Determinations of Proton Affinities of Methylated Cysteine and Serine Homologs

W&M ScholarWorks

Undergraduate Honors Theses Theses, Dissertations, & Master Projects

4-2019

Danielle Long

Follow this and additional works at: https://scholarworks.wm.edu/honorstheses

Part of the Analytical Chemistry Commons, and the Physical Chemistry Commons

Recommended Citation Long, Danielle, "Determinations of Proton Affinities of Methylated Cysteine and Serine Homologs" (2019). Undergraduate Honors Theses. Paper 1351. https://scholarworks.wm.edu/honorstheses/1351

This Honors Thesis is brought to you for free and open access by the Theses, Dissertations, & Master Projects at W&M ScholarWorks. It has been accepted for inclusion in Undergraduate Honors Theses by an authorized administrator of W&M ScholarWorks. For more information, please contact [email protected]. Scanned with CamScanner Abstract

Non-protein amino acids (NPAAs) are of interest to study for their potential to be incorporated into peptides and proteins, as well as for understanding their structure-energetics relationships. Studying these amino acids’ thermochemical properties such as their acidity and basicity allow for elucidation of the structure and bonding characteristics of these molecules. By examining these molecules in the gas-phase in mass spectrometers, we are able to determine their intrinsic thermodynamic properties without solvation effects. These thermochemical properties, in part, determine the structure, function and bonding of the molecules.

This study focused on determining the proton affinities of methylated cysteine and serine homologs by using Cook’s extended kinetic method with orthogonal distance regression analysis in a triple quadrupole mass spectrometer. The NPAAs selected have been shown to mis- incorporate into small peptides, making it of interest to determine how these changes affect the structures and energetics of the peptide.1 The NPAAs studied included α-methylcysteine, L- penicillamine (gem-dimethyl cysteine), α-methylserine, and 3-methylthrenonine (gem-dimethyl serine), and their proton affinities were determined to be 923.7 ± 11 kJ/mol, 925 ± 15 kJ/mol, 932

± 20 kJ/mol and 924 ± 15 kJ/mol, respectively. These experimental proton affinities are in excellent agreement with Boltmann-weighted computational proton affinities determined for these molecules. All of the methylated homologs had larger proton affinities than their respective protein amino acid, serine and cysteine, which have proton affinities of with proton affinities of 903 kJ/mol and 912 kJ/mol respectively.2

Table of Contents

Chapter 1: Introduction 1

1.1 Acid and Base Chemistry 1

1.2 Previous Non-Protein Amino Acid Studies 2

1.3 Kinetic Method 3

1.4 Quadrupole Mass Spectrometry 7

Chapter 2: Experimental Procedures 10

2.1 Experimentally Determined Proton Affinities 10

2.2 Computationally Determined Proton Affinities 12

Chapter 3: Results and Discussion 13

3.1 Computational Results 14

3.2 α-Methylcysteine Results 16

3.3 L-Penicillamine Results 20

3.4 α-Methylserine Results 24

3.5 3-Methylthreonine Results 28

3.6 Experimental Results Summary 32

Chapter 4: Conclusion and Future Work 34

References 35

Figures and Tables

Figure 1: Kinetic Method Equation 5

Figure 2: Competitive Dissociation 6

Figure 3: Schematics of Quadrupole Mass Spectrometry 8

Figure 4: α-Methylcysteine 13

Figure 5: L-Penicillamine (dimethyl cysteine) 13

Figure 6: α-Methylserine 13

Figure 7: 3-Methylthreonine (dimethyl serine) 13

Figure 8: Yields of Peptide Produced by Valine, 3-Methylthreonine and L-Penicillamine 14

Table 1: Calculated Proton Affinities 14

Figure 9: Low Energy Conformers of Cysteine Homologs 15

Figure 10: Low Energy Conformers of Serine Homologs 15

Table 2: α-Methylcysteine Experimental Data 16

Figure 11: α-Methylcysteine Plot 1 17

Figure 12: α-Methylcysteine Plot 2 18

Figure 13: α-Methylcysteine Effective Temperature 19

Figure 14: α-Methylcysteine ODR 20

Table 3: L-Penicillamine Experimental Data 20

Figure 15: L-Penicillamine Plot 1 21

Figure 16 L-Penicillamine Plot 2 22

Figure 17: L-Penicillamine Effective Temperature 23

Figure 18: L-Penicillamine ODR 24

Table 4: α-Methylserine Experimental Data 24

iii

Figure 19: α-Methylserine Plot 1 25

Figure 20: α-Methylserine Plot 2 26

Figure 21: α-Methylserine Effective Temperature 27

Figure 22: α-Methylserine ODR 28

Table 5: 3-Methylthreonine Experimental Data 28

Figure 23: 3-Methylthreonine Plot 1 29

Figure 24: 3-Methylthreonine Plot 2 30

Figure 253-Methylthreonine Effective Temperature 31

Figure 26: 3-Methylthreonine ODR 32

Table 6: Summary of Experimental versus Computational Results 33

Table 7: Proton Affinity of Cysteine and Serine 33

Acknowledgements

I would to first thank Dr. J.C. Poutsma for his guidance and encouragement throughout my years in IonLab. I would also like to thank the members of IonLab for their support and collaboration. I would like to especially thank Izzy Owens, whose teamwork on this project was invaluable, as well as Zach Smith and Gwendyln Turner, for all their work in the computational aspects of the project. I would also like to thank Anwar Radwan for his continuous positive outlook and friendship throughout our time together in IonLab.

I would like to thank both the College of William and Mary and the Charles Center for making this honors thesis possible.

Lastly, I would like to thank the National Science Foundation for funding this project.

Chapter 1: Introduction

1.1 Acid Base Chemistry

The acid-base properties of molecules are of interest to chemists for a variety of reasons.

Thermodynamic properties such as acidity and basicity are partially responsible for governing the structure, bonding, and function of molecules. By determining these properties, we are able to get a clearer picture of the molecule’s overall structure and bonding behavior. These properties are useful in many different fields of study, ranging from medicine and biology to chemistry and nutrition.3-7

Acids and bases follow a typical reaction, which can be seen in (1), where a protonated conjugate acid (BH+) dissociates into a gaseous base (B) and a free proton (H+).

!"#(%) → !(%) + "#; Δ" = -.; ∆0 = 0! (1)

In Equation 1, proton affinity (PA) is the enthalpy change (Δ") for the reaction, and gas basicity

(GB) is the Gibbs free energy (∆0) for the reaction. 8 This equation follows the Bronsted definition of acids and bases, where a base is a molecule that is capable of accepting a proton from a proton donor, while an acid is a molecule that is capable of donating a proton to a proton acceptor.9

Amino acids are of particular interest in acid-base chemistry study, as they are the building blocks of peptides and proteins. Although there are only twenty common protein amino acids that are incorporated into proteins in humans, there are many other non-protein amino acids that are of interest to study.10,11 Protein amino acids are those that are found in the main chains of peptides and proteins, while the non-protein amino acids are not found in protein main chains either because they do not have the specific transfer RNA and codon triplet needed, or because they do not have the correct post-translational modification.10 Studying these non-protein amino acids is important, as many of them have potential benefits in areas such as drug discovery, while others are toxic to

1 humans, secondary to their similar structures to protein amino acids.11 As these non-protein amino acids have similar structures to the common amino acids, it is of interest to see how these small differences in structure change the molecules bonding and acid-base properties.

While acid-base reactions can be studied both in solution and in the gas-phase, gas-phase study is of interest due to the reactions solely depending on the acid-base strength of the reacting species, rather than on solvation effects.12 These reactions follow gas-phase acidity and gas-phase basicity scales, which are fundamentally different from the aqueous solution scale of pKa, due to solvation effects.12 On the gas-phase acidity/basicity scale, a stronger acid will have a smaller gas- phase acidity, and a stronger base will have a larger gas-phase basicity.12

Many of the experiments in gas-phase acid-base chemistry rely on the proton to evaluate the strength of the base. Acid-base reactions are based on the transfer of a proton and it is of interest to be able to measure the proton affinity of molecules to evaluate how strongly the proton will be attracted to a molecule of interest. This is discussed in detail in section 1.3, the kinetic method.

1.2 Previous Non-Protein Amino Acid Studies

This lab has previously studied the thermochemical properties of various non-protein amino acids, including analogs of proline, lysine and arginine.5-7 All of these studies sought to determine how the thermochemical properties change with small structural changes to the amino acid.5-7 For proline, the study sought to determine the effects of increasing ring size on proton affinity.5 The proline analogs included L-azetidine-2-carboxylic acid, L-proline and L-pipecolic acid, which have 4-, 5-, and 6-membered ring nitrogen heterocycles, respectively.5 This study found that L-azetidine-2-carboylic acid, L-proline and L-pipecolic acid had proton affinities of

933 kJ/mol ± 6.3 kJ/mol, 941 kJ/mol ± 6.7 kJ/mol and 944 kJ/mol ± 6.7 kJ/mol, respectively.5 This

2 study was consistent with previous studies, which also found that increasing the size of a nitrogen heterocycle ring increased proton affinity.5,13,14

A second study by this lab looked at the shorter chain homologs of the protein amino acid lysine (n=5): ornithine (n=4), 2,4-diaminobutanoic acid (DABA, n=3) and 2,3-diaminopropanoic acid (DAPA, n=2).6 This study investigated if these molecules with shorter side chains formed as strong of internal hydrogen bonds in the gas phase as lysine does, which would lead to higher gas-phase bascities.6 They found that ornithine, DABA and DAPA had proton affinities of

1001.1 ± 6.6 kJ/mol, 975.8 ± 7.4 kJ/mol and 950.2 ± 7.2 kJ/mol, respectively.6 This data was compared to previously reported data for α,ω-diamines, also with varying side-chain lengths from n=2-5.6,13 Their results were consistent with the α,ω-diamines findings, as the species with n=4 and n=5 side chains have nearly identical proton affinity values, while the species with n=2 and n=3 have much lower proton affinities.6,13

A third study investigated oxyanalogues of arginine and ornithine: canavanine and canaline, respectively.7 While ornithine is not a protein amino acid, it is produced when arginine is metabolized in the urea cycle, making it vital to human metabolism.7 These oxyanalogues were of interest to study as they both have been observed as metabolites produced during the urea cycle.7 This study found that the canavanine and canaline had proton affinities of 1001 kJ/mol ± 9 kJ/mol and 950 ± 6 kJ/mol, respectively.7 For both molecules studied, the oxygen substitution lead to a decrease in proton affinity of 50 kJ/mol and 40 kJ/mol for canavanine and canaline, respectively, significantly decreasing the molecule’s basicity compared to that of the non-oxygen substituted molecule.7

1.3 Kinetic Method

One way to study the acidities and basicities of molecules is by using the kinetic method.

The kinetic method is a popular method for evaluating the relative thermochemistry of a molecule in a mass spectrometer, including a molecule’s proton affinity. A molecule’s proton affinity is “the negative of the enthalpy change in the gas phase reaction (real or hypothetical) between a proton and the chemical species concerned, usually an electrically neutral species to give the conjugate acid of that species,” as seen in equation (1).8 Because the site of proton attachment in the molecule, which is influenced by that site’s proton affinity, guides both the fragmentation of the molecule and its overall structure, studying proton affinity is of interest to many researchers, especially in regard to amino acids, peptides and proteins.14

The kinetic method of analysis was first introduced in 1977 by Cooks and Kruger.15 This method was used to replace earlier methods of determining proton affinities due to its simplicity in theory and in execution, as well as its sensitivity to small differences in base strength.15 Previous methods to determine proton affinity involved evaluating the position of equilibrium in the proton transfer reaction from one base to another, as seen in (2), where B1 is the molecule of interest and

15 B2 is the reference base. This reaction can be used to determine the relative proton affinity

15 between B1 and B2:

# # 15 !1" + !2 ⇋ !1 + !2" (2)

This equilibrium method, as shown in equation 2, evaluates which side of the reaction is favored after thousands of collisions in a mass spectrometer are allowed to occur.16 With this data, an equilibrium constant for proton transfer between the two molecules, B1 and B2, and the sign of the standard free energy can be found. With a known proton affinity of the reference molecule, the proton affinity of the analyte can be determined.16 This method, however, does have significant limitations. The molecules used must be pure and volatile, as the number densities of the neutral

4 molecule needs to be known, the molecules must be stable at the experiment’s temperature and the measurement of equilibrium must not be disturbed by other side reactions, like the molecule fragmenting or creating ion-bound clusters.17 The kinetic method takes advantage of the last of these limitations, and uses the decomposition of the ion-bound clusters to assess thermochemistry.17

In the kinetic method, the proton affinity is instead determined by evaluating the rate of fragmentation of a heterodimer composed of a reference base, B1, the molecule of interest, B2, and

15 + a proton, as seen in Figure 1. Analyzing the relative abundances of the fragment ions, B1H and

+ 15 B2H , allows the proton affinity to be determined. This can be simply visualized as which fragment ion is favored (has a lower activation energy) after decomposition of the heterodimer is allowed to occur in a mass spectrometer. As shown in Figure 1, the heterodimer can either fragment

+ into the protonated reference base with the neutral molecule of interest (B1H + B2) or protonated

+ molecule of interest with the neutral base (B2H + B1).

Figure 1: Kinetic Method Equation15

In the gas phase, proton affinity is related to how a protonated conjugate acid (BH+) dissociates into a gaseous base (B) and the free proton (H+) as shown previously in equation 1.18

This can be extended to an ideal gas, where the proton affinity of the molecule would be:18

-.(!) = Δ" = ∆4 + 5618 (3)

, where ∆4 is the difference in the total energies of the products and reactant of the reaction in (1),

R is the ideal gas constant and T is absolute temperature.18

A major advantage of the simple version of the kinetic method is that the measurements are independent of internal energy and source conditions, which implies that these competitive reactions (Figure 1) will have identical frequency factors, and that ultimately their rates are dictated by their relative activation energies (Figure 2).15 Assuming that there is no reverse activation energy, the differences in the entropy requirement for the competitive reactions is negligible and there are no isomeric forms that will produce inaccurate fragmentation ratios in the mass spectrometer, the difference in the forward activation energies of the fragments is exactly equal to the difference in the proton affinities.19 Another advantage of the kinetic method is that it can be used to study amino acids, as they are not volatile enough to make measurable amounts of their neutral forms, as would be needed if the equilibrium method was used.19

Figure 2: Competitive Dissociation20

Figure 2 visualizes how the activation energies play a role in determining the proton affinity. As more energy is put into the system, the heterodimer, AHB+ will be fragmented, and

6 the molecule with the lower activation energy will get the proton preferentially, and that molecule will therefore have the higher proton affinity.20 Because this is done in a mass spectrometer, this dissociation ratio of the heterodimer can be seen as the ratio of the intensity of BH+ to AH+.

A useful extension of the simple kinetic method is known as the extended kinetic method.

In the extended kinetic method, enthalpic and relative entropic values are determined by quantitatively examining the energy dependence of the heterodimer dissociation.21 This adds a new variable to the calculations, effective temperature, which is assumed to be a constant at each tested collision energy.21 This method ultimately corrects for errors in the affinities from not taking into account entropy and allows for the determination of entropy and a more accurate determination of the enthalpy.22 One advantage to this method is that previously it was essential that the reference solution used to be structurally similar to the molecule of interest to restrict these entropic effects.21 Because this approach explicitly measures entropy effects, it allows for reference bases that are not structurally similar to the molecule being used.21

A last extension of the kinetic method adds the statistical approach of orthogonal distance regression, introduced by Ervin and Armentrout in 2004.22 This method “minimizes the weighted sum of squared closest distances between calculated function and observed points.”22 The program gives statistical deviations and 95% confidences intervals for each of the fit parameters and treats all the data points of the original measurements equally.22 Compared to other methods to find the intersection point of the data, which guides proton affinity determinations, Ervin and Armentrout found that the ODR method gave the most accurate value due to the minimum orthogonal distance criterion.22

1.4 Quadrupole Mass Spectrometry

A mass spectrometer is an instrument that measures the relative abundance of the mass to charge ratios of ion species in the instrument. This instrument creates a mass spectrum when the ions reach the external detector which creates a series of ion signals dispersed in time, which can be used for data collection in experiments such as the kinetic method.23

A triple quadrupole mass spectrometer allows tandem mass spectrometry to be performed.

Tandem mass spectrometry is when two or more mass analyzers can be used together, allowing for multiple stages of mass selection.23 This is useful in experiments seeking to determine fragmentation patterns of molecules by allowing a parent ion to be selected, isolated, fragmented with a collision gas, and daughter fragments to be seen.23 Figure 3 shows a simplified scheme of a triple quadrupole mass spectrometer.24

Figure 3: Schematics of Triple Quadrupole Mass Spectrometry24

As Figure 3 shows, the sample is introduced into the instrument via an ionization source, like electrospray ionization (ESI), atmospheric pressure chemical ionization (APCI), or matrix- assisted laser desorption ionization (MALDI).25 Electrospray ionization is the most popular ionization technique for biological applications.25 To create the electrospray, a high voltage is applied to a flow of liquid at atmospheric pressure.25 This spray is then be directed into the vacuum opening of the mass spectrometer, where with a combination of heat, vacuum, and collisions with

8 buffer gas molecules, the droplets are de-solvated.25 The ions will then become separated from the droplets and are accelerated into the mass spectrometer, such as into quadrupole 1 (Q1) of a triple quadrupole instrument.25

Once inside the triple quadrupole mass spectrometer, the molecule enters Q1, a mass filter where ions are selected.26 Once an ion is selected in Q1, it passes through quadrupole 2, an RF- only quadrupole that is pressurized with a collision gas that allows for the ion to be fragmented.26

These fragments are transmitted to in quadrupole 3, another mass filter that shows the product ions from the parent selected in Q1.26 An electron multiplier at the end of the mass spectrometer is used for detection.26

One advantage of the triple quadrupole mass spectrometer is that due to its high sensitivity, it has proven to be useful in alluding to molecular structure.26 Because molecules tend to fragment in specific patterns, these patterns can be used to piece together the structure of the parent ion.26

Another advantage of the triple quadrupole mass spectrometer is its ability to analyze mixtures.26

Both of these advantages are useful when using the triple quadrupole mass spectrometer for the kinetic method, as it is of interest to see how the molecule fragments to see where the proton could be attached, and the heterodimer is part of a mixture of ions emerging from the ion source.

Chapter 2: Experimental Procedures

2.1 Experimentally Determined Proton Affinities

All experimental work was done on a Finnigan TSQ Quantum Ultra triple quadrupole mass spectrometer with an external electrospray ionization source. Solutions were injected using an external syringe pump at a rate of 60 µL/hr. The instrument’s heated capillary tube temperature was set at 125°C. The spray voltage was set at 4000 V. The nitrogen flow rate was set at 20 arbitrary units.

The molecules of interest used in this experiment were purchased by the Hartman group at

Virginia Commonwealth University (VCU) from Sigma-Aldrich and sent to this lab for proton affinity determinations. The heterodimer for the experiment was made by diluting the molecule of interest to a 10-3 M solution with a mixture of 50:50 methanol to Millipore water ratio with 1% formic acid. This solution was mixed with the reference base, diluted in the same manner, in a 1:1 ratio to form a heterodimer. This mixed solution was injected via the external syringe pump into the mass spectrometer to form the heterodimer by electrospray ionization, which is seen at the m/z value corresponding to the sum of the masses of the molecule of interest, reference base and a hydrogen atom.

Data was collected in MS/MS product mode after isolating the heterodimer. All data were taken at varying collision energies, from 3 eV to 30 eV, at 3 eV intervals over a series of several days. Argon was used as the collision gas. The data was then exported to an Excel® file, where the ratio of the intensity of protonated reference base to that of the protonated molecule of interest was determined. In case in which the primary dissociation products further fragmented to give secondary products, the intensities of the secondary fragment products were added to those of the appropriate primary product to give the final ratio. The average ratios of all the data for each

10 reference base at each different collision energy was then determined, and the natural logarithms of those averages were used in further calculations.

+ + Three plots were used to determine the proton affinity. Plot 1 is (ln[BiH /AH ]) versus PA-

PAavg where PA is the proton affinity of the reference base and PAavg is the average proton affinity of all the reference bases. By creating a line of best fit at each collision energy through the data for the different reference bases, an intersection point of the different trend lines can be observed.

Adding the x-coordinate of the intersection point to the PAavg will give the experimental proton affinity.

A second plot can be created to give a better indication of the intersection point and is also typically used to determine the range of collision energies to use in the final workup. The plot is created from the negative intercept of each line from plot 1 versus that line’s slope. This plot is expected to follow a typical linear curve. Deviations from linearity indicate non-statistical dissociations at low energies or unaccounted for secondary products at higher energies. Only energies in the linear region of the curve are used for the final analysis.

A third plot was created with this data to visualize the effective temperatures at the different collision energies, using equation (4). This plot gives the relationship between effective temperature (K) versus the collision energy (eV). Again, for every molecule of interest, this plot is expected to follow a typical linear curve and is also used to determine which collision energies are used in the final analysis.

47789:;<8 68=>8?@:A?8 = 1 (4) BCDEF∗H

In equation 4, R is the gas constant, 0.008314 IJ and slope is the slope of the line of best fit KDC∗L from plot 1.

A last step in evaluating the data is to perform an orthogonal distance analysis of the data, after using the orthogonal distance regression program developed by Ervin and Armentrout.22 This program takes into account entropic effects and finds the best intersection point for the lines while minimizing orthogonal distance.22 From these measurements, a Monte Carlo simulation is performed to provide an estimate of the error in the derived proton affinity due to the effects of random measurement error in reference affinities.22 The uncertainties in proton affinities for the

+ + references bases is ± 8 kJ/mol, while the uncertainty in the (ln[BiH /AH ]) ratio is ± 0.05. When reporting data, the ODR-derived proton affinity is the reported proton affinity.

2.2 Computational Determined Proton Affinities

Another student in the lab used PCModel and Gaussian09 to determine the theoretical proton affinities and entropy changes by looking at low-energy conformers of the amino acid derivatives. Predictions for the proton affinities of the four molecules were generated at the

B3LYP/6-311++G(d,p)//B3LYPP/6-31+G(d) level of theory. This data was initially used to help guide which reference bases to use, and then finally used to compare with the experimental results.

Chapter 3: Results and Discussion

In collaboration with a research team at VCU, we sought to determine the proton affinities of methylated serine and cysteine homologs, specifically α-methylcysteine, L-penicillamine, α- methylserine, and 3-methylthrenonine, whose structures can be seen in Figures 4 through 7.

Figure 4: α-methylcysteine Figure 5: L-penicillamine (gem-dimethyl cysteine)

Figure 6: α-methylserine Figure 7: 3-methylthreonine (gem-dimethyl serine)

The Hartman group at VCU investigated whether these non-protein amino acids are able to incorporate into peptide chains in place of protein amino acids, and were successful in showing that L-penicillamine and 3-methylthreonine incorporated into small peptides using valine’s t-RNA synthetase just as well as valine(Figure 8).1,27 This discovery made it of interest for our lab to study the molecules themselves, looking at where the molecules were likely to protonate, what their low- energy conformers look like, and to determine their thermochemical properties.

25.0

20.0

15.0

10.0

picomole peptide 5.0

0.0 Valine 3-methyl Thr Penicillamine

Figure 8: Yields of Peptide Produced by Valine, 3-Methylthreonine and L-

Penicillamine,27

3.1 Computational Results

Another member of the lab used a computational approach to determine calculated proton affinities for the methylated amino acids. Table 1 shows the data that they collected. In order to determine these proton affinities, as well as determine where the proton is attaching, they also determined the low energy conformers of the molecules (Figures 9 and 10).

Amino Acid Calculated Proton Affinity α-Methylcysteine 925 kJ/mol L-Penicillamine 920 kJ/mol α-Methylserine 933 kJ/mol 3-Methylthreonine 929 kJ/mol

Table 1: Calculated Proton Affinities (B3LYP/6-311++G(d,p)//B3LYP/6-31+G(d))

Figure 9: Low Energy Conformers of Cysteine Homologs

In these Figures, red atoms are oxygen, blue atoms are nitrogen, yellow atoms are sulfur, gray atoms are carbon, and white atoms are hydrogen. The solid lines represent molecular bonds, while dotted lines represent hydrogen bonding.

Figure 10: Low Energy Conformers of Serine Homologs 15

Our lab was successfully able to find the lowest energy structures of the molecules, which can be seen in Figures 9 and 10. These structures were of particular interest for the cysteine molecules, as it has been debated if the deprotonation site of cysteine is at the carboxylic acid

(COOH) or thiol (SH).28,29 This study was able to show that in both α-methylcysteine and l- penicillamine, the lowest energy conformer had the deprotonation at the thiol, consistent with prior calculations.28,29

3.2 α-Methylcysteine Results

For α-methylcysteine, the references bases used included N-ethylacetamide, pyridazine, ethylamine and propylamine, with their respective proton affinities as 898 kJ/mol, 907 kJ/mol, 912 kJ/mol and 917.8 kJ/mol.13 Due to there being unaccounted for secondary products at collision energies 27 and 30 eV, only data from collision energies 3-24 eV was used in the proton affinity determinations. Table 2 shows the experimental data collected for α-methylcysteine.

Reference Base Average Proton Affinity 905.36 kJ/mol Slope (plot 2) 17.7 Proton Affinity from plot 2 923.1 kJ/mol Intercept (plot 2) 3.1 Change in Entropy from plot 2 -25.97 kJ/mol ODR Proton Affinity 923.0 ± 11.3 kJ/mol ODR Entropy Change -29.4 ± 19 kJ/mol Calculated Proton affinity 925 kJ/mol

Table 2: α-Methylcysteine Experimental Data

From the data, a plot of the average fragmentation intensity ratio of reference base to

+ + molecule of interest (ln[BiH /AH ]) versus PA-PAavg was made. Lines of best fit were added at each collision energy (3 eV to 24 eV) in order to look for a point of intersection. Adding the x- coordinate of the crossing point to PAavg will give the experimental proton affinity. In Figure 10, this intersection can be seen between 15 and 20 of the PA-PAavg.

Plot 1: α-Methylcysteine 6

3 4 6 9 2 12 )

+ 15 0 18 )/I(refH

+ 21 24 -2 Linear (3)

Ln(I(amcys Linear (6) -4 Linear (9) Linear (15)

-6 Linear (18) Linear (21) Linear (24) -8 -15 -10 -5 0 5 10 15 20 25 PA-PAavg

Figure 11: α-Methylcysteine Plot 1

A second plot, Plot 2, was created to evaluate the relationship between the negative intercept and slope of the lines of best fit in plot 1. It is expected that as the collision energies increase, the data will follow a linear curve. For α-methylcysteine, the data generally followed a linear curve, with a slope of 17.7 and an R2 value of 0.9672 (Figure 11). The slope of this line indicates the intersection points of the lines in plot 1 that is used to determine the experimental proton affinity. The data from collision energies of 27 eV and 30 eV were excluded from the calculation, and they are seen in black on this plot. This data was excluded as it deviated from the linear curve, indicating that there were unaccounted for secondary products.

y = 17.7x - 3.1 Plot 2: α-Methylcysteine R² = 0.9672 3.5

2.5

1.5 Negative Intercept

0.5

0 0.22 0.24 0.26 0.28 0.3 0.32 0.34 0.36 Slope

Figure 12: α-Methylcysteine Plot 2

A third plot was created to evaluate the effective temperatures derived from the kinetic method experiments in the mass spectrometer. It is expected that as there are more energetic collisions in the mass spectrometer at higher collision energies, the effective temperature derived from the kinetic method data will rise in a linear fashion. As seen in Figure 12, this was seen in the α-methylcysteine data, with an R2 value of 0.9779. Again, this Figure shows that for collision energies of 27 eV and 30 eV, the data no longer followed a linear path.

Effective Temperature: α-methylcysteine y = 8.7323x + 337.58 R² = 0.9779 600

550

500

450 Effective Temperature Effective(K) Temperature

400

350 0 5 10 15 20 25 30 35 Collision Energy (eV)

Figure 13: α-Methylcysteine Temperature Effective

An orthogonal distance regression (ODR) analysis was performed, using a program developed by Ervin and Armentrout, on the data to force the lines of best fit in plot 1 to converge to a single point.22 Figure 13 shows the ODR results, which were used to determine the proton affinity for α-methylcysteine of 924 kJ/mol, with a 95% confidence interval of ± 10 kJ/mol.

6 ODR: α-Methylcysteine 4 3 6 2 9 12 )) + 0 15 18 /IamcysH + -2 21 24

-4Ln(IrefH ODR 3 -6 ODR 6 ODR 9 -8 ODR 12 ODR 15 -10 ODR 18 -20 -10 0 10 20 ODR 21

Figure 14: α-Methylcysteine ODR

3.3 L-Penicillamine Results

For L-penicillamine, the references bases used included N-ethylacetamide, thiazole, pyridazine, N,N-dimethylacetamide and benzylamine, with their proton affinities as 898 kJ/mol,

903 kJ/mol, 907.4 kJ/mol, 908 kJ/mol and 922.7 kJ/mol, respectively.13 Table 3 shows the experimental data collected for L-penicillamine.

Reference Base Average Proton Affinity 908 kJ/mol Slope (plot 2) 21.33 Proton Affinity from plot 2 929.3 kJ/mol Intercept (plot 2) 5.5 Change in Entropy from plot 2 -45.5 kJ/mol ODR Proton Affinity 925 ± 17 kJ/mol ODR Entropy Change -34 ± 4 kJ/mol Calculated Pproton Affinity 920 kJ/mol

Table 3: L-Penicillamine Experimental Data

+ + From the data, a plot of the average fragmentation intensity ratio (ln[BiH /AH ]) versus

PA-PAavg was made. Lines of best fit were added at each collision energy in order to look for a point of intersection, which can be seen on Figure 15 between 10 and 20 of the PA-PAavg.

20 Plot 1: L-Penicillamine

15 3 6 9 10 ))

+ 12 15 )/I(refH

+ 18 5 21 Linear (3)

Ln(I(dmcysH 0 Linear (6) Linear (9) Linear (12)

-5 Linear (15) Linear (21)

-10 -20 -10 0 10 20 30 40 50 PA - PAavg

Figure 15: L-Penicillamine Plot 1

Another plot, Plot 2, was created to evaluate the relationship between the slope and the negative of the intercept of the data from plot 1. For L-penicillamine the data generally followed a linear curve, with a slope of 17.695 and an R2 value of 0.9715 (Figure 16). This Figure includes the data from 24 eV to 30 eV that was excluded in calculations due to its deviation from linearity.

This deviation can be attributed to unaccounted for secondary products.

Plot 2: L-Penicillamine y = 17.695x - 4.2471 R² = 0.9715 2.5

1.5

0.5 Negative Intercept

-0.5

-1 0.23 0.25 0.27 0.29 0.31 0.33 0.35 0.37 0.39 Slope

Figure 16: L-Penicillamine Plot 2

A third plot was created to evaluate the effective temperature in the mass spectrometer. As seen in Figure 17, the L-penicillamine data is reasonably linear with an R2 value of 0.9691. Figure

17 again shows the deviations from linearity seen in the data from 24 eV to 30 eV.

Effective Temperature: L-Penicillamine y = 8.8669x + 308.46 550 R² = 0.9691

500

450

400 Effective Temperature Effective(K) Temperature

350

300 0 5 10 15 20 25 30 35 Collision Energy (eV)

Figure 17: L-Penicillamine Effective Temperature

After all the data is collected, it is further evaluated using the orthogonal distance regression technique developed by Ervin and Armentrout.22 This plot can be seen in Figure 18, as the technique forces the lines of best fit for each reference base to merge to a single point while minimizing orthogonal distance.22 This is used to determine the reported proton affinity, which for

L-penicillamine is 925 kJ/mol with a 95% confidence interval of ± 15 kJ/mol.

ODR plot: L-Penicillamine 6

3 4 6 9 2 )) + 12

0 15 )/I(refH + 18 -2 21 ODR 3

Ln(I(dmcysH -4 ODR 6 ODR 9 -6 ODR 12 ODR 15 -8 ODR 18 -15 -10 -5 0 5 10 15 20 PA - PAavg

Figure 18: L-Penicillamine ODR plot

3.4 α-Methylserine Results

For α-methylserine, the references bases used included ethylamine, 4-chloropyridine, propylamine and isobutylamine, with their respective proton affinities as 912 kJ/mol, 916 kJ/mol,

917.8 kJ/mol and 924.8 kJ/mol.13 Due to there being unaccounted for secondary products at collision energies from 21-30 eV, only data from collision energies 3-18 eV was used in the proton affinity determinations. Table 4 shows the experimental data collected for α-methylserine.

Reference Base Average Proton Affinity 917.7 kJ/mol Slope (plot 2) 11.1 Proton Affinity from plot 2 928.8 kJ/mol Intercept (plot 2) 1.8 Change in Entropy from plot 2 -15.27 kJ/mol ODR Proton Affinity 934.7 ± 17 kJ/mol ODR Entropy Change -39 ± 3.7 kJ/mol Calculated Proton Affinity 933 kJ/mol

Table 4: α-Methylserine Experimental Data

+ + From the data, a plot of the average fragmentation intensity ratio (ln[BiH /AH ]) versus

PA-PAavg was made. Lines of best fit were added at each collision energy in order to look for a point of intersection, which can be seen in Figure 19 around between 10 and 15 of the PA-PAavg.

Plot 1: α-Methylserine 6

4 3 6

+) 2 9 12 )/I(refH

+ 15 0 18 Linear (3)

Ln(I(amserH -2 Linear (6) Linear (9) Linear (12) -4 Linear (15) Linear (18)

-6 -10 -5 0 5 10 15 20

PA-PAavg

Figure 19: α-Methylserine Plot 1

Another plot, Plot 2, was created to evaluate the relationship between negative intercept and the slope of the data used in plot 1. For α-methylserine, the data generally followed a linear curve, with a slope of 11.12 and an R2 value of 0.9886 (Figure 20). This Figure shows that collision energies 21 eV through 30 eV deviate from linearity, likely due to unaccounted for secondary products.

Plot 2: α-Methylserine y = 11.119x - 1.8363 R² = 0.9886 2.5

1.5

0.5

0 Negative Intercept

-0.5

-1

-1.5 0.1 0.15 0.2 0.25 0.3 0.35 0.4 Slope

Figure 20: α-Methylserine Plot 2

A third plot was created to evaluate the effective temperature in the mass spectrometer. As seen in Figure 21, this was seen in the α-methylserine data is quite linear, with an R2 value of

0.9903. This Figure again shows that collision energies 21 eV through 30 eV deviate from linearity.

Effective Temperature: α-methylserine y = 22.294x + 287.41 R² = 0.9903 800

750

700

650

600

550

500

Effective Temperature Effective(K) Temperature 450

400

350

300 0 5 10 15 20 25 30 35 Collision Energy (eV)

Figure 21: α-Methylserine Effective Temperature

The final step to determine the reported proton affinity of α-methylserine is to do an orthogonal distance regression, finding the best intersection points for the lines of best fit from plot 1.22 This can be seen in Figure 21, which when the x value of the point of intersection was added to the PAavg, gave the proton affinity for α-methylserine of 932 kJ/mol, with a 95% confidence interval of ± 20 kJ/mol.

ODR: α-methylserine 6 3

4 6 9 2 12 15

I(amserH+)) 0 \ 18

-2 ODR 3

Ln(IrefH+) ODR 6 -4 ODR 9 ODR 12 -6 -10 -5 0 5 10 15 20 ODR 15 PA - PAavg ODR 18

Figure 22: α-Methylserine ODR

3.3 3-Methylthreonine Results

For 3-methylthreonine, the references bases used included ethylamine, propylamine, N,N- dimethylacetamide and isobutylamine, with their respective proton affinities as 912 kJ/mol, 917.8 kJ/mol, 908 kJ/mol and 924.8 kJ/mol.13 Due to there being non-statistical deviations at collision energies of 3 and 6 eV, and unaccounted for secondary products at collision energies 27 and 30 eV, only data from collision energies 9-24 eV was used in the proton affinity determinations. Table

5 shows the experimental data collected for 3-methylthreonine.

Reference Base Average Proton Affinity 915.7 kJ/mol Slope (plot 2) 13.8 Proton Affinity from plot 2 929.5 kJ/mol Intercept (plot 2) 1.04 Change in Entropy from plot 2 -8.6 kJ/mol ODR Proton Affinity 924.4 ± 8.7 kJ/mol ODR Entropy Change -4.7 ± 0.6 kJ/mol Calculated Proton Affinity 929 kJ/mol

Table 5: 3-Methylthreonine Experimental Data

+ + From the data, a plot of the average fragmentation intensity ratio (ln[BiH /AH ]) versus

PA-PAavg was made. Lines of best fit were added at each collision energy in order to look for a point of intersection. In Figure 23, this can be seen between 12 and 17 of the PA-PAavg.

Plot 1: 3-Methylthreonine 3

1 9 12

))) 0 15 + 18 -1 21 )/I(dmserH

+ 24 -2 Linear (9) -3 Linear (12) Ln(I(refH Linear (15) -4 Linear (18) Linear (21) -5 Linear (24)

-6 -8 -3 2 7 12 17 PA-PAavg

Figure 23: 3-Methylthreonine Plot 1

Another plot, Plot 2, was created to evaluate the relationship between the negative intercept and the slope of the lines of best fit in plot 1. For 3-methylthreonine, the data generally followed a linear curve, with an R2 value of 0.9641 (Figure 24). As seen in Figure 24, the data at collision energies 3 eV, 6 eV, 27 eV and 30 eV were excluded from further calculations. The deviations at the lower collision energies are likely due to non-statistical dissociations, while the deviations at the higher collision energies are likely due to unaccounted for secondary products.

Plot 2: 3-Methylthreonine y = 13.831x - 1.0381 R² = 0.9641 3.5

2.5

1.5 Negative Intercept

0.5

0 0.14 0.16 0.18 0.2 0.22 0.24 0.26 0.28 Slope

Figure 24: 3-Methylthreonine Plot 2

A third plot was created to evaluate the effective temperature in the mass spectrometer. As seen in Figure 25, the 3-methylthreonine data is reasonably linear in the range of 9-24 eV, with an

R2 value of 0.9749. The deviations from linearity seen in plot 2 can again be seen in this plot.

Effective Temperature: 3-Methylthreonine y = 22.562x + 290.43 R² = 0.9749 900

850

800

750

700

650

600 Effective (K) Temp

550

500

450

400 0 5 10 15Collision Energy20 (eV) 25 30 35

Figure 25: 3-Methylthreonine Effective Temperature

A last step to determining the experimental proton affinity involved performing an orthogonal distance regression (ODR), which finds the best intersection point for the reference base lines in plot 1 to give the proton affinity of the molecule of interest.22 This plot can be seen in Figure 26, which corresponds to an experimental proton affinity for 3-methylthreonine of 924 kJ/mol with a 95% confidence interval of ± 15 kJ/mol.

ODR: 3-Methylthreonine 2

1 9 12 0 15 )) + -1 18

)/I(refH 21 + -2 24

-3 ODR 9 ODR 12 Ln(I(dmserH -4 ODR 15 ODR 18 -5 ODR 21 -6 ODR 24 -8 -4 0 4 8 12 PA-PAavg

Figure 26: 3-Methylthreonine ODR

3.6 Experimental Results Summary

The proton affinities determined for α-methylcysteine, l-penicillamine, α-methylserine and

3-methylthrenonine can be compared with the proton affinity of the protein amino acids cysteine and serine. As the molecules are homologs of cysteine and serine, it is of interest to see how their proton affinities compare to these common amino acids. Table 6 shows a summary of the experimental versus computational results from this study, while Table 7 shows the previously determined proton affinities of cysteine and serine.13

Amino Acid PA (experimental) PA (calculated) α-methylcysteine 923.7 ± 11 kJ/mol 925 kJ/mol L-penicillamine 925 ± 17 kJ/mol 920 kJ/mol α-methylserine 932 ± 20 kJ/mol 933 kJ/mol 3-methylthreonine 924 ± 15 kJ/mol 929 kJ/mol

Table 6: Summary of Experimental versus Computational Results

Amino Acid Proton Affinity Cysteine 903 kJ/mol13 Serine 912 kJ/mol13

Table 7: Proton affinity of Cysteine and Serine

In all cases, the proton affinity of the methylated non-protein amino acids is higher than the proton affinity of the protein amino acid. This trend is consistent which previous studies that found methylating histidine, another protein amino acid, resulted in an increased proton affinity.30

This is likely due to increased polarization and inductive effects in the methylated molecules, which increases stabilization. This increase is significant and implies that when these molecules form peptides, they will act as stronger bases than their protein amino acid counterpart.

Chapter 4: Conclusions and Future Work

We were successfully able to determine the absolute proton affinities of α-methylserine, α- methylcysteine, L-penicillamine and 3-methylthrenonine using the extended kinetic method on our triple quadrupole mass spectrometer. These proton affinities were in excellent agreement with the computational values calculated from the lowest energy conformers using PCModel and

Gaussian09. The computational experiment also successfully showed that the lowest energy conformer of the cysteine homologs had the thiol group as the deprotonating group, consistent with previous studies.28,29 This data will be useful as we continue working with the Hartman group to learn more about the structure and energetics of these non-protein amino acids, and their potential for incorporation into peptides and proteins.

The gas-phase acidities for the methylated serine and cysteine homologs have already been computationally calculated by another member of the lab. In future kinetic method experiments, we will turn to determining the gas-phase acidities of these molecules in negative ion mode in the triple quadrupole mass spectrometer to compare these values. These studies will determine the gas phase acidities of these molecules and give more indications into the bonding and structure of these molecules.

We will also continue determining the proton affinities of proline containing dipeptides, a particular area of interest due to proline’s unique R group that includes a five-membered ring. This ongoing project, in collaboration with Dr. Jennifer Poutsma’s lab at Old Dominion University, is seeking to see how the ‘proline effect’ affects dipeptides with proline at either the C-terminal or

N-terminal.

References

[1] Iqbal, E. S., Dods, K. K., & Hartman, M. C. T. T. Ribosomal incorporation of backbone Modified amino acids via an editing deficient aminoacyl-tRNA synthetase. Org. Biomol. Chem. 2018, 16, 795.

[2] Hunter, E.P.; Lias, S.G., Evaluated Gas Phase Basicities and Proton Affinities of Molecules: An Update, J. Phys. Chem. Ref. Data, 1998, 27, 3, 413- 656, https://doi.org/10.1063/1.556018.

[3] Dunlop RA, Cox PA, Banack SA, Rodgers KJ The Non-Protein Amino Acid BMAA Is Misincorporated into Human Proteins in Place of l-Serine Causing Protein Misfolding and Aggregation. PLoS ONE. 2013. 8(9): e75376. https://doi.org/10.1371/journal.pone.0075376.

[4] Kanai Y. Family of neutral and acidic amino acid transports: molecular biology, physiology and medical implications. Current Opinion in Cell Biology. 1997. 9(4) pp 565-572. https://doi.org/10.1016/S0955-0674(97)80035-X.

[5] Kuntz AF, Boynton AW, David GA, Colyer KE, Poutsma, JC. The Proton Affinity of Proline Analogs Using the Kinetic Method with Full Entropy Analysis. Amer. Soc. Mass. Spec. 2002, 13, pp 72-81.

[6] Schroeder OE. Andriole EJ, Carver KL, Colyer KE, Poutsma JC. Proton Affinity of Lysine Homologues from the Extended Kinetic Method. J. Phys. Chem. A, 2004, 108 (2), pp 326–332 DOI: 10.1021/jp0359182.

[7] Andriole EJ, Colyer KE, Cornell E, Poustma JC. Proton Affinity of Canavanine and Canaline, Oxyanalogues of Arginine and Ornithine, from the Extended Kinetic Method. J. Phys. Chem. A, 2006, 110 (40), pp 11501–11508 DOI: 10.1021/jp063081f.

[8] Maksić, Z.B., Kovačević, B., and Vianello, R. Advances in Determining the Absolute Proton Affinities of Neutral Organic Molecules in the Gas Phase and Their Interpretation: A Theoretical Account. Chem. Rev. 2012. 112(10), pp 5240-5270. doi:10.1021/cr100458v.

[9] IUPAC. Compendium of Chemical Terminology, 2nd ed. (the "Gold Book"). Compiled by A. D. McNaught and A. Wilkinson. Blackwell Scientific Publications, Oxford . 1997. XML on-line corrected version: http://goldbook.iupac.org (2006-) created by M. Nic, J. Jirat, B. Kosata; updates compiled by A. Jenkins. ISBN 0-9678550-9- 8. https://doi.org/10.1351/goldbook.

[10] Hunt S. The Non-Protein Amino Acids. In: Barrett G.C. (eds) Chemistry and Biochemistry of the Amino Acids. Springer, Dordrecht. 1985.

[11] Bell, E.A. Nonprotein Amino Acids of Plants: Significance in Medicine, Nutrition and Agriculture. J. Agric. Food Chem. 2003. 51, pp 2854-2865.

[12] Raczynska, E.D., Gal, J.F., Maria, P.C., Szelag, M. Proton Transfer Chemistry in the Gas Phase. Is a Spontaneous ‘Neutralization’ Reaction a Myth or Reality? Corat. Chem. Acta, 2009, 82(1), pp 87-103.

[13] Hunter, E.P.; Lias, S.G., Evaluated Gas Phase Basicities and Proton Affinities of Molecules: An Update, J. Phys. Chem. Ref. Data, 1998, 27, 3, 413- 656, https://doi.org/10.1063/1.556018.

[14] Harrison, AG. The gas-phase basicities and proton affinities of amino acids and peptides. Mass. Spec. Rev. 1997. 16(4), pp 201-217.

[15] Cooks, R.G., Kruger, T.L. Intrinsic Basicity Determination Using Metastable Ions. J. Am. Chem. 1977. 99 (4), pp 1279-1291.

[16] Bohme, D.K., Hemsworth, R. Determination of Proton Affinity from the Kinetics of Proton Transfer Reactions. The Journal of Chemical Physics. 1973. 95(22), pp 512-7513.

[17] Decouzon, M., Gal, J.F., Herreros, M., Maria, P.C., Murrell, J., and Todd., J.F. On the Use of the Kinetic Method for the Determination of Proton Affinities by Fourier-transform IonCyclotron Resonance Mass Spectrometry. Rapid Communications in Mass Spectrometry. 1996. 10, pp 242-245.

[18] Maksić, Z.B., Kovačević, B., and Vianello, R. Advances in Determining the Absolute Proton Affinities of Neutral Organic Molecules in the Gas Phase and Their Interpretation: A Theoretical Account. Chem. Rev. 2012. 112(10), pp 5240-5270. doi:10.1021/cr100458v.

[19] Cooks, R.G., Wong, P.S. Kinetic Method of Making Thermochemical Determinations: Advances and Applications. Accounts of Chemical Research. 1998. 31 (7), pp 379-386.

[20] Poutsma J. Mass Spec Powerpoints. 2017. Powerpoint Presentation.

[21] Cooks, R.G., Koskinen, J.T., Thomas, P.D. The Kinetic Method of Making Thermochemical Determinations. J. Mass Spectrom. 1999. 34, pp. 85-92.

[22] Ervin, K.M., Armentrout, P.B. Systematic and random errors in ion affinities and activation entropies from the extended kinetic method. J. Mass Spectrom. 2004. 39(9).

[23] March, R.E. An Introduction to Quadrupole Ion Trap Mass Spectrometry. Journal of Mass Spectrometry. 1997. 32, pp 351-369. https://doi-org.proxy.wm.edu/10.1002/jms.682

[24] Ni, J., and Rowe, J. Microdosing Assessment to Evaluate Pharmacokinetics and Drug Metabolism Using Liquid Chromatography-Tandem Mass Spectrometry Technology, Topics on Drug Metabolism, James Paxton, IntechOpen, 2012. doi: 10.5772/28525. Available from: https://www.intechopen.com/books/topics-on-drug-

metabolism/microdosing-assessment-to-evaluate-pharmacokinetics-and-drug- metabolism-using-lc-mc-ms-technology.

[25] Emory College Mass Spectrometry Center. Mass Spectrometry Ionization Methods. http://chemistry.emory.edu/msc/tutorial/mass-spectrometry-ionization.html. Accessed April 10, 2019.

[26] Yost, R.A., and Enke, C.G. Triple Quadrupole Mass Spectrometry. Anal. Chem. 1979. 51(12), pp 1251A-1264A.

[27] Iqbal, E. Addition of L-Penicillamine and 3-methyl Threonine to in vitro translation. Powerpoint. 2019.

[28] DeBlase, A.F., Kass, S.R., and Johnson, M.A. On the character of the cyclic ion H-bond in cryogenically cooled deprotonated cysteine. Phys. Chem. Chem. Phys. 2014. 16, pp 4569 4575. doi: 10.1039/C3CP54117G.

[29] Tian Z, Pawlow A, Poutsma JC, Kass SR. Are Carboxyl Groups the Most Acidic Sites in Amino Acids? Gas-Phase Acidity, H/D Exchange Experiments, and Computations on Cysteine and Its Conjugate Base. J. Am. Chem. Soc., 2007, 129 (17), pp 5403–5407 DOI: 10.1021/ja0666194

[30] Rožman, M. Proton Affinity of Several Basic Non-Standarad Amino Acids. Chem. Phys. Letters. 2012. 543, pp 50-54. doi: https://doi.org/10.1016/j.cplett.2012.06.048.