the first instance is of lower energy, the coupling constant between the two nuclei is positive. When the second is of lower energy, the coupling constant is negative. Figure 1-2. An example of the effects of a spin-tickling experiment. The bottom spectrum represents a conventional ex periment while the top is from a double-resonance experiment where a spin-tickling pulse was applied to a transition connected to the second peak. Note the splitting of the second peak in the top spectrum.

PAGE 24

24 Figure 1-3. The pfg-se pulse sequence. In the absence of diffusion along the axis of the magnetic field, the spins are refocused and the signal is maximum. When the nuclei diffuse along the z axis, they experience a different gradient strength from the second gradient pulse and are not fully refo cused, resulting in signal attenuation.

PAGE 25

25 Figure 1-4. The pfg-ste pulse sequence. The 180o inversion pulse is replaced by two 90o pulses to store the nuclear m agnetization along the z axis during 2 and eliminate T2 relaxation effects. Figure 1-5. Representative cyclopropanes. All the molecules studied have at least three fluorines, and many are ethers.

PAGE 26

26 Figure 1-6. The amyloid precursor pr otein and the positions at which the th ree enzymes can cleave the protein to produce amyloid beta.

PAGE 27

27Gly Ser Asn Lys Gly Ala Ile Ile Gly Leu MetO O CH3CH2CH3CH NH3 +O NH2C CH2OH CH2CH2NH3 +C NH CH C NH CH C NH CH C NH CH2C NH CH C NH CH C O O O O O O O CH2CH2CH2CH2CH3NH CH C CH CH3CH2CH3NH CH2C NH O CH C NH CH C CH2CH CH3CH3O O-CH2CH2S CH3 25 26 27 28 29 30 31 32 33 34 35 Figure 1-7. The sequence of amyloid beta (25-35).

PAGE 28

28 CHAPTER 2 ANALYSIS OF THE RELATIVE SIGNS OF COUPLING CONSTANTS IN FLUORINA TED CYCLOPROPANES Introduction Protonated cyclopropanes have been well characterized by NMR26-32, but little work has been reported on fluor inated cyclopropanes. Sargeant33 reported the magnitude of fluorine-fluorine coupling constants in various tetraand pentafluorinated cyclopropanes, and W illiamson and Braman34 reported the signs and magnitudes of the coupling constants in trifluor inated cyclopropanes. Cavalli35 reported the signs and magnitudes of the c oupling constants in tetrafluor inated cyclopropanes and examined the effect of temperature on the magnitude of the coupling c onstants. In the present work, the signs of the vicinal fluorine-fl uorine coupling constant s in six fluorinated cyclopropanes were examin ed using double-resonance 19F NMR and the effect of temperature on the magnitude of the coup ling constants in ni neteen fluorinated cyclopropanes was examined. To fully understand the application of double-resonance NMR, a more thorough explanation of the experiment and its effect on the resulting spectrum is necessary. Figure 2-1 shows a theoretical energy level diagram for the transitions associated with a two-spin AM spin system where JAM = 0. In the absence of c oupling, the transition pairs for each nucleus (for instance and ) are of equal energy. This results in a spectrum containing two peaks corresponding to the frequency of each transition pair. However, when the nuclei ar e coupled together, the ener gy diagram changes and the spectrum includes four lines (Figure 2-2). The change in energy of an individual energy level depends on the str ength of the coupling:

PAGE 29

29 () A MAMEJmm (2-1) Following Eq. 2-1, two of the energy levels in the two spin system will increase by J and two will decrease by J This results in the frequency of one of the transitions for each nucleus increasing in energy by J and the other decreasing by J (Figure 2-2). The two pairs of spectral lines correspondi ng to each nucleus are therefore separated by J The energy-level diagram for a threespin AMX system (Figure 2-3) is more complex but the effect of coupling remain s the same. Up to now the sign of the coupling constant has been ignored, but it is evident from Eq. 2-1 that the sign will affect the energy diagram. Figure 2-4 shows the energy-level diagrams and resultant spectra for four sets of theoretical coupling-constant signs in a th ree-spin system. The similar transitions are color coded (for instance is navy blue across all four diagrams) to demonstrate the effect on the spectrum. The peaks and transitions for the upfield X nucleus are all left light blue for the sake of simplicity. In black and white all the spectra would be identical, and this is t he reason the sign of the coupling constant is often ignored. However, it is the variation in position of the colored peaks that is the basis for the application of double-resonance NMR. In a double-resonance (or spin-tickling) expe riment, a weak rf pulse is applied to the frequency of one peak in the spectrum (that is, to one of the transitions in the energy diagram). This pulse results in a splitting of the peaks of any trans ition connected to the tickled transition. The nature of the splitti ng depends on the way the two transitions are connected. When the shared energ y level lies between the two transitions, it is said to be a progressive connection, and when the shared energy level is above or below both transitions it is said to be a regressive connection (Figure 2-5)5. Progressive

PAGE 30

30 connections result in a rather broad splitting while regressive connections result in a relatively sharp splitting (Figure 2-5). When multiple spectra are acquired with the tickling pulse placed on different transiti ons, the energy level diagram can be pieced together via which transitions are connected. Since changing the signs of the coupling constants results in different transitions bei ng connected, the result ing diagram will only match the theoretical diagram calculated from one set of coupling constant signs. Experimental The cyclopropanes used for this research were synthesized decades ago in the laboratories of t he Dewey and Almy division of W. R. Grace and Co. The samples were dissolved in deuterated acetone at a concentration of 20% vol. All spectra were acquired on either a Varian Mercury-300 equi pped with a Varian ATB probe, or a Varian Inova-500 equipped with a Varian indirect -detection triple-resonance probe. The temperature for the variable temperature experiments was regulated using an FTS Systems XR401 Air-Jet equipped with a TC-84 te mperature control unit. Four, eight, or 16 scans were acquired for each experiment. The spectral window was typically set to approximately -130 to -180 ppm (relative to CFCl3) to maximize the di gital resolution. The number of points was typically maximize d to 128,000, resulting in acquisition times of around 5 seconds. For the double-resonanc e experiments, the decoupler modulation mode was set to continuous wave and the decoupler power was 1 dB. Doubleresonance spectra were acquired with the tickling pulse placed on every transition and the multiple spectra were analyzed until the possible sets of coupling constant signs was narrowed down to one.

PAGE 31

31 Results and Discussion Temperature Dependence of Coupling Constants Cavalli reported that an increase in temperature results in both cis and trans threebond, or vicinal, coupling constants becomi ng more positive. Observation of the coupling constants over a r ange of temperatures would th en help establish the sign, with negative coupling constants decreas ing in magnitude and positive coupling constants increasing. Table 2-1 shows the change in magnitude of the couplings constants for nineteen cyclopropanes over a range from -20 to 25oC. To simplify analysis and establish stereochemistry, numbers were assigned to each substituent position on the cyclopropyl ring (Figure 2-6). Substituent pairs one and two, three and four, and five and six are on the same ca rbons, and the odd and even positions are on the same sides of the ring. In general, most of the cis coupling constants have a positive temperature coe fficient and most of the trans have a negativ e temperature coefficient. This agrees with Cavallis results from CF2-CFCl-CFCl where he also found Jtrans to be negative and Jcis to be positive. Surprisingly, all of the J3-6 trans couplings, corresponding to a CFCl coupling with a CF(O R), had a positive temperature coefficient. Also of note are the negative J1-3 cis couplings found in molecules with a CCl2. However, this agrees with the work of Williamson and Br aman who found both Jtrans and Jcis to be negative in various CF2-CCl2-CFX cyclopropanes. Signs of Coupling Constants Double-resonance experiments were performed on select cyclopropanes to compare with the results from the temperature study. Figur e 2-7 shows the spectrum of CF2-CFCl-CF(OCCl3) with the four highlighted pea ks corresponding to one of the isomers. The effect on the spectrum with the application of a spin-tickling pulse to the

PAGE 32

32 most downfield peak is shown in the expans ions. A spin-tickling pulse was applied to each peak and a list of the connected transitions was compiled (Table 2-2). For simplicity, only three nuclei were considered at a time, as the energy-level diagram for a four-spin system becomes cumbersome to follow. Thus, two tables similar to Table 2-2 were constructed for each molecule/isomer. The connected transit ions were compared to the energy-level diagrams of the four possible combinat ions of coupling-constant signs (Figure 2-8). The correct signs coul d typically be ascertained by simple analysis of the results of only one of the spin-tickling s pectra. However, to avoid the possibility of misinterpretation of the slight splitting as re gressive or progressive, analysis of all of the connected transitions was done, and the correct energy-level diagram was verified. Analysis of Table 2-3 shows the sign and magnitude of the coupling constants for six cyclopropanes. The signs r eported are relative to t he two-bond, or geminal, CF2 coupling constant being positive10;34. All but one of the Jcis values were found to be positive. This agrees with their positive te mperature coefficients and Cavallis assertion that the coupling constants become more positi ve with an increase in temperature. The exception is J1-3 for CF2-CF(OCH3)-CCl2, this represents a positive coupling constant with a negative temperature coefficient. The Jtrans values present two deviations from the expected results. The values of J2-3 trans were always negative which agree with their negative temperature coefficient s. However, the values of J2-5 and J1-6 were found to be positive with negative temperat ure coefficients and the values of J3-6 were negative with positive temperature coefficients. The last compound in Table 2-3, CFClCClF-CF(OCH3), has no geminal CF2 to ascertain absolute si gn, but the three coupling constants were found to be of the same rela tive sign. Regardless of the absolute sign,

PAGE 33

33 at least one of the trends f ound in the other molecules fo r the signs of the coupling constants between particular substituent positions always having the same sign would be broken. All of these data point to the fact that the sign of coupling constants and their temperature coefficients do not nece ssarily correlate to bonding geometry and should not be used as the sole basis for assignment.

PAGE 34

34 Figure 2-1. A theoretical energy-level diagram and corresponding spectrum for the transitions of a two-spin first-order system with no coupling.

PAGE 35

35 Figure 2-2. A theoretical energy-level diagram and corresponding spectrum for the transitions of a two-spin first-order system with coupling.

PAGE 36

36 Figure 2-3. A theoretical energy-lev el diagram for a three-spin system.

PAGE 37

37 Figure 2-4. The effect of t he sign of the coupling constants in a three-spin AMX system on the theoretical spectrum. JAM is assumed to always be positive. A) a ll coupling constants are positive. B) JAX is negative and JMX is positive. C) JAX is positive and JMX is negative. D) both JAX and JMX are negative.

PAGE 38

38 Figure 2-5. The difference between progr essively and regressively connected transit ions and their effect on the spectrum of a double-irradiation experiment.

PAGE 39

39 12 35 6 4 Figure 2-6. The numbering scheme assigned to the substituent positions of the cyclopropane.

PAGE 40

40 Figure 2-7. The effect of a spin-tickling experiment on the 19F spectrum of CF2-CFCl-CF(OCCl3).

PAGE 41

41 Figure 2-8. The possible sets of coupling-constant signs and their corre sponding energy-level diagrams for CF2-CFClCF(OCCl3). The labeled transitions correspond to the results in the first row of Table 2-2.

PAGE 42

42 Table 2-1. The change in magnitude of the vicinal fluorine-fluorin e coupling constants (Hz) with a change in temperature from -20oC to 25oC. cis trans Compound J(1-3)J(1-5) or J(2-6) J(3-5) J(2-3) J(1-6) or J(2-5) J(3-6) CF2 CFCl CFCl + 0.54 + 0.54 0.13 0.13 CF2 CFCl CF(OCH3) + 0.85 + 0.47 + 0.59 0.09 0 CF2 CFCl C(OCH3)F + 0.51 nra 0.23 + 0.07 + 0.22 CF2 CFCl CF(OCHCl2) + 0.61 + 0.36 + 0.69 0.07 0.11 CF2 CFCl C(OCHCl2)F + 0.35 + 0.33 0.06 0.09 + 0.39 CF2 CFCl CF(OCCl3) + 0.42 + 0.51 + 0.39 0.12 0.05 CF2 CFCl C(OCCl3)F + 0.65 + 0.19 0.07 nr + 0.22 CF2 CFCl CF(OCHF2) + 0.48 + 0.39 + 0.39 0.06 0.15 CF2 CFCl C(OCHF2)F + 0.35 + 0.20 0.17 0.06 + 0.29 CF2 CFCl CF(OCH2F) + 0.27 + 0.78 + 0.47 0.06 0.09 CF2 CFCl C(OCH2F)F + 0.45 + 0.23 0.15 0 + 0.06 CF2 CFCl CF(OCF2Cl) + 0.45 + 0.50 + 0.39 0.14 0.06 CF2 CFCl C(OCF2Cl)F + 0.45 + 0.23 0.09 + 0.11 + 0.15 CF2 CF(CH3) CCl2 0 0.49 CF2 CF(CF3) CCl2 0.94 0.65 CF2 CF(OCH3) CCl2 0.30 + 0.18 CFCl CFCl CF(OCH3) + 0.52 CFCl CFCl C(OCH3)F + 0.04 CFCl CClF CF(OCH3) + 0.08 0.07b 0.39 anot resolved, bJ(1-4), The substituent positions 1-6 on the ring are defined in Figure 2-6. The chemical formula is written in su ch a way that the positions 1 and 2 are on the first carbon in sequential order, positions 3 and 4 on the second carbon, and 5 and 6 on the third

PAGE 43

43 Table 2-2. List of the connected transitions for CF2-CFCl-CF(OCCl3). Connected transitions Double-resonance peak Regressive Progressive 1 6, 9 8, 11 2 5, 9 7, 11 3 8, 10 6, 12 4 7, 10 5, 12 5 2a 4a 6 1, 12 3, 11 7 4, 9 2, 10 8 3a 1a 9 1, 7 2, 8 10 3, 8 4, 7 11 2, 5 1, 6 12 4, 6 3, 5 aThe effect of some spin-tickling pulses could not be resolved due to complication from coupling with the fourth fluori ne in the spin system or overlap with peaks resulting from the isotope effect. The peak numbering scheme can be found in Figure 2-8. Table 2-3. The sign and magnitude of the vicinal fluorine-fluorine coupling constants (Hz) in fluorinated cyclopropanes. cis trans Compound J(1-3) J(1-5) or J(2-6) J(3-5) J(2-3) J(1-6) or J(2-5) J(3-6) CF2 CFCl CF(OCH3) + 6.54 + 5.73 + 3.80 5.13 + 2.85 CF2 CFCl C(OCH3)F + 6.87 nra 5.85 + 2.40 4.37 CF2 CFCl CF(OCCl3) + 4.14 + 5.71 + 7.17 7.13 + 0.86 CF2 CFCl C(OCCl3)F + 1.81 + 1.44 7.06 nr 7.16 CF2 CF(OCH3) CCl2 + 2.50 5.50 CFCl CClF CF(OCH3) 2.28c 11.93b,c 4.43c aNot resolved, bJ(1-4) cThe compound contains no CF2 to ascertain absolute sign. The substituent positions 1-6 on the ring ar e defined in Figure 2-6. The chemical formula is written in such a way that the positions 1 and 2 are on the first carbon in sequential order, positions 3 and 4 on the second carbon, and 5 and 6 on the third.

PAGE 44

44 CHAPTER 3 ANALYSIS OF THE AGGREGATION OF AMYLOID BETA (25-35) Introduction There is a great deal of research involv ing Alzheimers Disease and the amyloid beta peptide. While the curr ent trend entails utilizing the whole 40-42 residue peptide, it is not uncommon to find recent articles involving A (25-35)36-42. In fact, A (25-35) is often used as a model for the behavior of th e full length peptide because it retains both the physical and biological properties, and is easier to synthesize and derivatize due to its relatively short 11 amino acid sequence43. The genesis of this research goes back to observations made during earlier work in this laboratory. Similar samples prepar ed by previous researchers appeared to show reversible aggregation upon placement in and re moval from the refrigerator. Samples were stored in the refrigerator and when removed for analysis they appeared cloudy and viscous. Thought to be ruined, they were left on the laboratory bench to be discarded. Upon warming to room temperature they became clear and liquid. This process was repeated with similar results. No quantitative anal ysis was performed. The goal of the current research was to us e NMR to analyze the effects of temperature and other sample conditions such as c oncentration on the apparent aggregation. We also hoped to measure the oligomerizati on equilibrium and estimate the size and shape of the oligomers. Experimental The synthetic A (25-35) peptide was obtained from Dr. John West at Florida A&M University. A 1 mM stock solution was made using 16.6 mg peptide and deionized water with the pH adjusted to 3.4 using deuter ium chloride. Attemp ts to make samples

PAGE 45

45 using unadjusted pH water were unsuccessful. All samples made with water from pH 57 became very viscous gels within five minutes and were unsuitable for liquid NMR analysis. Samples of concentrations gr eater than 1 mM were made on an individual basis using low-pH deionized water. Un less otherwise noted, deuterated water was added to all samples such that H2O:D2O was 90%:10% volume to volume. All experiments were performed on a Va rian Inova-500 MHz spectrometer equipped with a Varian indirect-detection pulsed-field-gradient probe. All standard proton spectra were acquired using the pr esat pulse sequence to suppress the water peak. A saturation delay of 1.00 second and a saturation power of 15 dB was used. The acquisition time was 2.90 seconds and the recycle delay was 1.00 second. Depending on the concentration of the sample, either 64 or 128 scans were acquired. For the pulsed-field-gradient experiments, the gradient was arrayed from 20-60 G/cm using either 10 or 15 steps with intervals such that the squares of the gradient strengths were linearly spaced. Either 16 or 32 scans were acquired for each step. The gradients were applied for 1250 s with a diffusion time of 0.10 seconds. The twodimensional experiments were all acquired with an acquisition time of 0.36 seconds, a recycle delay of 1.00 second, and a mixing time of 150 ms. The TOCSY (or total correlation spectroscopy) experiments we re acquired using the tnTOCSY pulse sequence with a saturation delay of 1.00 se cond and a saturation power of 10 dB, the number of increments was ei ther 256 or 400, and 32 scans were acquired for each increment. The ROESY (or rotating frame nuclear Overhauser effect spectroscopy) experiments were acquired using the tnRO ESY pulse sequence with a saturation delay

PAGE 46

46 of 1.00 second and a saturation power of eit her 2 or 5 dB; the number of increments was either 512 with 20 scans per increm ent, or 768 with 16 scans per increment. Gradient Calibration The strength of the gradient pulses had to be calibrated using a compound with a known diffusion coefficient. The acquisi tion software generates an array of the experimental parameter dac_p1 (dac units). This value is sent to a digital-to-analog converter and is translated to an actual gradient strength, G (gauss/cm), by multiplication with another parameter, grad_p_coef ((gauss/cm)/dac units): dac_p1 grad_p_coef = gradient strength G (3-1) By observing the signal intensity of a reference compound over an array of dac_p1, the appropriate value of grad_p_coef can be calculated with Eq. 3-2 using the expected diffusion coefficient. 22 01 ln()ln()(()) 3iiAADG (3-2) Equation 3-2 relates the observed intensity, A and the gradient strength, G to the diffusion coefficient, D The diffusion time, and the length of t he gradient pulses, are both experimental parameters; and as previously noted is the magnetogyric ratio. Lysozyme from chicken egg white was obtained from Sigma-Aldrich. A 10 mg/mL lysozyme solution in D2O was used to calibrate the gradient strength. A value of 11.0 10-7 cm2/s 44 corrected by a factor of 1.23 for the increased viscosity of D2O45;46 was used for the diffusion coefficient of lysozyme. The intensity of three regions in the lysozyme spectrum was observed over a range of fifteen gradient strengths from 20 to 60 G/cm. The natural log of these three sets of intens ities was plotted against the square of the gradient strength (Figure 3-1) and a linear r egression was applied. From

PAGE 47

47 Eq. 3-2, the slope of the line is proportional to the diffusion coefficient. Using the known value for the diffusion coeffi cient, the expected slope can be calculated with Eq. 3-3. 223 mD (3-3) The value of grad_p_coef was adjusted, changing the values of G2 in Figure 3-1, so that the average slope of the three regre ssion lines equaled the expected slope. The calibration was verified using bovi ne serum albumin obtained from SigmaAldrich. A 10 mg/mL so lution was prepared in D2O. The intensit ies of the peaks corresponding to the methyl side-chain protons in alanine, isoleucine, and leucine were observed over a range of fifteen gradient strengths from 20 to 60 G/cm. The natural log of these sets of intensities was plotted against the square of the gradient strength (Figure 3-2). A linear regre ssion was applied to the three data sets and the slopes were used to calculate D using Eq. 3-3. The average measured diffusion coefficient, corrected by a factor of 1.23 fo r the increased viscosity of D2O, was 6.09 10-7 3 10-9 cm2/s compared to the liter ature value of 5.8 10-7 cm2/s 47;48. Temperature Regulation The standard method of regul ating the temperature of the sample in an NMR spectrometer is not appropriate for diffusion measurements. A coil, slightly below the sample (Figure 3-3), intermittently heats th e nitrogen flowing into the probe to regulate the temperature. This can result in a te mperature gradient across the sample from top to bottom. Since the diffu sion coefficient is calculated based on movement of nuclei along the z axis, the presence of a tem perature gradient must be avoided. A temperature gradient will result in convecti ve flow and lead to additional displacement of spins along the z axis and to an overestimation of the diffusion coefficient49-52.

PAGE 48

48 To avoid convection, the temperature of the nitrogen flowing into the probe was regulated upstream using an external vari able-temperature unit (Figure 3-3), and the heating coil in the probe was disabled. To account for thermal conduction over the longer flow path of the nitrogen, the regulator on the exter nal unit had to be set below the desired sample temperature when running below room temperature and above the desired temperature w hen running above room temperature. The samples were left to equilibrate for a minimum of th irty minutes after the temper ature readings from both the external control unit and the thermocouple in the probe were stable. The temperature remained constant to a tenth of a degree celsius during all experiments. Chemical Shift Referencing When reporting chemical shift data it is imperative to use a suitable standard compound as a reference. This is especially true with biomolecul ar NMR where it is common to make assumptions about the sec ondary structure of pe ptides and proteins based on chemical shifts. Unfortunately ther e is a wide array of reference compounds, many of which have chemical shifts that vary with changes in pH, temperature, and solvent composition53-56. For this work all reported chemical shifts are relative to internal 2,2-dimethyl-2-silapent ane-5-sulfonic acid (DSS). The chemical shift of DSS has been shown to be insensitive to changes in solvent, temperature, and pH55. Results and Discussion Chemical Shift Evaluation and Assignment To verify the sequence of the synthetic peptide was correct, all peaks had to be assigned to their respective residues and the order the residues are connected had to be analyzed. The reasons for this are two-fold. Due to an error in synthesis, we have two versions of the solid A(25-35) peptide, one with the correct sequence and one

PAGE 49

49 missing an isoleucine. To ensure t he proper peptide was used, the sequence needed to be verified. Also, to analyze any changes in chemical shifts, it was necessary to know which peaks belong to which residues. The peaks were assigned using a combi nation of proton, TOCSY, and ROESY spectra. Due to significant overlap, as is typical with biomolecular NMR, it was impossible to assign and calculate chemical sh ifts for many peaks using only the proton spectra (Figure 3-4). Assignment was done following the procedure known as the sequential assignment strategy developed by Wthrich et al.57 and further outlined by Cavanagh et al.58. Since each residue represents an individual spin system, a peptide of N residues has N distinct backbone-based spin systems58. The first stage of assignment typically involves a two-dimens ional scalar correlation experiment such as COSY or TOCSY. Each type of amino acid gives rise to a characteristic peak pattern based on its side-chain protons. Analysis starts in t he fingerprint region of the spectrum, where the cross peaks resulting fr om the correlation of the alpha and side chain protons to the amide protons can be obser ved (Figure 3-5). Distinct horizontal or vertical lines, depending on which fingerprint region is used, connect the peaks of each individual residue at the chemical shift of the amide proton for that residue. Examination of the peak patte rn yields a straight-forward assignment of which amino acid the peaks represent. The amide protons of the asparagine side chain are not part of the same spin system as the backbone amide and alpha protons. In the same that way each residue forms an individual spin system isolated by the backbone amide bonds, the amide bond in the asparagine side chain isolates the amide protons. Howe ver, these protons

PAGE 50

50 appear at distinctive chemical shifts, 6.8 and 7.5 ppm (Figure 3-5) and the only overlap is with the amine proton of the lysine side chain, which is easily distinguished by its correlation to the other lysine protons. Al so, the asparagine side chain amide protons show correlation with each other via chemical exchange, making the assignment apparent. Analysis of the TOCSY spectrum helps assign each peak to a particular residue but yields little information about the placem ent of the residue in the peptide sequence. Determination of the sequence and discrimination of degenerat e residues is step two of the sequential assignment strategy57;58. Assignment of Duplicate Residues The second stage of the sequential assignm ent strategy involv es distinguishing the duplicate glycine and isoleucine residues of A (25-35), and requires the use of nuclear Overhauser enhancement and exc hange spectroscopy (NOESY). Twodimensional NOESY spectra yield cross peaks for nuclei in close spatial proximity57. The assignment of ambiguous residues, and verification of the whole peptide sequence itself, is established by observing cross peaks (off-diagonal peaks) between the amide and alpha protons of neighboring residues 1N1 1HHii While the intensity of the cross peak, for a given experimental mixing time, is relative to the distance between two nuclei, t here are limits to that intensity imposed by the spectrometer frequency, o, and the molecular correlation time c 58;59. The homonuclear nuclear Overhauser enha ncement, as a function of c, ranges from 0.5 for small molecules to -1 for macromolecules (Figure 3-6)5. Unfortunately, at the frequency

PAGE 51

51 of modern high field spectromet ers, the correlation time of peptides (on the order of nanoseconds) corresponds to the range where the NOE is nearly zero. Rotating frame Overhauser effect spec troscopy (ROESY) was developed by Bothner-By et al.. to overcome this problem59. Originally termed CAMELSPIN (crossrelaxation appropriate for minimolecules emulated by locked spins), ROESY has the advantage that the rotating fram e Overhauser effect (ROE) is positive for all correlation times, with a minimum of 0.38 for small molecules and a maximum of 0.68 for macromolecules58;59. Therefore, ROESY is useful in the examination of peptides, or any molecules with similar correlation time s, where the laborat ory frame NOEs are nearly zero58. ROESY was used to differentiate glycine29 and glycine-33 (the assignment of glycine-25 was apparent due to its placement at the N-terminus of the peptide), and isoleucine-31 and isoleucine-32. Figure 3-7 shows an expa nsion of the alpha-amide region above the diagonal of the ROESY spectrum of 0.69 mM A (25-35) in 80% D2O and 20% d3-TFE acquired with a mixing time of 150 ms. Many 1N1 -1H-Hii cross peaks can be seen, but the cross peak at the chemical sh ift of an ambiguous glycine amide proton (8.35 ppm) and the lysine-28 alpha proton (4.28 ppm) was used to assign glycine-29. Similarly, the cross peaks between the is oleucine-31 amide proton (7.96 ppm) and alanine-30 alpha proton (4.34 ppm), an d glycine-33 amide proton (8.27 ppm) and isoleucine-32 alpha proton (4.17 ppm) were used to assign the isoleucines and remaining glycine respectively. A cross peak can also been seen at the chemical shift of the isoleucine-32 amide (8.00 ppm) and isoleucine-31 alpha (4.17 ppm) protons, further confirming the assignment of the two isoleucines. Figure 3-8 presents a similar

PAGE 52

52 picture, this time an expansio n of the alpha-amide region below the diagonal, of the ROESY spectrum of 6.5 mM A (25-35) in 90% H2O and 10% D2O. However, this spectrum better demonstrates some of the diffi culties in assigning pep tide spectra. Of note, the alpha protons of glycine-29 and gl ycine-33 are not resolvable, nor are the amide and alpha protons of isoleucine-31 and isoleucine-32. The cross peak at the chemical shift of the glycine-29 amide pr oton (8.39 ppm) and lysine-28 alpha proton (4.29 ppm) was used to distinguish the glyc ines. The combination of the TOCSY and ROESY spectra was used to assign each peak and verify the peptide was in fact the known A(25-35) sequence. Chemical Shift Results Complete assignments of A (25-35) at two concentrations in 90% H2O and 10% D2O are shown in Tables 3-1 and 3-2. T here are many ways to analyze and interpret peptide chemical shifts60-64, but perhaps the most straig ht-forward approach is the technique developed by Dalgarno et al.60. The technique utilizes the finding that alpha protons shift upfield for residues in -helix regions and downfield for -sheet regions60;65-67. The chemical shift of the alpha protons is very dependent on the orientation of the C-H bond axis relative to the adjacent C=O bond axis68. As a result, the chemical shifts of the alpha protons ca n be used to estimate local secondary structure. This technique relies on the co mparison of observed chemical shifts to known random-coil shifts (Eq. 3-4). RCObs (3-4) It has been shown that the alpha protons can sh ift by as much as 1 ppm from their expected random-coil chemical shift60;67 and that the mean shifts in helices and sheets

PAGE 53

53 differ by as much as 0.8 ppm with little overlap64. Many tables of random-coil chemical shifts can be found in literature53-55;69;70. For this work, the table suggested by Wishart and Nip69 was used. Figure 3-9 shows the plot for the alpha protons of 6.71 mM A (25-35) in 90% H2O and 10% D2O at 2.53. All of the chemical sh ifts are very near their expected random-coil chemical shift, with the largest shift being 0.06 ppm for the N-terminal glycine. This indicates the observable A (25-35) in a solution of 90% H2O and 10% D2O adopts a random-coil conformation. This also validates our chemical-shift calculation and assignment meth od. The results (not shown) for the alpha protons of 3.47 mM A (25-35) in similar sample conditions were comparable, with the largest shift being -0.06 for the serine26 proton. It should, howe ver, be noted that the alpha protons for glycine-25 and as paragine-27 were not observed due to solvent exchange and overlap with the solvent peak, respectively. Effect of TFE 2,2,2-trifluoroethanol (TFE) has been shown to mimic the lipid environment of the cell membrane and promote -helices71. It has also been shown that TFE does not promote new structures in small peptides, bu t rather stabilizes helices inherent to the entire protein72. An -helical A structure has been suggested as an intermediate on the way from a mostly unstructured peptide to -sheet-rich A fibrils73-77. Deuterated TFE (d3-TFE) was used as a co-solvent and the chemical shifts were again measured. Table 3-3 shows the complete assignment of 0.69 mM A (25-35) in 80% H2O and 20% d3-TFE. The results are remarkably similar to the chemical shifts found in a mixture of H2O and D2O except for the amide protons This is most likely due

PAGE 54

54 to the sensitivity of the am ide protons to the change in hydrogen bonding with the addition of TFE. A mixture of TFE and water has a reduced capacity for forming hydrogen bonds compared to just water76. There is also little change in the plot (Figure 3-10) compared to A in just water, with the la rgest shift from the expected random-coil value being -0.07 fo r both serine-26 and leucine-34. All of the chemical shifts for the alpha protons after the addition of 20% d3-TFE are within 0.03 ppm of the results without TFE except fo r methionine-35, which changed by 0.07 ppm. This is most likely a result of its position on the C-te rminus, and the sensitiv ity of the carboxylic acid to the change in hydrogen bonding. These results, indicating soluble A (25-35) remains a random-coil even in the presence of low concentrations of TFE, contradict previously published results. Using circular dichroism and NMR, Lee et al.. found A (25-35) has a random-coil st ructure in water and adopts a helical structure from isoleucine-31 to methionine-35 in the presence of TFE78. This is perhaps a result of the higher TFE content in their so lvent. They published the fu ll assignment of chemical shifts for 1 mg A (25-35) in 0.4 mL of 1:1 (vol/vol) d3-TFE:H2O at pH 4.0. Surprisingly, virtually all of their chemical shifts for each proton (including the side chain protons) on every residue are significantly more downfield (as much as 0.11 ppm and on average 0.06 ppm) than the results found here, contra dicting the thought that the protons of helical residues shift upfield. Kohno et al.. also concluded that the C-terminal region (lysine-28 through methionine-35) of A (25-35) adopted a helical conformation in a membrane mimicking environment79. They published the chemical shift results for 2 mM A (25-35) in 90% H2O and 10% D2O at pH 4.0 with 250 mM LiDS-d25. Their alpha proton chemical shifts

PAGE 55

55 for serine-26 and asparagine-27 were simila r to the results found here, but lysine-28 through methionine-35 were all shifted upfiel d, as expected for helical regions, by an average of 0.11 ppm. Sticht et al.. examined 2.5 mM A (1-40) with 40% TFE at pH 2.880. Their chemical shift results were similar to those published by Kohno et al.., and they concluded A forms a helix from lysine-28 to valine-40. The absence of a C-terminus helix in this work is presumably the result of the relatively low percentage of TFE used. A mixture of 80% H2O and 20% TFE was chosen due to the prohibitive cost of d3-TFE. This conclusion is supported by the work of DUrsi et al.21. They examined A (25-35) with varying ratios of hexafluoroisopropanol (HFIP), another membrane-mimicking solven t, and water. They found that A (25-35) adopted a random-coil conformation up to 40 % HFIP and the helix content increased almost linearly up to 80% HFIP. Effect of metal ions Transition metals, such as Cu(II) and Zn(II), have been implicated in the development of Alzheimers disease81-83. They are found in elevated concentrations in amyloid plaques22;84;85 and have been shown to promot e or even induce aggregation8690. The current consensus in the literature suggests the binding site is near the Nterminus of amyloid peptide, with the hist idine-6,13, and 14 side chains, the N-terminus itself, and the carboxylate groups of asparti c acid-1 and glutamic acid-11 being potential ligands85-87;91. Interestingly, A (25-35) has no known metal binding sites, but aggregates more readily than other fragments and maintains the toxicity of the full length peptide.

PAGE 56

56 Copper(II) and zinc(II) were added at various molar ratios and the chemical shifts of A (25-35) were measured (Tables 3-4 through 37). All of the results are remarkably similar, and comparison to the results with no metal (Table 3-3) reveals little change. The largest shift was 0.05 ppm for the amide pr otons of the N-terminal glycine-25 in the sample with 3:1 zinc-to-peptide molar ratio. These protons readily exchange with the solvent and in some instances were not even observed. The comparatively large shift could easily be a result of error in chemical -shift calculation due to the small signal-tonoise ratio of the peak. The mean shift fo r all the metal samples compared to the sample with no metal was 0.0009 ppm, an insignificant amount. The lack of any observed changes in chemical shift indicate s there was no change in the peptide with addition of the metal ions. The spectra were also examined for line broadening. Gaggelli et al. observed large chemical-shift variations and selective line broadening for the previously mentioned ligand protons with th e addition of Zn(II) to a me mbrane mimicking solution of A (1-28)87. Similar results have also been observed with A (1-40), A (1-28), and A (1-16)92-95. Not surprisingly, no broadening was observed. This confirms the chemical shift result, that ther e is no specific binding of A (25-35) with the metal ions. This defies the qualitative obser vation on the disposition of the samples. Figure 311 shows a picture of three samples made at the same time with similar sample conditions, 1.84 0.06 mM A (25-35) in 90% H2O and 10% D2O. The samples which appear cloudy include a 3:1 molar ratio of Cu (II) or Al(III) and the sample which appears clear has no added metal ions. The appearance of the samples indicates some sort of interaction between the peptide and the metal ions. When this observation is combined

PAGE 57

57 with the lack of change in the NMR data, it suggests some of the peptide aggregated into a high-order oligomer that was uno bservable via liquid NMR and the peptide remaining in solution was unchanged. The pr esumed equilibrium of the peptide (Figure 3-12) includes soluble monomer and low-order oligomers along with unobservable insoluble high-order oli gomers. Since all of the spectra include fairly well resolved and sharp peaks, it is assumed we observed random-coil monomer and once aggregation was triggered it went all the way to high-order oligomers. This is confirmed by the fact that the only observable change between spectra of clear samples and cloudy samples acquired under identical conditions was a decr ease in signal intensity (Figure 3-13). Effect of temperature The effect of temperature on the chemical shifts of the amine protons was also examined. The temperature dependence of the chemical shifts of amine protons has a somewhat dubious past when used as the on ly evidence for suggesting peptides exist in particular conformations, but they are o ften used to support the findings of other techniques. The amine protons of random-c oil peptides have been shown to display /T values of -7.8 1.2 ppb/oC in aqueous solutions70;96, while the amine protons of structured peptides typically di splay a less negative (<-4 ppb/oC) temperature dependency97;98. The change in the chemical shifts of the amine protons of 2.82 mM A (25-35) in 90 H2O and 10% D2O was linear in the observed range of 5 to 65oC (Figure 3-14) and the average /T value was -7.6 ppb/oC with the lowest being -5.8 0.1 ppb/oC for alanine-30 (Table 3-8). These re sults further suggest the presence a random-coil peptide.

PAGE 58

58 Diffusion Results In the absence of aggregation, a plot of log D versus 1/T should follow the Arhhenius relationship and have a linear slope99: 0ln()ln()a i iE DD RT (3-5) To verify experimental setup and parameters the diffusion coefficient of the same lysozyme sample used to calibrate the grad ient strength was acquired in five degree increments over a range of 5-55oC (Figure 3-15). Ilyina et al. reported lysozyme does not aggregate at this concentration over this temperature range47. The linearity of the data in Figure 3-15 confirms this. All diffusion data reported here is the resu lt of integrating the peaks corresponding to the side-chain methyl protons on the l eucine and isoleucines (Figure 3-16). The diffusion coefficient of 2.82 mM A (25-35) was measured in five degree increments from 5-70oC (Figure 3-17). To further verify what appears to be a deviation from linearity around 22oC (or 3.39/K), data were m easured approximately every 2.5oC from 5oC to 25oC. A linear regression was appl ied over the whole range of temperatures and a second linear regression was applied from 25-70oC and the residuals were plotted (Figure 3-17). The pa rabolic behavior of the residuals from the regression applied over the whole range of temperatures indica tes the data are not linear. Additionally, a linear regression wa s applied to the lower temperatures (22oC to 5oC, or 3.39/K to 3.60 000/K) and the upper-95%-conf idence-limit slope (-2.40) was compared to the lower-95%-confidence-limit slope of the regression applied to the higher temperatures (-2.10). The change in linearity over the two temperature ranges appears to be statistically significant. This could be the result of the peptide self

PAGE 59

59 aggregating or a conformational change which increased the hydrodynamic radius at lower temperatures. The effect of concentration on the diffu sion over a range of temperatures was examined. The diffusion coefficients of three concentrations of the peptide were measured from 5-55oC (Figure 3-18, Table 3-9). Th ere appears to be no dependence of the diffusion on concentrations in the mM range. All of the values are in good agreement and, if nothing else, validate the reproducibility of the experiment. The diffusion was also measured at two pH values with a concentration of 2.82 mM (Figure 3-19, Table 3-10). There were slight differe nces at lower temperat ures, but all of the values were within error of one another. As previously mentioned, attempts to make samples with pH values closer to neutra l were unsuccessful and the solutions quickly turned to viscous gels unsuitable for diffusion measurements. While there is no mention of similar observations about neutral pH samples becoming highly viscous, the common trend in literature is to use pH values around four21;78;79. The diffusion was measured with the addition of Zn(II) at a 3:1 molar ra tio as well (Figure 3-20, Table 3-11). There were slight differences at higher temperatur es, but the low temperature values were all in good agreement. In an attempt to observe a change in the measured D, samples of 50 and 100 M were made. The diffusion coefficients of these samples were measured at 25oC and compared with the higher millimolar concentration samples (Figure 3-21). There appears to be no change from 50 M up to 6.75mM. Additionally, formic acid has been shown to dissolve protein aggregates and help in the analysis of peptides that are not easily solvated100. 2.3 mg of the raw amyloid powder was dissolved in formic acid. The

PAGE 60

60 solution was left to sit for 15 mi nutes and the formic acid was blown to leave slightly wet crystals. The resulting solid was then dissolved in 90% H2O and 10% D2O and the D of the resulting 2.6 mM solution was measured at 25oC (Figure 3-21). This resulted in no change of the observed diffusion coefficient. The lack of any significant difference in the measured D over all the different sample conditions, combined with the chemic al shift result, suggests low-order soluble oligomers were never observed. The signal that was observed must have been from monomeric peptide and the aggregation that occurred resulted in high order oligomers which could not be detected using liquid NMR. Systematic error, such as that which could result from not allowing sufficient ti me to equilibrate the temperature of the sample, is ruled out because the lysozyme data, which was acquired under similar conditions, did not deviate from linearity at lower temperatures. The observation of visible aggregation accompanied by no change in the diffusion coefficient is supported by the recently published findings of resear chers working with the whole 40 residue amyloid-beta peptide101. Filippov et al. noted that it is surprising that the visually observed state of the sample and the self-diffusion do not completely correspond to one another. They came to a similar conclusion: In the course of pept ide aggregation, no signal attributed to dimers arises, while th e aggregation itself is easy to observe visually. They also noted that it has pr eviously been found that the low-order oligomers are not stable intermediate products in the course of A aggregation102;103. Collectively, the NMR data presented in th is chapter indicate that the visible aggregates (Figure 3-11) are too large to be observed by high-resolution NMR spectra acquired under the given conditions due to their excessively short T2 relaxation times104.

PAGE 61

61 Furthermore, the lack of any change in the di ffusion coefficient under conditions which induce visible aggregates also indicates that it is unlikely any significant amounts of loworder oligomers such as dime rs and trimers are formed. Wh ile the low-order oligomers could result in peaks unseen upon visual inspec tion of the spectra, they would affect the diffusion measurements. The signal intensities, A values, used to calculate D (Eq. 3-2) represent the total intensities across a given chemical shift range. If low-order oligomers were present, the calculated D would represent a weighted average of the diffusion coefficients based on the relative concentrations of the monomer and loworder oligomers. However, as is shown in the following, this possibility is excluded by the close agreement between the observed diffu sion coefficient and the theoretically predicted value for monomeric A(25-35). The diffusion coefficient of a particle is re lated to its size by the Stokes-Einstein equation: 6B HkT D r (3-6) where kB is the Boltzmann constant, T is temperature, is solvent viscosity, and rH is hydrodynamic radius. The rH of a peptide is given by105;106: 1 3 211 03() 4HMVV r N (3-7) where M is mass, V1 and V2 are the specific volumes of the particle and solvent, 1 is the solvent fraction bonded to the particle, and No is Avogadros number. The selfdiffusion coefficient of a particle with mass M is given by101:

PAGE 62

62 1 3 0 2114 63()BN kT D MVV (3-8) where is the form factor and is particle density. When calculating D, the following assumptions were made: th e monomer has a globular c onformation, particles are spherical ( =1), and the effects of interpar ticle interactions can be ignored101. The expected diffusion coefficient for A (25-35) at 25oC was calculated using the following values from the literature, = 1300 kg/m3 104, = 9.0882 10-4 kg/ms (the viscosity of water corrected for the 10% volume fraction of D2O), and 1 = 0.328107. The calculated diffusion coefficient was 2.71 10-6 cm2/s, which compares very well with the experimentally determi ned value of 2.69 10-6 cm2/s. This result is consistent with the absence of any liquid-NMR-observable low-order oligomers. A minimum mass for the non-observable high-order oligomers was estimated using an assumed maximum observable spectr al line width. The full-width-at-halfmaximum NMR line width, FWHM, for a nucleus dominated by spin-spin relaxation due to dipolar interactions, as would be expected for protons in the visible aggregates, is given by: 21FWHMT (3-9) where3: 12 23 3(0)5()2(2) 20 TbJJJ (3-10) The normalized spectral density function J is given by3: 22() 1c cJ (3-11)

PAGE 63

63 where c is the rotational correlation time The dipolar coupling constant, b from eq. 310 is given by3: 2 0 34 b r (3-12) where r is the distance between the spins. For th is work a distance of 0.2 nm was used, corresponding to a b/2 (for conversion from radi ans to Hz) of -15.012 kHz 3. The rotational correlation time, c, is given by108: 12( )cxyzDDD (3-13) According to the Perrin elli psoidal shape approximation109;110, the diffusion tensor elements Dx, Dy, and Dz are given by: B l lkT D f (3-14) where: 22 2216() 3()yz x yzaa f aQaR (3-15) 22 2216() 3()xz y zxaa f aRaP (3-16) 22 2216() 3()xy z xyaa f aPaQ (3-17) Here, ax, ay, and az are the semi-axes of a general ellipsoid. The parameters P Q, and R are the following ellipsoidal integrals: 0 23221 ()()()xyzPd s asasas (3-18)

PAGE 64

64 0 23221 ()()()yzxQd s asasas (3-19) 0 23221 ()()()zxyRd s asasas (3-20) Combining equations 3-9, 3-10, 3-13, and 3-14 relates aggregate size to spectral line width. Three geometric models were used to estimate line width, a 2D beta-sheet topology, a spherical model, and an ellip soidal model. In all the models n is the number of monomer units in the aggregat e, the length of the distended A (25-35) monomer was assumed to be 4.4 nm in length and 1 nm in thickness. The hydration layer was taken to be 2.8 nm per side104. Approximating the 2D betasheet as an ellipsoid with a constant thickness in the y axis, the dim ensions (in cm) along t he x, y, and z axis increase with n as: 84410za 8(10.022.28)10ya 8()(10.022.28)10xann For the ellipsoid model where ax= ay: 84410za 1 8 2()()(10.022.28)10xyanann For the spherical model the hydrodynamic r adius was taken to be 0.886 nm (from eq. 37). The volume of the monomer (in cm3) is given by: 834 (8.8610) 3mV

PAGE 65

65 Taking the volume of the oligomer to be proportional to n : mVVn The radius increases with n as: 1 33 ()()() 4m xyznV ananan In all three models, a line width of ov er 100 Hz is obtained after only around 10 monomer units (giving a total mass of ~10 kDa) have aggregat ed (Figure 3-22).

PAGE 66

66 Figure 3-1. Amplitude of 10 mg/mL lysozyme signal for thr ee spectral regions over a range of gradient strengths.

PAGE 67

67 Figure 3-2. Amplitude of 10 mg/mL BSA signal for three spectral regions over a range of gradient strengths.

PAGE 68

68 Figure 3-3. Diagram of external variable-temperature unit used to regulate the sample temperature.

PAGE 69

69 Figure 3-4. 1H spectrum of 0.69 mM A (25-35) in 80% H2O and 20% d3-TFE at 10oC referenced to internal DSS. Spectrum represents 256 scans acquired with the pr esat pulse sequence. Spectrum has been baseline corrected and processed with 0.5 Hz line broadening.

PAGE 70

70 Figure 3-5. TOCSY spectrum of 6.71 mM A (25-35) in 90% H2O and 10% D2O at 25oC. Processed with a Gaussian weighting function in both dimensions such that the signal decays to zero before the end of the fid.

PAGE 71

71 Figure 3-6. Dependence of the homonuc lear nuclear Overhauser enhancement ( ) on the spectrometer frequency ( o) and correlation time ( c) for A) NOESY and B) ROESY.

PAGE 72

72 Figure 3-7. Expansion of t he ROESY spectrum of 0.69 mM A (25-35) in 80% H2O and 20% d3-TFE. Shows a portion of the amide-proton region in the f1 dimension (x axis) and a portion of the alpha-proton region in the f2 dimension (y axis).

PAGE 73

73 Figure 3-8. Expansion of t he ROESY spectrum of 6.71 mM A (25-35) in 90% H2O and 10% D2O. Shows a portion of the alpha proton region in the f1 dimension (x axis) and a portion of the amide proton region in the f2 dimension (y axis).

PAGE 74

74 Figure 3-9. plot for the alpha protons of 6.71 mM A (25-35) in 90% H2O and 10% D2O at pH 2.53 compared to random-coil chemical shifts.

PAGE 75

75 Figure 3-10. plot for the alpha protons of 0.69 mM A (25-35) in 80% H2O and 20% d3-TFE at pH 3.23 compared to random-coil chemical shifts.

PAGE 76

76 Figure 3-11. Two visibly cloudy A (25-35) samples and one clear sample.

PAGE 77

77 Figure 3-12. The equilibrium of A (25-35) in solution, an d the effect of each transition. The high order oligomers are not observable by solution phase NMR.

PAGE 78

78 Figure 3-13. Two spectra showing the decrease in A (25-35) signal intensity of the sa mples with visible aggregation. The intensities of the spectra were normalized using the peaks at 1.1 and 3.5 ppm which represent an unknown impurity present in the solid pept ide as received. A) 3.26 mM A (25-35), no visible aggregat ion. B) 3.13 mM A (25-35) with 5:1 molar ratio of Zn(II), visible aggregation.

PAGE 79

79 Figure 3-14. The chemical shifts of the amide prot ons of 1.88 mM A (25-35) in 90% H2O and 10% D2O over a range of temperatures. Glycine-25 was not observable due to solvent exchange.

PAGE 80

80 Figure 3-15. Diffusion of lysozyme over a range of temperatures. Error bars represent st andard deviation of the mean (n=3).

PAGE 81

81 Figure 3-16. Typical results from a pfg experiment. The methyl peaks displayed wi th relative intensity over 15 gradient strengths.

PAGE 82

82 Figure 3-17. Diffu sion of 2.82 mM A (25-35) over a range of temperatures with linear regression applied over all temperatures (yellow) and from 25-70oC (red). The error bars represent st andard deviation of the mean (n=3). The plots below represent the residuals resu lting from the corresp onding regression.

PAGE 83

83 Figure 3-18. The diffusion coefficient of A (25-35) for three concentrations in 90% H2O and 10% D2O from 5-55oC. The error bars represent standard dev iation of the mean (n=3).

PAGE 84

84 Figure 3-19. The diffusion coefficient of 2.82 mM A (25-35) with a change in pH over a range of temperatures. The error bars represent the standard deviation of the mean (n=3).

PAGE 85

85 Figure 3-20. The diffusion coefficient of A (25-35) with the addition of 3:1 molar rati o Zn(II) over a range of temperatures. The error bars represent standar d deviation of the mean (n=3).

PAGE 86

86 Figure 3-21. The diffusion coefficient of A (25-35) at 25oC over a large range of concentrations. The error bars represent standard deviation of the mean (n=3). The yellow point correspond s to the diffusion coefficient measured after dissolving the peptide in and blowing off formic acid followed by dissolving in a 90% H2O and 10% D2O.

PAGE 87

87 Figure 3-22. The calculated line widt hs over a range of degrees of oligomerization fo r three geometric models.

PAGE 88

88 Table 3-1. Chemical sh ifts (ppm) of 6.71 mM A (25-35) in 90% H2O and 10% D2O at pH 2.53 Residue HN H H H H H other Gly-25 8.05a 3.90 Ser-26 8.68 4.52 3.89, 3.85 Asn-27 8.63 4.75a 2.85, 2.80 NH2 7.61, 6.94 Lys-28 8.43 4.29 1.87, 1.77 1.45 1.68 3.00 NH3 + 7.52 Gly-29 8.39 3.94 Ala-30 8.11 4.36 1.38 Ile-31 8.12 4.16 1. 88 1.50, 1.22 0.87a CH3 0.89a Ile-32 8.12 4.16 1. 88 1.50, 1.22 0.92a CH3 0.93a Gly-33 8.42 3.94 Leu-34 8.05 4.38 1.62 1.63 0.93, 0.88a Met-35 8.37 4.52 2.18, 2.05 2.63, 2.53 2.10 aChemical shift could not be accurately determined from 1H spectrum due to overlap or solvent exchange, value was determined from TOCSY spectrum with an error of 0.02 ppm. Table 3-2. Chemical sh ifts (ppm) of 3.47 mM A (25-35) in 90% H2O and 10% D2O at pH 2.60 Residue HN H H H H H other Gly-25 N.O. N.O. Ser-26 8.69 4.53a 3.89, 3.85 Asn-27 8.64 N.O. 2.85, 2.80 NH2 7.61, 6.94 Lys-28 8.43 4.30a 1.87, 1.77 1.45 1.68 3.00 NH3 + 7.52 Gly-29 8.43 3.94 Ala-30 8.07 4.32a 1.37 Ile-31 8.17 4.16 1. 85 1.49, 1.20 0.87a CH3 0.89a Ile-32 8.23 4.16 1. 86 1.49, 1.20 0.91a CH3 0.92a Gly-33 8.39 3.94 Leu-34 8.07 4.32a 1.63 1.63 0.93, 0.88a Met-35 8.37 4.51a 2.18, 2.03 2.62, 2.54 2.10 aChemical shift could not be accurately determined from 1H spectrum due to overlap or solvent exchange, value was determined from TOCSY spectrum with an error of 0.02 ppm. N.O. = not observed due to solvent exchange or overlap with solvent peak.

PAGE 89

89 Table 3-3. Chemical sh ifts (ppm) of 0.69 mM A (25-35) in 80% H2O and 20% d3-TFE at pH 3.23 Residue HN H H H H H other Gly-25 8.14a 3.91 Ser-26 8.65 4.54 3.90, 3.87a Asn-27 8.58 4.76a 2.87, 2.82 NH2 7.60, 6.89 Lys-28 8.38 4.28 1.89, 1.78 1.46 1.70 3.01 NH3 + 7.57a Gly-29 8.35 3.93a Ala-30 7.98 4.34 1.39 Ile-31 7.96 4.17 1. 90 1.51, 1.20 0.90a CH3 0.91a Ile-32 8.00 4.17 1. 89 1.51, 1.22 0.94a CH3 0.96a Gly-33 8.27 3.94a Leu-34 7.88 4.41 1.64 1.64 0.96, 0.89a Met-35 8.05 4.45a 2.17, 2.03 2.60, 2.54 2.10 aChemical shift could not be accurately determined from 1H spectrum due to overlap or solvent exchange, value was determined from TOCSY spectrum with an error of 0.02 ppm. Table 3-4. Chemical sh ifts (ppm) of 0.69 mM A (25-35) in 80% H2O and 20% d3-TFE at pH 3.19 with 1:1 molar ratio of Zn2+ Residue HN H H H H H other Gly-25 8.15a 3.92a Ser-26 8.65 4.54 3.91, 3.86a Asn-27 8.58 4.76a 2.87, 2.82 NH2 7.60, 6.89 Lys-28 8.38 4.29 1.89, 1.78 1.45 1.70 3.01 NH3 + 7.57a Gly-29 8.35 3.93a Ala-30 7.98 4.34 1.39 Ile-31 7.96 4.16 1. 90 1.51, 1.21 0.90a CH3 0.91a Ile-32 8.00 4.16 1. 88 1.51, 1.22 0.92a CH3 0.94a Gly-33 8.26 3.95a Leu-34 7.88 4.41 1.64 1.64 0.94, 0.89a Met-35 8.06 4.46 2.18, 2.04 2.60, 2.53 2.10 aChemical shift could not be accurately determined from 1H spectrum due to overlap or solvent exchange, value was determined from TOCSY spectrum with an error of 0.02 ppm.

PAGE 90

90 Table 3-5. Chemical sh ifts (ppm) of 0.60 mM A (25-35) in 80% H2O and 20% d3-TFE at pH 3.24 with 3:1 molar ratio of Zn2+ Residue HN H H H H H other Gly-25 8.09a 3.88 Ser-26 8.65 4.54a 3.90, 3.87a Asn-27 8.58 4.76a 2.87, 2.82 NH2 7.60, 6.89 Lys-28 8.38 4.29 1.89, 1.79 1.47 1.70 3.01 NH3 + 7.58a Gly-29 8.35 3.94a Ala-30 7.98a 4.34 1.40 Ile-31 7.97a 4.17 1.89 1.51, 1.21 0.90a CH3 0.91a Ile-32 8.00a 4.17 1.89 1.51, 1.21 0.92a CH3 0.93a Gly-33 8.27 3.95a Leu-34 7.88 4.41 1.64 1.64 0.94, 0.89a Met-35 8.01a 4.42 2.16, 2.02 2.59, 2.54 2.10 aChemical shift could not be accurately determined from 1H spectrum due to overlap or solvent exchange, value was determined from TOCSY spectrum with an error of 0.02 ppm. Table 3-6. Chemical sh ifts (ppm) of 0.69 mM A (25-35) in 80% H2O and 20% d3-TFE at pH 3.23 with 0.1:1 molar ratio of Cu2+ Residue HN H H H H H other Gly-25 8.13a 3.91 Ser-26 8.65 4.54a 3.92, 3.86a Asn-27 8.58 4.76a 2.87, 2.82 NH2 7.60, 6.89 Lys-28 8.38 4.29 1.90, 1.79 1.47 1.70 3.01 NH3 + 7.57a Gly-29 8.34 3.93a Ala-30 7.98 4.34a 1.39 Ile-31 7.97 4.17 1. 89 1.51, 1.21 0.87a CH3 0.91a Ile-32 8.00 4.17 1. 89 1.51, 1.22 0.90a CH3 0.94a Gly-33 8.26 3.95a Leu-34 7.88 4.42a 1.64 1.64 0.94, 0.89a Met-35 8.04 4.44a 2.17, 2.04 2.60, 2.54 2.10 aChemical shift could not be accurately determined from 1H spectrum due to overlap or solvent exchange, value was determined from TOCSY spectrum with an error of 0.02 ppm.

PAGE 91

91 Table 3-7. Chemical sh ifts (ppm) of 0.69 mM A (25-35) in 80% H2O and 20% d3-TFE at pH 3.08 with 1:1 molar ratio of Cu2+ Residue HN H H H H H other Gly-25 8.14a 3.88a Ser-26 8.65 4.55 3.91, 3.87 Asn-27 8.57 4.76 2.88, 2.83 NH2 7.60, 6.89 Lys-28 8.38 4.29 1.89, 1.79 1.47 1.70 3.02 NH3 + 7.57a Gly-29 8.34 3.94 Ala-30 7.98 4.34 1.40 Ile-31 7.96 4.18 1. 89 1.52, 1.22 0.87a CH3 0.90a Ile-32 8.04 4.18 1. 89 1.52, 1.22 0.90a CH3 0.93a Gly-33 8.26 3.96 Leu-34 7.87 4.42 1.64 1.64 0.94, 0.90a Met-35 8.05 4.45 2.20, 2.02 2.58a 2.11 aChemical shift could not be accurately determined from 1H spectrum due to overlap or solvent exchange, value was determined from TOCSY spectrum with an error of 0.02 ppm. Table 3-8. Temperature dependence of the amide chemic al shifts of 2.82 mM A (2535) in 90% H2O and 10% D2O. Residue T (ppb/oC) Ser-26 -6.2 0.2 Asn-27 -6.8 0.1 Lys-28 -7.6 0.1 Gly-29 -8.2 0.2 Ala-30 -5.8 0.1 Ile-31 -8.5 0.2 Ile-32 -9.3 0.2 Gly-33 -7.0 0.2 Leu-34 -7.6 0.1 Met-35 -9.4 0.4 Glycine-25 was not observabl e due to solvent exchange.

PAGE 92

92 Table 3-9. The diffusion coefficien ts of three concentrations of A (25-35) in 90% H2O and 10% D2O from 5-55oC. Diffusion Coefficient 106 (cm2/s) Temperature (oC) 1.88 mM 3.76 mM 6.34 mM 5 1.41 0.02 1.42 0.05 1.42 0.06 15 1.95 0.07 1.93 0.11 1.93 0.05 25 2.51 0.04 2.58 0.04 2.54 0.04 35 3.24 0.04 3.17 0.06 3.18 0.07 45 3.96 0.05 3.99 0.03 3.96 0.10 55 4.76 0.06 4.76 0.04 4.71 0.03 Table 3-10. The diffusion coefficient s of two pH values of 2.82 mM A (25-35) in 90% H2O and 10% D2O from 5-70oC. Diffusion Coefficient 106 (cm2/s) Temperature (oC) pH = 2.91 pH = 3.93 5 1.38 0.14 1.47 0.09 7 1.54 0.09 1.65 0.08 10 1.65 0.16 1.77 0.16 12 1.78 0.09 1.99 0.17 15 1.93 0.16 2.16 0.29 18 2.10 0.13 2.35 0.25 20 2.27 0.14 2.44 0.06 22 2.39 0.12 2.59 0.24 25 2.69 0.12 2.91 0.09 30 3.02 0.20 3.11 0.10 35 3.26 0.16 3.60 0.16 40 3.77 0.32 4.02 0.09 45 4.17 0.10 4.29 0.12 50 4.47 0.09 4.74 0.13 55 4.86 0.14 5.16 0.13 60 5.52 0.23 5.63 0.30 65 6.07 0.13 5.89 0.13 70 6.56 0.12 6.40 0.15

PAGE 93

93 Table 3-11. The diffusion coefficients 2.82 mM A (25-35) in 90% H2O and 10% D2O with and without zinc from 5-55oC. Diffusion Coefficient 106 (cm2/s) Temperature (oC) no Zinc Zinc 5 1.38 0.14 1.36 0.05 7 1.54 0.09 1.52 0.10 10 1.65 0.16 1.66 0.06 12 1.78 0.09 1.70 0.06 15 1.93 0.16 1.90 0.05 18 2.10 0.13 2.05 0.13 20 2.27 0.14 2.16 0.10 22 2.39 0.12 2.38 0.05 25 2.69 0.12 2.51 0.13 30 3.02 0.20 2.82 0.21 35 3.26 0.16 3.12 0.13 40 3.77 0.32 3.43 0.10 45 4.17 0.10 3.72 0.20 50 4.47 0.09 4.08 0.09 55 4.86 0.14 4.40 0.08

PAGE 94

94 CHAPTER 4 CONCLUSION The coupling constants of fluorinated cyclop ropanes were examined. In general it was found that three-bond cis couplings were positive and three-bond trans couplings were negative. However, when one of the fluorines was geminal to an oxygen, the trans couplings were found to be positive. Addi tionally, the effect of temperature on the magnitude of the coupling cons tants was examined. The pr eviously reported trend that the coupling constants increas e in magnitude with increas ing temperature was found to be violated in multiple instances. The si gns of the coupling c onstants should not be used to determine relative geometry of fluorines in cyclopropanes, and the change in magnitude of coupling constant s with a change in temperature cannot reliably be used to predict the correct sign of the coupli ng constants. The aggregation of amyloid bet a (25-35) was examined extensively with different sample conditions. The chemical shi fts were measured and compared to compiled tables of known random-coil chemical shifts. In all sample cond itions the observable peptide was found to maintain a random-coil conformation. T he chemical shifts of the amide protons were measured over a r ange of temperatures and again the values indicated the peptide exists as a random-c oil under the sample conditions used. Pulsed-field-gradient NMR was used to measure the diffusion coefficient of the peptide. There was little change with many changes in sample conditions. The pH was adjusted from approximately three to f our with little effect on the measured diffusion coefficient despite the observation that samples made at near neutral pH rapidly turned to a very viscous gel. Zinc was added to the solution and, although the samples became visibly cloudy, there was still no change in the obser ved diffusion coefficient. Finally, the

PAGE 95

95 peptide was dissolved in formic acid. The formic acid was blown off and the resulting solid dissolved in water. This too re sulted in no change of the observed diffusion coefficient. Despite the samples visibly aggregating, the observed diffusion coefficient never changed. This indicates we only obs erved monomeric peptide in solution and the aggregation resulted in large unobs ervable high-order oligomers. Suggestions for future work include increasing the ratio of trifluoroethanol. Other researchers have observed alpha helical structures in samples with as high as 50% TFE or other membrane mimicking solvents. Additionally, acquiring ROESY spectra with an array of mixing times might resu lt in the observati on of NOE peaks between non-neighboring residues. However, this woul d demand a great deal of instrument time even at relatively high concentrations. A nother possible study would be measuring the diffusion coefficient from day to day over a long period of time to see if there are any changes as the sample settles.

PAGE 96

96 LIST OF REFERENCES (1) Pauli, W. Naturwissenschaften 1924, 12, 741-743. (2) Rabi, I. I.; Millman, S.; Kusch, P.; Zacharias, J. R. Physical Review 1939, 55, 0526-0535. (3) Levitt, M. H. Spin Dynamics ; John Wiley & Sons Ltd: New York, NY, 2001. (4) Ramsey, N. F.; Purcell, E. M. Physical Review 1952, 85, 143-144. (5) Becker, E. D. High resolution NMR theor y and chemical applications ; Academic Press: San Diego, CA, 2000. (6) Bloch, F. Physical Review 1954, 96, 944. (7) Maher, J. P.; Evans, D. F. Proceedings of the Chem ical Society of London 1961, 208. (8) Freeman, R.; Anderson, W. A. Journal of Chemical Physics 1962, 37, 20532073. (9) Freeman, R.; Whiffen, D. H. Molecular Physics 1961, 4 321-325. (10) Evans, D. F. Molecular Physics 1962, 5 183-187. (11) Evans, D. F. Discussions of the Faraday Society 1962, 139-142. (12) Stejskal, E. O.; Tanner, J. E. Journal of Chemical Physics 1965, 42, 288-292. (13) Freund, A. Monatshefte fur Chemie / Chemical Monthly 1882, 3 625-635. (14) Lucas, G. H. W.; Henderson, V. E. Canadian Medical Association Journal 1929, 21, 173-175. (15) Dimakopoulos, A. C. Curr.Alzheimer Res. 2005, 2 19-28. (16) Campiglia, P.; Esposit o, C.; Scrima, M.; Gomez-Mont errey, I.; Bertamino, A.; Grieco, P.; Novellino, E.; D'Ursi, A. M. Chemical Biology & Drug Design 2007, 69, 111-118. (17) Hardy, J. A.; Higgins, G. A. Science 1992, 256, 184-185. (18) Selkoe, D. J. Alzheimers Disease: A Comp endium of Current Theories 2000, 924, 17-25. (19) Esler, W. P.; Wolfe, M. S. Science 2001, 293, 1449-1454.

PAGE 97

97 (20) Danielsson, J.; Andersson, A.; Jarvet, J .; Graslund, A. Magnetic Resonance in Chemistry 2006, 44, S114-S121. (21) D'Ursi, A. M.; Armen ante, M. R.; Guerrini, R.; Sa lvadori, S.; Sorrentino, G.; Picone, D. Journal of Medicinal Chemistry 2004, 47, 4231-4238. (22) Ricchelli, F.; Drago, D.; Filippi, B.; Tognon, G.; Zatta, P. Cellular and Molecular Life Sciences 2005, 62, 1724-1733. (23) Jarvet, J.; Damber g, P.; Bodell, K.; Eriksson, L. E. G.; Graslund, A. Journal of the American Chemical Society 2000, 122, 4261-4268. (24) Kayed, R.; Head, E.; Thompson, J. L.; McIn tire, T. M.; Milton, S. C.; Cotman, C. W.; Glabe, C. G. Science 2003, 300, 486-489. (25) Quist, A.; Doudevski, L.; Lin, H.; Az imova, R.; Ng, D.; Frangione, B.; Kagan, B.; Ghiso, J.; Lal, R. Proceedings of the Nati onal Academy of Sci ences of the United States of America 2005, 102, 10427-10432. (26) Graham, J. D.; Rogers, M. T. Journal of the American Chemical Society 1962, 84, 2249-&. (27) Grant, D. M. Journal of Chemical Physics 1961, 34, 699-700. (28) Closs, G. L.; Closs, L. E. Journal of the American Chemical Society 1960, 82, 5723-5728. (29) Smidt, J.; Deboer, T. J. Recueil des Travaux Chimiques des Pays-Bas-Journal of the Royal Netherlands Chemical Society 1960, 79, 1235-1243. (30) Williamson, K. L.; Lanf ord, C. A.; Nicholso, C. R. Journal of the American Chemical Society 1964, 86, 762-&. (31) Hutton, H. M.; Schaefer, T. Canadian Journal of Chem istry-Revue Canadienne de Chimie 1963, 41, 2774-&. (32) Schrumpf, G.; Luttke, W. Tetrahedron Letters 1969, 2635-&. (33) Sargeant, P. B. Journal of Organic Chemistry 1970, 35, 678-&. (34) Williamson, K. L.; Braman, B. A. Journal of the American Chemical Society 1967, 89, 6183-&. (35) Cavalli, L. Organic Magnetic Resonance 1970, 2 233-244. (36) Pena, F.; Ordaz, B.; Balleza-Tapia, H.; Bernal-Pedr aza, R.; Marquez-Ramos, A.; Carmona-Aparicio, L.; Giordano, M. Hippocampus 2010, 20, 78-96.

PAGE 98

98 (37) Kaminsky, Y. G.; Marlatt, M. W.; Smith, M. A.; Kosenko, E. A. Experimental Neurology 2010, 221, 26-37. (38) Tomas, M.; Garcia, N.; Sant afe, M. M.; Lanuza, M.; Tomas, J. Journal of Alzheimers Disease 2009, 18, 877-884. (39) Kong, L. P.; Wu L. Z.; Zhang, J.; Liao, Y. P.; Wang, H. Q. Neural Regeneration Research 2009, 4 405-412. (40) Fu, Z. M. ; Luo, Y.; Derreumaux P.; Wei, G. H. Biophysical Journal 2009, 97, 1795-1803. (41) Arif, M.; Chikuma, T.; Ahmed, M. M.; Nakazato, M.; Smith, M. A.; Kato, T. Neuroscience 2009, 164, 1199-1209. (42) Liu, R.; Gao, M.; Qiang, G. F.; Zhang, T. T.; Lan, X.; Ying, J.; Du, G. H. Neuroscience 2009, 162, 1232-1243. (43) Hughes, E.; Burke, R. M.; Doig, A. J. Journal of Biological Chemistry 2000, 275, 25109-25115. (44) Mikol, V.; Hirsch, E.; Giege, R. Journal of Molecular Biology 1990, 213, 187-195. (45) Cho, C. H. ; Urquidi, J.; Singh, S .; Robinson, G. W. Journal of Physical Chemistry B 1999, 103, 1991-1994. (46) Longsworth, L. G. Journal of Physical Chemistry 1960, 64, 1914-1917. (47) Ilyina, E.; Roongta, V.; P an, H.; Woodward, C.; Mayo, K. H. Biochemistry 1997, 36, 3383-3388. (48) Wattenbarger, M. R.; Bloomfiel d, V. A.; Bu, Z.; Russo, P. S. Macromolecules 1992, 25, 5263-5265. (49) Momot, K. I.; Kuchel, P. W. Concepts in Magnetic Resonance Part A 2006, 28A 249-269. (50) Johnson, C. S. Progress in Nuclear Magnetic Resonance Spectroscopy 1999, 34, 203-256. (51) Sorland, G. H.; Seland, J. G.; Krane, J.; Anthonsen, H. W. Journal of Magnetic Resonance 2000, 142, 323-325. (52) Jerschow, A.; Muller, N. Journal of Magnetic Resonance 1997, 125, 372-375. (53) Szilagyi, L. Progress in Nuclear Magnet ic Resonance Spectroscopy 1995, 27, 325-443.

PAGE 99

99 (54) Wishart, D. S.; Sykes, B. D. Nuclear Magnetic Resonance, Pt C 1994, 239, 363392. (55) Wishart, D. S.; Bigam, C. G.; Holm, A.; Hodges, R. S.; Sykes, B. D. Journal of Biomolecular Nmr 1995, 5 67-81. (56) Wishart, D. S.; Bigam, C. G.; Yao, J.; Abildgaard, F. ; Dyson, H. J.; Oldfield, E.; Markley, J. L.; Sykes, B. D. Journal of Biomolecular Nmr 1995, 6 135-140. (57) Wuthrich, K. NMR of Proteins and Nucleic Acids ; Wiley-Interscience: New York, 1986. (58) Cavanagh, J.; Fair brother, W. J.; Palmer III, A. R.; Skelton, N. J. Protein NMR Spectroscopy Principles and Practice ; Academic Press, Inc.: New York, NY, 1996. (59) Bothnerby, A. A.; Stephens, R. L.; Lee, J. M.; Warr en, C. D.; Jeanloz, R. W. Journal of the American Chemical Society 1984, 106, 811-813. (60) Dalgarno, D. C.; Levine, B. A.; William s, R. J. P. Bioscience Reports 1983, 3 443-452. (61) Pastore, A.; Saudek, V. Journal of Magnetic Resonance 1990, 90, 165-176. (62) Wishart, D. S.; Syke s, B. D.; Richards, F. M. Febs Letters 1991, 293, 72-80. (63) Wishart, D. S.; Syke s, B. D.; Richards, F. M. Biochemistry 1992, 31, 1647-1651. (64) Osapay, K.; Case, D. A. Journal of Biomolecular Nmr 1994, 4 215-230. (65) Szilagyi, L.; Jardetzky, O. Journal of Magnetic Resonance 1989, 83, 441-449. (66) Williamson, M. P. Biopolymers 1990, 29, 1428-1431. (67) Wishart, D. S.; Syke s, B. D.; Richards, F. M. Journal of Molecular Biology 1991, 222, 311-333. (68) Clayden, N. J.; Williams, R. J. P. Journal of Magnetic Resonance 1982, 49, 383396. (69) Wishart, D. S.; Nip, A. M. Biochemistry and Cell Biology 1998, 76, 153-163. (70) Merutka, G.; Dys on, H. J.; Wright, P. E. Journal of Biomolecular Nmr 1995, 5 1424. (71) Urry, D. W.; Masotti, L.; Krivacic, J. R. Biochimica et Biophysica Acta 1971, 241, 600-612.

PAGE 100

100 (72) Dyson, H. J.; R ance, M.; Houghten, R. A.; Wright, P. E.; Lerner, R. A. Journal of Molecular Biology 1988, 201, 201-217. (73) Walsh, D. M.; Lomakin, A.; Benedek, G. B.; Condron, M. M.; Teplow, D. B. Journal of Biological Chemistry 1997, 272, 22364-22372. (74) Walsh, D. M.; Hart ley, D. M.; Kusumoto, Y.; Fezoui, Y.; Condron, M. M.; Lomakin, A.; Benedek, G. B.; Selk oe, D. J.; Teplow, D. B. Journal of Biological Chemistry 1999, 274, 25945-25952. (75) Kirkitadze, M. D.; C ondron, M. M.; Teplow, D. B. Journal of Molecular Biology 2001, 312, 1103-1119. (76) Lee, S.; Kim, Y. Bulletin of the Korean Chemical Society 2004, 25, 838-842. (77) Otvos, L.; Sz endrei, G. I.; Lee, V. M. Y.; Mantsch, H. H. European Journal of Biochemistry 1993, 211, 249-257. (78) Lee, S.; Suh, Y. H.; Kim, S.; Kim, Y. Journal of Biomolecular Structure & Dynamics 1999, 17, 381-391. (79) Kohno, T.; Kobayashi, K.; M aeda, T.; Sato, K.; Takashima, A. Biochemistry 1996, 35, 16094-16104. (80) Sticht, H.; Bayer, P.; Willbold, D.; Dames, S.; Hilbic h, C.; Beyreuther, K.; Frank, R. W.; Rosch, P. European Journal of Biochemistry 1995, 233, 293-298. (81) Bush, A. I. Trends in Neurosciences 2003, 26, 207-214. (82) Adlard, P. A.; Bush, A. I. Journal of Alzheimer's Disease 2006, 10, 145-163. (83) Gaggelli, E.; Kozlowski, H.; Valensin, D.; Valensin, G. Chemical Reviews 2006, 106, 1995-2044. (84) Lau, T. L.; Gehm an, J. D.; Wade, J. D.; Perez, K. ; Masters, C. L. ; Barnham, K. J.; Separovic, F. Biochimica et Biophysi ca Acta-Biomembranes 2007, 1768, 24002408. (85) Faller, P.; Hureau, C. Dalton Transactions 2009, 1080-1094. (86) Talmard, C.; Y ona, R. L.; Faller, P. Journal of Biological Inorganic Chemistry 2009, 14, 449-455. (87) Gaggelli, E.; Janicka -Klos, A.; Jankowska, E.; Kozl owski, H.; Mig liorini, C.; Molteni, E.; Valensin, D.; Va lensin, G.; Wieczerzak, E. Journal of Physical Chemistry B 2008, 112, 100-109.

PAGE 101

101 (88) Bush, A. I.; Pettingell, W. H.; Multhaup, G.; Paradis, M. D.; Vonsattel, J. P.; Gusella, J. F.; Beyreuther, K.; Ma sters, C. L.; Tanzi, R. E. Science 1994, 265, 1464-1467. (89) Huang, X. D.; Atwood, C. S.; Moir, R. D.; Hartshorn, M. A.; Vonsattel, J. P.; Tanzi, R. E.; Bush, A. I. Journal of Biological Chemistry 1997, 272, 26464-26470. (90) Raman, B.; Ban, T.; Yamaguchi, K.; Sakai, M.; Kawai, T.; Naiki, H.; Goto, Y. Journal of Biological Chemistry 2005, 280, 16157-16162. (91) Gaggelli, E.; Grzonka, Z.; Kozlowski, H.; Migliorini, C.; Molteni, E.; Valensin, D.; Valensin, G. Chemical Communications 2008, 341-343. (92) Syme, C. D.; Viles, J. H. Biochimica et Biophysica Ac ta-Proteins and Proteomics 2006, 1764 246-256. (93) Zirah, S.; Kozin, S. A.; Mazur, A. K.; Blond, A.; Cheminant, M.; Segalas-Milazzo, I.; Debey, P.; Rebuffat, S. Journal of Biological Chemistry 2006, 281, 2151-2161. (94) Danielsson, J.; Pierattell i, R.; Banci, L.; Graslund, A. Febs Journal 2007, 274, 4659. (95) Curtain, C. C.; Ali, F.; Volitakis, I.; Cherny, R. A .; Norton, R. S.; Beyreuther, K.; Barrow, C. J.; Master s, C. L.; Bush, A. I.; Barnham, K. J. Journal of Biological Chemistry 2001, 276, 20466-20473. (96) Andersen, N. H.; Neidi gh, J. W.; Harris, S. M.; Lee, G. M.; Liu, Z. H.; Tong, H. Journal of the American Chemical Society 1997, 119, 8547-8561. (97) Jimenez, M. A.; Nieto, J. L.; Rico, M.; Santoro, J. ; Herranz, J.; Bermejo, F. J. Journal of Molecular Structure 1986, 143, 435-438. (98) Zimmermann, G. R.; Legault, P.; Selsted, M. E.; Pardi, A. Biochemistry 1995, 34, 13663-13671. (99) Mills, R. Journal of Physical Chemistry 1973, 77, 685-688. (100) Hazeki, N.; Tukamoto, T.; Goto, J.; Kanazawa, I. Biochemical and Biophysical Research Communications 2000, 277, 386-393. (101) Filippov, A. V.; Suleimanova, A. V.; Grobner, G.; Antsutkin, O. N. Colloid Journal 2008, 70, 501-506. (102) Narayanan, S.; Reif, B. Biochemistry 2005, 44, 1444-1452. (103) Filippov, A.; Sulejmanova, A.; Antzutkin, O.; Grobner, G. Applied Magnetic Resonance 2005, 29, 439-449.

PAGE 102

102 (104) Danielsson, J.; Jarvet, J.; Damberg, P.; Graslund, A. Magnetic Resonance in Chemistry 2002, 40, S89-S97. (105) Cantor, C. R .; Schimmel, P. R. Biophysical Chemistry, Te chniques for the Study of Biological Structure and Function ; W.H. Freeman: San Francisco, CA, 1980. (106) Tanford, C. Physical Chemistry of Macromolecules ; Wiley: New York, NY, 1963. (107) Krishnan, V. V. Journal of Magnetic Resonance 1997, 124, 468-473. (108) Ryabov, Y. E.; Geraghty, C.; Varshney, A.; Fushman, D. Journal of the American Chemical Society 2006, 128, 15432-15444. (109) Perrin, F. Journal de Physique et Le Radium 1934, 5 497-511. (110) Perrin, F. Journal de Physique et Le Radium 1936, 7 1-11.

PAGE 103

103 BIOGRAPHICAL SKETCH David Warren Richardson was born in Orang e Park, Florida, in 1980. His father, originally from the Boston area of Massachusetts, was stationed her e while he was a Captain in the United St ates Navy. David grew up in this suburb of Jacksonville with his parents, older brother, and y ounger sister. He graduated fr om Middlebur g High School in 1998. In the fall of 1998, David began his academic ca reer at the University of Florida, where he studied chemistry. During this time, he worked in the NMR lab under the advisement of Dr. Wallace Brey. In December of 2001, David graduat ed with a Bachelor of Science in Chemistry. David continued to work in the NMR lab at UF for two years post baccalaureate. He returned to school by entering the graduate program at the Univer sity of Florida in August of 2003. Following the guidance of Dr. Brey and Dr. Ion Ghiviriga, he conducted his research and trained fellow researchers how to operate NMR instrumentation.