<<

Wright State University CORE Scholar

Browse all Theses and Dissertations Theses and Dissertations

2015

Computational Estimation of the PKa's of and Related Compounds

Kara L. Geremia Wright State University

Follow this and additional works at: https://corescholar.libraries.wright.edu/etd_all

Part of the Chemistry Commons

Repository Citation Geremia, Kara L., "Computational Estimation of the PKa's of Purines and Related Compounds" (2015). Browse all Theses and Dissertations. 1610. https://corescholar.libraries.wright.edu/etd_all/1610

This Thesis is brought to you for free and open access by the Theses and Dissertations at CORE Scholar. It has been accepted for inclusion in Browse all Theses and Dissertations by an authorized administrator of CORE Scholar. For more information, please contact [email protected]. Computational Estimation of the pKa’s of Purines and Related Compounds

A thesis submitted in partial fulfillment

of the requirements for the degree of

Master of Science

By

Kara L Geremia

B.S., Wright State University, 2014

2015

Wright State University

i

WRIGHT STATE UNIVERSITY

GRADUATE SCHOOL

November 23, 2015

I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY

SUPERVISION BY Kara Geremia ENTITLED Computation Estimation of the pKa’s of Purines and Related Compounds BE ACCEPTED IN PARTIAL FULFILLMENT OF THE REQUIRMENTS FOR THE DEGREE OF Master of Science.

______Paul Seybold, Ph.D. Thesis Director

______Eric Fossum, Ph.D. Thesis Co-Director

______David A. Grossie, Ph.D. Chair, Department of Chemistry Committee on Final Examination

______Paul Seybold, Ph.D.

______Eric Fossum, Ph.D.

______Kenneth Turnbull, Ph.D.

______Rachel Aga, Ph.D.

______Robert E. W. Fyffe, Ph.D. Vice President for Research and Dean of Graduate School

ii

ABSTRACT

Kara L. Geremia M.S., Department of Chemistry, Wright State University, 2015. Computational

Estimation of the pKa’s of Purines and Related Compounds.

The dissociation value pKa of a compound provides important information regarding the neutral and ionic forms of the compound present under different conditions. However, for many compounds experimental pKa values are not available, and for others the measured values are sometimes uncertain. Consequently, having a computational means for estimating pKa values can be of benefit for biochemical, pharmaceutical, polymer, and many other studies.

Purines form a key group of bioactive compounds whose acid/ activities are of considerable interest. The class includes such diverse compounds as the nucleic acid bases and , the gout-related compound , the stimulant caffeine, and the leukemia drug 6- mercaptopurine. We have developed a computational method for estimating purine pKas using a quantitative structure property relationship (QSAR) approach. A group of 31 purines and related compounds was first assembled and then examined using the semi-empirical quantum chemical method RM1. This was followed by an ab initio analysis based on density functional theory at the B3LYP/ 6-31+G** level.

iii

Table of Contents Chapter 1—Purines and Related Compounds

1.1 Introduction...... [1]

1.2 Historical Background...... [1]

1.3 Profiles of the Compounds Studied...... [2]

1.4 References...... [13]

Chapter 2--Computational Analysis of Cationic, Neutral, and Anionic Purine Tautomers

2.1 Abstract...... [20]

2.2 Introduction...... [20]

2.3 Computational Methods...... [22]

2.4 Results and Discussion...... [23]

2.4.1 Purine...... [23]

2.4.2 ...... [27]

2.4.3 6-Mercaptopurine...... [31]

2.4.4 Xanthine...... [35]

2.4.5 Adenine...... [39]

2.4.6 Guanine...... [42]

2.5 Conclusion...... [45]

2.6 References...... [45]

Chapter 3-- Computational Analysis of Tautomers for and Related Compounds

3.1 Abstract...... [48]

3.2 Introduction...... [48]

3.3 Computational Method...... [49]

iv

3.4 Results and Discussion...... [49]

3.4.1 Indole ...... [49]

3.4.1a. Neutral Tautomers...... [49]

3.4.1b. Cationic Tautomers...... [50]

3.4.2. ...... [52]

3.4.2a. Neutral Tautomers...... [52]

3.4.2b. Cationic Tautomers...... [55]

3.4.3 ...... [56]

3.4.3a. Neutral Tautomers...... [56]

3.4.3b. Cationic Tautomers...... [57]

3.4.4 ...... [59]

3.4.4a. Neutral Tautomers...... [59]

3.4.4b. Cationic Tautomers...... [60]

3.4.5 7-Azaindole...... [61]

3.4.5a. Neutral Tautomers...... [61]

3.4.5b. Cationic Tautomers...... [62]

3.4.6 Caffeine...... [63]

3.4.6a. Neutral Species...... [63]

3.4.6b. Cationic Species...... [63]

3.4.6c. Anionic Species...... [64]

3.5 Conclusion...... [66]

3.6 References...... [66]

Chapter 4-- Computational Estimation of the pKa’s of Purine and Purine Related Compounds

4.1 Abstract...... [69]

v

4.2 Introduction...... [69]

4.3 Computational Methods...... [71]

4.4 Results and Discussion...... [74]

4.5 Conclusion...... [86]

4.6 Appendix...... [86]

4.7 References...... [87]

vi

List of Figures

Chapter 2--Computational Analysis of Cationic, Neutral, and Anionic Purine

Tautomers

2.1 An illustration of the intrinsic energies and the adjustment to the energies as a result [21] from being placed in a solvent model......

2.2 Four possible neutral tautomers of purine with energy differences calculated in [23] vacuum......

2.3 Molecular electrostatic potential maps of the 9H-P and 7H-P tautomers of purine...... [24]

2.4 Cationic tautomers with decreasing stabilizations as found in vacuum...... [26]

2.5 The deprotonation of the favored tautomer (9H-P) in vacuum to its anionic form...... [27]

2.6 Hypoxanthine tautomers and their relative energies in vacuum...... [29]

2.7 Cationic tautomers resulting from the 1,9(H)-HX neutral tautomer in vacuum...... [29]

2.8 Relative energies for deprotonation of the two neutral hypoxanthine tautomers, a) 1,9- [30] hypoxanthine and b) 1,7-hypoxanthine, to their anionic forms as found in a water solvent...

2.9 Thione-thiol tautomers of 6-mercaptopurine as calculated in vacuum...... [32]

2.10 Illustrations showing bond lengths for the favored thione tautomer 1,7(H)-MP and the [33] high-energy 3,9(H)-MP as found using the water solvent model......

2.11 Cationic tautomers from the 1,7(H)-MP neutral tautomer as found in vacuum...... [34]

2.12 Anionic tautomers for 6-mercaptopurine as found in vacuum...... [34]

2.13 The two lowest-energy keto tautomers of xanthine calculated in vacuum...... [35]

2.14 Cationic tautomers for xanthine from the vacuum calculations...... [37]

2.15 Anionic tautomers for xanthine found in vacuum...... [37]

2.16 Natural charges for the stable neutral tautomer in solution with the possible [38] dissociation sites highlighted in red......

2.17 Tautomer energies of adenine determined in vacuum...... [39]

2.18 Cationic tautomers with decreasing stabilities found in vacuum...... [41]

2.19 Two anionic tautomers for adenine with relative energies found in vacuum...... [41]

vii

2.20 Neutral keto-amine tautomers of guanine as determined in vacuum...... [42]

2.21 Cationic tautomers of guanine found in vacuum...... [44]

2.22 Three anionic tautomers that should be probable in solution for guanine as found in [45] vacuum......

Chapter 3--Computational Analysis of Tautomers for Indole and Related Compounds

3.1 Neutral tautomers for indole as found in vacuum...... [50]

3.2 Cationic tautomers for indole as found in vacuum...... [51]

3.3 The deprotonation scheme to the neutral tautomer of indole...... [52]

3.4 Neutral tautomers for isoindole as found in vacuum...... [53]

3.5 Molecular electrostatic potential maps of the a) 3C-Iso and b) 2H-Iso tautomers of [54] isoindole......

3.6 Cationic tautomers for isoindole as found using a water solvent model, SM8...... [55]

3.7 Dissociation scheme for isoindole in vacuum acting as a carbon acid...... [55]

3.8 Neutral tautomers for indazole as found in vacuum...... [57]

3.9 Cationic tautomers for indazole as found in vacuum...... [58]

3.10 Neutral tautomers for benzimidazole as found in vacuum...... [59]

3.11 Cationic tautomers for benzimidazole as found using a water solvent model, SM8...... [60]

3.12 Neutral tautomers for 7-azaindole as found in vacuum...... [62]

3.13 Cationic tautomers for azaindole as found in vacuum...... [62]

3.14 The singular, neutral tautomer of caffeine. The numbering scheme is different from [63] indole derivatives as noted above......

3.15 The cationic tautomer of caffeine...... [64]

3.16 Dissociation of caffeine to its anionic form...... [64]

viii

Chapter 4--Computational Estimation of the pKa’s of Purines and Related Compounds

4.1 Comparison of the experimental and calculated Gibbs energy changes for dissociation [76] of the neutral species to the anionic form and H+......

4.2 A plot of ΔEvac against the experimental (NIST) Gibbs energies...... [78]

4.3 The QSAR plot for experimental pKa2 against the QC descriptor ΔEVAC...... [79]

4.4 The bimolecular hydrogen bonding of 7-azaindole, which results in dimerization...... [80]

4.5 The QSAR plot for experimental pKa1 against the QC descriptor ΔEH2O...... [81]

4.6 The QSAR plot for experimental pKa2 against the QC descriptor ΔEH2O...... [82]

4.7 The QSAR plot for experimental pKa2 against the QC descriptor ΔQNH...... [83]

4.8 The QSAR plot for experimental pKa2 against the QC descriptors ΔEH2O and ΔQNH...... [84]

ix

List of Tables

Chapter 2--Computational Analysis of Cationic, Neutral, and Anionic Purine

Tautomers

2.1 Relative energies and solvation energies of the neutral, cationic, and anionic tautomers [25] for purine......

2.2 Relative energies and solvation energies of the neutral, cationic, and anionic tautomers [28] for hypoxanthine......

2.3 Relative energies and solvation energies of the neutral, cationic, and anionic tautomers [31] for 6-mercaptopurine......

2.4 Relative energies and solvation energies of the neutral, cationic, and anionic tautomers [36] for xanthine......

2.5 Relative energies and solvation energies of the neutral, cationic, and anionic tautomers [40] for adenine......

2.6 Relative energies and solvation energies of the neutral, cationic, and anionic tautomers [43] for guanine......

Chapter 3--Computational Analysis of Tautomers for Indole and Related Compounds

3.1 Relative energies and salvation energies of the neutral and cationic tautomers for indole [51]

3.2 Relative energies and salvation energies of the neutral and cationic tautomers for [54] isoindole......

3.3 Relative energies and salvation energies of the neutral and cationic tautomers for [56] indazole......

3.4 Relative energies and salvation energies of the neutral and cationic tautomers for [59] benzimidazole......

3.5 Relative energies and salvation energies of the neutral and cationic tautomers for 7- [61] azaindole......

Chapter 4--Computational Estimation of the pKa’s of Purines and Related Compounds

4.1 Experimental pK values for purine and related compounds along with theoretical a [73] values obtained from ACD/Phys Chem Suite......

4.2 Experimental and calculated Gibbs energy changes used in Figure 4.1...... [75]

x

4.3 The experimental Gibbs energies correlated with the calculated ΔEVAC...... [77]

4.4 Theoretically calculated pK ’s using the QSAR regressions shown in the figure 4.5, 4.6, a [85] and 4.8 above......

xi

Acknowledgement

I would like to thank Dr. Paul Seybold for his guidance, patience, and supervision. I would also like to thank Dr. Eric Fossum for his many, many moments. I cannot extend my gratitude enough to my husband, mother, father, daughter, and brothers. Completing this project would not have happened if it were not for my supportive family.

I am also very appreciative of Dr. David Dolson for all of his advice throughout my undergraduate and graduate careers.

“I touch the sky when my knees hit the ground”

--Hillsong United

xii

CHAPTER ONE

Introduction: Purines and Related Compounds

1.1 Introduction

Purines and purine-related compounds comprise a large class of biochemically active compounds with a great diversity of roles. This thesis is devoted to a computational study of these compounds with an emphasis on the relative stabilities of the tautomers of these compounds and their acidities.

This chapter is devoted to an introductory description of the uses and biochemical roles of the compounds examined. Chapter Two explores the tautomerization and implications on dissociations for six purine related compounds. Chapter Three is an overview of the tautomerization of five indole related compounds plus caffeine. Chapter Four focuses on the estimation of the pKa’s of 31 purines and related compounds.

1.2 Historical Background

The purines are N-heterocycles that have an unusual numbering scheme arising from the fact that the parent compound is treated as a modified ring.1 Most indole-related compounds such as the , on the other hand, do adhere to the normal numbering pattern for

N-heterocycles.

1

Purine numbering scheme Indole numbering scheme

Purine-related compounds have been studied since the 1770s when Scheele and

Bergmann isolated uric acid from human urine.1 By the 1880s most naturally occurring purines had been isolated. This has led to their utilization in many research areas such as pharmaceuticals,2 industrial applications,3 environmental chemistry,4 and polymer chemistry.5

1.3 Profiles of the Compounds Studied

Thirty-one compounds were compiled for computational analysis. To facilitate a thorough understanding of these compounds a literature search was performed on each compound to find its current uses, historical background, and so forth. The pertinent information obtained for each compound ultimately assists in interpreting the theoretical results.

Adenine, one of the five nucleic , is a crucially important compound found in the human body. It is involved in DNA, RNA, and the energy releasing compounds adenosine triphosphate (ATP) and nicotinamide adenine dinucleotide (NAD).6 Due to its ability to phosphorylize other compounds ATP is central to energy exchange in cells.7 The total energy released from a phosphodiester bond cleavage on an ATP molecule is -7.3 kcal/mol.8

2

The azaindole series, 4-, 5-, 6-, and 7-azaindole, are heterocycles that act similarly to and , yet their

9 range in pKa leads to a broader range of activities and uses. In

2014 a 4-azaindole derivative was reported to be a candidate for tuberculosis treatment.10 Following in the pharmaceutical trend, 5-azaindole and 7-azaindole were seen as promising Cdc-7 inhibitors for cancer treatment. It can be noted that Cdc-7 is a cell division cycle-7 kinase that initiates DNA replication during the S-phase.11 Finally, 6-azaindole has been incorporated in the core of the first small molecule to be used in a treatment of the

Human Immunodeficiency Virus type I.2 7-Azaindole is the most studied of the series. This compound possesses the unique capability to dimerize due to intermolecular , which causes it to preferentially dimerize.12

Benzimidazole is a six membered ring fused to a five membered ring containing two . This compound rose to fame in the mid twentieth century after it was discovered to be involved in the synthesis of vitamin B12. Anthelmintic applications were soon discovered using derivatives of this compound.13 Presently it is being incorporated into polymeric systems as a bidentate monomer accompanied by transition metals to be used as the impetus for self- assembling supramolecules.5

Benzimidazoline, the saturated form of benzimidazole, has been integrated into a select few areas of chemistry. This highly reducing organic compound that has been noted to react with o-phenyldiamine.14 Benzimidazoline gained interest due to its ability to undergo exergonic reactions under mild conditions to release hydrogen that can be used in hydrogen storage. Hydrogen release under conditions described by

3

Schwarz et al.15 provides future hopes for thermodynamically reversible hydrogen evolution/devolution.

Carbazole is a tricyclic ring that has been applied to many areas such as environmental, dye production, reagents, insecticides, and medicinal chemistry.16 Carbazole knowledge is vital due to the ratio at which it is found in crude oil where its derivatives comprise the majority of all containing compounds. Carbazole can radicalize easily and become toxic; therefore the removal of residues in oil is a priority.

Researchers have been looking into bacteria which can metabolize this compound. These bacteria include Pseudomonas resinovorans, Sphingomonas sp. CB3, and various others.17

Caffeine is the mostly widely consumed stimulant in the world.18 This methylated xanthine acts on the central nervous system to replace adenosine on the receptors of nerve cells.19

Caffeine is a trimethylated xanthine that is degraded in the liver to the following three compounds paraxanthine (1,7-dimethyl), theobromine (3,7-dimethyl), and theophylline (1,3-dimethyl) in decreasing concentrations, respectively.20 Due to caffeine’s ability to excite and stimulate nerve cells many studies have tried to correlate its properties to enhanced performances, exceptional creative abilities, and so forth. A study in the Journal of the American Medical Association has successfully correlated an increase in caffeine intake to a decrease in Parkinson’s disease in elderly Japanese men.21

Diphenylamine has been known to have antioxidant properties which have made it prominent in industrial

4 applications. A commercial use for diphenylamine has been to prevent apple scald by dipping harvested apples into the compound.3 When apples develop the peel disorder the compounds commonly associated with the lesions and peel pittings are α-farnesene trienols. Treatment with diphenylamine ameliorates the oxidative damage and allows for the superficial appearances to be maintained.22

Guanine is a purine found in DNA and RNA that is crucial to the make-up of human life. Guanine is used in energy transfer within cells. During this process guanine takes the form of guanine triphosphate, GTP, and follows the same mechanism as ATP.23 The keto- tautomer has been studied extensively going back to the discoverers of DNA Watson and Crick. The lower energy, more thermodynamically stable keto tautomer is found frequently, however, the higher energy enol tautomer has been hypothesized to lead to spontaneous mutagenesis.24 Mismatching of the hydrogen bonding between the complementary or can lead to base pair wobble.

Hypoxanthine is a purine derivative that is biologically important due to its presence in tRNA. Hypoxanthine is a product of purine and nucleotide degradation. One common pathway that has been mapped is through fish muscle where ATP dephosphorylates to

AMP so that it becomes deaminated to inosine (the nucleoside of hypoxanthine). This compound becomes hydrolyzed to hypoxanthine which can be oxidized to xanthine and finally oxidized to the final product uric acid. Therefore, it has been proposed that the concentration of hypoxanthine within fish meat can be an indicator of quality.25

5

Indazole is an aromatic heterocycle. This particular species is rarely found naturally, but has been found in many complex pharmaceutical agents. Even when it is found in nature it is substituted.26 Indazole’s framework has been utilized in a variety of biological activities such as estrogen receptors, HIV protease inhibition, and anti-tumor activity.27 The indazole skeleton has also been documented in synthetic cannabinoids as psychoactive drugs.28

Indole is a metabolism product of tryptophan. It is produced naturally by many bacteria, both Gram-positive and Gram-negative, which apply the compound in intercellular signaling, biofilm production, and plasmid stability.29 Gut bacteria can produce ample quantities of indole, which is why indole metabolites can be directly related back to the bacteria found in the gut.30 Indole has not been found to act as a nitrogen acid in the protonated form, however indole protonates on the 3(C) position preferentially.31

Indoline is the saturated form of indole. While a wide array of information can be found on indole the research on indoline is selective.

It was recently shown that N-benzyl tricyclic indoline was used successfully as an antibiotic for methicillin-resistant Staphylococcus aureus (MRSA).32 Indoline has been incorporated many times into dye-solar cells as part of the third generation photovoltaic technology.33

Isatin was first discovered in 1840 by Laurent and

Erdmann.34,35 The compound is found in plants under the genus

Isatis.36 Isatin can also be seen as a component of coal tar and

6 fungi.37,38 There have been many areas of chemistry in which isatin has been utilized.

Remarkably, it has been known for its ability to be used as a dye in hair, as well as its ability to be used in biological analytical techniques.

The presence of isoguanine in a genome has been studied extensively due to its ability to form 23 different possible tautomers. For the five in RNA and DNA only certain tautomers are more dominant in the formation of the natural polymers.39 The idea of minor tautomers showing up in base pairing has led to the discussion of mutagenesis and base pair wobble in the genome. Subsequently, Watson-Crick like hydrogen bonding with isocytosine has been proposed unique complexes and expanded the genomic alphabet.40

Isoindole is an of indole and far less stable. This results from the compound’s desire to auto-oxidize and proceed to self-condensation reactions.41

Using the isoindole moiety, many research groups have assessed the compound’s ability to be utilized in antiproliferative drugs with anti- tumor activity.42-44 The isoindole skeleton has also been utilized in many dyes, which have been incorporated into fluorophore and nano-technology.45,46

Oxindole was prepared in 1866 and 1868 by Baeyer.47

The tautomerism of oxindole was explored and it was concluded that the lactam tautomer was the favorable structure based on

7 spectroscopic studies performed by Ramart and Biquard.48 Since the discovery oxindole has been patented for use in anti-cancer treatments due to its toxicity levels to cancer cell lines.49

Oxindole is an intermediate compound in the conversion of indole to isatin.

Paraxanthine is a dimethylated xanthine derivative. It is not a naturally occurring , but can be found in animals as a by-product from caffeine metabolism. Due to the compound’s ability to act as a central nervous system stimulant it has been proposed to be found as a metabolite for cancer cachexia and cancer related anorexia.50 As previously stated, paraxanthine was the primary product of caffeine degradation as well as caffeine playing a suggestive role in decreasing the occurrence of Parkinson disease.21,22 Guerrero et al not only agreed with the findings for caffeine and Parkinson disease, but also mechanistically described paraxanthine as an inhibitor of neurodegeneration.50

Phthalimide is an connected to a six member ring.

This compound has been incorporated in many facets of chemistry due to its imide functionality. A derivative has recently been used as a drug delivery system for neo-growth factors in regenerative skin repair polymers. Guadalupe et al found that phthalimide neovascular factor-1 (CAS#: 10442-95-2) showed promising characteristics for angiogenic effects.51 The nucleotide ATP is essential in the transduction of energy in human systems,52 whereas the reduction of ATP concentrations in humans has been linked to ischemia and various other clinical conditions.53 A polymer that fixed phthalimide to the backbone was synthesized for colorimetric technology in ATP detection. This phthalimide

8 moiety was essential because the imide mimicked the hydrogen bonding capabilities adenine has with its complementary base pair, thymine.54

Purine, the core structure for this study, was first discovered by Emil Fischer.55 The anomalous numbering scheme of purine follows as if it were a pyrimidine derivative though it is not.56 Purine has the capability to form four tautomers. The 7H tautomer has been isolated as the primary product when purine is in a crystalline form; however, in solution the 7H and 9H tautomers are in equivalent amounts.57

6-Mercaptopurine is a leukemia treatment discovered by Elion and Hitchings in

1950.58 6-Mercaptopurine is used in combination treatment usually with methotrexate. Toxicity can occur by cellular uptake and conversion of guanine nucleotides to thioguanine nucleotides. The conversion ultimately causes the DNA to become fragile, inhibiting replication, and causing apoptosis.59 This mechanism may be the appropriate action for malignant cells, but the attacks happen to healthy cells as well. Be it as it may, 6-mercaptopurine is still one of the leading treatments in childhood leukemia.

Saccharin, the long time controversial compound, was discovered in the late 1860s by Constantine Fahlberg60 by happenstance. This opened the door for artificial sugars to be used by diabetics because of the zero caloric impact that has on the body. Saccharin is excreted from the body and into the

9 waste water systems in its natural form. Subedi et al followed the fate of saccharin in these systems to show that it can be easily removed; for example, in Albany, New York approximately

1100 g of saccharin are removed from waste-water per day.4

3-Methylindole, more commonly known as skatole, is a unique compound in that its concentration plays a major role in the selectivity for its uses. Skatole in high concentrations has been patented as a malodorant for the United States Military for non-lethal weaponry.61 Contrary to this, skatole has been known to have a floral scent found in teas62 and naturally occurring as an odorant in butter.63

Theobromine is the second most common byproduct of caffeine degradation.20 This 3,7-dimethylxanthine is also commonly found in cocoa plants, specifically Theobroma cacao.64 Theobromine has been researched extensively for its use as an antioxidant in chocolate; however it is now being considered as a drug release system for a poly(dimethysiloxane) based matrix.65 Many pharmaceutical drugs have a unique amine or functionality thereby allowing theobromine to be a qualifying candidate.

Theophylline is the least common byproduct of caffeine degradation.20 It was originally extracted from tea in 1888.1 Theophylline is commonly used as a bronchodilator in asthma treatment.66 Recently it has been thought to have a mechanistic approach in antagonizing a

10 tumor’s necrosis factor alpha and also increase the pro-inflammatory effects of prostaglandin.67

It is because of the multi-faceted functionality of theophylline that a new approach was given to this old pharmaceutical. It was shown to lower the lipid levels in rats in experimentally induced hyperlipidemia.68

Uric acid is the final stage in purine metabolism for humans. It can come from dietary sources or endogenously.

For many years uric acid has been a suspected player in gout, kidney stone formation and many other ailments that affect the urinary tract system. The salt form of uric acid, urate, is released in the human body as a protective response when it experiences hypoxia due to its antioxidant capabilities.69 Further conclusions were that there was no evolutionary symbiotic affect for urate’s antioxidant capabilities, where as it developed as a biochemical coincidence.70

Unusually high levels of uric acid in the blood has been linked to many diseases such as hyperuricemia, cardiovascular disease, and Lesch-Nyan disease.71 On the contrary, low levels of uric acid has been linked to diabetes mellitus, liver cirrhosis, and xanthinuria.72

9-Methyluric acid has been used in many research areas. 9-Methyluric acid has been noted to be less soluble in aqueous solution than uric acid. This decreases in solutions is a mode for forming urinary calculi. Various analytical techniques have been proposed for the study of these structures including HPLC with UV/VIS spectrophotometry.73 Uric acid derivatives have been known to assist in peroxynitrite radical elimination as endogenous scavengers. Heightened peroxynitrite concentrations have led to an increase in pathogenesis of multiple sclerosis and

11 experimental allergic encephalomyelitis.74 An immunological proposal has been made to increase the amount of uric acid and uric acid derivatives in a system to alleviate multiple sclerosis symptoms.75

Xanthine is the second to last product in human purine metabolism. The xanthine oxidase converts the carbon at the 8 position on the five membered ring from an sp2 hybridized carbon to a . Xanthine was originally discovered in 1902, where it was thought to be an reductase.76 The xanthine oxidase protein is the target site for many pharmaceutical agents that treat gout and hyperuricemia.77 Xanthine oxidase has a molybdenum center and this center is conventionally established to be the location of the binding of xanthine.78

Methylated purine derivatives have been proposed to play a large part in biological chemistry. 9-Methylxanthine is a naturally occurring purine that can be coordinated with platinum complexes to facilitate in cancer treatment and tumor identification.79 N-substituted purine derivatives were shown to have a substantial inhibitory effect on cyclin-dependent kinases which are invariably used in all phases of the cell division cycle.80

It became apparent during the literature search that purine and purine related compounds have a unique capability to tautomerize. A tautomer is a constitutional isomer in which the compound retains the same number of atoms, however the connectivity is different. Specifically for the N-heterocyclic systems studied, prototropic tautomerization was displayed not only on the

12 aromatic nitrogens, but the exo amines, mercapto group, and carbonyl groups. Furthermore, the formation of tautomers is not limited to the neutral species but extends to ionized forms. In conclusion if one knows the probable tautomeric forms of the neutral, cationic, and anionic species it is inherent to knowing the structure the compound takes in different environments, i.e. pKa.

1.4 References

1Henderson, J. F. The Position of the Glycosidic Bond in Purine Nucleosides: The Conservative Influence of a Convention of Chemical Nomenclature. Annals of Science, 1978, 35, 299-323. 2 Bender, J. A.; Yang, Z.; Eggers, B.; Gong, Y.; Lin, P.; Parker, D. D.; Rahematpura, S.; Zheng, M.; Meanwell, N. A.; Kadow, J. F. Inhibitors of HIV-1 Attachment. Part 11: The Discovery and Structure-Activity Relationships Associated with 4,6-Diazaindole Cores. Bioorg. Med. Chem. Lett., 2013, 23, 218-222. 3 Ingle, M.; Morris, J. C.; D’Souza, M. C. Soil Contamination Inconsistently Affects Performance of Diphenylamine as a Superficial Scald Inhibitor. Hort Science, 1990, 25, 1414- 1415. 4 Subedi, B.; Kannan, K. Fate of Artificial Sweeteners in Wastewater Treatment Plants in New York State, U. S. A. Environ. Sci. Technol., 2014, 48, 13668-13674. 5 Beck, J.; Ineman, J. M.; Rowan, S. J. Metal/Ligand-Induced Formation of Metallo- Supramolecular Polymers. Macromolecules, 2005, 38, 5060-5068. 6 Bruice, P. Y. Organic Chemistry, 5th ed.; Pearson Prentice Hall: Upper Saddle River, NJ, 2007. 7 Six, E.; Lagresle-Peyrou, C.; Susini, S.; De Chappedelaine, C.; Sigrist, N.; Chouteau, M.; Cagnard, N.; Fontenay, M.; Hermine, O.; Chomienne, C.; Reynier, P.; Fischer, A.; André- Schmutz, I.; Gueguen, N.; Cavazzana, M. AK2 Deficiency Compromises the Mitochrondrial Energy Metabolism Required for Differentiation of Human Neutrophil and Lymphoid Lineages. Cell Death and Diseases, 2015, 6, e1856. 8 Lodish, H.; Berk, A.; Kaiser, C.; Kreiger, M.; Bretscher, A.; Ploegh, H.; Matsudaira, P. Molecular Cell Biology, 6th ed.; W. H. Freeman and Company: NewYork, 2008. 9 Adler, T.; Albert, A. Diazaindenes (“Azaindoles”). Part 1. Ionization Constants and Spectra. J. Chem. Soc., 1960, 1794-1797.

13

10 Chatterji, M.; Shandil, R.; Manjunatha, M. R.; Solapure, S.; Ramachandran, V.; Kumar, N.; Saralaya, R.; Panduga, V.; Reddy, J.; Prabhakar, K. R.; Sharma, S.; Sadler, C.; Cooper, C. B.; Mdluli, K.; Iyer, P. S.; Narayanan, S.; Shirude, P. S. 1,4-Azaindole, a Potential Drug Candidate for Treatment of Tuberculosis. Antimicrobial Agents and Chemotherapy, 2014, 58, 5325-5331. 11 Bryan, M. C.; Falsey, J. R.; Frohn, M.; Reichelt, A.; Yao, G.; Bartberger, M. D.; Bailis, J. M.; Zalameda, L.; San Miguel, T.; Doherty, E. M.; Allen, J. G. N-substituted Azaindoles, as Potent Inhibitors of Cdc7 Kinase. Bioorg. Med. Chem. Lett., 2013, 23, 2056-2060. 12 Koyama, T.; Wakisaka, A. Molecular Self-Assembly Composed of Aromatic Hydrogen-Bond Donor-Acceptor Complexes. J Chem. Soc., Faraday Trans., 1997, 21, 3813-3817. 13 Preston, P. N. The Chemistry of Heterocyclic Compounds; Wiley and Sons: New York, 1981. 14 Kondo, T.; Yang, S.; Huh, K.-T.; Kobayashi, M.; Kotachi, S.; Watanabe, Y. Ruthenium Complex-Catalyzed Facile Synthesis of 2-Substituted Benzo-azoles. Chem. Lett., 1991, 7, 1275- 1278. 15 Schwarz, D. E.; Cameron, T. M.; Hay, P. J.; Scott, B. L. Tumas, W.; Thorn, D. L. Hydrogen Evolution from Organic “Hydrides”. Chem. Commum., 2005, 47, 5919-5921. 16 Reddy, P. N.; Padmaja, P. Applications of Aminocarbazoles in Heterocyclic Synthesis. ARKIVOC, 2015, 244-268. 17 Vazquez-Duhalt, R. In Petroleum Biotechnology Developments and Perspectives Delmon, B. Yates, J. T. Centi, G., Eds.; Studies in Surface Science and 151; Elsiver: San Diego, 2004. 18 Daly, J. W.; Holmén, J.; Fredholm, B. B. Is Caffeine Addictive? The most Wodely Used Psychoactive Substance in the World Affects the Same Parts of the Brain as Cocaine. Lakartidningen, 1998, 95, 5878-5883. 19 Fisone, G.; Borgkvist, A.; Usiello, A. Caffeine as a Psychomotor Stimulant: Mechanism of Action. Cellular and Molecular Life Sciences, 2014, 61, 857-872. 20 Martínez-López, S.; Sarriá, B.; Baeza, G.; Mateos, R.; Bravo-Clemente, L. Pharmacokinetics of Caffeine and its Metabolites in Plasma and Urine After Consuming a Green/Roasted Blend by Healthy Individuals. Food Research International, 2014, 64, 125-133. 21 Ross, G. W.; Abbott, R. D.; Petrovitch, H. Association of Coffee and Caffeine Intake With the Risk of Parkinson Disease. JAMA, 2000, 20, 2674-2679.

14

22 Leisso, R.; Buchanan, D.; Lee, J.; Mattheis, J.; Rudell, D. Cell Wall, Cell Membrane, and Volatile Metabolism are Altered by Antioxidant Treatment, Temperature Shifts, and Peel Necrosis During Apple Fruit Storage. J. Agric. Food Chem., 2013, 61, 1373-1387. 23 Berg, J. M.; Tymoczko, J. L.; Stryer, L. Biochemistry, 7th ed.; W. H. Freeman and Company, New York: 2012. 24 Wang, W.; Hellinga, H. W.; Beese, L. S. Structural Evidence for the Rate Tautomer Hypothesis of Spontaneous Mutagenesis. Proc Natl Acad Sci U S A, 2011, 43, 17644-17648. 25 Botta, J. R. Evaluation of Seafood Freshness Quality; VCH Publishing: New York, 1995. 26 Atta-ur-Rahman; Malik, S.; Cun-heng, H; Clardy, J. Tetrahedron Lett., 1995, 36, 2759-2762. 27 Inamoto, K.; Katsuno, M.; Yoshino, T.; Arai, Y.l Hiroya, K.; Sakamoto, T. Synthesis of 3- Substituted and Benzoisoxazoles via Pd-Catalyzed Cyclization Reactions: Application to the Synthesis of Nigellicine. Tetrahedron, 2007, 63, 2695-2711. 28 Shevyrin, V.; Melkozerov, V.; Nevero, A.; Etsov, O.; Baranovsky, A.; Shafran, Y. Synthetic Cannabinoids as Designer Drugs: New Representatives of Indol-3-Carboxylates as a Novel Group of Cannabinoids. Forensic Sci Intl, 2014, 244, 263-275. 29 Lee, J.-H.; Lee, J. Indole as an Intercellular Signal in Microbial Communities. FEMS Micribiol Rev, 2010, 34, 426-444. 30 Wykoff, W. R.; Anfora, A. T.; Liu, J.; Schultz, P. G.; Lesley, S. A.; Peters, E. C.; Siuzdak, G. Metabolomics Analysis Reveals Large Effects of Gut Microflora on Mammalian Blood Metabolites. PNAS, 2009, 10, 3698-3703. 31 Alata, I.; Bert, J.; Broquier, M.; Dedonder, C.; Feraud, G.; Grégoire, G.; Soorkia, S.; Marceca, E.; Jouvet, C. Electronic Spectra of the Protonated Indole Chromophore in the Gas Phase. J. Phys. Chem. A, 2013, 117, 4420-4427. 32 Barbour, P. M.; Podoll, J. D.; Marholz, L. J.; Wang, X. Discovery and Initial Structure- Activity Relationships of N-Benzyl Tricyclic Indolines as Antibacterials for Methicillin- Resistant Staphylococcus aureus. Bioorg. Med. Chem. Lett., 2014, 24, 5602-5605. 33 Caranzi, L.; Pace, G.; Guarnera, S.; Canesi, E. V.; Brambilla, L.; Raavi, S. S. K.; Petrozza, A.; Caironi, M. Photoactive Molecular Junctions Based on Self-Assembled Monolayers of Indoline Dyes. Appl. Mater. Interfaces, 2014, 6, 19774-19782. 34 Larent, A. Reserches sur L’indigo. Annales de Chimie et de Physique, 1840, 3, 393-434.

15

35 Erdmann, O. L. Untersuchungen über den Indigo. Journal für Praktische Chemie, 1840, 19, 321-362. 36 da Silva, J. F. M.; Garden, S. J.; Pinto, A. C. The Chemistry of Isatins: a Review from 1975- 1999. J. Braz. Chem. Soc., 2001, 3, 273-324. 37 Yan, Y.; Li, G.; Wang, F.; Mao, W. Huadong Huagong Xueyuan Xuebao, 1992, 18, 192. 38 Grafe, U.; Radics, L. J. Antibiotics, 1986, 29, 162. 39 Watson, J. D.; Crick, F. H. C. Nature, 1953, 171, 964. 40 Blas, J. S.; Luque, F. J.; Orozco, M. Unique Isomeric Properties of Isoguanine. J. Am. Chem. Soc., 2004, 126, 154-164. 41 Bonnett, R.; Brown, R. F. C. Isoindole. J. C. S. Chem. Comm., 1972, 7, 393-395. 42 Parrino, B.; Ciancimino, C.; Carbone, A.; Spanò, V.; Montalbano, A.; Barraja, P.; Cirrincione, G.; Diana, P. Synthesis of Isoindolo[1,4]benzoxazinone and Isoindolo[1,5]benzoxazepine: Two New Ring Systems of Pharmaceutical Interest. Tetrahedron, 2015, 71, 7332-7338. 43Barraja, P.; Spanò, V.; Patrizia, D.; Carbone, A.; Cirrincione, G.; Vedaldi, D.; Salvador, A.; Viola, G.; Dall’Acqua, F. Pyrano[2,3-e]insoindol-2-ones, New Angelicin Heteroanalogues. Bioorg. Med. Chem. Lett., 2009, 19, 1711-1714. 44 Spanò, V.; Montalbano, A.; Carbone, A.; Parrino, B.; Diana, P.; Cirrincione, G.; Barraja, P. Convenient Synthesis of Pyrrolo[3,4-g]indazole. Tetrahedron, 2013, 69, 9839-9847. 45 Alberto, M. E.; De Simone, B. C.; Mazzone, G.; Quartarolo, A. D.; Russo, N. Theoretical Determination of Electronic Spectra and Intersystem Spin-Orbit Coupling: The Case of Isoindole-BODIPY Dyes. J. Chem. Theory Comput., 2014, 10, 4006-4013. 46 Kim, B.; Yalaz, C.; Pan, D. Synthesis and Characterization of Membrane Stable Bis(arylimino)isoindole Dyes and Their Potential Application in Nano-Biotechnology. Tetrahedron Letters, 2012, 53, 4134-4137. 47 Sumpter, W. C. The Chemistry of Oxindole. Chem. Rev., 1945, 37, 443-479. 48 Ramart-Lucas, M.; Biquard, M. The Changes in Color Which Accompany the Transformation of Amino Acids and into Lactams in the Series. Bull. Soc. Chim., 1935, 2, 1383-1388. 49 Abrams, T.; Murray L.; Pryer, N.; Cherrington, J. M. Combination Administration of an Indolinone with a Chemotherapeutic Agent For Cell Proliferation Disorders. WO2004/045523 A2. June 3, 2004.

16

50Guerreiro, S.; Toulorge, D.; Hirsch, E.; Marien, M.; Sokoloff, P.; Michel, P. P. Paraxanthine, the Primary Metabolite of Caffeine, Provides Protection Against Dopaminergic Cell Death via Stimulation of Ryanodine Receptor Channels. Mol. Pharmacol., 2008, 4, 980-989. 51 Guadalupe, E.; Ramos, D.; Shelke, N. B.; James, R.; Gibney, C.; Kumbar, S. G. Bioactive Polymeric Nanofiber Matrices for Skin Regeneration. J. Appl. Polym. Sci., 2015, 132, 41879. 52 Kornberg, A. DNA Replication. J. Biol. Chem., 1998, 263, 1-4. 53 Bush, K. T.; Keller, S. T.; Nigam, S. K. Genesis and Reversla of the Ischemic Phenotype in Epithelial Cells. J. Clin. Invest., 2000, 106, 621-626. 54 Kato, A.; Fukushima, Y. Colorimetric Chemosensor for ATP Based on Phthalimide- Appended Poly(2,5-dimethoxyaniline). Polym. Bull., 2013,70, 3519-3527. 55 Fischer, E. Synthesis on the Purine Group. Berichte, 1899, 32, 435-504. 56 Eicher, T.; Hauptmann, S. The Chemistry of Heterocycles: Structures, Reactions, Synthesis, and Applications, 2nd Ed; Wiley-VCH: Weinheim, 2003. 57 Joule, J. A.; Mills, K. Heterocyclic Chemistry, 5th Ed; Wiley&Sons: West Sussex, UK, 2010. 58 Hitchings, G. H.; Elion, G. B.; Falco, E. A.; Russell, P. B.; VanderWerff, H. Studies on Analogs of Purines and . Annals of the New York Academy of Sciences, 1950, 1318- 1355. 59 Vang, S. I.; Schmiegelow, K.; Frandsen, T.; Rostoj, S. Nersting, J. Mercaptopurine Metabolite Levels are Predictors of Bone Marrow Toxicity Following High-Dose Methotrexate Therapy of Childhood Acute Lymphoblastic Leukaemia. Cancer Chemo. and Pharm., 2015, 75. 60Warner, D. J. Ira Remsen, Saccharin, and the Linear Model. Ambix, 2008, 1, 50-61. 61 Pinney, V. R. Non-Lethal Weaponry Systems. US 6242489, May 14, 2002. 62 Kim, T. H.; Lee, S. M.; Kim, Y-S.; Kim, K. H.; Oh, S.; Lee, H. J. Aroma Dilution Method Using GC Injector Split Ratio for Volatile Compounds Extracted by Headspace Solid Phase Microextraction. Food Chem., 2003, 83, 151-158. 63 Mallia, S.; Escher, F.; Schlichtherle-Cerny, H. Aroma-Active Compounds of Butter: a Review. Eur. Food Res. Technol., 2008, 226, 315-325. 64 Kobori, K.; Maruta, Y.; Mineo, S.; Shigematsu, T.; Hirayama, M. Polyphenol-Retaining Decaffeinated Cocoa Powder Obtained by Supercritical Carbon Dioxide Extraction and its Antioxidant Activity. Foods, 2013, 2, 462-477.

17

65 Panou, A. I.; Papadokostaki, K. G.; Sanopoulou, M. Release Mechanisms of Semipolar Solutes from Poly(dimethylsiloxane) Elastomers: Effect of a Hydrophilic Additive. J. Appl. Polym. Sci., 2014, 131, 40782. 66 Barnes, P. J.; Pauwels, R. Theophylline in the Management of Asthma: Time for Reappraisal. Rut. Trdpit. J., 1994, 7, 579-591. 67 Barnes, P. J. Theophylline: New Perspective for Old Drug. Am. J. Respir. Crit. Care Med., 2003, 167, 813-818. 68 Parmar, G.; Jain, S. Study of Theophylline Against Experimentally Induced Hyperlipidemia. J. of Pharmaceutical and Scientific Innovation, 2014, 5, 474-477. 69 Maxwell, S. R.; Thomason, H.; Sandler, D. Antioxidant Status in Patients with Uncomplicated Insulin-Dependent and Non-Insulin-Dependent Diabetes Mellitus. Eur. J. Clin. Invest., 1997, 27, 484-490. 70 Baillie, J. K.; Bates, M. G. D.; Waring, W. S.; Partridge, R. W.; Simpson, A.; Maxwell, S. R. J. Endogenous Urate Production Augments Plasma Antioxidant Capacity in Healthy Lowland Subjects Exposed to High Altitude. Chest, 2007, 131, 1473-1478. 71 Huang, S-H.; Shih, Y-C.; Wu, C-Y.; Yuan, C-J.; Yang, Y-S.; Li, Y-K.; Wu, T-K. Detection of Serum Uric Acid Using the Optical Polymeric Enzyme Biochip System. Biosensors and Bioelectronics, 2004, 19, 1627-1633. 72 Fathallah-Shaykh, S. A.; Cramer, M. T. Uric Acid and the Kidney. Pediatr. Nephrol., 2014, 29, 999-1008. 73 Safranow, K.; Machoy, Z. Simultaneous Determination of 16 Purine Derivatives in Urinary Calculi by Gradient Reversed-Phase High-Performance Liquid Chromatography with UV Detection. J. Chromatogr. B, 2005, 819, 229-235. 74 Kean, R. B.; Spitsin, S. V.; Mikheeva, T.; Scott, G. S.; Hooper, D. C. The Peroxynitrite Scavenger Uric Acid Prevents Inflammatory Cell Invasion into the Central Nervous System in Experimental Allergic Encephalomyelitis through Maintenance of Blood-Central Nervous System Barrier Integrity. The J. of Immunology, 2000, 165, 6511-6518. 75 Scott, G. S.; Spitsin, S. V.; Kean, R. B.; Mikheeva, T.; Koprowski, H.; Hooper, D. C. Therapeutic Intervention in Experimental Allergic Encephalomyelitis by Administration of Uric Acid Precursors. Proc. Natl. Acad. Sci. U S A, 2002, 99, 16303-16308.

18

76 Schardinger, F. Detection of Heated Milk by Means of Methylene Blue. Untersuch. Nahrungs. Genussmittel, 1902, 5, 1113-1121. 77 Elion, G. B. The Purine Path to Chemotherapy. Science, 1989, 244, 41-47. 78 Nishino, T.; Okamoto, K.; Eger, B. T.; Pai, E. F.; Nishino, T. Mammalian Xanthine Oxidoreductase-Mechanism of Transition from Xanthine Dehydrogenase to Xanthine Oxidase. FEBS Journal, 2008, 275, 3278-3289. 79 Roat, R. M.; Jerardi, M. J.; Kopay, C. B.; Heath, D. C.; Clark, J. A.; DeMars, J. A.; Weaver, J. M.; Bezemer, E.; Reedijk, J. Platinum(II) Complexes Catalyze Reactions Between Platinum(IV) Complexes and 9-Methylxanthine. J. Chem. Soc., Dalton Trans., 1997, 19, 3615-3621. 80 Veselý, J.; Havliček, L.; Strnad, M.; Blow, J. J.; Donella-Deana, A.; Pinna, L. Inhibition of Cyclin-Dependent Kinases by Purine Analogues.. Eur. J. Biochem., 1994, 224, 771-786.

19

CHAPTER TWO

Computational Analysis of Cationic, Neutral, and Anionic Purine Tautomers

2.1 Abstract:

The tautomers for neutral, cationic, and anionic purine, 6-mercaptopurine, adenine, guanine, hypoxanthine, and xanthine were examined using quantum chemical B3LYP/6-31+G** calculations both in vacuum and with the SM8 solvent model to estimate the most stable forms present. Keto-enol, amine-imine, and thione-thiol tautomerisms were investigated in the neutral forms. The corresponding cationic and anionic species were examined.

2.2 Introduction:

The tautomeric states of purines and related compounds have been discussed extensively for decades.1-9 In most of the tautomeric studies an important feature was the nature of not only the neutral species, but also of the protonated and deprotonated forms of the compounds.10 The importance of understanding tautomer behavior is apparent in DNA structure, since when a higher energy, less stable tautomer bonds to a complementary base the resulting mismatch of the base pairs can lead to mutagenesis.

The goal of the present report is to examine the cationic, anionic, and neutral tautomers of six purines and related compounds—purine, hypoxanthine, 6-mercaptopurine, xanthine, adenine, and guanine—in order to establish the groundwork for a later theoretical pKa study.

While 6-mercaptopurine is not a naturally occurring purine derivative it is used commonly in the treatment of childhood leukemia,11 and thus is an important medicinal compound. Although

20 many groups have examined specific tautomers of purines, we are not aware of any comprehensive, detailed study being presented.

A secondary goal of this report was to understand the relative populations of the different tautomers of these compounds in aqueous solution. There are two main factors that affect the populations: the intrinsic stability displayed in the vacuum energy of each tautomer and the adjustments to these energies resulting from the compound’s placement within a solvent, i.e. the solvation energies. The solvation energy is determined by the interaction of the compound with the solvent via the prevailing intermolecular forces. These forces consist of hydrogen bonding,

London dispersion forces, electrostatic forces, etc. The solvation energies are important to understand because they strongly affect tautomer stabilities in different media. One expects greater solvation energies in water to occur for more polar species. An illustration of this can be seen in Figure 2.1

Figure 2.1 An illustration of the intrinsic energies and the adjustment to the energies as a result from being placed in a solvent model.

We note that the global dipole moment of a compound can be viewed as a gross measure of the compound’s polarity, but a more relevant measure for aqueous solvation would involve the microscale point-to-point charge heterogeneity. It should also be noted that the solvation

21 energies are generally larger in magnitude for compounds with more electronegative atoms, and even more so for species with ionic charges.

Tautomers of a compound are constitutional such that the individual compounds consist of the same atoms; but with shifts in the hydrogen positions. The tautomerism of the compounds mentioned is of great importance because it is expected that the more stable tautomeric forms will play the greatest roles in physiological systems. The Boltzmann distribution equation was utilized to calculate the population ratios one would expect to find for the various tautomers.

2.3 Computational Methods:

Theoretical calculations were performed on each compound using the Spartan’10 computational package (Wavefunction, Inc., Irvine, CA). Initial geometry optimizations were performed using the semi-empirical RM1 method.12 This was followed by optimization using density functional theory with the B3LYP functional13,14 and the 6-31+G** basis set. Each compound was optimized first under vacuum conditions and then re-optimized with the SM815 water solvent model. All reasonably plausible tautomers were examined using the aforementioned calculations. Once the most stable neutral tautomer was established, related cationic tautomers were constructed by appropriate additions of protons. The neutral form was also employed to form anionic tautomers by removing protons from relevant positions.

As a reminder the purine numbering system is shown below:

22

2.4 Results and Discussion:

2.4.1 Purine:

Purine is an ubiquitous compound in chemistry. Experimentalists have examined this compound for its acid-base behavior because it is the main skeleton for many compounds found in biological systems, polymer science, medicinal chemistry, and environmental chemistry.

Experimentally, the two tautomeric forms that are of most concern for purine are the 7H-P and

9H-P isomers. The naming chosen for the compounds indicates the atoms that hold the protons.

Many authors agree that, in solution, purine will exist in roughly equal concentrations of these two tautomers. However, in a crystalline form the 7H-P tautomer is the only one present.5,6

Figure 2.2. Four possible neutral tautomers of purine with energy differences calculated in vacuum.

In contrast to the experimental observations in solution, the present B3LYP/6-

31+G**/SM8 calculations yielded the tautomer distributions as shown in Table 2.1 and Figure

2.2. The contrast is that the calculations indicate a greater difference in stability of the 9H-P and

7H-P tautomers, both in vacuum and in solution. The vacuum calculations place the 9H-P tautomer roughly 16 kJ mol-1 below the 7H-P tautomer.

23

In contrast, the aqueous solution (SM8) results place the 7H-P tautomer roughly 16 kJ mol-1 below the 9H-P tautomer as a result of its greater solvation energy. The vacuum calculations above coincide with a more uniform electron density map as shown in Figure 2.3.

The larger solvation energy of the 7H-P tautomer comes as a result of the more heterogeneous charge distribution in this tautomer, which leads it to interact more strongly with the aqueous solvent.

Figure 2.3 Molecular electrostatic potential maps of the 9H-P and 7H-P tautomers of purine.

The present calculations for purine were optimized using density functional theory with a diffuse function placed on the heavy atoms only and polarization functions on both light and heavy atoms. Earlier theoretical calculations7 were performed on purine at lower computational levels. Nowak and group used Hartree-Fock calculations, HF/6-31(d,p), where the vacuum ab initio calculations favored the 9H-P tautomer by approximately 16 kJ mol-1 over the next stable

7H-P tautomer. It was argued that in solution the 7H-P should be more stable. The present calculations reach the same conclusion as shown in Table 2.1.

24

Table 2.1. Relative energies and solvation energies of the neutral, cationic, and anionic tautomers for purine.

Purine – Neutral Tautomers

-1 -1 -1 Compound ΔEVAC (kJ mol ) ESOLV. (kJ mol ) ΔEH2O (kJ mol )

9H-P 0.00 -48.7 16.6

7H-P 16.2 -82.1 0.00

3H-P 41.2 -64.5 42.2

1H-P 54.9 -92.3 28.1

Purine Cation Tautomers

-1 -1 -1 Compound Vacuum ΔEVAC (kJ mol ) ESOL. (kJ mol ) Compound Water ΔEH2O (kJ mol )

1,9(H)-P+ 0.00 -271. 1,9(H)-P+ 0.00

3,7(H)-P+ 13.3 -269. 1,7(H)-P+ 0.48

1,7(H)-P+ 21.0 -292. 7,9(H)-P+ 3.47

7,9(H)-P+ 28.0 -296. 3,7(H)-P+ 16.2

3,9(H)-P+ 46.5 -282. 3,9(H)-P+ 35.3

+ + 9(H2)-P 208. -317. 7(H2)-P 156.

+ + 7(H2)-P 227. -342. 9(H2)-P 164.

Purine Anion*

P- 0.00 ---- P- 0.00

As shown above, depending on the environment, neutral purine should exist predominately as one stable tautomer, whereas each neutral tautomer will have its own cationic tautomer. The results are displayed in Table 2.1. It should be noted the most stable cationic tautomer for both neutral species places the protonation site on the pyrimidine ring (1,7(H)-P+ and 1,9(H)-P+). As a result of tautomeric equilibrium, the 1,9(H)-P+ form should prevail in vacuum and in aqueous solution.

25

Figure 2.4. Cationic tautomers with decreasing stabilizations as found in vacuum.

Based on the relative energy differences, a major conclusion can be made. It should be recognized that protonated are more energetically stable, by far, than protonated secondary amines. Whereas in both cases the nitrogens have two lone electrons and are sp2 hybridized the orientation of the electrons are crucial to cationic stabilities. The imine holds the lone electrons in an sp2 orbital. These electrons do not participate in the resonance and aromaticity of the neutral species. The amine holds the lone pair electrons perpendicular to the compound in a 2pz orbital. Conversely, the electrons in this orbital can participate in resonance and aromaticity of the neutral species. Interruption of any electrons involved in a conjugated

+ system will result in a much less stable ionized species. This is why the relative energy of 9(H2) is so great.

The anionic purine tautomer relative energies are displayed in Table 2.1. It was previously mentioned that one stable neutral tautomer will exist depending on the environment;

26 likewise each environment will have its own anionic tautomers. The following figure shows the dissociation pattern for the stable neutralanion purine tautomer in vacuum.

Figure 2.5. The deprotonation of the favored tautomer (9H-P) in vacuum to its anionic form.

2.4.2 Hypoxanthine:

Hypoxanthine is the third to final metabolic intermediate in purine metabolism for humans and higher primates. This nucleobase can also be formed from the deamination of adenine8 As shown in Figure 2.6, hypoxanthine can be represented by eight tautomers, with four keto and four enol forms. The solvent energies for the eight different tautomers are listed in

Table 2.2.

In both vacuum and the aqueous solvent, there was only a small energy difference of 1.30 kJ mol-1 between 1,9(H)-HX and 1,7(H)-HX, with 1,9(H)-HX found to be more stable in solution. This indicates that both low energy tautomers should be present as a mixture. The small energy difference between the stable tautomers is lower than RT (~2.48 kJ mol-1). Also supporting this conclusion are various NMR studies in different solvents at varying temperatures that showed a nearly 1:1 ratio of the two tautomers.16,17 However, the tautomeric populations in aqueous solutions still remain undetermined.16,18,19 As expected from previous studies,5,6,11 the enol tautomers were found, in general, to lie higher in energy than the keto tautomers.

27

Table 2.2. Relative energies and solvation energies of the neutral, cationic, and anionic tautomers for hypoxanthine.

Hypoxanthine -Neutral

-1 -1 -1 Compound ΔEVAC (kJ mol ) ESOLV. (kJ mol ) ΔEH2O (kJ mol )

1,7(H)-HX 0.00 -77.7 1.30

1,9(H)-HX 3.42 -82.3 0.00

3,7(H)-HX 31.7 -101. 10.1

3,9(H)-HX 88.5 -135. 32.0

Hypoxanthine Cations

1,7,9(H)-HX+ 0.00 -302. 0.00

1,3,9(H)-HX+ 65.0 -332. 35.0

1,1,9(H)-HX+ 194. -311. 185.

1,9,9(H)-HX+ 207. -340. 169.

Hypoxanthine Anions

1H-HX- 0.00 -306. 0.00

9H-HX- 27.3 -331. 35.0

28

Figure 2.6. Hypoxanthine tautomers and their relative energies in vacuum.

The relative energies for the cationic tautomers of hypoxanthine are listed in Table 2.2.

Since the neutral form of hypoxanthine exists as two statistically probable tautomers in solution, it is assumed each of these would be capable of ionizing into a cationic form. However, both of the two dominant neutral tautomers lead to the same cationic form. The cationic tautomers as found in vacuum are depicted in Figure 2.7.

Figure 2.7. Cationic tautomers resulting from the 1,9(H)-HX neutral tautomer in vacuum.

29

Gogia et al.9 viewed the ionization sites of hypoxanthine using surface enhanced raman (SERS) and theoretical calculations to correlate the experimental findings.

Accordingly, hypoxanthine was found to exist as one major tautomeric form, where the sp2 hybridized imine in the ring holds the positive charge from the cation. The experimental results were the same as those calculated in Table 2.2.

The anionic tautomers for hypoxanthine exist in two probable forms. The relative energies are displayed in Table 2.2. As before, since there are two stable neutral tautomers in solution one should expect two find four unique anionic tautomers. This is due to the low energies of all solvent-stabilized anionic tautomers. Experimentally, through the use of NMR spectroscopy, more than one tautomer has been found in solution.16,20 It should be noted that this is not expected in vacuum due to the calculated relative energy difference being 27.3 kJ mol-1.

Figure 2.8. Relative energies for deprotonation of the two neutral hypoxanthine tautomers, a)

1,9-hypoxanthine and b) 1,7-hypoxanthine, to their anionic forms as found in a water solvent.

30

2.4.3 6-Mercaptopurine:

An important purine derivative, 6-mercaptopurine (MP) is not found naturally in biological systems, but has been utilized as a chemotherapy agent since the mid 20th century. Its structure is similar to hypoxanthine, but with the replaced by a thione group.

Table 2.3. Relative energies and solvation energies of the neutral, cationic, and anionic tautomers for 6-mercaptopurine.

6-Mercaptopurine -Neutral

-1 -1 -1 Compound ΔEVAC (kJ mol ) ESOLV. (kJ mol ) ΔEH2O (kJ mol )

1,7(H)-MP 0.00 -71.3 0.00

1,9(H)-MP 13.1 -75.3 9.15

3,7(H)-MP 36.4 -41.9 26.0

3,9(H)-MP 103. -118. 57.0

6-Mercaptopurine Cations

1,7,9(H)-MP+ 0.00 -291. 0.00

1,3,7(H)-MP+ 16.6 -295. 13.3

1,7,7(H)-MP+ 200. -326. 165.

1,1,7(H)-MP+ 204. -316. 180.

6-Mercaptopurine Anions

7H-MP- 0.00 -280. 2.74

1H-MP 1.92 -285. 0.00

In analogy to the keto-enol equilibrium, the thione-thiol equilibrium for the neutral molecule appears to favor the thione forms. As seen in Table 2.3 and Figure 2.9, 1,7(H)-MP is theoretically determined to be the most stable. The tautomerism of neutral MP has also been examined through NMR spectroscopy, which showed that the same neutral tautomer was

31 favored.16,17 This was surprising because of the interruption of resonance from the thione functionality.

Figure 2.9. Thione-thiol tautomers of 6-mercaptopurine as calculated in vacuum.

The extremely high energy calculated for the 3,9(H)-MP tautomer in both vacuum and solution warrants some additional discussion. In contrast to the thiol tautomers one might expect this thione tautomer to have a lower energy, but it does not. The high energy of this tautomer has several origins. First, the two close-lying NH groups on the meta secondary amines on the lower side in 3,9(H)-MP increase repulsive electrostatic interactions in this compound. This is reflected in the longer NH bond lengths in 3,9(H)-MP as shown in Figure 2.10. Second, additional repulsive interactions occur between the two top imine nitrogen lone pair electrons and the lone pairs on the sulfur atom.

32

Figure 2.10. Illustrations showing bond lengths for the favored thione tautomer 1,7(H)-MP and the high-energy 3,9(H)-MP as found using the water solvent model.

The relative cationic tautomer energies for 6-mercaptopurine are listed in Table 2.3.

Protonated imines are again shown to be more stable than the corresponding protonated amines.

This observation holds true for all nitrogen acids shown here. Between the two possible imine sites to protonate the imidizole ring is more probable because the ring does not lose its aromaticity upon the protonation. While these compounds do not have resonance through both rings, the most stable cationic tautomer can be explained by the imidazole ring not losing resonance upon protonation. The two unfavored cationic tautomers can be explained through basicity. 1,7,7(H)-MP+ would result in a non-aromatic imidazole ring while 1,1,7(H)-MP+ is less basic because of resonance with the thione group.

33

Figure 2.11. Cationic tautomers from the 1,7(H)-MP neutral tautomer as found in vacuum.

The relative energies for the anionic tautomers of 6-mercaptopurine can be seen in Table

2.3. The compound possesses two different dissociations sites leading to the formation of two anionic tautomers. Both of the anionic tautomers are predicted to exist in appreciable amounts because of the small energy difference calculated between them in both vacuum and solution.

Figure 2.12. Anionic tautomers for 6-mercaptopurine as found in vacuum.

There have, in the past, been conflicting experimental results regarding the evidence for stable anionic tautomers. Lichtenberg19 and colleagues used NMR to reveal NH(1) to be more acidic than NH(7), allowing for this to be the predominant anionic form. On the other hand,

Pazderski21 and colleagues used crystalline mercaptopurine in alkaline solution to evaluate the

34 anionic tautomers using 15N NMR spectroscopy. They observed the more prominent anionic tautomer to be 7H-MP-. It is our conclusion that the solvent used in the NMR studies affects the tautomeric populations, especially true when comparing polar protic and aprotic solvents.

2.4.4 Xanthine:

Xanthine exists as a neutral diketo-purine derivative that participates in many biochemical processes. It is also formed through the deamination of guanine. As previously discussed, enolic tautomers are not normally stable species; they were therefore not included in subsequent energy calculations. Figure 2.13 illustrates the two lowest energy keto tautomers for xanthine in vacuum. It can be seen that even with the two lowest lying tautomers, only one would be found in solution or in vacuum. The stability for 1,3,7(H)-X over 1,3,9(H)-X lies in electronic factors. The NH(9) hydrogen would interact repulsively with the NH(3) hydrogen.

However, the 1,3,7(H)-X tautomer would be electronically stabilized by the lone pairs on the nearby oxygen and the proton on the 7-position.

Figure 2.13. The two lowest-energy keto tautomers of xanthine calculated in vacuum.

35

Table 2.4. Relative energies and solvation energies of the neutral, cationic, and anionic tautomers for xanthine.

Xanthine -Neutral

-1 -1 -1 Compound ΔEVAC (kJ mol ) ESOLV. (kJ mol ) ΔEH2O (kJ mol )

1,3,7(H)-X 0.00 -75.2 0.00

1,3,9(H)-X 39.0 -94.2 20.0

Xanthine Cations

1,3,7,9(H)-X+ 0.00 -337. 0.00

1,1,3,7(H)-X+ 109. -294. 152.

1,3,3,7(H)-X+ 114. -308. 143.

1,3,7,7(H)-X+ 170. -360. 147.

Xanthine Anions

1,3(H)-X- 0.00 -273. 0.00

1,7(H)-X- 40.8 -313. 1.03

3,7(H)-X- 69.9 -338. 5.70

The cationic tautomers for xanthine exist in four probable forms, where the most probable protonation site exists on the imidazole ring. The relative energies are listed in Table

2.4.

36

Figure 2.14. Cationic tautomers for xanthine from the vacuum calculations.

Examination of the cationic tautomers for xanthine exposes just how stabilizing electrons that participate in aromaticity can be. There is only one sp2 hybridized imine and it can be reasoned one plausible cationic tautomer will exist where the imine becomes protonated. The vacuum and solvent model calculations reveal the same trend.

Figure 2.15. Anionic tautomers for xanthine found in vacuum.

There are three probable forms for anionic xanthine. The relative energies are listed in

Table 2.4. In vacuum and aqueous solution the most stable anionic tautomer has the dissociation site on the imidazole ring. However, based on experimental results from several groups the actual dissociation site on xanthine remains a mystery. Based on the two electron-withdrawing carbonyl groups one would assume the NH(1) position to dissociate more readily. Cavalieri et

37 al.22 initially studied the tautomers of anionic xanthine using ultraviolet spectroscopy in 1953 to place the dissociation at the NH(3) position. Correlating to that work, Pfeiderer23 et al used the same instrumentation and found the same dissociation product in 1961. Contrary to this,

Bergmann and colleagues used NMR and UV-Vis in 1970 to find similar results to ours. Lastly, in 2009 Gogia et al.9 used ultraviolet resonance Raman to uncover the same trend as Cavalieri and Pfeiderer. Based on this evidence, and the relative energies in solution listed in Table 2.4, there is a distinct possibility that all of these tautomers exist in solution. The intrinsic properties of the compound are theoretically explained through the vacuum energies, where we propose the

1,3(H)-X- tautomer to be the more stable product. As mentioned before, the natural charges on the nitrogen atoms can yield insight to the stabilization of tautomers. These are highlighted in

Figure 2.16 where the values in red are the natural charges on the nitrogen atoms. Since the

NH(7) position has the more positive charge it can be assumed that the bond strength (nitrogen- hydrogen) will be weaker.

Figure 2.16. Natural charges for the stable neutral tautomer in solution with the possible dissociation sites highlighted in red.

38

2.4.5 Adenine:

Adenine and guanine are the two purine nucleobases incorporated into DNA and RNA.

Watson, Crick, and others proposed that if the nucleobases did not exist in their most energetically favorable forms the genomic alphabet would drastically expand.24 Adenine has the potential for prototropic shifts to transform its exo primary amine into an imine. The six tautomeric arrangements for adenine are shown in Figure 2.17.

Figure 2.17. Tautomer energies of adenine determined in vacuum.

It is clear that the imine tautomers fall higher in energy than the amine tautomers and therefore are less likely to be important. While the existence of the higher energy tautomers is theoretically possible the likelihood for these tautomers to appear in DNA has been debated.

There has yet to be one dominant conclusion on the importance of the higher-energy tautomeric

39 forms. The two contending arguments for their formation lie between a proton shift via quantum tunneling or a classical barrier hop.25 On the rare occasions that the less favored tautomer might exist it would lead to a spontaneous point and a coding error. One can also hypothesize that a person’s environment might impact the likelihood of having a mutation occur via either route. Future research could therefore include a focus on the environmental route in which the mutagenic tautomers occur as opposed to a mechanistic route. This may yield greater insight to medical and clinical implications.

Table 2.5. Relative energies and solvation energies of the neutral, cationic, and anionic tautomers for adenine.

Adenine –Neutral

-1 -1 -1 Compound ΔEVAC (kJ mol ) ESOLV. (kJ mol ) ΔEH2O (kJ mol )

9H-A 0.00 -63.0 0.00

3H-A 34.1 -80.6 16.5

7H-A 37.2 -85.5 14.7

1H-A 80.2 -115. 28.0

Adenine Cations

1,9(H)-A+ 0.00 -267. 0.00

7,9(H)-A+ 6.36 -297. 4.06

3,9(H)-A+ 34.7 -268. 9.27

+ 9H2-A 192. -338. 120.

Adenine Anions

A- 0.00 -295. 0.00

10H-A- 79.1 -310. 64.5

40

It is essential to understand the ionized tautomers of biologically relevant compounds because adenine forms hydrogen bonds to thymine via the N1 position and the primary amine. If this position is occupied, either by being exposed to an acidic condition (pH< 3) or through oxidation, bonding mismatch is likely to cause .

Figure 2.18. Cationic tautomers with decreasing stabilities found in vacuum.

Adenine has the capability to form two anionic tautomers and the relative energies for these are listed in Table 2.5. The predominant form results from the dissociation of the NH(9) proton. A rare anionic tautomer could result in the dissociation from the exo amine.

Figure 2.19. Two anionic tautomers for adenine with relative energies found in vacuum.

Quantum chemical descriptors are used to explain the theoretical findings. The natural charges on the 9H and exo amine are -0.581 and -0.826, respectively. It should be noted the

41 charge on an atom is not an actual property that can be determined, because electronic contributions cannot be treated as point charges, unlike a nucleus. It has been found in the past that the natural charge is a valid descriptor for the compounds examined. The bond lengths for the nitrogen-hydrogen bonds for the 9H and exo amine are 1.023 and 1.015, respectively as well.

In conclusion, the lower negative charge and shorter bond length for the exo primary amine makes dissociation at that position unlikely.

2.4.6 Guanine:

Guanine is interesting when examining possible tautomeric arrangements. Guanine exists in its neutral form as an amine-keto tautomer. As described for adenine the amine-imine tautomer equilibrium generally finds the primary amine form to be more stable, in the same way as keto-enol tautomerization favors the keto form. Since guanine has both moieties, one might anticipate that the energetically favorable form would be an amine-keto tautomer. There should be four neutral tautomers for guanine, as depicted in Figure 2.20.

Figure 2.20. Neutral keto-amine tautomers of guanine as determined in vacuum.

Guanine is one of the unique compounds where the most stable tautomer is not the same in vacuum and in solution; the relative energies are listed in Table 2.6. The results suggest that a mixture of 1,9(H)-G and 1,7(H)-G would exist in both vacuum and aqueous solution, with the

42 predominant form depending on the state. That is to say 1,7(H)-G will be the predominant form in vacuum while 1,9(H)-G is predominant in solution.

Table 2.6. Relative energies and solvation energies of the neutral, cationic, and anionic tautomers for guanine

Guanine – Neutral Tautomers

-1 -1 -1 Compound ΔEVAC (kJ mol ) ESOLV. (kJ mol ) ΔEH2O (kJ mol )

1,7(H)-G 0.00 -96.5. 4.05

1,9(H)-G 2.34 -103. 0.00

3,7(H)-G 29.6 -122. 8.20

3,9(H)-G 84.2 -157. 28.1

Guanine Cation Tautomers

-1 -1 -1 Compound ΔEVAC (kJ mol ) ESOLV (kJ mol ) Compound Water ΔEH2O (kJ mol ) Vacuum

1,7,9(H)-G+ 0.00 -299. 1,7,9(H)-G+ 0.00

1,3,7(H)-G+ 22.1 -305. 1,3,7(H)-G+ 16.1

1,3,9(H)-G+ 75.4 -337. 1,3,9(H)-G+ 37.4

1,1,9(H)-G+ 169. ---- 1,7,7(H)-G+ 156.

1,1,9(H)-G+ 189. -332. 1,9,9(H)-G+ 158.

1,1,7(H)-G+ 194. -319. 1,1,7(H)-G+ 174

1,9,9(H)-G+ 198. -339. 1,1,9(H)-G+ ----

Guanine Anion Tautomers

7H-G- 0.00 -336. 9H-G- 0.00

1H-G- 2.18 -335. 7H-G- 2.03

9H-G- 12.2 -351. 1H-G- 6.11

Guanine is the most ionizable of the nucleobases.26 As stated above the cationic tautomer stabilities are important to understand because of the hydrogen bonding arrangements the purine

43 nucleobases form with pyrimidines. Upon protonation, the stable cationic tautomers in aqueous solution and vacuum are resonance forms of each other. This implies that regardless of the medium for the protonated form, only one probable form will exist.

Figure 2.21. Cationic tautomers of guanine found in vacuum.

It can be seen that the tendency for the proton to attack the imine on the imidazolic ring takes precedence. As shown in Figure 2.20 there is an electron-donating amino group ortho to an sp2 hybridized imine, the electron withdrawing carbonyl group is highly effective at decreasing the electron density on the imine nitrogen atom. The end result is protonation on the imine nitrogen atom of the imidazole ring.

Guanine can form three anionic tautomers of interest for the current research and their relative energies are listed in Table 2.6. In solution both neutral tautomers contribute to the anionic tautomer populations. Conversely, in vacuum, any anionic tautomer should be formed from the predominant neutral tautomer.

44

Figure 2.22. Three anionic tautomers that should be probable in solution for guanine as found in vacuum.

In vacuum and solution the dissociation site takes place on the six-member ring. This can be rationalized due to the electron withdrawing carbonyl substituent and electron donating amine. The amine pushed electronic density through the pi system to the N(3) position. This causes the NH(1) site to be more electropositive. Also, the charge build up around the N(1) position site can be distributed through the different donor/acceptor groups and peptide bond resonance. If the charge was placed on the imidazolic ring it would not be distributed as readily.

2.5 Conclusion:

The six compounds studied indicate the necessity to fully understand the tautomer distributions and how these distributions might affect the experimental biochemical processes.

Once the tautomer distributions are fully understood an understanding of how the pKa’s vary may become clearer.

2.6 References

1 Watson, J. D.; Crick, F. H. C. Genetic Implications of the Structure of Deoxyribonucleic Acid. Nature, 1953, 171, 964-967. 2 Woolley, E. M.; Wilton, R. W.; Hepler, L. G. Thermodynamics of Proton Dissociations from Imidazolium , 6-Uracilcarboxylic Acid, and Protonated Hypoxanthine in Aqueous Solution. Canadian Journal of Chemistry, 1970, 48, 3249-3252.

45

3 Löwdin, P.-O. Proton Tunneling in DNA and its Biological Implications. Rev. Mod. Phys., 1963, 35, 724-732. 4 Kumar, A.; Sevilla, M. D. Proton-Coupled Electron Transfer in DNA on Formation of Radiation –Produced Ion Radicals. Chem Rev., 2010, 110, 7002-7023. 5 Eicher, T.; Hauptmann, S. The Chemistry of Heterocycles: Structures, Reactions, Synthesis, and Applications, 2nd Ed; Wiley-VCH: Weinheim, 2003. 6 Joule, J. A.; Mills, K. Heterocyclic Chemistry, 5th Ed; Wiley&Sons: West Sussex, UK, 2010. 7 Nowak, M. J.; Rostkowska, H.; Lapinski, L. Tautomerism N(9)H N(7)H of Purine, Adenine, and 2-Chloroadenine: Combined Experimental IR Matrix Isolation and ab Initio Quantum Mechanical Studies. J. Phys. Chem., 1994, 98, 2813-2816. 8 Nguyen, T.; Brunson, D.; Crespi, C. L.; Penman, B. W.; Wishnok, J. S.; Tannenbaum, S. R. DNA Damage and Mutation in Human Cells Exposed to Nitric Oxide in Vitro. Proc. Natl. Acad. Sci. USA, 1992, 89, 3030-3034. 9 Gorgia, S.; Jain, A.l Puranik, M. Structures, Ionization Equilibria, and Tautomerism of 6- Oxopurines in Solution. J. Phys. Chem. B, 2009, 113, 15101-15118. 10 Costas, M. E.; Acevedo-Chávez, R. Density Functional Study of the Monocationic Hypoxanthine Tautomers. J. Molec. Structure (Theochem), 1999, 489, 73-85. 11 Hitchings, G. H.; Elion, G. B.; Falco, E. A.; Russell, P. B.; VanderWerff, H. Studies on Analogs of Purines and Pyrimidines. Annals of the New York Academy of Sciences, 1951, 1318- 1355. 12 Rocha, G. B.; Freire, R. O.; Simas, A.; Stewart, J. J. P. J. Comput. Chem. 2006, 27, 1101. 13 Becke, A. D. J. Chem. Phys., 1993, 98, 5648-5652. 14 Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B, 1998, 37, 785-789. 15 Marenich, A. V.; Olson, R. M.; Kelly, C. P.; Cramer, C. J.; Truhlar, D. G. J. Chem. Theory Comput., 2007, 3, 2011. 16 Bartl, T.; Zacharová, Z.; Sečkářová, P.; Kolehmainen, E.; Marek, R. NMR Quantification of Tautomeric Populations in Biogenic Purine Bases. Eur. J. Org. Chem., 2009, 2009, 1377-1383. 17 Chenon, M. T.; Pugmire, R. J.; Grant, D. M.; Panzica, R. P.; Townsend, L. B. Carbon-13 Magnetic Resonance. XXVI. A Quantitative Determination of the Tautomeric Populations of Certain Purines. J. Am. Chem. Soc., 1975, 97, 4636-4642. 18 Sun, X. J.; Lee, J. K. Acidity and Proton Affinity of Hypoxanthine in the Gas Phase Versus in Solution: Intrinsic Reactivity and Biological Implications. J. Org. Chem., 2007, 72, 6548-6555. 19 Lichtenberg, D.; Neiman, Z.; Bergmann, F. New Observations on Tautomerism and Ionization Processes in Hypoxanthine and 6-Thiopurines. Isr. J. Chem., 1972, 10, 805-817. 20 Villa-Monsalve, J. P.; Noria, R.; Spiridoula, M.; Péon, J. On the Accessibility to Conical Intersections in Purines: Hypoxanthine and its Singly Protonated and Deprotonated Forms. J. Am. Chem. Soc., 2012, 134, 7820-7829. 21 Pazderski, L.; Toušek, J.; Sitkowski, J.; Kozerski, L.; Szlyk, E. NMR Studies of Tautomerism and Protonation Patterns in Bis(6-Purinyl) Disulfide and Ionic Forms of 6-Mercaptopurine. Polish J. Chem., 2007, 81, 193-210. 22 Cavalieri, L.; Fox, J. J.; Stone, A.; Chang, N. On the Nature of Xanthine and Substituted Xanthines in Solution. J. Am. Chem. Soc., 1954, 76, 1119-1122. 23 Pfleiderer, W.; Nübel, G. Zur Struktur des Xanthin und Seiner N-Methyl-xanthin-derivate. Ann. Chem., 1961, 647, 155-160. 24 Blas, J. S.; Luque, F. J.; Orozco, M. Unique Isomeric Properties of Isoguanine. J. Am. Chem. Soc., 2004, 126, 154-164.

46

25 Godbeer, A. D.; Al-Khalili, J. S.; Stevenson, P. D. Modelling Proton Tunnelling in the Adenine-Thymine Base Pair. Phys. Chem. Chem. Phys., 2015, 17, 13034-13044. 26 Steenken, S.; Jovanovic, S. V. How Easily Oxidizable is DNA? One-Electron Reduction Potentials of Adenosine and Guanosine Radicals in Aqueous Solution. J. Am. Chem. Soc., 1997, 119, 617-618.

47

CHAPTER THREE

Computational Analysis of Tautomers for Indole and Related Compounds

3.1 Abstract:

This report deals with the tautomer stabilities of the cationic, neutral, and anionic forms of five indoles and related compounds—indole, isoindole, indazole, benzimidazole, and 7- azaindole—plus caffeine. To evaluate the most stable neutral tautomers of each compound quantum chemical calculations were first compared with literature data. From these results the stability of the corresponding cationic forms was explored. All of the compounds possess only one potential site for deprotonation, leading to the anionic forms.

3.2 Introduction:

Indoles and related compounds have a wide range of uses in forensic chemistry,1 polymer science,2 pharmaceuticals,3 as well as playing a key role in biochemical signaling for bacteria.4

These compounds display a variety of tautomeric forms in aqueous solution. This report focuses on the tautomerism of indole, isoindole, indazole, benzimidazole, 7-azaindole, and caffeine, as studied through quantum chemical calculations in vacuum and aqueous solution.

Our interest ultimately lies in the populations of the different tautomers of a compound in aqueous solution. In addition to tautomeric forms a compound can exist in different charge states. Each of the charge states, neutral, cationic, and anionic, can display its own set of tautomeric forms. Two factors determine the relative proportions of the tautomeric forms: the intrinsic stabilities, as exhibited by the vacuum energies of each tautomer, and the adjustment to this energy resulting from the compound’s placement within a solvent, i.e. the solvation energy.

This latter term is determined by the prevailing intermolecular forces, such as hydrogen bonding,

48

London dispersion forces, and electrostatic forces. It should be noted that ionic species have considerably larger solvation energies than neutral species due to their stronger electrostatic interactions.

3.3 Computational Methods:

Five compounds were examined for their energetic stabilities using Spartan ’10

(Wavefunction, Inc. Irvine, CA). Indole, isoindole, indazole, benzimidazole, and caffeine were optimized using a semi-empirical method, RM1.5 The geometry was further optimized using density functional theory at the level B3LYP/6-31+G(d,p).6,7 Each compound was optimized in vacuum followed by an application of a water solvent model, SM8.8 All neutral tautomers were placed under the aforementioned parameters. Upon completion, the tautomers were made into cations and anions. The dissociation of a proton from a cationneutral indicates pKa1 and neutralanion indicates pKa2. The numbering scheme for indole is pictured below.

3.4 Results and Discussion:

3.4.1 Indole

3.4.1a. Neutral tautomers:

Indole is a stable compound due to its aromaticity. Its stabilization energy has been estimated to be only slightly less than that of .9 The three neutral tautomers of

49 indole, 1H-I, 3C-I, and 2C-I, were examined (see Figure 3.1) and their relative energies are listed in Table 3.1. Based on the calculations performed it would be expected to find only one dominant tautomer in vacuum or in solution, which can be explained through aromaticity.

Tautomer 1H-I is fully conjugated, follows Hückel’s 4n+2 rule, and has sp2 hybridization of each atom in the system. Tautomers 3C-I and 2C-I have 8π electrons, non-continuous conjugation, and one carbon atom with sp3 hybridization.

Figure 3.1. Neutral tautomers for indole as found in vacuum.

3.4.1b. Cationic tautomers

Much like pyrrole, indole acts as a weak base, where the protonation at the nitrogen atom of 1H-I would result in the loss of aromaticity as a result of the change in hybridization from sp2 to sp3.10 Accordingly, the protonation occurs at the C3 position. The relative energies listed in

Table 3.1 and the structures shown in Figure 3.2 depict how impactful a change in hybridization can be.

50

Table 3.1. Relative energies and solvation energies of the neutral and cationic tautomers for indole.

-1 -1 -1 Compound ΔEvac (kJ mol ) Esolv (kJ mol ) ΔEH2O (kJ mol )

Indole –Neutral

1H-I 0.00 -21.3 0.00

3C-I 40.7 -25.5 36.6

2C-I 122. -35.7 107.

Indole Cations

3C-I+ 0.00 -264. 0.00

2C-I+ 21.1 -227. 58.0

1H-I+ 59.8 -248. 75.4

Figure 3.2. Cationic tautomers for indole as found in vacuum.

Alata and colleagues11 used electron spectroscopy of cold protonated indole in the gas phase to experimentally produce the same trend in tautomer stability. Although we are not the first group to indicate that protonated indole behaves as a carbon acid, it is our intention to provide a clear-cut, unambiguous depiction of the pathway for protonation and deprotonation of indole (see Figure 3.3).

51

Figure 3.3. The deprotonation scheme to the neutral tautomer of indole.

As discussed previously, multiple factors influence tautomer stability, one of which is hybridization. More compelling evidence for the scheme depicted in Figure 3.3 is that the hybridization of N(1) remains sp2. If the scheme were to allow for the deprotonation of 1H-

I+1H-I the nitrogen would go from sp3 to sp2. Adding to this, the indole compound features an enamine functionality, which leaves a partial positive on the nitrogen as well as a partial negative on the C(3) position via resonance contributors.

3.4.2 Isoindole

3.4.2a. Neutral tautomers

Isoindole derivatives have been used in dyes.12 for many years, yet it has remained difficult to synthesize that do not carry a substituent on the nitrogen atom.13 Isoindole is far less stable than its indole counter-part, in part due to its lack of resonance contributors and incomplete benzenoid unit.14 Because of this, we propose a semi-stable neutral tautomer of isoindole, 3C-Iso. Accordingly, DFT calculations were performed, the results of which are listed in Table 3.2, that show a 29% tautomeric contribution of 3C-Iso in vacuum conditions.

Conversely, in aqueous solutions 3C-Iso would be the predominant tautomer with only 11% in population from 2H-Iso. It is understandable to find that the non-conjugated 3C-Iso tautomer is more stable, simply because of its complete benzenoid unit.

52

Figure 3.4. Neutral tautomers for isoindole as found in vacuum.

Joule14 stated tautomeric rearrangements can occur due to solute-solvent effects.

In this case, it has been shown that aprotic solvents favor 2H-Iso, which correlates with the vacuum calculations, while protic solvents have been shown to favor imine arrangements, such as the 3C-Iso neutral tautomer. If a protic solvent were to be placed around the 2H-Iso tautomer a large solvent cage would result inhibiting the accessibility to or the release of the proton. To mimic the experimental results the two tautomers were run under the same parameters using the

SM8 solvent model where dimethyl sulfoxide (DMSO) was used as the solvent. This polar aprotic solvent favored the 2H-Iso arrangement.

The different tautomeric stabilities are also evident in the molecular electrostatic potential maps as shown in Figure 3.5. The more evenly distributed electrostatic charges indicate a higher difficulty of solvation. In other words the interactions between the polar aqueous solvent and the tautomer would not be as stable as a non-uniform molecular electrostatic potential map, as shown in Figure 3.5b.

53

a) b)

Figure 3.5 Molecular electrostatic potential maps of the a) 3C-Iso and b) 2H-Iso tautomers of isoindole.

The solvation energies correspond to the rearrangements from the most stable gas phase neutral tautomer (or aprotic solvent) to the most stable aqueous phase neutral tautomer. These are listed in Table 3.2. As previously mentioned the solvation energy is the adjustment energy from being solvated, it can also be used in this case as the adjustment energy from being in a non-hydrogen bonding environment to one that is.

Table 3.2 Relative energies and solvation energies of the neutral and cationic tautomers for indole.

-1 -1 -1 Compound ΔEvac (kJ mol ) Esolv (kJ mol ) ΔEH2O (kJ mol )

Isoindole –Neutral

2H-Iso 0.00 -21.5 5.46

3C-Iso 3.26 -30.5 0.00

Isoindole Cations

3C-Iso+ 0.00 -233. 0.00

2H-Iso+ 178. -253. 159.

54

3.4.2b. Cationic tautomers

One stable form of protonated isoindole was predicted using the calculations described.

The relative energies listed in Table 3.2 correspond to the tautomers shown in Figure 3.6. In the two neutral tautomers, only 3C-Iso would become ionized to a protonated state, 3C-Iso+. In chapter two it was shown that protonated imines lie lower in energy than protonated amines.

Again this has to do with the stabilizing effect of retaining hybridization on the basic atom.

Figure 3.6. Cationic tautomers for isoindole as found using a water solvent model, SM8.

Adding to this, 3C-Iso+ also has a complete benzenoid unit, which results in 3C-Iso+ being far more stable than 2H-Iso+. The scheme for isoindole acting as a carbon acid is shown in Figure 3.7. This is proposed to occur in both vacuum and solution.

Figure 3.7. Dissociation scheme for isoindole in vacuum acting as a carbon acid.

55

3.4.3 Indazole

Table 3.3. Relative energies and solvation energies of the neutral and cationic tautomers for indazole.

-1 -1 -1 Compound ΔEvac (kJ mol ) Esolv (kJ mol ) ΔEH2O (kJ mol )

Indazole – Neutral

1H-Ind 0.00 -23.7 0.00

2H-Ind 20.1 -30.0 13.7

3C-Ind 80.6 -32.4 71.9

7aC-Ind 219. -34.8 208.

Indazole Cations

+ 1Hm-Ind 0.00 -237. 0.00

+ 3Cm-Ind 69.3 -235. 72.3

+ 3Co-Ind 91.1 -240. 88.4

+ 1Ho-Ind 91.2 -246. 82.3

+ 2Hm-Ind 171. -253. 155.

3.4.3a. Neutral tautomers

Indazole rarely exists in nature and when it is found it is usually substituted.15 Much like isoindole, a common tautomer of indazole, 2H-Ind, contains an unstable, incomplete benzenoid unit. 1H-Ind can be presumed to be the most stable tautomer because both rings are involved in the overall resonance. An NMR16 spectrum (1H) shows all peaks to have a shift greater than 6 ppm, indicating all atoms are involved in aromaticity. 1H-Ind has been correlated

56 experimentally to be the most stable neutral tautomer in solution and the gas phase through microwave and ultra-violet spectroscopy.17-20

In addition, two other rare tautomers were explored for the purpose of this report. The first rare tautomer, 3C-Ind, was shown to be the third most stable tautomer in vacuum and solution. This can easily be explained though lack of conjugation in the ring. The other rare tautomer, 7aC-Ind, has an incomplete benzenoid unit, incomplete conjugation, and non- planarity, which has a dihedral angle of 141.87 °. This yielded a tautomer that was found to be statistically improbable.

Figure 3.8. Neutral tautomers for indazole as found in vacuum.

3.4.3b. Cationic tautomers

It is not difficult to understand the results from the cationic tautomers listed in Table 3.3.

+ 1Hm-Ind retains the conjugated and hybridized stability of the compound while being protonated. Furthermore, the compound does not adopt a quinoid feature, which is generally less stable than a benzenoid feature. Previous infra-red gas phase21 studies on protonated indazole

57

+ concluded that 1Hm-Ind was the only probable tautomeric form of cationic indazole. The same group also performed theoretical calculations in which they found approximately a 100 kJ mol-1

+ + energy difference between 1Hm-Ind and the next most stable tautomer, 3Cm-Ind .

Figure 3.9. Cationic tautomers for indazole as found in vacuum.

58

3.4.4. Benzimidazole

Table 3.4. Relative energies and solvation energies of the neutral and cationic tautomers for benzimidazole.

-1 -1 -1 Compound ΔEvac (kJ mol ) Esolv (kJ mol ) ΔEH2O (kJ mol )

Benzimidazole –Neutral

1H-B 0.00 -37.6 0.00

2C-B 130. -51.3 116.

Benzimidazole Cations

+ 1Himine-B 0.00 -244. 0.00

2C-B+ 154. -253. 145.

+ 1Hamine-B 178. -275. 147.

3.4.4a. Neutral Tautomers

It has been a theme throughout the tautomeric stabilities of the indole related compounds that

structures that contain quinoid or an incomplete benzenoid unit tend to be less stable. This

observation is again on display with the tautomeric forms of benzimidazole.

Figure 3.10. Neutral tautomers for benzimidazole as found in vacuum.

59

3.4.4b. Cationic tautomers

+ 1Himine-B was found to be the most stable cationic tautomer for benzimidazole due to the stability it retains following protonation. This cationic tautomer, unlike the others, maintains sp2 hybridization in both rings and follows Hückel’s rule. There were two other possible cationic tautomers explored, which were found to be statistically improbable. Anderson22 and colleagues used polarized two-photon excitation spectra of benzimidazole and its cation and found the same structure for the neutral and cationic state of benzimidazole. The associated energies for all cationic tautomers are listed in Table 3.4 and their structures can be seen in Figure 3.11.

Figure 3.11. Cationic tautomers for benzimidazole as found using a water solvent model, SM8.

60

3.4.5. 7-Azaindole

Table 3.5. Relative energies and salvation energies of the neutral and cationic tautomers for 7- azaindole.

-1 -1 -1 Compound ΔEvac (kJ mol ) Esolv (kJ mol ) ΔEH2O (kJ mol )

7-Azaindole – Neutral

1H-A 0.00 -30.4 0.00

3C-A 59.9 -28.7 43.5

7-Azaindole Cations

+ 1Himine-A 0.00 -232. 0.00

7H,3C-A+ 45.8 -241. 37.0

1H,3C-A+ 74.3 -261. 45.5

+ 1Hamine-A 130. -259. 103.

3.4.5a. Neutral Tautomers

7-Azaindole is a widely used compound. Its skeleton has been incorporated into pharmaceutical structures,23 coordination chemistry,24,25 and used to tune photochromic compounds.26 The compound is fully aromatic with ten π-electrons in the most stable tautomer

(1H-A). 7-Azaindole can form two plausible tautomers where the unstable tautomer (3C-A) would be considered rare in gas phase or solution. The calculated relative energies can be seen in Table 3.5.

61

Figure 3.12. Neutral tautomers for 7-azaindole as found in vacuum.

3.4.5b. Cationic Tautomers

The relative stabilities for the cationic tautomers of 7-azaindole are listed in Table 3.5.

The results are as expected where the protonation occurs on the imine nitrogen of 1H-A. The rare tautomer results can be explained just as the aforementioned compounds have. First,

7H,3C-A+ shows a greater stability than 1H,3C-A+ because the charge is delocalized over the

+ six-member ring. Secondly, the hybridization of the pyrrole nitrogen in 1Hamine-A is compromised revealing a great loss in stability and large relative energy difference.

Figure 3.13. Cationic tautomers for azaindole as found in vacuum.

62

3.4.6. Caffeine

3.4.6a. Neutral Species

Caffeine is one of the most popular compounds in the world. Not only is it widely used in chemistry, it is commonly discussed in the general public. Caffeine is also the most consumed stimulant in the world.27 It is a trimethylated xanthine that acts on the central nervous system to replace adenosine on the receptors of nerve cells.28 When caffeine is degraded in the liver the following three compounds are found in decreasing concentrations: paraxanthine (1,7-dimethyl), theobromine (3,7-dimethyl), and theophylline (1,3-dimethyl).29

Figure 3.14. The singular, neutral tautomer of caffeine. The numbering scheme is different from indole derivatives as noted above.

3.4.6b. Cationic Species

The protonated form of caffeine has been known and understood for many years.30-32

The protonation takes place on the N(9) position resulting in a protonated imine as determined theoretically and experimentally.

63

Figure 3.15. The cationic tautomer of caffeine.

3.4.6c. Anionic Species

Caffeine exists in only one tautomeric form which has the capability to lose a proton only at the C(8) position as determined theoretically.

Figure 3.16. Dissociation of caffeine to its anionic form.

The anionic dissociation was puzzling for this compound. The dissociation constant, at standard conditions, is 14.33-35 While this may not seem unusual at first, it is quite bewildering.

In order to obtain a dissociation constant where the neutral species dissociates to produce an anion the C(8) proton would have to abstracted. The issue our group takes is not on carbon acids, but the probability of a carbon acid having a low dissociation constant with no major

64 electron withdrawing groups or resonance stabilization. In fact, the two nitrogen atoms on the five-member ring should donate electron density through resonance. Simply due to the fact that there are no other protons to dissociate, caffeine would need to act like a carbon acid. However, we assert that the dissociation constant published may be incorrect and the numerical value has been reproduced many times over without a re-examination on the experimental techniques.

Quantum descriptors were examined to assist in correlating the dissociation of the sp2 CH proton. At times the natural charges can be used in predicting an outcome for the dissociation patterns. In this case the charges for N(7), C(8), H(10), and N(9) were -0.361, 0.241, 0.264, and

-0.546, respectively. Since the dissociation site holds charges of the same sign one would predict there to be repulsion between the carbon and hydrogen, however many indole and purine compounds (known nitrogen acids) have this motif. Therefore, it was concluded that the natural charges no longer applied as a model for predicting the behavior on this compound.

The energy change under the solvent model can act as an excellent surrogate in predicting how a compound will dissociate. An example compound can be examined to assist in the understanding of the dissociation of caffeine. Benzimidazole has been shown to act as a nitrogen acid. The energy change from the neutral species to the anionic species in water is 1213 kJ mol-

1. This correlates to most other N-heterocyclic nitrogen acids. However, under the computational methods described above if benzimidazole were to be a carbon acid, losing a proton from the C(2) position, the energy change in water would be 1367 kJ mol-1. The findings were similar to caffeine where the energy change was 1365 kJ mol-1. This corresponds to caffeine acting as a carbon acid, but this should be followed up by further experimental data.

65

3.5 Conclusion:

The neutral and cationic tautomers of five indole related compounds along with caffeine were evaluated for their relative stabilities. From this study one can assume that most N- heterocycles will lose a proton from the nitrogen. Surprisingly, certain compounds studied were shown to act as carbon acids. This is especially true if the nitrogen atom were to undergo a hybridization change. This anomaly leads to bond rearrangements for the cationic formation of indole, isoindole, and the anionic formation of caffeine. This was interesting to uncover due to the unexpected outcome and the implications it has on the wide array of systems these compounds are involved in.

3.6 References

1 Shevyrin, V.; Melkozerov, V.; Nevero, A.; Etsov, O.; Baranovsky, A.; Shafran, Y. Synthetic Cannabinoids as Designer Drugs: New Representatives of Indol-3-Carboxylates as a Novel Group of Cannabinoids. Forensic Sci Intl, 2014, 244, 263-275. 2 Beck, J.; Ineman, J. M.; Rowan, S. J. Metal/Ligand-Induced Formation of Metallo- Supramolecular Polymers. Macromolecules, 2005, 38, 5060-5068. 3 Inamoto, K.; Katsuno, M.; Yoshino, T.; Arai, Y.l Hiroya, K.; Sakamoto, T. Synthesis of 3- Substituted Indazoles and Benzoisoxazoles via Pd-Catalyzed Cyclization Reactions: Application to the Synthesis of Nigellicine. Tetrahedron, 2007, 63, 2695-2711. 4 Lee, J.-H.; Lee, J. Indole as an Intercellular Signal in Microbial Communities. FEMS Micribiol Rev, 2010, 34, 426-444 5 Rocha, G. B.; Freire, R. O.; Simas, A.; Stewart, J. J. P. J. Comput. Chem. 2006, 27, 1101. 6 Becke, A. D. J. Chem. Phys., 1993, 98, 5648-5652. 7 Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B, 1998, 37, 785-789. 8 Marenich, A. V.; Olson, R. M.; Kelly, C. P.; Cramer, C. J.; Truhlar, D. G. J. Chem. Theory Comput., 2007, 3, 2011.

66

9 Hess, B. A.; Schaad, L. J.; Holyoke, C. W. Tetrahedron, 1972, 28, 3657. 10 Sundberg, R. J. Indoles Academic Press, Inc. San Diego, 1996. 11 Alata, I.; Bert, J.; Broquier, M.; Dedonder, C.; Feraud, G.; Grégoire, G.; Soorkia, S.; Marceca, E.; Jouvet, C. Electronic Spectra of the Protonated Indole Chromophore in the Gas Phase. J. Phys. Chem. A, 2013, 117, 4420-4427. 12 Alberto, M. E.; De Simone, B. C.; Mazzone, G.; Quartarolo, A. D.; Russo, N. Theoretical Determination of Electronic Spectra and Intersystem Spin-Orbit Coupling: The Case of Isoindole-BODIPY Dyes. J. Chem. Theory Comput., 2014, 10, 4006-4013. 13 Bonnett, R.; Brown, R. F. C. Isoindole. J. C. S. Chem. Comm., 1972, 7, 393-395. 14 Joule, J. A.; Mills, K. Heterocyclic Chemistry, 4th Ed; Wiley&Sons: West Sussex, UK, 2000. 15 Atta-ur-Rahman; Malik, S.; Cun-heng, H; Clardy, J. Tetrahedron Lett., 1995, 36, 2759-2762. 16 Indazole: http://sdbs.db.aist.go.jp (National Institute of Advanced Industrial Science and Technology, 10/26/2015). 17 Elguero, J.; Marzin, C.; Katritzky, A. R.; Linda, P. The Tautomerism of Heterocycles, Academic Press, New York, 1976. 18 Elguero, J. and Their Benzo Derivatives, in Comprehensive Heterocyclic Chemistry, Pergamon Press, Oxford, 1984, pp 291-297. 19 Velino, B.; Cané, E.; Trombetti, A. Microwave Spectrum and ab Initio Calculations of Indazole. J. Molec. Spec., 1992, 155, 1-10. 20 Catalán, J.; de Paz, J. L. G.; Elguero, J. Importance of Aromaticity on the Relative Stabilities of Indazole Annular Tautomers: an ab Initio Study. J. Chem. Soc., Perkin Trans. 2, 1996, 57- 60. 21 Oomens, J.; Meijer, G.; von Helden, G. An Infrared Spectroscopic Study of Protonated and Cationic Indazole. Int. J. Mass. Spec., 2006, 249/250, 199-205. 22 Anderson, B. E.; Jones, R. D.; Rehms, A. A.; Ilich, P.; Callis, P. R. Polarized Two-Photon Fluorescence Excitation Spectra of Indole and Benzimidazole. Chem. Phys. Lett., 1986, 125, 106-112. 23 Bryan, M. C.; Falsey, J. R.; Frohn, M.; Reichelt, A.; Yao, G.; Bartberger, M. D.; Bailis, J. M.; Zalameda, L.; San Miguel, T.; Doherty, E. M.; Allen, J. G. N-substituted Azaindoles, as Potent Inhibitors of Cdc7 Kinase. Bioorg. Med. Chem. Lett., 2013, 23, 2056-2060.

67

24 Koyama, T.; Wakisaka, A. Molecular Self-Assembly Composed of Aromatic Hydrogen-Bond Donor-Acceptor Complexes. J Chem. Soc., Faraday Trans., 1997, 21, 3813-3817. 25 Dufour, N.; Dartiguenave, Y.; Dartiguenave, M.;Lebuis, A.-M.; Bélanger-Gariépy, F.; Beauchamp, A. Crystal Structures of 7-Azaindole, an Unusual Hydrogen-Bonded Tetramer, and of Two of its Methylmercury(II) Complexes. Can. J. Chem., 1990, 68, 193-201. 26 Sun, Z.; Li, H.; Liu, G.; Fan, C.; Pu, S. Photochromisn of New Unsymmetrical Diarylethenes Based on the Hybrid of Azaindole and Moieties. Dyes and Pigments, 2014, 106, 94- 104. 27 Daly, J. W.; Holmén, J.; Fredholm, B. B. Is Caffeine Addictive? The most Wodely Used Psychoactive Substance in the World Affects the Same Parts of the Brain as Cocaine. Lakartidningen, 1998, 95, 5878-5883. 28 Fisone, G.; Borgkvist, A.; Usiello, A. Caffeine as a Psychomotor Stimulant: Mechanism of Action. Cellular and Molecular Life Sciences, 2014, 61, 857-872. 29 Martínez-López, S.; Sarriá, B.; Baeza, G.; Mateos, R.; Bravo-Clemente, L. Pharmacokinetics of Caffeine and its Metabolites in Plasma and Urine After Consuming a Green/Roasted Blend by Healthy Individuals. Food Research International, 2014, 64, 125-133. 30 Sondheimer, E.; Covitz, F.; Marquisee, M. J. Association of Naturally Occurring Compounds, the Chlorogenic Acid-Caffeine Complex. Arch. Bioch. Biophys., 1961, 93, 63-71. 31 Marta, R. A.; Wu, R.; Eldridge, K. R.; Martens, J. K.; McMahon, T. B. Infrared Vibrational Spectra as a Structural Probe of Gaseous Formed by Caffeine and Theophylline. Phys. Chem. Chem. Phys., 2010, 12, 3431-3442. 32 Bahrami, H.; Tabrizchi, M.; Farrokhpour, H. Protonation of Caffeine: A Theoretical and Experimental Study. Chemical Physics, 2013, 415, 222-227. 33 Dean, J.A. Lange’s Handbook of Chemistry, 13th Ed; McGraw-Hill: New York, 1985. 34 Perera, V.; Gross, A. S.; McLachlan, A. J. Measurement of CYP1A2 Activity: A Focus on Caffeine as a Probe. Current Drug Metabolism, 2012, 13, 667.678. 35 Marra, M. C.; Silva, P. L.; Muñoz, R. A. A.; Richter, E. M. Ultra-Fast Determination of Scopolamine, Orphenadrine, Mepyramine, Caffeine, Dipyrone, and Ascorbic Acid by Capillary Electrophoresis with Capacitively Coupled Contactless Conductivity Detection. J. Braz. Chem. Soc., 2014, 25, 913-919.

68

CHAPTER FOUR

Computational Estimation of the pKa’s of Purine and Purine Related Compounds

4.1 Abstract:

Quantitative structure-activity relationships provide powerful insight for physical properties that have not been determined experimentally. The acid dissociation constant, pKa, is especially useful because it allows for one to know the ionic form a compound may take in different environments. Thirty-one purine and related compounds were examined using quantum chemical calculations to estimate the acid dissociation constant for those which were not readily available. Various descriptors were used to evaluate linear free energy relationships and a multivariable linear free energy relationship.

4.2 Introduction:

The acid dissociation value pKa of a compound provides important information regarding the neutral and ionic forms of the compound present under different conditions. However, for many compounds experimental pKa values are not available, and some others the measured values may be uncertain. Consequently, having a computational means for estimating pKa values can be beneficial for biochemical, pharmaceutical, polymer, and many other studies.

The acid dissociation constant can be derived through the following generic :

[1]

69 where the equilibrium constant, Ka, is a ratio of the product activities to the reactant activity.

More frequently, concentrations are used in place of activities as approximations for practical purposes. The pKa of the acid is then described by a logarithmic relationship to Ka.

[2]

In principle, a thermodynamic relationship can then be obtained using the previous equation:

[3] where ΔGo is the standard Gibbs energy change for the dissociation, R is the ideal gas constant, T is the absolute temperature in kelvins. For the purposes of the present research the temperature is assumed to be 298.15 K. Thus there is a direct linear relationship between the Gibbs energy

1 change and the pKa.

There are many theoretical approaches to estimate the pKa of a compound. One such route is through first principles. This involves a purist methodology that does not rely on any previous experimental values.2 However, this approach is computationally demanding and requires high levels of accuracy. The absolute technique requires calculation of the Gibbs

-1 energy change to within 5.69 kJ mol in order to obtain solution pKa values to within one pKa unit.3,4 Another method commonly used in commercial programs is a fragment based approach that uses empirical values to theoretically derive pKa’s. Groups of starting compounds are

‘fragmented’ and Hammett constants are applied. In this way a theoretical pKa is derived by giving more or less weight to each fragment depending on its relevance to the parent compound.

A weakness of this methodology is that it requires suitable pre-programmed fragments. The final methodology is a quantitative structure-activity relationship (QSAR).

70

QSAR’s are centrally focused on finding a mathematical relationship between one or more molecular features and a specific molecular behavior.2 The mathematical relationship takes on the form:

[4]

In this expression Pj is a molecular property of interest, the Xi’s are molecular descriptors, and the ci’s are the coefficients of these descriptors. The QSAR technique first employs a literature search to gather published data on the desired physical property for different compounds.

Quantum chemical (QC) calculations are performed to derive descriptors that yield ‘clues’ to the compound’s behavior. It is assumed that each descriptor adds a characteristic Gibbs energy

0 change, ΔGi to the property of interest. Accordingly, expression [4] is sometimes called a linear free energy relationship (LFER).2

4.3 Computational Methods:

We have developed a QSAR approach for estimating the pKas of purines and related compounds. A group of thirty-one compounds was first assembled to find literature values for their dissociation constants. These compounds have two dissociation constants which we will call pKa1 and pKa2:

cation  neutral form neutral form  anion

pKa1 pKa2 The collected literature values can be found in Table 4.1, which also contains computational values obtained using Advanced Chemical Development’s (ACD) Percepta Phys Chem Suite commercial program. The dissociation constants were recorded at 25°C and in aqueous solution.

71

Along with the solution pKa values, vapor phase Gibbs energy changes for deprotonation reactions of some of these compounds have been experimentally determined and can be found on the National Institute of Standards and Technology’s (NIST) ChemwebBook website.5

As previously described in Chapters 2 and 3, these compounds can exist in a variety of tautomeric forms. For the present study, only the most stable tautomeric forms of each compound were considered. The most stable neutral tautomers of each compound were first built and then examined using the semi-empirical quantum chemical method RM1.6 This was followed by an ab initio analysis using density functional theory with the B3LYP functional of

Becke,7 Lee, Yang, and Parr8 and the basis set 6-31+G**, i.e. at the B3LYP/6-31+G** level.

Each compound was optimized in vacuum followed by a calculation applying the water solvent model, SM8 of Marenich et al.9 Having obtained the neutral forms the associated possible cation and anionic forms were determined and calculated in the same manner as above. Quantum chemical descriptors for each of the compounds were determined and analyzed. These included

N-H bond lengths, RNH, the energy differences between the highest occupied molecular orbital

(HOMO) and the lowest unoccupied molecular orbital (LUMO), ΔεHL, relevant atomic charges,

Qi, bond orders, BOi, energy differences between the parent compound and the product species,

ΔEi, etc. Tables were compiled from these calculations. Mathematical relationships were sought of the form:

pKa = . [5]

Various statistics were recorded when examining the QSAR models. The statistics allow one to identify the best correlations between pKa’s and the descriptors, the uncertainties associated with each descriptor, and other statistical measures. It is also important that outliers—

72 compounds falling outside the regression models-- be examined. These can be insightful because they may shed light on shortcomings of either the model or the experimental values.2

Table 4.1. Experimental pKa values for purine and related compounds along with theoretical values obtained from ACD/Phys Chem Suite.

Compound Formula pKa1 pKa2 ACD pKa1 ACD pKa2

10,11 10, 11 1 Adenine C5H5N5 4.2 9.8 4.4±0.1 9.9±0.1

12 2 4-Azaindole C7H6N2 6.9 6.9±0.1 14.7±0.3

12 3 5-Azaindole C7H6N2 8.3 8.2±0.3 15.2±0.3

12 4 6-Azaindole C7H6N2 7.9 7.8±0.1 14.7±0.3

12 5 7-Azaindole C7H6N2 4.6 4.9±0.1 13.9±0.3

10 11, 13,14 11 6 Benzimidazole C7H6N2 5.3 , 5.5 12.0 , 5.3±0.1 12.6±0.1 12.310,12.813, 12.914

7 Benzimidazoline C7H8N2 6.9±0.2

15 15 8 Carbazole C12H9N -4.9 17.1 -2.4±0.5 17.0±0.3

16 17 9 Caffeine C8H10N4O2 0.5 14.0 0.5±0.7

11,18 10 Diphenylamine C12H11N 0.9 0.8±0.2

11, 19 11,19 11 Guanine C5H5N5O 3.3 9.2 , 3.4±0.3 9.6±0.2 12.311, 12.419

10 20 10 10 12 Hypoxanthine C5H4N4O 1.19 , 2.20 , 8.8 , 12.0 2.2±0.3 8.9±0.5

15 15 13 Indazole C7H6N2 1.3 13.9 1.3±0.1 14.0±0.1

14 10 14 Indole C8H7N -3.6 16.97 -2.4±0.5 17.0±0.3

14 15 Indoline C8H9N 4.9 5.2±0.2

16 Isatin C8H5NO2 -1.9±0.2 10.3±0.2

73

18 18 17 Isoguanine C5H5N5O 4.5 9.0 3.7±0.2 10.8±0.4

18 Isoindole C8H7N

19 Oxindole C8H7NO 0.9±0.2 14.8±0.2

23 20 Paraxanthine C7H9N4O2 8.6 0.2±0.7 8.5±0.7

11 21 Phthalimide C8H5NO2 9.9 -5.2±0.2 10.4±0.2

11 10,11 22 Purine C5H4N4 2.5 8.9 2.7±0.1 9.2±0.1

21 23 6-Mercaptopurine C5H4N4S 7.77 2.3±0.2 10.9±0.2

11 24 Saccharin C7H5NO3S 2.3 -12.8±0.2 1.6±0.1

10 25 Skatole C9H9N 16.6 -1.6±0.5 17.3±0.3

11 11 16 26 Theobromine C7H8N4O2 0.68 7.9 , 8.8 0.4±0.7 9.9±0.7

19 27 Theophylline C7H8N4O2 8.6 1.8±0.7 8.6±0.5

10 11 28 Uric Acid C5H4N4O3 5.6 , 5.5 -1.9±0.2 5.6±0.3

22 29 9-Methyluric Acid C6H6N4O3 5.7 -1.9±0.2 5.6±0.3

18 10 30 Xanthine C5H4N4O2 0.8 7.53 1.0±0.7 7.6±0.5, 13.9±0.2

22 11 31 9-Methylxanthine C6H6N4O2 2.0 6.3 , 0.1±0.7 6.3±0.7 10.522,23

4.4 Results and Discussion:

An initial step in evaluating the results is to verify that in the absence of a solvent the level of approach (functional and basis set) is adequate for the dissociation reaction in the absence of a solvent. This was examined using Gibbs energy changes for deprotonation reactions found on the NIST webpage (http://webbook.nist.gov) and comparing with the corresponding values determined at the B3LYP/6-31+G** level. The Gibbs energy change for the dissociation can be found using the following generic equations based on expression [1].

74

[6]

A few things should be noted regarding the calculations and plot. First, the Gibbs energy could not be calculated within the solvent due to coding limitations of the Spartan ’10 program, and therefore all values apply to vacuum conditions. Second, the Gibbs energy Go of a proton was determined to be ~ -26.3 kJ mol-1. The derivation for this can be found in the appendix. The values used in the Gibbs energy plot can be seen in Table 4.2. It should be noted the calculated values all fell within the experimental uncertainties.

Table 4.2 Experimental and calculated Gibbs energy changes used in Figure 4.1.

-1 -1 Compound ΔGrxn NIST (kJ mol ) ΔGrxn Calc. (kJ mol )

Adenine 1372. 1382.

Indazole 1425. 1425.

Indole 1431. 1432.

Carbazole 1412. 1413.

Diphenylamine 1438. 1435.

75

ΔGexp vs ΔGcalc 1450

1440

1430 y = 1.1991x - 284.11 1420 R² = 0.9962

1410

) ) Experimental 1 - 1400

(kJ mol (kJ 1390

rxn

G 1380 Δ 1370

1360 1370 1380 1390 1400 1410 1420 1430 1440 -1 ΔGrxn (kJ mol ) Calculated

Figure 4.1. Comparison of the experimental and calculated Gibbs energy changes for dissociation of the neutral species to the anionic form and H+.

There were five experimental values in the NIST database that corresponded to the present study. They were adenine, indole, indazole, carbazole, and diphenylamine. The resulting comparison between experimental and calculated values is summarized in the following expression:

-1 ΔGexp = 1.199±0.043 ΔGcalc – 284.1±60.5 kJ mol [7]

n = 5, R2 = 0.996, s = 1.86, F = 789.

The excellent regression statistics above show that the level of theory used, in the vacuum model, is sufficient for accurate results. This is evident in the high R2 value, the large Fisher statistic, and the high T-test value (28.9) of the slope.

76

Because it is not possible to calculate Go within a solvent model we would like to find a surrogate QC descriptor for the pKa regression calculations. The energy change ΔE has been used in earlier studies and would seem to be a reasonable candidate.24,25 The experimental and calculated values applied to this model can be seen in Table 4.3.

Table 4.3 The experimental Gibbs energies correlated with the calculated ΔEVAC.

-1 -1 Compound ΔGrxn NIST (kJ mol ) ΔGrxn Calc. (kJ mol )

Adenine 1372. 1439

Indazole 1425. 1489.

Indole 1431. 1496.

Carbazole 1412. 1476.

Diphenylamine 1438. 1501.

Accordingly the correspondence between ΔEVAC and ΔGexp was examined with the results shown in Figure 4.2 and below:

-1 ΔGexp = 1.047±0.019 ΔEVAC - 134.9±28.3 kJ mol [8]

n = 5, R2 = 0.999, s = 0.956, F = 2295.

Ironically, the statistics provided in expression [8] were better than those in expression [7]. This encourages the suggestion that ΔE might act as a good regression descriptor, because of the relationship

. [9]

77

ΔGexp vs ΔEVAC 1450 1440 y = 1.0472x - 134.88 1430 R² = 0.999

1420

) 1 - 1410

1400

(kJ mol (kJ

rxn 1390 G Δ 1380 1370 1360 1430 1440 1450 1460 1470 1480 1490 1500 1510 -1 ΔEVAC (kJ mol )

Figure 4.2. A plot of ΔEvac against the experimental (NIST) Gibbs energies.

Following this it was desirable to test the ΔE descriptor’s ability to model the aqueous pKa values. A first test was to see if the ΔEVAC is sufficient for modeling these pKa’s. In other words, do the calculations need to be taken to the additional step of applying the SM8 solvent model? This was tested using pKa1.

Four compounds were removed from this regression calculation (indole, indoline, 7- azaindole, and diphenylamine) because they were considered to be outliers. (See further explanation below.) As seen, ΔEVAC performs surprisingly well. Figure 4.3 and the regression statistics below suggest that there is considerable correlation without using the solvent model.

pKa1 = 0.0824±0.0104 ΔEVAC – 77.12±10.12 [10]

n = 15, R2 = 0.8285, s = 1.46, F = 62.8

78

pKa1 vs. ΔEvac 10

8

6 y = 0.0824x - 77.118 R² = 0.8285

4

a1 2 pK 0

-2

-4

-6 870 890 910 930 950 970 990 1010 1030 1050 -1 ΔEVAC (kJ mol )

Figure 4.3. The QSAR plot for experimental pKa2 against the QC descriptor ΔEVAC.

The experimentally determined dissociation constants were measured in aqueous solutions. The ΔEVAC descriptor does not take into account potential solvent-solute interactions.

The next test was to apply the solvent model to see if that led to a better correlation. The results for pKa1 are shown below and in Figure 4.4.

pKa1 = 0.1042±0.0070 ΔEVAC – 119.92±8.24 [11]

n = 15, R2 = 0.9445, s = 0.828, F = 221.

The statistics from expression [11] indicate that the addition of a solvent model assists the regression calculations. In this case the coefficient of determination, R2, increases. This indicates that there was a 94% correlation between the experimental values and our model. Of the original thirty-one compounds nineteen had experimental pKa1 values reported. Four compounds were regarded as outliers and removed from the regression calculations. The

79 compounds removed were 7-azaindole, indole, indoline, and diphenylamine. Indole act as carbon acids and thus fall outside the nitrogen acid scheme. (Indole was discussed in chapter three.) Indoline is an aniline derivative—no other compound is similar to this in the set and therefore indoline is an outlier.

7-Azaindole is known to dimerize in solution due to bimolecular hydrogen bonding. This is shown below in Figure 4.4.

Figure 4.4. The bimolecular hydrogen bonding of 7-azaindole, which results in dimerization.

The phenomenon has been observed experimentally many times.26-28 In fact, Koyama et al.26 suggested that 7-azaindole will not be found as a singular unit unless forced through extreme high or low temperatures. Therefore, our model was not suitable for this compound.

Diphenylamine was also considered an outlier. Surprisingly though, it correlated into our Gibbs energy plots well. It is proposed that when compared directly with 31 compounds, it fell outside of the scope of our data set because it was the only non-planar ring system.

80

pKa1 vs. ΔEH2O 10

8

6 y = 0.1042x - 119.37 R² = 0.9445

4

a1 2 pK

0

-2

-4

-1 ΔEH2O (kJ mol ) -6 1090 1110 1130 1150 1170 1190 1210

Figure 4.5. The QSAR plot for experimental pKa1 against the QC descriptor ΔEH2O.

For the compounds presented in this study each can undergo multiple dissociations. Thus far the only experimental dissociation values examined are for pKa1. However, the loss of a proton that forms an anionic compound needs to be evaluated, i.e. pKa2.

pKa2 = 0.131±0.009 ΔEH2O – 147.23 ±10.79 [12]

n = 18, R2 = 0.929, s = 1.17, F = 212.

Based on the regression calculations three compounds were eliminated from the original data set. These compounds were caffeine, paraxanthine, and theobromine. Caffeine was eliminated from the data set for acting like a carbon acid as discussed in chapter three. When the experimental techniques were reviewed for paraxanthine and theobromine the determined pKa’s

81 appeared correct. However, there seems to be some mechanism at play that is not mimicked in our model. While there is not a clear cut answer at this point, it can be assumed that the compounds may act similarly to caffeine because each of them is a caffeine metabolite.

pKa2 vs. ΔEH2O

18 16 y = 0.131x - 147.23 R² = 0.9298 14 12

10

a2

pK 8 6 4 2 0 1100 1120 1140 1160 1180 1200 1220 1240 1260 -1 ΔEH20 (kJ mol )

Figure 4.6. The QSAR plot for experimental pKa2 against the QC descriptor ΔEH2O.

A different QC descriptor was approached at this point. The change in the natural charge on the nitrogen atom ΔQNH as it undergoes the dissociation is intuitively a reasonable descriptor.

This was correlated against the pKa2 values of each compound. As described in expression [12] and Figure 4.6, caffeine, paraxanthine, and theobromine were eliminated from this data set.

pKa2 = 157.44±43.85 ΔQN-N(an.) – 3.302 ±1.983 [13]

n = 18, R2 = 0.4462, s = 3.27, F =12.9

82

The regression calculations shown in expression [13] provide a coefficient of determination that is far from great. However, there are obvious outliers that could be removed, but were not for consistency with the other pKa2 regression calculations.

ΔQNH for pKa2 18 16 14 y = 157.44x + 3.3024 R² = 0.4462 12

10 a2

pK 8 6 4 2 0 -0.01 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08

ΔQNH

Figure 4.7. The QSAR plot for experimental pKa2 against the QC descriptor ΔQNH.

At times, a QC descriptor may be mediocre in describing a physical property (as shown in Figure 4.8). In spite of this, when an unreliable descriptor is combined with a high-impact descriptor a successful multi-variant regression calculation may be acquired. This is displayed in

Figure 4.8 and expression [14] when the change in energy on the nitrogen atom for the neutral species to its anionic form along with the change in energy with the solvent model is linearly applied. As such, the following multi-variant LFER was formed using the QC descriptors: ΔQNH and ΔEH2O for the pKa2.

83

Figure 4.8. The QSAR plot for experimental pKa2 against the QC descriptors ΔEH2O and ΔQNH.

pKa2 = 0.1146±0.0070 ΔEH2O + 55.60±12.09 ΔQN-N(an.) – 129.86±8.11 [14]

n = 18, R2 = 0.971, s = 0.775, F = 250.

The success of expression [14] as shown in the statistics gives the opportunity to provide estimation of the pKa2 for the compounds missing experimental values. This is also evident when the calculated dissociation constants are compared to those derived through ACD-

PhysChem Suite. The values calculated using expressions [11], [12], and [14] are shown in

Table 4.2.

84

Table 4.4. Theoretically calculated pKa’s using the QSAR regressions shown in the figure 4.5,

4.6, and 4.8 above.

Compound pKa1 [Eq. 11] pKa2 [Eq. 12] pKa2 [Eq.14]

4-Azaindole ---- 14.5 15.2

5-Azaindole ---- 14.6 14.9

6-Azaindole ---- 16.0 15.9

7-Azaindole ---- 14.8 15.2

Benzimidazoline 3.05 21.4 21.2

Diphenylamine ---- 20.0 20.2

Indoline ---- 24.5 24.4

Isatin -11.2 10.3 10.9

Isoindole -10.5 14.3 14.2

Oxindole -8.87 13.5 14.0

Paraxanthine 0.47 ------

Phthalimide -12.7 ------

6-Mercaptopurine 2.42 ------

Saccharin -14.0 ------

Skatole -4.78 ------

Theophylline 0.86 ------

Uric Acid -10.4 ------

9-Methyluric Acid -10.0 ------

Other QC descriptors were examined and failed to provide valuable plots. These included the bond order and bond length. This is not atypical for QSAR plots using pKa’s as the

85 physical property. This is because the bond order and bond lengths tend to stay within a small range while the pKa values can be cover a wide range. A similar explanation can be used for the change in energy from the HOMO to the LUMO.

4.5 Conclusion:

The ultimate goal for this study was to design a model allowing us to estimate the pKa’s of purines and related compounds, with a secondary goal of estimating values for those compounds in the collection that did not have experimental values found in the literature. This was successfully performed using three different regression calculations on eighteen compounds.

As described in Chapter 1 purines and related compounds are used in a wide variety of applications. For many purposes it is essential to have a known pKa value for each compound to understand the form it will take in a given environment. The present study provides estimated pKa values for eighteen compounds that were previously without experimental values. This can provide insights that may be useful for some biochemical, pharmaceutical, and environmental applications.

4.6 Appendix: Determination of the Gibbs energy Go for H+

 Use the thermodynamic expression for enthalpy as the sum of the internal energy and the product of pressure and volume.

 Apply the ideal gas law where the number of moles is assumed to be one.

 Express the internal energy Uo using the classical law of equipartition of energy for a monatomic gas that has only translational energy.

86

 Substitute the internal energy and ideal gas law expressions into the enthalpy expression.

 The Sakur-Tetrode equation, with M equals the molar mass in grams and T equals the absolute temperature kelvins, leads to a value for the entropy of a monatomic gas of molar mass = 1 g.

 At T = 298.15 K this leads to the Gibbs energy of H+:

4.7 References

1 Perrin, D. D.; Dempsey, B.; Sergeant, E. P. pKa Prediction for Organic Acids and Bases, Chapman & Hall: New York, 1981. 2 Shields, G. C.; Seybold, P. G. Computational Approaches for the Prediction of pKa Values, CRC Press: Boca Raton, 2014. 3 Liptak, M. D.; Shields, G. C.; Accurate pKa Calculations for Carboxylic Acids Using Complete Basis Set and Gaussian-n Models Combined with CPCM Continuum Solvation Methods. J. Am. Chem. Soc., 2001, 123, 7314-7319. 4 Liptak, M. D.; Gross, K. C.; Seybold, P. G.; Feldgus, S.; Shields, G. C. Absolute pKa Determinations for Substituted Phenols. J. Am. Chem. Soc., 2002, 124, 6421-6427. 5Lindstrom, P. J.; Mallard, W. G., Eds. NIST Chemistry WebBook, NIST Standards Reference Database Number 69, National Institute of Standards and Technology, Gaithersburg, MD, 20899, http://webbook.nist.gov. 6Rocha, G. B.; Freire, R. O.; Simas, A.; Stewart, J. J. P. J. Comput. Chem. 2006, 27, 1101. 7Becke, A. D. J. Chem. Phys., 1993, 98, 5648-5652. 8Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B, 1998, 37, 785-789. 9Marenich, A. V.; Olson, R. M.; Kelly, C. P.; Cramer, C. J.; Truhlar, D. G. J. Chem. Theory Comput., 2007, 3, 2011.

87

10 Serjeant, E.P.; Dempsey, B. In Ionization Constants of Organic Acids in Aqueous Solution; IUPAC Chemical Data Series-No. 23; International Union of Pure and Applied Chemistry; Oxford, UK, 1979. 11 Dean, J.A. Lange’s Handbook of Chemistry, 13th Ed; McGraw-Hill: New York, 1985. 12 Adler, T.; Albert, A. Diazaindenes (“Azaindoles”). Part 1. Ionization Constants and Spectra. J. Chem. Soc., 1960, 1794-1797. 13 Walba, H.; Isensee, R. Acidity Constants of Some Arylimidizaoles and Their Cations. J. Org. Chem. 1960, 26, 2789-2791. 14 Katritzky, A. R.; Ramsden, C. A.; Joule, J. A.; Zhdankin, V. V. Handbook of Heterocyclic Chemistry, 3rd Ed; Elsevier: New York, 2010. 15 Eicher, T.; Hauptmann, S. The Chemistry of Heterocycles: Structures, Reactions, Synthesis, and Applications, 2nd Ed; Wiley-VCH: Weinheim, 2003. 16 Sondheimer, E.; Covitz, F.; Marquisee, M. J. Association of Naturally Occurring Compounds, the Chlorogenic Acid-Caffeine Complex. Arch. Bioch. Biophys., 1961, 93, 63-71. 17 Spiller, G. Caffeine; CRC Press: New York, 1998. 18 Albert, A.; Serjeant, E. P. Ionization Constants of Acids and Bases: A Laboratory Manual, Wiley&Sons: New York, 1962. 19 Dawson, R.; Elliott, D.; Elliott, W.; Jones, K. Data for Biochemical Research, 3rd Ed; Oxford Science: New York, 1986. 20 Milletto, F.; Storchi, L.; Goracci, L.; Bendels, S.; Wagner, B.; Kansey, M.; Cruciani, G. Extending pKa Prediction Accuracy: High-throughput pKa Measurements to Understand pKa Modulation of New Chemical Series. European Journal of Medicinal Chemistry, 2010, 45, 4270-4279. 21Fox, J. J.; Wempen, I.; Hampton, A.; Doerr, I. L. Thiation of Nucleosides. I. Synthesis of 2- Amino-6-mercapto-9-β-D-ribofuranosylpurine (“Thioguanosine”) and Related Purine Nucleosides. JACS, 1958, 80, 1669-1675. 22Paragi, G.; Kovacs, L.; Kupihar, Z.; Szolomajer, J.; Penke, B.; Guerra, C.; Bickelhaupt, F. Neutral and Positively Charged Purine Tetramer Structures: a Computational Study of Xanthine and Uric Acid Derivatives. New J. Chem., 2011, 35, 119-126.

88

23Kulikowska, E.; Kierdaszuk, B.; Shugar, D. Xanthine, Xanthosine and its Nucleotides: Solution Structures of Neutral and Ionic Forms, and Relevance to Substrate Properties in Various Enzyme Systems and Metabolic Pathways. Acta Biochimica Polonica, 2004, 51, 493-531. 24Seybold, P. G.; Kreye, W. C. Theoretical Estimation of theAcidities of Alcohols and Azoles in Gas Phase, DMSO, and Water. International Journal of Quantum Chemistry, 2012, 112, 3769- 3776. 25 Hunter, N.; Seybold, P. G. Theoretical Estimation of the Aqueous pKas of Thiols. MolPhys, 2014, 112, 340-348. 26 Koyama, T.; Wakisaka, A. Molecular Self-Assembly Composed of Aromatic Hydrogen-Bond Donor-Acceptor Complexes. J Chem. Soc., Faraday Trans., 1997, 21, 3813-3817. 27 Dufour, N.; Dartiguenave, Y.; Dartiguenave, M.;Lebuis, A.-M.; Bélanger-Gariépy, F.; Beauchamp, A. Crystal Structures of 7-Azaindole, an Unusual Hydrogen-Bonded Tetramer, and of Two of its Methylmercury(II) Complexes. Can. J. Chem., 1990, 68, 193-201. 28 Pogozhev, D.; Baudron, S. A.; Hosseini, M. W. From Insertion of Rhodium Acetate Paddlewheels into Functionalized 7-Azaindole Hydrogen-Bonded Dimers to infinite Architectures. Dalton Trans., 2011, 40, 7403-7411.

89