Investigating the Hydrophobic Collapse and Secondary Structure Formation of Model Polypeptides in Aqueous Salt Solutions

The Pennsylvania State University

The Graduate School

Department of Chemistry

INVESTIGATING THE HYDROPHOBIC COLLAPSE AND SECONDARY

STRUCTURE FORMATION OF MODEL POLYPEPTIDES IN AQUEOUS

SALT SOLUTIONS

A Thesis in

Chemistry

Seung-Yi Lee

Submitted in Partial Fulfillment

of the requirements

for the Degree of

Master of Science

May 2015 The thesis of Seung-Yi Lee was reviewed and approved* by the following:

Paul Cremer Professor of Chemistry Thesis Adviser

Carsten Krebs Professor of Chemistry

Scott Showalter Associate Professor of Chemistry

J. Gregory Ferry Professor of Biochemistry and Molecular Biology

Barbara J. Garrison Head of the Chemistry Department Shapiro Professor of Chemistry

*Signatures are on file in the Graduate School

Abstract

The first part of this thesis discusses the unique hydration of model polymers in

H2O and D2O in order to expand the understanding of the relationship between hydrophobic hydration and solvent hydrogen bonding strength. Two model polymers, poly(dimethyl acrylamide) (PDMA) and poly(diethyl acrylamide) (PDEA), were compared. The uncollapsed state of PDMA was found to be more stable in H2O, while PDEA was more stable in D2O. To examine the origin of these phenomena, the effects of pH, solvent transfer, and ion-specificity on polymer collapse were investigated. Experiments were conducted utilizing light scattering, differential scanning calorimetry (DSC), Fourier transform infrared spectroscopy (FT-IR), and Raman with multivariate curve resolution (MCR). Among the variables tested, the dominant effects were found to be highly dependent on the nature of the model polymer and its interaction with the solvent. PDMA was found to be less hydrophobically hydrated than PDEA and real proteins.

The second part of this thesis describes the specific effects of calcium ions on the secondary structure of elastin-like polypeptides (ELPs). Two ELPs, V-120 and V5A2G3-

120, were employed in this study. The phase transition temperatures of these ELPs were measured by light scattering measurements and correlated with spectroscopic signatures from circular dichroism (CD). At molar concentrations of calcium chloride, the amount of

β-turn structure was found to increase more for V5A2G3 than V-120. This may be due to the fact that V5A2G3 has less number of valine residues in its polypeptide sequence. These molecular level results may contribute to the understanding of the formation of bones in the body and may also aid in the development of the medical and bioengineer materials.

iii

TABLE OF CONTENTS

LIST OF FIGURES··········································································· iv LIST OF TABLES············································································· vi ACKNOWLEDGEMENT··································································· vii CHAPTER I: INTRODUCTION··························································· 1 1. Hydrophobic Effect ································································ 1 2. Effects of Ions on Proteins ························································ 3 3. H2O vs. D2O as a Solvent ························································ 4 4. Model Polymers: acrylamides ···················································· 5 5. Model Polymers: ELPs ···························································· 7 CHAPTER II: ISOTOPE EFFECTS ON HYDRATION ······························ 9 1. Introduction ········································································· 9 2. Experimental········································································ 9 3. Results and Discussion ···························································· 11 4. Conclusion·········································································· 37 CHAPTER III: DIVALENT CATION EFFECTS ON PROTEINS················· 38 1. Introduction········································································· 38 2. Experimental········································································ 39 3. Results and Discussion····························································· 41 4. Conclusion··········································································· 61 CHAPTER IV: KOSMOTROPIC EFFECTS OF PHYTIC ACID··················· 62 1. Introduction········································································· 62 2. Experimental········································································ 64 3. Results and Discussion····························································· 64 4. Conclusion··········································································· 69 REFERENCES················································································· 70

LIST OF FIGURES Figure 1-1 (a) Structure of poly(dimethyl acrylamide) (b)Structure of poly(diethyl acrylamide)····················································································· 6 Figure 1-2 Structure of elastin-like polypeptides (VPGVG unit) ······················ 8

Figure 2-1 LCSTs of PNIPAM in Na3PO4 solutions····································· 14 Figure 2-2 LCSTs of PDEA in Na3PO4 solutions······································· 15 Figure 2-3 LCSTs of PDMA in Na3PO4 solutions······································· 16 Figure 2-4 LCSTs of PDEA in NaSCN solutions········································ 17 Figure 2-5 LCSTs of PDEA in NaCl solutions··········································· 18 Figure 2-6 LCSTs of PDMA in NaCl solutions··········································· 19 Figure 2-7 Differential scanning calroimetry thermograms of 10mg/mL PDEA solutions in D2O and H2O···································································· 21 Figure 2-8 Differential scanning calroimetry thermograms of 10mg/mL PDMA 22 solution in D2O and H2O·····································································

Figure 2-9 FTIR spectrum of PDEA in H2O at 20℃··································· 25

Figure 2-10 FTIR spectrum of PDEA in D2O at 20℃··································· 26

Figure 2-11 FTIR spectrum of PDMA in H2O at 20℃·································· 27

Figure 2-12 FTIR spectrum of PDMA in D2O at 20℃·································· 28

Figure 2-13 Bulk H2O Raman spectra at various temperature·························· 32

Figure 2-14 SC spectra of PDMA in H2O at various temperature······················ 33

Figure 2-15 Bulk D2O Raman spectra at various temperature·························· 34

Figure 2-16 SC spectra of PDMA in D2O at various temperature······················ 35 Figure 2-17 SC spectra of n-butanol························································ 36 Figure 3-1 The LCSTs of V5A2G3-120 in divalent cation chloride salt solutions·· 42 Figure 3-2 The LCSTs of V5-120 in divalent cation salt solutions···················· 43

Figure 3-3 The LCSTs of various model polymers in CaCl2 solutions················ 44

Figure 3-4 The LCSTs of various model polymers in MgCl2 solutions··············· 45

Figure 3-5 The LCSTs of various model polymers in SrCl2 solutions················· 46

Figure 3-6 V5-120 in different concentrations of CaCl2································· 49

Figure 3-7 V5-120 in different concentrations of MgCl2 ································ 50

Figure 3-8 V5-120 in different concentrations of SrCl2 ································· 51

Figure 3-9 V5-120 in different concentrations of NaCl·································· 52

Figure 3-10 V5A2G3-120 in different concentrations of CaCl2 ························· 53

Figure 3-11 V5A2G3-120 in different concentrations of MgCl2························· 54

Figure 3-12 V5A2G3-120 in different concentrations of SrCl2 ·························· 55

Figure 3-13 V5A2G3-120 in different concentrations of urea··························· 56

Figure 3-14 V5A2G3-120 in different concentrations of NaCl·························· 57

Figure 3-15 V5A2G3-120 at various temperature········································· 58 Figure 3-16 Diagram of the hypothesis for Ca2+ binding to VPGGG unit············ 59 2+ Figure 3-17 Diagram of the hypothesis for Mg binding for V5A2G3-120··········· 60 Figure 4-1 The structure of phytic acid. Blue figures: pKas ····························· 63 Figure 4-2 The LCST values at three representative pH conditions····················· 66 Figure 4-3 The LCST of poly(N-isopropyl acrylamide) in sodium phytate solutions at various pH ························································ 67 Figure 4-4 Correlation of the negative charges of phytate and the slope of the LCST data ················································································ 68

LIST OF TABLES

Table 2-1 Summary of LCST measurements in Na3PO4 solution······························· 13 Table 2-2 Normalized intensity at each water peak and its ratio································ 29

vii

ACKNOWLEDGEMENT

I would like to pass many, many thanks to my advisor Prof. Paul Cremer who has been a wonderful mentor

I could ever meet. He was consistently supportive throughout the years with patience and great advising. I appreciate my committee members as well. Dr. Krebs led me to analytical thinking in analyzing data. Dr.

Showalter taught me to critically think as a chemist. Dr. Ferry helped me to look at a bigger picture as a scientist and gave me a chance to think how I could contribute to the society. Also, I thank the members of

Cremer group for help, cooperation, feedback, great discussion, and being together. Special thanks to the

Hofmeister family, Dr. Halil A. Okur and Dr. Kelvin Rembert II. They helped me from the first day of the joining the group until their last day. I expand thanks to Dr. Tinglu Yang, who was always gentle and kind as a colleague and mentor. Our philosophical discussion about life was a good refreshment out of science.

My friends, Dr. Bora Lee, Moon-Jung Kim, Minjung Chae, and Bumsoo Kim, deserve more than thanks.

They became my State College family. They took the best care of me for years. This achievement may have not been possible without their care and love. Along with, I would like to thank Father Jaesik Yang, and the Penn State Korean Catholic Church Community. I appreciate that you prayed for me the whole time.

I also thank my sweet American parents and friends, Dimitri Lazo and Dianne Lazo. They provided me with home to go for holidays, hugs when I was blue, and precious moments to share (also, lots of American food). I am grateful to have you.

I pass indescribable appreciation to my dear friend B, who has fully trusted, supported and encouraged me along the journey.

Endless thanks and love to my dearest mother, Sujin Park, who always respects my choice, and tells me that she has faith in me. Lastly, I thank my family back in South Korea for support.

viii

CHAPTER I: INTRODUCTION

1. Hydrophobic Effect and Hydrophobic Hydration of Proteins

The hydrophobic effect is an important concept in biochemistry to understand the environment that support the life. Since Kauzmann established the insight that protein folding is driven by hydrophobic effect in 1959, it has been widely accepted that the hydrophobic effect is the major force which drives the formation of membrane and micelles, and folding of proteins.1

This idea has been leading many qualitative and quantitative studies and its application.2–7

The hydrophobic effect is a phenomenon that nonpolar substances aggregate to one another when they are transferred into aqueous solutions. A common example is an oil-water mixture. At the interface of oil and water, oil molecules disrupt the hydrogen bonding of water.

Consequently, water excludes the oil molecules and the oil is separated from water. In contrast, some nonpolar molecules preferentially interact with polar molecules when they are in the right size. For instance, a noble gas is slightly soluble in water, which can be described by the term hydrophobic hydration. One of the hydrophobic hydration feature is that it is solute size dependent.

When small molecules like noble gases interact with water, the volume of the gas molecule are small enough to stabilize in the cavities of the water network. This results in solvation of gas in water. Also, the hydrogen bonding of water is enhanced. In other words, water molecules around the hydrophobic solutes are more highly ordered than the water molecules in bulk.8 Thus, the solution becomes enthalpically stable. This hydrophobic hydration is preferred if the volume of a solute is small enough to afford the thermodynamic cost (< 0.5nm in diameter and ideally spherical, the curvature area < 1nm2).5,9 However, the larger scale of the solutes ruptures the interactions between water molecules, and separates the solutes from water.

Another feature is that hydrophobic hydration is the temperature dependent.10 For instance, inert gases can better dissolve in cold water than in hot water. At low temperature, water molecules can form stronger hydrogen bonding than at high temperature, thus they are more capable of surrounding the nonpolar solutes than the hot water molecules. As David Chandler claimed, this argument is valid only if the nonpolar solutes are smaller than 0.5 nm in diameter.

However, proteins which are in a much larger scale undergo denaturation at a low temperature.

This phenomenon is called cold denaturation. Even though Kauzmann argued that the hydrophobic effect stabilizes the native structure of proteins by the force that buries the hydrophobic residues and minimizes the contact of the hydrophobic moieties to adjacent water, he noticed that denaturation of ovalbumin by urea is negatively correlated with temperature below

20ºC. This is the opposite tendency to the experiments conducted above the room temperature.

Therefore, he concluded that the mechanism of cold denaturation is different from that of heat denaturation.

In the early twentieth century, some models to explain cold denaturation were proposed.

Some of them declared that columbic force or inter/intra-hydrogen bonding were the major driving forces.11 Finally, it was shown that the hydrophobic hydration is the dominant interaction that promotes denaturation of proteins, and that ΔH (enthalpy) of protein solvation/unfolding is positive by calorimetry and thermodynamic analyses.12–15 This proves that the hydrophobic hydration of protein (unfolded state) is enthalpically favorable whereas dehydration of protein

(folded state) is entropically favorable at the given conditions. At the molecular level, water molecules that are adjacent to the nonpolar residues of proteins form stronger hydrogen bonding to each other. It results in the stronger hydration layers. Despite that the water molecules lose

some hydrogen bonds, the formation of stronger hydrogen bonding compensates the loss. Once temperature is increased, the entropic effect exceeds the enthalpic effect. Continuously, it causes protein folding and increase in entropy of the water molecules that were strictly constrained at the interfaces. In addition, protein folding accompanies inter/intra-hydrogen bonding, which contributes to reducing overall the enthalpy change of the system.

2. Effects of ions on proteins

Ions specifically interact with proteins in an aqueous solution, regulating salting-out or salting-in effects. Salting-out is an effect where the solubility of a certain protein or crystal substances decreases by salt addition. Salting-in is an effect that increases the solubility of the substances upon salt concentration increase. The recurring effect is known as the Hofmeister series.16

+ 4+ + + + + + 2+ 2+ N(CH3)4 > NH > Cs > Rb > K > Na > Li > Ca > Mg

2- 2- 2------CO3 > SO4 > S2O3 > H2PO4 > F > Cl > Br >NO3 > I > ClO4 >SCN

The species on the left have the tendency to salt-out proteins whereas the species on

17 - - the right of the series tend to salt-in proteins. Ions such as SCN and ClO4 have the low charge- to-volume ratio, which gives a smaller number of hydration and low surface tension at the

2- 2- hydrophobic/water interface. In contrast, ions like SO4 and HPO4 have the high charge to volume ratio, which results in a higher number of hydration and high surface tension. At the water/hydrophobic moiety interface, weakly hydrated ions increase the phase transition temperature by partitioning to the water/hydrophobes interface.18 Indeed, Rembert et al. demonstrated that the weakly hydrated ions such as SCN- interact with the backbone of polypeptide via α-carbon.19 On the other hand, the strongly hydrated ions better exclude water

molecules from the hydrophobes. As a result, the temperature of the phase transition of proteins decreases with the ion concentration increase. In many studies, it has been shown that strongly hydrated ions linearly decrease the phase transition temperature of a biomolecule as a function of

14,20 3- the ion concentration. In this work, PO4 was introduced for the phase transition measurement of the model polymers.

3. H2O vs. D2O as a Solvent

D2O is a common substitute for H2O in protein stability studies. Most physical

21– properties such as boiling point, compressibility, dielectric constant of H2O and D2O are similar.

23 However, other properties between these two solvents are quite different. For example, the viscosity of D2O is 23% higher, temperature of the maximum density is 7.25℃ higher, and the

23–25 heat capacity is 12% higher than of H2O. These indicate that D2O is more structured and deuterium bonding is stronger than hydrogen bonding. The hydrogen bond of D2O is about 5%

25 stronger than hydrogen bond in H2O. Additionally, H2O is a better solvent than D2O for most ions.9 In the most cases, the isotopic effects on the solubility of salts decrease with temperature increase.

The hydrophobic effects in H2O and D2O has been compared by many experiments in the hope of understanding hydrophobic hydration interacting with different strengths of hydrogen bonding. Kreshechk showed that the critical micelle concentration (cmc) of dedecyldimethylphosphine oxide, nonionic surfactant, was found to be always lower in D2O than

26 in H2O. Also, the enthalpy change of the micelle formation was always greater in D2O solutions

26 from 15℃ to 40℃. Moreover, predictably, the solubility of inert gases is higher in D2O than in

27 H2O. These results support that the hydrophobic hydration is stronger in D2O than in H2O. In

addition, Zhang and coworkers measured the ΔH of the phase transition of PNIPAM in H2O and

20 D2O by differential scanning calorimetry (DSC). They found that the ΔH(collapse) of poly(N- isopropyl acrylamide) (PNIPAM) to be greater in D2O. Similarly, the ΔH upon folding of elastin-

14 like peptides (ELPs) was found to be greater in D2O than in H2O.

4. Model polymers: acrylmides

The model polymers are employed in order to study the isotope effects of water (solvent) on hydrophobic collapse. They are poly(dimethyl acrylamide) (PDMA) and poly(diethyl acrylamide) (PDEA). Their structure is depicted in Figure 1-1 (a) and (b). These models are specifically chosen to compare their behavior base on hydrophobicity as the only difference is the length of adjacent alkyl group to amide group.

PDEA undergoes phase transition around 31ºC in pure water. The temperature where

PNIPAM undergoes phase transition is called the lower critical solution temperature (LCST). Its phase transition is sharp enough to be described as coil-to-globule transition.28,29 Therefore, in this study, only uncollpased state and collapsed state of model polymers were considered.

PDMA has high solubility in pure water and its LCST is above 100 ºC .30 From the structure and LCST, one can already find that the hydrophobicity of PDMA is much lower than

PDEA. One method to see the phase transition below 100 ºC is adding salts. Salt addition was used in this work to lower the LCST of PDMA. Other than salts, LCST of model polymers are also affected by the property of solvents, pressure, and its concentration.

CH3 CH3 H C H3C n 3 n

CH3 O O N N CH3

CH3 CH3 Figure 1-1 (a) Structure of poly(dimethyl acrylamide) (b) Structure of poly(diethyl acrylamide)

5. Model Polymer: Elastin-Like Peptides (ELPs)

Elastin-like peptides are thermo-responsive proteins that consist of five amino acid repeats, (VPGXG)n where X can be any specified amino acids except proline and n is the number of repeats (Figure 1-2). For example, V5A2G3-120 represents polypeptides with 5:2:3 valin, aniline, and glycine in the X position, whose total peptide number is 600.

Unlike random aggregation of PDEA and PDMA, ELPs form the specific secondary structures upon folding. The majority of the secondary structure of ELPs is β-turn structure which the first and fourth amides are connected through hydrogen bonding. Thus, ELPs have stronger intra/intermolecular interactions.

The LCSTs of ELPs are also affected by the chain lengths, concentration of itself, solvent, salt identity and concentration.31 In the past several decades, the effects of ion specificity following the series known as the Hofmeister series on biomolecules have been extensively studied.

The specific anion effects of the Hofmeister series on ELPs were investigated by Cho and coworkers.31,32 However, the effects of cations of Hofmeister series is not explicitly known.

Particularly, divalent cations draw our attention since they play the vital roles in human body. For instance, calcium composes the majority of bones and teeth in body, which will be discussed in the Chapter III.

Figure 1-2 Structure of elastin-like polypeptides (VPGVG unit)

CHAPTER II: ISOTOPIC EFFECTS ON HYDRATION

1. Introduction

The hydrophobic effect is the fundamental idea in application of drug delivery studies, membrane bioactivity studies, and applied protein sciences.33–36 It is the driving force of micelle formation, lipid bilayer formation, and specific three-dimensional protein folding. Hence, understanding the hydrophobic effects has been a central issue in biophysics, pharmaceutical, and chemistry.

Briefly, the stability of biopolymers in body is determined by hydrophobic hydration.

Hydrophobic hydration is stabilized at a low temperature on enthalpic ground of water hydrogen bonding, whereas hydrophobic groups aggregate at high temperature on entropic ground.

In this study, two model polymers, PDMA and PDEA, with different alkyl side chain lengths were investigated. The length of alkyl groups determines the opposite stability preference in H2O and D2O. Molecular mechanisms were studied by spectroscopic measurement.

2. Experimental

A. Lower Critical Solution Temperature (LCST) Measurements

The turbidity of the polymer solution was measured on an automated apparatus with a built-in digital camera that captures and records the real-time images of the light scattering of the samples as temperature increase at the ramp rate of 0.5℃/푚푖푛 for PDMA 1℃/푚푖푛 for PDEA

(Optimelt MPA 100—automated melting point system). The Meltview software was used to display and process the data. Each data point represents an average of the six measurements.

B. DSC Measurements

Differential scanning calorimeter data was taken with MicroCal VP-DSC at a temperature ramping rate of 0.5℃/푚푖푛 . Each 5mg/mL, 10mg/mL, and 20mg/mL of polymer solutions were dissolved and refrigerated at 4℃ at least for 24 hours prior to the measurements.

The 0.25M solutions of Na3PO4 in H2O and D2O were used as reference. For the PDMA, pure H2O and D2O were used. The samples were cooled to 5℃ lower than their LCSTs and incubated for 30 minutes before each scan. The temperature range was 5℃ below the LCST to 30℃ above the

LCST to cover the wide range of phase transition. The base line of the thermogram was subtracted.

The heat capacity unit (y-aix) was converted into cal/℃ ∙ 푔.

C. Amide I Band ATR/FTIR Measurements

Fourier transform infrared (FTIR) spectra were taken with a Nicolet 470 FTIR spectrometer that was equipped with a Pike Miracle attenuated total reflection (ATR) attachment with single bounce geometry on a diamond coated ZnSe crystal (Pike Technologies, Madison WI) and liquid nitrogen MTC detector (Thermo ElectronCorporation, Madison, WI). 20mg/mL of

PDMA and PDEA in pure H2O and D2O solutions were prepared at least 24 hours in advance of the measurement and stored at room temperature. 64 scans were accumulated per spectrum from

4000cm-1 to 400cm-1 at 2cm-1 resolution. Background spectra were every time before the collection of the sample spectra and automatically subtracted from the sample spectra by the software. The

D2O solution samples were covered with a plastic cap during the measurement to minimize H/D exchange at the air/water interface.

D. Raman Measurements

In monomer concentration, 250mM of PDMA and PDEA were prepared in H2O and

D2O 24 hours prior to the measurements. The spectra were collected using a home-built back- scattering micro-Raman system with an air-cooled Ar-ion laser (MellesGriot 514.532 nm). It was fiber-coupled to an imaging spectrograph (Acton SpectraPro 300i) with a 600 gr/mm grating and a liquid N2-cooled CCD detector (Princeton Instruments, 1024 X 256 pixel). The spectrum of helium gas were used as the reference peaks. The spectra were collected twice each measurement with a total scan time of five minutes per spectra. Each sample were measured 4-6 times.

Background water spectra were taken before the sample spectra each time.

PDEA was purified by inverse transition cycling (ITC). However, the purity did not reach that the spectra contained aggravated background noise. The SMCR and thermodynamic analysis was performed using Igor Pro 6.22A (WaveMetrics). The more detailed protocol for

Raman measurements can be found in the literature.37,38

3. Results and Discussion

A. LCST measurements

The phase transition temperatures of PNIPAM and PDEA are similar in H2O and D2O solutions (Figure 2-1, 2-2). The LCST of PNIPAM is 1℃ higher in D2O and it is ~0.4℃ higher in

D2O for PDEA. Structurally, one would expect that PNIPAM would be more sensitive to

H2O→D2O transfer than PDEA due to the presence of the N-H group. Indeed, PDEA without the

N-H group is less sensitive to the D2O substitution. Curiously, the LCST behavior of these

polymers in salt solutions is similar. LCST is higher in D2O and the difference in LCST in H2O and D2O increases as the salt concentration increases with both polymers.

However, PDMA presents the opposite trend with the salt concentration increase

(Figure 2-3). As opposed to PNIPAM and PDEA, the calculated value shows that PDMA has a higher LCST in pure H2O. That is, the uncollapsed state of PDMA is more stable in H2O or the collapsed state of PDMA is more stable in D2O. Since the strongly hydrated ions, phosphate and chloride, barely bind to the polymers the LCST and ion concentration can be linearly related as following equation. 20

LCST = To + c[M]

To is LCST in pure solvent, [M] is salt concentration, and c is coefficient that relates LCST and salt concentration. Although the LCSTs of PDMA in pure H2O and D2O are not measureable, they can be extrapolated. Opposite to PNIPAM or PDEA, the difference between the LCSTs of PDMA in H2O and D2O solutions decreases with the salt concentration increase.

One common trend is that absolute c value is always larger (steeper slope) in H2O than in D2O. In general, upon the collapse of the hydrophobic polymer with temperature increase, water-polymer hydrogen bonds break, which consumes energy (endothermic process). However, the degree of freedom that water molecules gain which were the vicinity of the hydrophobes overwhelms the enthalpy increase(∆퐺 < 0). As the strongly hydrated ions reduces the amount of water molecules at the interfaces, less water molecules are available for the polymer hydration, and therefore the LCSTs decrease. The steeper slope in H2O solution involves stronger correlation between the LCSTs and the salt concentrations. The reason may be that tribasic phosphate ion is more strongly hydrated by H2O molecules than D2O molecules. The solubility of most salts is

39 higher in H2O than in D2O at a wide temperature range(5℃ − 90℃). Hence, the interaction

between H2O and phosphate molecules is possibly more pronounced than between D2O and phosphate ions. The LCST measurement comparison is summarized in Table 2-1.

Table 2-1 Summary of the LCST measurements in Na3PO4 solutions PDEA PDMA H2O D2O H2O D2O LCST (℃) 31.5 < 31.9 165 > 154 | ΔΔ H2O - D2O | increases decreases | c | (slope) H2O > D2O H2O > D2O

Similar LCST behavior can be also seen in NaCl and NaSCN solutions. Therefore the phenomenon is non-ion specific. Additionally, it shows that pH and the specific ion may not be the dominant factors of the trend. Regardless of the specific ion effects, the difference in LCSTs between H2O and D2O at the same concentration of a certain salt diverged with the concentration increase. Similarly, the LCSTs of PDMA were measured in NaCl solutions. As it does in sodium

phosphate solution, the ΔLCST(H2O-D2O) decreases with the salt concentration increase.

o 32 1 C D O 2 28 H O 2

O 24

LCST ( 16

o 12 3.5 C

0 50 100 150 200 250 Na PO (mM) 3 4

Figure 2-1 LCSTs of PNIPAM in Na3PO4 solutions. PNIPAM is better hydrated in

D2O solution than in H2O solution. The difference of the LCSTs diverges more with decrease in LCST and increase in salt concentration. c value is -13.90 ℃⁄푚푀 for D2O and -15.17 ℃⁄푚푀 for H2O.

30 D O 28 2 H O 26 2 24

C) O 22 20 18 LCST ( 16 14 0 40 80 120 160 200 Na PO (mM) 3 4

Figure 2-2 LCSTs of PDEA in Na3PO4 solutions. D2O is a better solvent for PDEA as for PNIPAM. The LCST diverges as salt concentration increases. The observed trend as a function of temperature and salt concentration resembles that of PNIPAM. c value is -

85.02 ℃⁄푚푀 for D2O and -87.67 ℃⁄푚푀 for H2O.

88 5oC 80 D O 2 72 H O 2

O 64

56 1.5oC 48

LCST ( 40

200 220 240 260 280 300 320 340 Na PO (mM) 3 4

Figure 2-3 LCSTs of PDMA in Na3PO4 solutions. Contrary to PNIPAM and PDEA,

H2O is a better solvent for PDMA. Also, with increase in salt concentration and temperature decrease, LCSTs in H2O and D2O solutions immerge. c value is -

362.17 ℃⁄푚푀 for D2O and -375.56 ℃⁄푚푀 for H2O.

34 D O 2 H O 32 2

C) o 30

LCST (

0.0 0.5 1.0 1.5 NaSCN (M)

Figure 2-4 LCSTs of PDEA in NaSCN solution. Data points are connected with lines as guide to the eye.

28 D O 2 H O 24 2

C) o 20

LCST ( 12

8 0.0 0.5 1.0 1.5 2.0 NaCl (M)

Figure 2-5 LCSTs of PDEA in NaCl solutions. Data points are connected with lines as

guide to the eye. c value is -11.66 ℃⁄푚푀 for D2O and -12.46 ℃⁄푚푀 for H2O.

88 3.6oC D O 80 2 H O 2 72

o 64

LCST (

o 48 1.5 C 2.5 3.0 3.5 4.0 4.5 5.0 NaCl (M)

Figure 2-6 LCSTs of PDMA in NaCl solutions. Data points are connected with lines as

guide to the eye. c value is -13.90 ℃⁄푀 for D2O and -15.17 ℃⁄푀 for H2O.

B. Differential scanning calorimetry Thermograms

DSC can be used to follow the enthalpy changes (ΔH) of the phase transition of polymer collapse. PDEA in pure water and pure heavy water were scanned. The phase transition temperature were 31.9℃ in D2O, and 31.5℃ in H2O. The higer phase transition temperature in

D2O agrees well with the LCST measurements. The ΔH(collpase) is positive and is ~0.5℃ higer in H2O (161.0 cal/g in H2O, 155.3 cal/g in D2O). The endothermic process occurs mainly due to the dissociation of more ordered water molecules around the hydrophonic groups. It indicates that the hydrophobic collapse of PDEA is entropically driven.

The thermograms in Figure 2-7 have asymmetric profiles. Since Cp is mostly detected by breaking the hydrogen-bonds around hydrophobic groups the decay of Cp seems to be greatly engaged in hydrogen bond breaking rate. In this perspective, the dueterium bonds break faster with the temperature incrase than hydrogen bond. It was claimed in the early 1960s that the number of deuterium bonds change faster with temperature change than that of hydrogen bonds.25 The relationship between temperature and the number of hydrogen/deuterium bonds are also seen in

MCR-Raman spectra in Figure 2-13 and 2-15 where strongly hydrogen bonded D2O clusters diminish faster than H2O with temperature increase.

Strikingly, the PDMA thermograms (Figure 2-9) present ΔH (collapse) ≈0 when the aggregation of PDMA were observed by the light scattering measurements (LCST measurements).

Hence, the aggregation of PDMA is much less hydrophobic effect than PDEA. It even possibly implies that PDMA is hydrated not primarily by the hydrophobic hydration of the methyl groups.

Other functional groups such as the carbonyl groups and backbones could play a bigger role in

PDMA, which manipulates the dynamics of the PDMA in solutions. To verify this hypothesis, further experiments were carried out.

0.25

PDEA (D O) 2 0.20 H= 1.55 PDEA (H O) 2

C 0.15

0.10

Cp ( cal / cal/ Cp( H=1.61

0.05

0.00 30 35 40 45 50 55 Temperature (OC)

Figure 2-7 Differential scanning calroimetry thermograms of 10mg/mL PDEA solutions in D2O

(green) and H2O (orange). Phase transition themperature of the polymer is slightly is higher in

D2O.

0.25

0.20

PDMA (D O) 0.15 2 PDMA (H O)

g) 2

O 0.10

Cp ( cal / cal/ Cp( 0.05

0.00

65 70 75 80 85 90 95 Temperature (OC)

Figure 2-8 Differential scanning calroimetry thermograms of 10mg/mL PDMA solution in D2O

(green) and H2O (orange). 250mM tribasic sodium phosphate solution was used as solution and reference. The phase transition is not observed by DSC under these samples’ conditions.

C. FT-IR Spectrum

To probe the molecular mechanisms of the hydration of PDEA and PDMA, FTIR was employed. Amide bands contain carbonyl stretch, CN stretch, and NH bend.40 The amide I band of PDEA and PDMA arises only from the carbonyl stretch. The amide I band of model polymers in each pure H2O and D2O was measured by FTIR.

-1 -1 Two peaks were seen in the PDMA spectrum in D2O at 1610cm and 1627cm . The first peak can be assigned to the 2-hydrogen bonded carbonyl and the second peak can be assigned

-1 to the 1-hydrogen bonded carbonyl. In pure H2O, two major peaks appeared at 1604cm and

-1 1622cm . One would expect that D2O would form stronger hydrogen bonding with the carbonyl group since it forms stronger hydrogen bonds in general. Unexpectedly, however, red-shift of carbonyl stretch in H2O indicates stronger hydrogen bonding between the PDMA carbonyl group

-1 -1 and H2O molecules than D2O molecules. The peak shift was 4cm and 5cm respectively. The dominant peak was 2-hydrogen bonded carbonyl stretch both in H2O and D2O.

-1 Likewise, two peaks were observed in the PDEA spectrum in D2O at 1600cm ,

-1 -1 -1 1617cm . In pure H2O, two peaks appeared at 1598cm and 1615cm . The spectrum also showed the red-shift upon D2O→H2O solvent transfer. The ratio of two peaks was about 50:50 and the peak shift was 2cm-1 for both peaks.

Earlier, I hypothesized that the hydrophobic hydration would play a minor role in

PDMA hydration based on the DSC scan result. From the PDMA results, the first peak (2- hydrogen bonded carbonyl stretch) shift is greater than the shift in PDEA. Besides, the fact that the first peak dominates in the PDMA spectrum corresponds with the LCST measurements opposite to PDEA. The LCST of PDMA is presumably higher in pure H2O than D2O. The first peak shift is smaller in the PDEA spectrum and it is not the most dominant peak or factor in the

hydration of the carbonyl group of the polymer. The shift of the second peak, on the other hand, is about the same in the PDMA and PDEA spectra. Thus, the isotope effects on carbonyl group hydration is more pronounced in between the carbonyl group and two hydrogen bonded water molecules.

So far, several experimental results imply that the PDMA has some unique features in the light of the hydration. However, PDEA and PDMA were not compared in the same condition in the experiments described above due to the diverse physical properties. The biggest challenge was to experiment two polymers in pure water at the same temperature. Thus, IR spectrum collection was not sufficient to illuminate isotope effects of water on hydrophobic hydration of the model polymers. In the next section, PDMA were examined at various temperature in pure water.

Figure 2-9 FTIR spectrum of PDEA in H2O at 20℃. Two hydrogen bonded carbonyl (peak at 1598) and one hydrogen bonded carbonyl (peak at 1615) arise at similar intensity.

Figure 2-10 FTIR spectrum of PDEA in D2O at 20℃. Two hydrogen bonded carbonyl

(peak at 1600) and one hydrogen bonded carbonyl (peak at 1617) arise at similar intensity.

Figure 2-11 FTIR spectrum of PDMA in H2O at 20℃. Two hydrogen bonded carbonyl (peak at 1604) is dominant compared to one hydrogen bonded carbonyl (peak at 1622).

Figure 2-12 FTIR spectrum of PDMA in D2O at 20℃. Two hydrogen bonded carbonyl (peak at 1610) is dominant compared to one hydrogen bonded carbonyl (peak at 1627). Carbonyl group peak shift is larger in PDMA solution than in PDEA solution.

D. Raman-MCR Spectrum

Raman scattering measurements with multivariate curve resolution (Raman-MCR) measurements provides solute-correlated spectra which arises from solute-induced perturbation of water vibrations.41 This new technique allows to produce spectra of not only perturbed water molecules by small ions but also of hydrophobic hydration shell.37 The hydration shell of PDMA was investigated at various temperatures in pure water.

Firstly, pure water isotope effects were compared as a function of temperature. (Firgure

2-13, 2-15). The low frequency component arises from more ordered water clusters. The peak at higher frequency is attributed to less ordered water clusters42,43. The normalized intensity at each peak was compared for quantitative analysis (Table 2-2). The ratio of the more ordered water cluster over less ordered water cluster as functions of temperatures is presented as ΔW. At 20℃

ΔW of D2O is higher than of H2O as D2O has stronger hydrogen bonding. As the temperature increases, ΔW decreases faster in D2O. In other words, the hydrogen bonding in D2O appear to break faster than H2O. Also, the peaks slightly blue-shift with temperature. Therefore, temperature may play an important role in isotope effects of water on the hydrophobic hydration.

Table 2-2 Normalized intensity at each water peak and its ratio

H2O D2O H2O D2O *Intesity at * Intesity at * Intesity at * Intesity at ΔW ΔW more ordered less ordered more ordered less ordered water peak water peak water peak water peak ퟐퟎ℃ 0.7525 0.9486 0.8730 0.8993 0.7933 0.9707 ퟑퟕ℃ 0.7324 0.9795 0.8203 0.9203 0.7476 0.8914 ퟔퟎ℃ 0.6859 1.008 0.7730 0.9508 0.6803 0.8130 ퟖퟎ℃ 0.6419 1.068 0.7059 1.0438 0.6010 0.6762 *Intensity is scaled to isobathic point.

Sequentially, solute correlated (SC) spectra of PDMA and PDEA were compared at

-1 -1 four temperature. Specifically, the perturbed hydration shell in 2200cm – 2800cm for D2O and

-1 -1 3000cm – 3700cm in H2O were analyzed. Since water spectra were subtracted from the solution sample spectra, the spectra obtained in the range is presumably caused by the polymer hydration.

The spectra were scaled to CH stretch band area. Hence, the hydration shell area implies the degree of perturbation by the molecule per C-H in the solution. It is unexpected yet intriguing that the hydration shell of PDMA is positively correlated with temperature. It is because lately Dor Ben-

41 Amotz reported SC spectrum of n-butanol in both H2O and D2O as a function of temperature . As n-butanol is a hydrophobic molecule, the hydration shell area decreased with temperate increase.

Instead, a new peak arises at a higher frequency than both of the water peaks.

Looking into the spectra of PDMA in H2O, ΔW varies roughly from 0.79 to 0.45 over the temperature range. In comparison, ΔW in the bulk H2O steadily decreases from 0.79 to 0.60, which falls in the similar range. Secondly, the solute-correlated (SC) spectra transform at higher temperatures. A new peak arises at ~3550cm-1, indicating the formation of the more loosely interacting water molecules around the hydration shell. Besides, the intensity of the hydration shell increases with temperature, which contrasts to the traditional understanding of the cold denaturation.5,8,10,44 The hydrophobic hydration is enthalpically stable at lower temperatures. The hydration water separates from the hydrophobes upon temperature increases due to the entropy increase.

ΔW of PDMA solution in D2O was also calculated. The peak at the low frequency rapidly diminishes and the peak of more loosely bound water clusters seem to rise at ~2635cm-1 from 37℃ (Figure 2-15). Compared to the bulk water spectra, the intensity fraction of more ordered water around PDMA decreases more quickly in D2O.

Overall, more structured water clusters at the hydration shell are better maintained in the presence of PDMA compared to the bulk water. Yet, in D2O/PDMA solution, more structured water clusters seem to disappear fast with temperature. In turn, the peak of less structured water clusters rises rapidly. The hydrophobic hydration can be stabilized by the strengthened hydrogen bonding of water molecules surrounding the hydrophobes. D2O with the stronger hydrogen bonds stabilizes the uncollapsed/unfolded states of biopolymers in many cases. However, the spectrum shows that structured D2O molecules stay less at the hydration shell than in bulk. It also shows that the methyl groups of the model polymer are better stabilized by less structured water clusters of

H2O with temperature increase.

The PDMA spectra were compared with the n-butanol SC spectra from the literature due to the experimental challenge of the PDEA solution spectrum collection and the current lack of Raman-MCR data (Figure 2-17).41 When PDMA is less hydrophobic, n-butanol could represent a hydrophobic molecule. In the SC spectra of n-butanol (in H2O), the hydration water peaks, especially less ordered water peak at 3200cm-1, are quenched with the temperature increase.

Quickly, it transforms and produces a new peak that suggests loosely bound water clusters at the hydration shell. The Raman scattering measurement of PDEA contained the strong background noise which overlaps with the intensity of water.

In the future, the spectra of PDMA with improved signal to noise should be taken. Also, collecting SC-spectra of acrylamide with longer di-alkyl group would give the direct evidence of the different hydration aspect of PDMA and PDEA. However, it is dubious that whether the driving force of the isotope effect of water on hydrophobic hydration could be explained by this method.

1.0 H O 2 20oC 0.8 37oC 60oC 0.6 80oC

0.4

Normalizedintensity(a.u.) 0.2

0.0

2800 3000 3200 3400 3600 3800 wavenumber (cm-1)

Figure 2-13 Bulk H2O Raman spectra at various temperature

1.0 PDMA in H O 2 o 0.8 20 C 37oC o 0.6 60 C 80oC

0.4

Normalizedu.) intensity(a. 0.2

0.0

2800 3000 3200 3400 3600 3800 wavenumber (cm-1)

Figure 2-14 SC spectra of PDMA in H2O at various temperature

1.0 D O 2 20oC 0.8 37oC 60oC 0.6 80oC

0.4

Normalizedintensity(a.u.) 0.2

0.0 2000 2200 2400 2600 2800 3000 3200 wavenumber (cm-1)

Figure 2-15 Bulk D2O Raman spectra at various temperature

1.0 PDMA in D O 2 20oC o 0.8 37 C 60oC o 0.6 80 C

0.4

Normalizedu.) intensity(a. 0.2

0.0 2000 2200 2400 2600 2800 3000 3200 wavenumber (cm-1)

Figure 2-16 SC spectra of PDMA in D2O at various temperature

41 Figure 2-17 SC spectra of n-butanol in H2O and D2O at various temperatures .

4. Conclusion

The isotope effects of water on hydrophobic hydration of the PDMA and PDEA were investigated. I suspect that the overall hydration difference originates from the nature of polymers and isotope effect from temperature. Regardless of the identity of ions and pH, the trend of the isotope effects remained the same. PDMA seemed to be the less hydrophobic polymer than PDEA.

I speculate that the uncollapsed state of PDMA is better stabilized in H2O at a high temperature due to the low sensitivity of hydrogen bond strength to heat than deuterium bond. On the other hand PDEA stabilizes in water mainly through the hydrophobic hydration. The hydrophobic hydration enhances hydrogen bonding at the hydration shell, resulting in the better uncollapsed state in D2O. However, the series of data neither advocate nor disprove my hypothesis. Comparing

PDMA and PDEA may not reveal why some hydrophobic hydration is more stable in H2O than in

D2O. Also, the conditions of the experiment need to be simplified. In the future, polyethylene glycol (PEG) could be used. PEG with 900K shows phase transition at high temperature without salts, but its LCST is higher in H2O. PEG will be an interesting polymer to study in order to figure out the relationship between hydrophobic effect and hydrogen bonding strength.

CHAPTER III: DIVALENT CATIONS EFFECTS ON PROTEINS

1. Introduction

Calcium is one of the most abundant inorganic element in human body. An adult contains approximately 1kg of calcium in body.45 99% of calcium is stored in bones and teeth as hard tissue. The rest is in serum, plasma and cells. Calcium ion in plasma is only about 1.7mM and [extracellular]/[intracellular]=104-105 in the scale of micro to nanomolar.46 Approximately

60 - 70 % of bone is inorganic calcium deposits which are hydroxyapatite (Ca10(PO4)6(OH)2) and some amorphous calcium phosphate. 30 - 40 % is organic component which is mostly collagen protein.47–49 Even though the majority of calcium is in skeleton form, calcification, the process of accumulation of calcium salts in a body, is not clearly understood at the molecular level.

Abnormal calcification is a critical issue since it can occur on soft tissues and can cause health issues such as kidney stones, gall stones, and heterotopic ossification. Besides, calcium related diseases are becoming more common these days50.

Abnormal biological calcification can be caused by the interaction of imbalanced calcium and proteins in serum. It is known that calcium ions can be caught by elastic fibers or mucopolysarccharides. However, the mechanism of abnormal calcification also remains elusive.

In order to provide fundamental insight of calcification in human body, it is necessary to understand calcification in a molecular level. Besides, understanding the mechanisms may allow chemistry to contribute to biomedical and bioengineering applications.

It is believed that bone is formed by the following process through osteoblast, involving collagen and matrix proteins by the following reaction: 51

2- 2+ + 6 HPO4 + 2 H2O + 10 Ca ⇌ Ca10(PO4)6(OH)2 + 8 H

Nonetheless, this mineralization course cannot rationalize the hydronium ion removal and pH maintenance in osteoblast.

Herein, I am interested in investigating the effect of ion on the hydrophobic collapse and the secondary structure formation of ELPs specific interactions between calcium ion and elastin-like polypeptides (ELPs). ELPs contain type II β-turn structure with which every first and fourth residues are hydrogen bonded. ELP is an appropriate model protein in this study as β-turn structured collagen in body is preferred on the aspect of proline hydroxylation.

Recently, Okur et al. spectroscopically and quantitatively demonstrated that

Hofmeister series cations such as Ca2+, Mg2+ and Li+ can only weakly bind to carbonyl group of butyamide monomer in an aqueous solution.52 However, application of the knowledge from this study to biology has limitation since their biomolecule is a simple monomer. In addition, at the interface where calcium phosphate inorganic mineral is formed, calcium is perhaps locally greatly concentrated from millimolar possibly up to molar range.

Accordingly, in this work I investigated specific divalent cations effects on ELPs by using high concentration of salt solutions. Surprisingly, I found distinct interactions between

ELPs and divalent cations.

2. Experimental

A. ELP Preparation

The pET-25b(+) plasmids were employed to construct ELPs by directional recursive ligation (RDL) as described above and literature.53 In the solutions of supplemented medium

(TBdry) culture with ampicillin, the plasmids in BLR/DE3 E. coli was expressed at 37℃ for 24

hours. The E. coli was sonicated twice and ELPs underwent inverse transition cycling (ITC) twice to be further purified. The concentration of purified ELP solutions were determined by measuring absorbance at 280nm where ɛ =5690 M-1 cm-1. The ELP solutions were dialyzed in deionized water for two days with fresh water twice a day to remove residual salts in the solution. The samples were dried by speed vacuum concentrator at 54℃ for 2 hours and stored at room temperature until use.

B. LCST Measurements

The LCSTs were measured as described in Chapter II experimental section. The concentration of V5A2G3-60, V5A2G3-120, V5A2G3-330, VG-128, V5-120 was 10ml/mL. measurements were made from 4 to 6 time for each data point. Each concentration of CaCl2,

MgCl2, and SrCl2 solution was added to the dry ELP sample to make the ELP concentration

10mg/mL.

C. Circular Dichroism Spectra Measurements

The measurement was performed on a Jasco-1500 between 200nm-270nm. Data interval was

0.025nm, band width was 1nm, and scanning speed was 50nm/min with 4 accumulation signals.

The ELP concentration was 1mg/ml. The measurement was made up to 5 times. The conversion of CD signal was done by following the equation below.

100∆푚° [휃] = 표푏푠푑 퐶푛푙

∆푚° is sample CD signal-background CD signal, C is the polypeptide concentration (mM), n is the number of residues in the polypeptide, and l is the path length (cm).54

3. Results and Discussion

A. LCST Measurements

With high concentration of CaCl2, MgCl2, and SrCl2 solutions, the LCST trend of ELPs is significantly different from what has been shown in literature to date. As expected, high concentration of ions causes a salting-out effect on polymers by hydrating water molecules around water/biopolymer interfaces and promoting hydrophobic effect collapse. In CaCl2 solutions, the

LCST steadily decreases and sharply drops after 2.8M (Figure 3-1, 3-3). In MgCl2 solutions, the

LCST of V5A2G3-120 decreases with salt concentration up to 1.6M (Figure 3-1, 3-4). Surprisingly, in the range of 1.6M – 4.0M LCST increases back, which indicates salting-in effect. The LCST trend linearly decreases with concentration in SrCl2 solutions (Figure 3-1, 3-5). The effect of chloride is not attributed to abrupt drop and increase of LCST in CaCl2 and MgCl2 solutions since chloride binding to biopolymer is negligible and it only linearly decreases the LCSTs.19,20,32 Also, linear behavior is an indicator of non-specific behavior of Sr2+ on ELPs. Hence, this data is the macroscopic evidence of that Mg2+ and Ca2+ have specific effects on ELPs. More interestingly,

2+ 2+ Mg specific effect was not distinguished from Ca with V5-120 which is more hydrophobic polypeptides than V5A2G3-120 (Figure 3-2). Hydrophobicity of the two polypeptides can be measured and compared by circular dichroism. More hydrophobic ELP, V5-120 has more random

14 coil and less secondary structure upon collapse than V5A2G3-120 . To study cation specific effects on the secondary structure of the ELPs, circular dichroism spectra were collected.

44 Vwild-1205A2G3 -120 42 40 38 36

C) 34

O 32 30 LCST ( 28 26 CaCl2 24 MgCl2 22 SrCl2

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 Salt Concentration (M)

Figure 3-1 The LCSTs of V5A2G3-120 in divalent cation chloride salt solutions. Data points are connected with lines as a guide to the eye.

30 CaCl2 V -120 5 MgCl2 25 SrCl2

C) O 15

LCST (

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 Salt Concentration (M)

Figure 3-2 The LCSTs of V5-120 in divalent cation salt solutions. Data points are connected with lines as a guide to the eye.

50 VG-128 V A G -60 45 5 2 3 V A G -120 5 2 3 V A G -330 40 5 2 3 V -120 5

o 30

LCST ( LCST 25

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 CaCl concentration(M) 2

Figure 3-3 The LCSTs of various model polymers in CaCl2 solutions. The LCST behavior is similar in V5A2G3 type ELPs regardless of the chain length, and VG-128. Only

V5-120 displays different trend.

o 30 VG-128 V A G -60 5 2 3

LCST ( LCST V A G -120 25 5 2 3 V A G -330 5 2 3 20 V -120 5

10 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 MgCl concentration (M) 2

Figure 3-4 The LCSTs of various model polymers in MgCl2 solutions. The LCST behavior is similar in V5A2G3 type ELPs regardless of the chain length, and VG-128. Only V5-120 displays different trend.

V -120 40 5 V A G -120 5 2 3 35

LCST ( LCST 15

0.0 0.5 1.0 1.5 2.0 SrCl concentration (M) 2

Figure 3-5 The LCSTs of various model polymers in SrCl2 solutions. Since the LCST behavior is similar in V5A2G3 type ELPs regardless of the chain length, and VG-128, only V5-120 and

V5A2G3-120 were measured to observe LCST trend with SrCl2.

V5A2G3-120 and V5-120 were chosen in order to investigate the relation of specific cation effects on polymers with different degree of hydrophobicity. V5A2G3 type ELPs showed the same trend in salt solutions regardless of the chain length.

B. Circular Dichroism Spectra The secondary structure of the ELPs was probed by circular dichroism in 200nm-

240nm range. Type II β-turn gives rise to an absorbance peak around 212nm as a positive band.

This peak comes from 푛 → п ∗ transition of the amide bond.

2+ V5-120 spectra is similar in trend in all salt solutions (Figure 3-6 ~ 3-9). As the Ca concentration increases, the peak at 212nm increases. The more ions are hydrated around the

ELP/water interface, the ELPs collapse and the more β-turn structure is pronounced. This appear to accord with the LCST measurements. In the MgCl2 and SrCl2 solutions, the peak at 212nm increases, which indicates that more β-turn is formed with the LCST decrease. The solution of V5-

120 with more than 2M salt concentration are not shown since they produced poor signal to noise spectra due to the high salt concentration.

V5A2G3-120 in CaCl2, however, shows the incredible secondary structure increase as

Ca2+ concentration increases (Figure 3-10). When the Mg2+ concentration increases, the β-turn peak seems to disappear (Figure 3-11). The peak at 212nm is lost and it flattens out with more salts. The rise of β-sheet or the increase of random coil portion may give a flat feature to the spectrum. However a clear interpretation is difficult since the peak is lost. What is obvious in these

2+ spectrum is that high Mg concentration does not promote β-turn structures in V5A2G3-120.

Correlating this to the LCST data, it is a reasonable since the LCSTs increase with the

2+ MgCl2 concentration increase. From LCST data, it seems that ELPs salt-in by Mg . Like mentioned earlier, well-hydrated metal ions only weakly binds to carbonyl group of amide of butyramide monomer.52 However, my preliminary aresult implies a possibility that Ca2+ and Mg2+

2+ may specifically bind to V5A2G3-120 at high concentrations. My hypothesis is that Ca binds to the multiple carbonyl groups of the backbone. The diagrams of hypothesis are shown in Figure 3-

16 and 3-17. When the bulk concentration is increased, the number of ions at the polypeptide

backbone/water interface also increases. Hence, ions like Ca2+ are more likely to interact with the carbonyl groups. Moreover, Ca2+ can ligate with 6 to 8 ligands, preferably oxygen atoms in body.55

Mg2+ has the smaller coordination number than Ca2+ at all cases. Also, Mg2+ requires higher energy to be dehydrated. When Ca2+ would be more easily dehydrated and more likely to interact with multiple carbonyl groups and promotes the salting-out effect, Mg2+ would be more difficult to dehydrate and tend to interact with one carbonyl group and promotes the salting-in effect.

-2

-4

-6

/dmol) -8

2 -10

-12

(deg. cm (deg.

3 -14

θ x10θ -16 Ca 0.0M -18 Ca 0.5M -20 Ca 1.0M

-22 200 205 210 215 220 225 230 235 240 wavelength (nm)

Figure 3-6 V5-120 in different concentrations of CaCl2

-2

-4

-6

/dmol) -8

2 -10

-12

(deg. cm (deg.

3 -14 Mg 0.0M

θ x10θ -16 Mg 0.5M Mg 1.0M -18

-20

-22 200 205 210 215 220 225 230 235 240 wavelength (nm)

Figure 3-7 V5-120 in different concentrations of MgCl2

-2

-4

-6

-8

/dmol)

2 -10

-12

(deg. cm (deg.

3 -14

θ x10θ -16 Sr 0.0M Sr 0.5M -18 Sr 1.0M -20 Sr 1.5M

200 205 210 215 220 225 230 235 240 wavelength (nm)

Figure 3-8 V5-120 in different concentrations of SrCl2

-2

-4

-6

/dmol) -8

-10

-12

(deg. cm (deg.

3 -14 NaCl 0M θ x10θ -16 NaCl 0.5MM NaCl 1.0M -18

-20

200 210 220 230 240 wavelength (nm)

Figure 3-9 V5-120 in different concentrations of NaCl

-2

-4

/dmol) 2 -6

-8

(deg. cm (deg.

3 -10

θ x10θ Ca 0M -12 Ca 0.5M Ca 1.0M -14 Ca 2.0M

-16 200 205 210 215 220 225 230 235 240 wavelength (nm)

Figure 3-10 V5A2G3-120 in different concentrations of CaCl2

-2

-4

/dmol)

2 -6

(deg. cm (deg. -8 3 Mg 0M Mg 0.5M

θ x10θ -10 Mg 1.0M Mg 2.0M -12 Mg 3.0M

-14 200 205 210 215 220 225 230 235 240 wavelength (nm)

Figure 3-11 V5A2G3-120 in different concentrations of MgCl2

-1

-2

-3

-4

/dmol)

2 -5

-6

(deg. cm (deg. 3 -7

θ x10θ -8 Sr 0M Sr 0.5M -9 Sr 1.0M -10 Sr 1.5M

-11 200 205 210 215 220 225 230 235 240 wavelength (nm)

Figure 3-12 V5A2G3-120 in different concentrations of SrCl2

-2

-4

/dmol)

-6

(deg. cm (deg.

3 -8 Urea 0M Urea 1M

θ x10θ -10 Urea 2M Urea 4M Urea 6M -12

205 210 215 220 225 230 235 240 245 250 255 260 wavelength (nm)

Figure 3-13 V5A2G3-120 in different concentrations of urea

Urea is a well-known denaturant. It increases LCSTs with concentration increase. Peak at 212nm steadily decreases with urea concentration.

-2

-4

/dmol)

2 -6

-8

(deg. cm (deg.

3 -10 NaCl 0M θ x10θ NaCl 0.5M -12 NaCl 1.0M

-14

200 205 210 215 220 225 230 235 240 wavelength (nm)

Figure 3-14 V5A2G3-120 in different concentrations of NaCl

The spectrum with NaCl was measured for comparison purpose. NaCl only linearly decreases the LCSTs. The peak at 212nm hardly shift with decrease in the LCST or increase in the salt concentrations.

-1

0 -2 2 C

/dmol) 2 70C 120C -3 170C

(deg. cm (deg. 0 3 22 C -4 270C θ x10θ 320C -5 370C 420C

-6 200 210 220 230 240 250 260 wavelength (nm)

Figure 3-15 V5A2G3-120 at various temperature. The spectra remain fairly constant at all temperature range. The spectra were smoothed for comparison.

198nm indicates the random coil of ELPs. This is from the п → п ∗ transition of the amide bond. This transition usually more strongly absorbs light than the 푛 → п ∗ transition.

Unfortunately, this range was excluded in these measurements because chloride absorption was so high below 200nm that signal was saturated. The spectrum below 200nm was indiscernible from the background. The temperature effects were minimized by lowering the temperature far below the polymer’s phase transition temperatures. The cell temperature was kept at 0℃ for all the measurements for V5-120 and 5℃ for V5A2G3-120. The background was measured at the same temperature and salt concentrations. The background was subtracted from the polymer spectrum.

The data is reported as mean residue ellipticity.

Figure 3-16 Diagram of the hypothesis for Ca2+ binding to VPGGG unit

2+ Figure 3-17 Diagram of the hypothesis for Mg binding for V5A2G3-120

4. Conclusion

The specific binding of divalent cations on ELPs with β-turn structures was investigated. V5A2G3-120 forms more β-turn structures in a highly concentration calcium chloride solutions. The LCST data shows salting-out effect of Ca2+ and salting-in effects of Mg2+.

Molecular level mechanism was not revealed by CD spectrum.

I hypothesize that Ca2+ binds to the multiple carbonyl groups of the proteins once the bulk solution is saturated. Ca2+ may bind to the multiple carbonyl groups of the proteins, enhancing the secondary structure. Mg2+may bind to the single carbonyl group due to higher emery boundary for dehydration.

To validate the hypothesis, one can investigate whether Ca2+ is trapped in the collapsed state. This could possibly achieved by precipitating the ELPs in a high concentration of CaCl2 solution, and by analyzing the precipitate with atomic absorption spectrometer. To examine molecular level hypothesis, carbonyl group of the ELPs could be probed by Fourier transform infrared spectrometer. If Ca2+ is associated with carbonyl group, amide I band should shift.

CHAPTER IV: KOSMTROPIC EFFECT OF PHYTIC ACID

1. Introduction

1, 2, 3, 4, 5, 6 hexakis (di-hydrogen phosphate) myo-inositol or phytic acid is an interesting natural occurring molecule (Figure 4-1). Its primary role is known as phosphorous and inositol storage in plant seeds, grains and animal cells56–58. It has been reported that phytic acid has anticarcinogenic/antineoplastic properties and antioxidant functions59–61. In addition, it has been used in many industrial applications. Most of the applications are carried out in solution forms where phytic acid are in its salts form, phytate. Although phytic acid seems to benefit medical community and other industries, it is anti-nutrient in agriculture62–64. It has been believed that phytate inhibits protein digestion forming complex with metal ions65–67.

Recently, an alternative aspect of phytate is recognized, which phytate is a strong kosmotrope ion68–70. As shown in earlier chapters, kosmotropic ions or strongly hydrated ions with a high ratio of charge to volume decrease the solubility of proteins. More curiously, phytic acid has 12 protons, therefore, 12 pKas (1.1, 1.5, 1.5, 1.7, 2.1, 6.85, 7.6, 10, 10, 12, 12)71. This property should alter phytic acid from a weak kosmostrope to strong kosmotrope depending on pH.

In this short chapter, the kosmotropic effects of phytate as a function of pH was investigated. At a low pH, phytate has a smaller amount of negative charges, which should result in a weak kosmotropic effect. At a high pH, phytate is deprotonated, obtaining a large amount of negative charges. It would manifest a stronger kosmotrope effect than any kosmotropes in

Hofmeister series.

1.1

6.85 12 1.5 1.5

1.1 1.1 1.1 1.1 12

1.1 10 2.1 1 .5 2.1 1.1 101.1 7.6 1.1 1.7 1.1 1.1 Figure 4-1 The structure of phytic acid. Blue figures: pKas 1.1

2. Experimental

Poly(N-isopropyl acrylamide) M.W. 19,000 was lyophilized and prepared as described in chapter I. Phytic acid sodium salt hydrate from rice (Sigma-Aldrich) was used for solution preparation. For each pH condition, 10mL of 100mM stock solutions were prepared. They were diluted by 10mM sodium phosphate buffer at the same pH condition. For example, 0mM sample only contains 10mM sodium phosphate. The LCST was measured as the previous chapters at pH

1.5, 2.2, 7.2, 8.2, 9.0, 10.5, 11.5, and 12.6. Each data point represents one measurement.

3. Results and Discussion

As Expected, the LCST of PNIPAM linearly decreases as a function of sodium phytate concentration (Figure 4-2). In the Hofmeister series, kosmotropes reduce the solubility of proteins by dehydrating the microenviroment of proteins. The linear correlation (slope of the graph) of the

LCST and the salt concentration is a typical characteristic of a kosmotrope. Intriguingly, the decrease in LCST becomes more rapid with pH increase. At all pH conditions, the slope does not appear to linearly relate with the pH values (Figure 4-3). Since the pKa of phytate is also sporadic in numbers, the number of charges at each pH were correlated with the slope of the LCST data

(Figure 4-4). Overall, the slope is steeper as the negative charge gets higher. In the pH range between 7 and 11.5, the slope values remain constant. This could be simple fluctuation of the measurement from the sample preparation. Also, the protonation constant may not match in the experimental set up since the constants are sensitive to experimental conditions72–74. Another possibility is that PNIPAM collapase mechanism at these pHs may involve the interactions among phytate, cations, and PNIPAM. However, it is too early to picture specific interactions.

The result demonstrated the kosmotropic effects of phytic acid. It also showed the effects can be modified as a function of pH. In order to confirm the trend, more measurement must be taken. If the trend strictly agree with the amount of phytate charges, the LCST measurement can potentially be used as a means to measure the pH of phytate. In other words, knowing the pH of a certain protein with phytate at a given condition, it would be possible to estimate the solubility of the protein..

pH 1.5 32 pH 7.2 pH 12.6 28

o 24

LCST ( LCST 20

16 0 10 20 30 40 50 Sodium Phytate Concentration (mM)

Figure 4-2 The LCST values at three representative pH conditions

buffer: phosphate 10mM 32 polymer: PNIPAM 19K 10mg/ml 30 28

C) 26 pH 1.5

O pH 2.2 24 pH 7.2 22 pH 8.2

LCST ( LCST pH 9 20 pH 10.5 pH 11.5 18 pH 12.6 16 0 10 20 30 40 50 Sodium phytate (mM) Figure 4-3 The LCST of poly(N-isopropyl acrylamide) in sodium phytate solutions at various pH

12 280

Negative Charges of Phytate Slope Pnipam-Phytate 10 Pnipam-phytate slope 240

8 200

6 160 4

Negative Charges 120 2

2 4 6 8 10 12 14 pH

Figure 4-4 Correlation of the negative charges of phytate and the slope of the LCST data

4. Conclusion

Kosmotropic effect of phytic acid on PNIPAM collapse was investigated as a function of pH. The number of charges were correlated with the slope of the LCST data. Overall, it appears to agree. Yet, the measurements must be repeated to confirm the result. Since the pH affects the

LCSTs of proteins by phytic acid, one information (of pH and solubility) can be used to determine the other.

In the future, the cation effects on the ability of phytic acid to precipitate proteins should studied. With high negative charge on one molecule, phytic acid must electrostatically attract counter cations. Some cations in the Hofmeister series and the Irving-Williams series are crucial element in body. Hence, exploring specific ion effects with phytic acid can also contribute to agriculture, medical science, and industry for applications. Another interesting work is to compare the kosmotropic effect of phytic acid inositol pentakisphosphate, inositol triskisphosphate, and phosphate.

References

(1) Kauzmann, W. In Advances in Protein Chemistry; C.B. Anfinsen, M. L. A., Kenneth Bailey and John T. Edsall, Ed.; Academic Press, 1959; Vol. 14, pp 1–63. (2) Guo, Z.; Zhou, F.; Hao, J.; Liu, W. J. Am. Chem. Soc. 2005, 127 (45), 15670–15671. (3) Wei, P.; Cook, T. R.; Yan, X.; Huang, F.; Stang, P. J. J. Am. Chem. Soc. 2014, 136 (44), 15497–15500. (4) Dill, K. A.; Truskett, T. M.; Vlachy, V.; Hribar-Lee, B. Annu. Rev. Biophys. Biomol. Struct. 2005, 34 (1), 173–199. (5) Chandler, D. Nature 2005, 437 (7059), 640–647. (6) Némethy, G.; Scheraga, H. A. J. Chem. Phys. 1962, 36 (12), 3401–3417. (7) Sharp, K. A.; Madan, B. J. Phys. Chem. B 1997, 101 (21), 4343–4348. (8) Dill, K. A. Biochemistry (Mosc.) 1990, 29 (31), 7133–7155. (9) Stillinger, F. H. Science 1980, 209 (4455), 451–457. (10) Dill, K. A.; Bromberg, S. Molecular Driving Forces: Statistical Thermodynamics in Biology, Chemistry, Physics, and Nanoscience, 2nd Edition, 2nd edition.; Garland Science: London ; New York, 2010. (11) Privalov, P. L. Crit. Rev. Biochem. Mol. Biol. 1990, 25 (4), 281–306. (12) Pl, P.; Sj, G. Adv. Protein Chem. 1987, 39, 191–234. (13) Makhatadze, G. I.; Privalov, P. L. J. Mol. Biol. 1993, 232 (2), 639–659. (14) Cho, Y.; Sagle, L. B.; Iimura, S.; Zhang, Y.; Kherb, J.; Chilkoti, A.; Scholtz, J. M.; Cremer, P. S. J. Am. Chem. Soc. 2009, 131 (42), 15188–15193. (15) Streicher, W. W.; Makhatadze, G. I. Biochemistry (Mosc.) 2007, 46 (10), 2876–2880. (16) Hofmeister, F. Arch. Für Exp. Pathol. Pharmakol. 1888, 25 (1), 1–30. (17) Collins, K. D. Methods 2004, 34 (3), 300–311. (18) Jungwirth, P.; Tobias, D. J. Chem. Rev. 2005, 106 (4), 1259–1281. (19) Rembert, K. B.; Paterová, J.; Heyda, J.; Hilty, C.; Jungwirth, P.; Cremer, P. S. J. Am. Chem. Soc. 2012, 134 (24), 10039–10046. (20) Zhang, Y.; Furyk, S.; Bergbreiter, D. E.; Cremer, P. S. J. Am. Chem. Soc. 2005, 127 (41), 14505– 14510. (21) Fernandez-Prini, R.; Harvey, A. H.; Palmer, D. A. Aqueous Systems at Elevated Temperatures and Pressures: Physical Chemistry in Water, Steam and Hydrothermal Solutions; Academic Press, 2004. (22) Rodnikova, M. N. J. Mol. Liq. 2007, 136 (3), 211–213. (23) Greenwood, N. N.; Earnshaw, A. Chemistry of the Elements 2nd Edition; Butterworth-Heinemann, 1997. (24) Hardy, R. C.; Cottington, R. L. J. Res. Natl. Bur. Stand. 1949, 42, 573–578. (25) Némethy, G.; Scheraga, H. A. J. Chem. Phys. 1964, 41 (3), 680–689. (26) Kresheck, G. C. J. Phys. Chem. B 1998, 102 (34), 6596–6600. (27) Maczynski, A.; Shaw, D. G.; Goral, M.; Wisniewska-Goclowska, B.; Skrzecz, A.; Owczarek, I.; Blazej, K.; Haulait-Pirson, M.-C.; Hefter, G. T.; Maczynska, Z.; Szafranski, A.; Young, C. L. J. Phys. Chem. Ref. Data 2005, 34 (2), 657–708. (28) Wang, X.; Qiu, X.; Wu, C. Macromolecules 1998, 31 (9), 2972–2976. (29) Farasat, R.; Vyazovkin, S. Macromol. Chem. Phys. 2014, 215 (21), 2112–2118. (30) Aseyev, V.; Tenhu, H.; Winnik, F. M. In Self Organized Nanostructures of Amphiphilic Block Copolymers II; Müller, A. H. E., Borisov, O., Eds.; Advances in Polymer Science; Springer Berlin Heidelberg, 2010; pp 29–89. (31) Meyer, D. E.; Chilkoti, A. Biomacromolecules 2004, 5 (3), 846–851. (32) Cho, Y.; Zhang, Y.; Christensen, T.; Sagle, L. B.; Chilkoti, A.; Cremer, P. S. J. Phys. Chem. B 2008, 112 (44), 13765–13771.

(33) Mou, F.; Chen, C.; Zhong, Q.; Yin, Y.; Ma, H.; Guan, J. ACS Appl. Mater. Interfaces 2014, 6 (12), 9897–9903. (34) Wang, H.; Sun, D.; Zhao, N.; Yang, X.; Shi, Y.; Li, J.; Su, Z.; Wei, G. J. Mater. Chem. B 2014, 2 (10), 1362. (35) Phan, H. T. T.; Yoda, T.; Chahal, B.; Morita, M.; Takagi, M.; Vestergaard, M. C. Biochim. Biophys. Acta BBA - Biomembr. 2014, 1838 (10), 2670–2677. (36) Chang, Y.-C.; Bowie, J. U. Proc. Natl. Acad. Sci. 2014, 111 (1), 219–224. (37) Davis, J. G.; Rankin, B. M.; Gierszal, K. P.; Ben-Amotz, D. Nat. Chem. 2013, 5 (9), 796–802. (38) Perera, P.; Wyche, M.; Loethen, Y.; Ben-Amotz, D. J. Am. Chem. Soc. 2008, 130 (14), 4576–4577. (39) Rabinovich, I. B. Influence of Isotopy on the Physicochemical Properties of Liquids; Springer: New York, 1995. (40) Barth, A.; Zscherp, C. Q. Rev. Biophys. 2002, 35 (04), 369–430. (41) Davis, J. G.; Gierszal, K. P.; Wang, P.; Ben-Amotz, D. Nature 2012, 491 (7425), 582–585. (42) Santucci, S. C.; Fioretto, D.; Comez, L.; Gessini, A.; Masciovecchio, C. Phys. Rev. Lett. 2006, 97 (22), 225701. (43) Cammarata, L.; Kazarian, S. G.; Salter, P. A.; Welton, T. Phys. Chem. Chem. Phys. 2001, 3 (23), 5192–5200. (44) Wiggins, P. M. Phys. Stat. Mech. Its Appl. 1997, 238 (1–4), 113–128. (45) Emsley, J. The Elements, 3 edition.; Oxford University Press: Oxford : New York, 1998. (46) Hofer, A. M.; Brown, E. M. Nat. Rev. Mol. Cell Biol. 2003, 4 (7), 530–538. (47) University of Cambridge. Structure and composition of bone http://www.doitpoms.ac.uk/tlplib/bones/structure.php. (48) Tamburro, A. M.; Guantieri, V.; Daga-Gordini, D.; Abatangelo, G. Biochim. Biophys. Acta 1977, 492 (2), 370–376. (49) Tamburro, A. M.; Guantieri, V.; Daga-Gordini, D.; Abatangelo, G. J. Biol. Chem. 1978, 253 (9), 2893– 2894. (50) Scales, C. D.; Smith, A. C.; Hanley, J. M.; Saigal, C. S.; Urologic Diseases in America Project. Eur. Urol. 2012, 62 (1), 160–165. (51) Brown, W. E.; Chow, L. C. Annu. Rev. Mater. Sci. 1976, 6 (1), 213–236. (52) Okur, H. I.; Kherb, J.; Cremer, P. S. J. Am. Chem. Soc. 2013, 135 (13), 5062–5067. (53) Meyer, D. E.; Chilkoti, A. Nat. Biotechnol. 1999, 17 (11), 1112–1115. (54) Myers, J. K.; Nick Pace, C.; Martin Scholtz, J. Protein Sci. 1998, 7 (2), 383–388. (55) Bertini, I.; Gray, H. B.; Valentine, J. S.; Lippard, J. Bioinorganic Chemistry; University Science Books: Mill Valley, Calif, 1994. (56) Konietzny, U.; Jany, K.; Greiner, R. J. Fuer Ernaehrungsmedizin 2006, 8, 18–28. (57) Raboy, V. Phytochemistry 2003, 64 (6), 1033–1043. (58) Crea, F.; De Stefano, C.; Milea, D.; Sammartano, S. Coord. Chem. Rev. 2008, 252 (10–11), 1108– 1120. (59) Marie Minihane, A.; Rimbach, G. Int. J. Food Sci. Technol. 2002, 37 (7), 741–748. (60) Doria, E.; Galleschi, L.; Calucci, L.; Pinzino, C.; Pilu, R.; Cassani, E.; Nielsen, E. J. Exp. Bot. 2009, ern345. (61) Fox, C. H.; Eberl, M. Complement. Ther. Med. 2002, 10 (4), 229–234. (62) Grases, F.; Costa-Bauzá, A. Anticancer Res. 1999, 19 (5A), 3717–3722. (63) Kumar, V.; Sinha, A. K.; Makkar, H. P. S.; Becker, K. Food Chem. 2010, 120 (4), 945–959. (64) Woyengo, T. A.; Nyachoti, C. M. Can. J. Anim. Sci. 2013, 93 (1), 9–21. (65) Martin, C. J.; Evans, W. J. J. Inorg. Biochem. 1986, 27 (1), 17–30. (66) Bohn, T.; Davidsson, L.; Walczyk, T.; Hurrell, R. F. Am. J. Clin. Nutr. 2004, 79 (3), 418–423.

(67) Kim, O.-H.; Kim, Y.-O.; Shim, J.-H.; Jung, Y.-S.; Jung, W.-J.; Choi, W.-C.; Lee, H.; Lee, S.-J.; Kim, K.-K.; Auh, J.-H.; Kim, H.; Kim, J.-W.; Oh, T.-K.; Oh, B.-C. Biochemistry (Mosc.) 2010, 49 (47), 10216– 10227. (68) Moraes, M. L.; Maki, R. M.; Paulovich, F. V.; Rodrigues Filho, U. P.; de Oliveira, M. C. F.; Riul, A.; de Souza, N. C.; Ferreira, M.; Gomes, H. L.; Oliveira, O. N. Anal. Chem. 2010, 82 (8), 3239–3246. (69) Cowieson, A.; Cowieson, N. Proc. 22nd Annu. Aust. Poult. Sci. Symp. 2011, 22. (70) Bye, J. W.; Cowieson, N. P.; Cowieson, A. J.; Selle, P. H.; Falconer, R. J. J. Agric. Food Chem. 2013, 61 (2), 290–295. (71) Costello, A. J.; Glonek, T.; Myers, T. C. Carbohydr. Res. 1976, 46 (2), 159–171. (72) Crea, F.; Crea, P.; Milea, D.; Porcino, N.; Sammartano, S. Ann. Chim. 2007, 97 (8), 635–645. (73) Concetta De Stefano, D. M. J. Chem. Amp Eng. Data 2002, 48 (1), 114–119. (74) Crea, P.; de Robertis, A.; de Stefano, C.; Sammartano, S. Biophys. Chem. 2006, 124 (1), 18–26.