8-Alkyl Adenines and Their Nucleoside Derivatives by Maha Abouelghit A dissertation submitted to the Graduate Faculty of Auburn University in partial fulfillment of the requirements for the Degree of Doctor of Philosophy Auburn, Alabama August 6, 2011 Copyright 2011 by Maha Abouelghit Approved by Stewart W. Schneller, Chair, Professor of Chemistry and Biochemistry Edward J. Parish, Professor of Chemistry and Biochemistry Susanne Striegler, Associate Professor of Chemistry and Biochemistry Orlando Acevedo, Assistant Professor of Chemistry and Biochemistry ii Abstract S-adenosylhomocysteine hydrolase has become an important target in the design of antiviral drugs. It is the only known enzyme in eukaryotes that is responsible for the catabolism of S- adenosylhomocysteine into adenosine and L-homocysteine. S-adenosylhomocysteine is both a product and potent feedback inhibitor of methylation reactions involving S-adenosylmethionine as a methyl donor. By inhibition of the methylation reactions responsible for formation of the mRNA 5?-cap required by most viruses for proper translation, the viral replication process is inhibited as well. Two carbocyclic nucleosides: aristeromycin and neplanocin have been shown to possess broad spectrum antiviral properties with severe toxicity. They have been used as templates for design of antiviral agents that retain the antiviral properties without the toxic side effects. Purine ribofuranosyl nucleosides substituted in the 8-position have shown usefulness as probes for the study of nucleoside structural conformations. Based on the nature of the purine 8- susbtituent, the nucleoside can exist in a syn/anti equilibrium that shifts from a preferred anti arrangement towards the syn conformation that may become purely syn depending on the 8- substituent. In this research, a group of purine carbocyclic nucleosides with variable alkyl substituents were studied theoretically and synthetically with the intent of exploring their conformational parameters that could lead to correlation with biological activity. A theoretical iii methodology, the density functional theory with various basis sets, was used to predict the structural parameters: phase angle of pseudorotation, the degree of pucker and the glycosyl torsion angle of the carbocyclic nucleosides. The results were compared to experimental data either obtained from the literature or generated within this research. The investigation was aimed at identifying the conformations that attribute to the observed experimental data as well as assessing the contribution of each conformation to the overall structure. The research also focused on studying the relationship between syn/anti conformation and the pseudorotation of the cyclopentyl moiety in carbocyclic nucleosides and extrapolating the findings from the 8-alkyl substituted carbocyclic nucleoside. A final aspect was providing the basis for uncovering a correlation between the syn/anti conformation and the biological activity of purine derived carbocyclic nucleosides, which was beyond the scope of this dissertation research. To avail authentic carbocyclic nucleoside samples for this study, synthetic procedures were developed that called upon pyrimidine and cyclopentane/cyclopentene precursors. In that effort the Mitsunobu and Luche reactions and the Grubbs metathesis procedures were employed. The required compounds were obtained in overall reasonable yields and their structures verified by x- ray crystallography and a thorough NMR analysis using contemporary, multi-dimensional methods. iv Acknowledgments For this work to be accomplished, a lot of people put in a lot of their time and knowledge to help me reach this point of my life. Allowing me to join his research group in 2006, Dr. S.W. Schneller has opened a door for me into an enriched life with a lot of expectations that there will be something new to learn each and every passing day. His kind help and support combined with his immense knowledge about science as well as his philosophy about life created a journey for me that entailed a lot of learning that will shape the rest of my life. And for all that and more, he has my deepest gratitude. Stepping into this experience, allowed me the company of a lot of people whom without thinking and with a lot of energy gave me a lot of their help. I have to express my utmost gratitude to Dr. Michael McKee and his student Tae Bum Lee. With their help and guidance, a large proportion of this work has been accomplished. I would like to express my gratitude to Drs. Susanne Striegler, Holly Ellis, Orlando Acevedo, and Edward Parish for helping me with this project. The knowledge I received in the classrooms of the Chemistry and Biochemistry Department as well as the Harrison School of Pharmacy has been immense and I thank each and everyone who contributed to this learning experience mainly Dr. Randall Clark to whom I owe a lot both on a professional as well as personal levels. I would like to thank Dr. Smita Mohanty for her help with the NMR experiments run on the 600 MHz machine. I would also like to thank Dr. John Gordon for his help with the X-ray crystallography. I would like also to thank Drs Wei Ye, Chong Liu, Olena Musiienko and Qi Chen for their help v with my lab experience. I would also like to thank Volodymyr Musiienko and Sherif Hamad for lot of help with my lab techniques. The knowledge I received in the NMR lab and the Mass Spectroscopy labs under the guidance of Drs. Michael Meadows, and Yonnie Wu is greatly appreciated. I extend my gratitude to the Department of Chemistry and Biochemistry and the National Institutes of Health for their financial support. Finally without the love and encouragement of my family and my husband, Tamer Awad, I would not be writing this today. vi Table of Contents Abstract ......................................................................................................................................... ii Acknowledgments ....................................................................................................................... iv List of Tables .............................................................................................................................. vii List of Figures ............................................................................................................................ viii List of Schemes ............................................................................................................................. x List of Graphs ............................................................................................................................. xii Introduction ................................................................................................................................... 1 Biological Background ............................................................................................... 1 Computational Background ...................................................................................... 16 Rational for Research ................................................................................................ 22 Chapter 1 Theoretical Investigation of Carbocyclic Nucleosides .............................................. 25 Chapter 2 Synthesis of Required Precursors .............................................................................. 43 Synthesis of the Carbocycle ...................................................................................... 43 Synthesis of the Heterocyclic Nucleobase ................................................................. 50 Chapter 3 Synthesis of 8-Alkylaristeromycin ............................................................................. 61 Chapter 4 Synthesis of 8-Susbtituted 4?-Norneplanocin Analogues ........................................... 68 Chapter 5 Analysis of 8-Ethylaristeromycin (2) ......................................................................... 74 Conclusion ................................................................................................................................ 84 Experimental Details ................................................................................................................. 86 vii References ............................................................................................................................... 106 Appendix ................................................................................................................................ 114 viii List of Tables Table 1 Results of initial geometry optimization of 4 compounds ............................................. 28 Table 2 Summary of the molecules from B3LYP/3-21G ........................................................... 29 Table 3 PROSIT output for (A) after single point energy in solution ........................................ 30 Table 4 Results shown for AA1 and AA2 based on single point energy in solution ................. 32 Table 5 Results shown for BB1, BB3 and BB4 based on single point energy in solution ......... 33 Table 6 Results shown for CC1 and CC4 based on single point energy in solution .................. 34 Table 7 Results shown for DD3 and DD4 based on single point energy in solution ................. 35 Table 8 A Comparison of experimental and theoretical spin-spin coupling constants for (A) . 37 Table 9 A Comparison of experimental and theoretical spin-spin coupling constants for (B) ... 38 Table 10 A Comparison of experimental and theoretical spin-spin coupling constants for (C) 39 Table 11 A Comparison of experimental and theoretical spin-spin coupling constants for (D) 40 Table 12 Initial conformers chosen for pseudorotation study of (F) .......................................... 74 Table 13 Conformations based on single point energy calculations in solution of (F) .............. 76 Table 14 A Comparison of experimental and theoretical spin-spin coupling constants for (F) 80 ix List of Figures Figure 1 Nucleotides vs. nucleosides ............................................................................................ 1 Figure 2 Naturally occurring nucleosides ..................................................................................... 2 Figure 3 Stereochemistry of (A) ?-nucleoside, (B) ?-nucleoside ................................................. 3 Figure 4 Carbocyclic Nucleosides as a modification of ribofuranoside nucleosides ................... 3 Figure 5 Aristeromycin and Neplanocin A ................................................................................... 5 Figure 6 Structure of S-adenosyl-L-methionine (AdoMet) .......................................................... 6 Figure 7 AdoMet metabolism ....................................................................................................... 7 Figure 8 First generation AdoHcy hydrolase inhibitors ............................................................ 12 Figure 9 Second generation AdoHcy hydrolase inhibitors ......................................................... 14 Figure 10 Purine Nucleoside numbering .................................................................................... 17 Figure 11 Pseudorotation cycle ................................................................................................... 18 Figure 12 Adenosine showing syn and anti conformations ........................................................ 20 Figure 13 Classical staggered rotational isomers around C4?-C5? .............................................. 21 Figure 14 Targets 1, 2 ................................................................................................................. 23 Figure 15 DHCA Analogues ....................................................................................................... 24 Figure 16 Compounds involved in the study .............................................................................. 25 Figure 17 Starting six geometries for all 4 compounds, example shown is compound A .......... 27 x Figure 18 Grubbs catalysts.......................................................................................................... 47 Figure 19 Structure of 32 from X-ray analysis. .......................................................................... 60 Figure 20 Structure of 1 from X-ray analysis. ............................................................................ 66 Figure 21 Structure of 7 from X-ray analysis ............................................................................. 73 Figure 22 Initial optimization of 8-ethylaristeromycin (F) ......................................................... 75 Figure 23 Final geometries for F after NMR calculations .......................................................... 77 Figure 24 Portion of 1HNMR spectrum of F in D2O showing overlapping H2? ......................... 78 Figure 25 Portion of 1HNMR spectrum of F in D2O (327 oK) showing resolved peaks ............ 79 Figure 26 Structure of 2 from X-ray analysis ............................................................................. 81 Figure 27 Summary of correlation study between theoretical and experimental data ................ 82 xi List of Schemes Scheme 1 Effect of phosphorylases on natural and modified nucleosides ................................... 4 Scheme 2 Toxicity of Ari and NpcA due to phosphorylation ...................................................... 5 Scheme 3 Structure of 5?-cap of mRNA ....................................................................................... 8 Scheme 4 Formation of AdoHcy as a byproduct of methylation reaction .................................... 9 Scheme 5 Mechanism of action of AdoHcy hydrolase ............................................................... 11 Scheme 6 Mechanism of Type I mechanism-based AdoHcy hydrolase inhibitors .................... 13 Scheme 7 Mechanism Type II mechanism-based AdoHcy hydrolase inhibitors ....................... 14 Scheme 8 Convergent retrosynthesis towards target compounds and their precursors .............. 44 Scheme 9 The synthesis of cyclopentenone 15 from D-ribose ................................................... 46 Scheme 10 Reaction mechanism of RCM .................................................................................. 48 Scheme 11 Synthesis of precursor 17 ......................................................................................... 49 Scheme 12 Synthesis of precursor 18 ......................................................................................... 49 Scheme 13 Purine synthesis based on various reagents.............................................................. 51 Scheme 14 Synthesis of 6-chloro-8-methylpurine 21 ................................................................. 52 Scheme 15 Amination of 4,6-dichloropyrimidin-5-amine.......................................................... 53 Scheme 16 Synthesis of 28 using carboxylic acid ...................................................................... 54 Scheme 17 Synthesis of 28 using acid chloride .......................................................................... 54 Scheme 18 Synthesis of 28 using orthoesters ............................................................................. 55 Scheme 19 Intermediates 26 and 27 from the orthoester reaction .............................................. 56 xii Scheme 20 Proposed reaction mechanism for the formation of 28 from intermediates 26, 27 .. 57 Scheme 21 Synthesis of 29 ......................................................................................................... 58 Scheme 22 Proposed mechanism for formation of 45 ................................................................ 59 Scheme 23 Synthesis of 6-chloro-8-t-butylpurine using acid chloride ....................................... 59 Scheme 24 Synthesis of 1 based on unpublished results using the linear route ......................... 62 Scheme 25 Proposed mechanism for side products .................................................................... 63 Scheme 26 Retro synthesis of 8-alkyl substituted Aristeromycin .............................................. 64 Scheme 27 Synthesis of compound 1 by the convergent route .................................................. 65 Scheme 28 Synthesis of compound 2 by the convergent route .................................................. 67 Scheme 29 Retrosynthesis of targets 3, 4, 5, 6 and 7 using the convergent route ...................... 69 Scheme 30 Synthesis of targets 3, 4, 5 and 6 .............................................................................. 71 Scheme 31 Convergent synthesis of target compound 7 ............................................................ 72 xiii List of Graphs Graph 1 Relative energy in solution for A .................................................................................. 32 Graph 2 Relative energy in solution for B .................................................................................. 33 Graph 3 Relative energy in solution for C .................................................................................. 34 Graph 4 Relative energy in solution for D .................................................................................. 35 Graph 5 Relative energy in solution for F .................................................................................. 76 1 Introduction Biological Background What are Nucleosides? Nucleic acids, elements of heredity and agents of genetic information transfer are formed from linear polymers of nucleotides. Nucleotides are biological molecules that are composed of a heterocyclic base, a five-carbon sugar (pentose) and a phosphate group (Figure 1).1 The two main classes of nucleic acids are deoxyribonucleic acid (DNA) and ribonucleic acid (RNA). The sugar moiety for the former is 2-deoxyribose and for the latter is ribose. Nucleosides form upon attachment of the sugar to a heterocyclic nucleobase through a ?-glycosidic linkage. Figure 1. Nucleotides vs. nucleosides 2 The bases are planar aromatic heterocyclic molecules that are divided into two groups: the pyrimidine bases thymine and cytosine and in DNA the purine bases adenine and guanine with thymine being replaced with Uracil in RNA2 (Figure 2). Figure 2. Naturally occurring nucleosides. The bond between the sugar and the base is known as the glycosidic bond and its stereochemistry is ? for the natural nucleic acids, which means that the base and the 4?- hydroxymethyl group are above the plane of the sugar2 i.e. cis to each other (Figure 3). If the base is trans to the 4?-hydroxymethyl group, the nucleoside is considered ?. Naturally occurring nucleosides and their analogues have been widely studied for their antiviral and anticancer properties.3 Purine nucleosides, in particular are of great interest chemically and pharmacologically, especially those compounds that are structurally related to adenosine (Ado) and possess 3 biological activity as antiviral, antitumor agents, and enzyme inhibitors.4 Adenosine is an intracellular modulator that mediates a wide variety of physiological functions including vasodilatation, vasoconstriction in the kidney, inhibition of platelet aggregation, inhibition of insulin release, and potentiating of histamine release.5 Figure 3. Stereochemistry of (A) ?-nucleoside, (B) ?-nucleoside What are Carbocyclic Nucleosides? Figure 4. Carbocyclic nucleosides as a modification of ribofuranoside nucleosides Carbocyclic nucleosides are analogues of naturally occurring nucleosides in which the furanose ring oxygen has been replaced with a methylene group (Figure 4). This modification transforms the glycosidic bond from an unstable hemiaminal to a more stable tertiary amine.6 4 This gives carbocyclic nucleosides potent metabolic stability due to resistance to phosphorylases and hydrolases that cleave the glycosidic bond of their natural counterparts (Scheme 1.).7 These nucleoside analogues display a wide range of biological activities.8 The most important of which are being antiviral, antiparasitic, fungicidal and anti-tumor activities.9 Their activity in part is due to being recognized by the same enzymes that recognize the natural nucleosides. Scheme 1. Effect of phosphorylases on natural and modified nucleosides Two carbocyclic nucleosides stand out; aristeromycin (Ari) and neplanocin A (NpcA) (Figure 5). Aristeromycin was first described by Shealy and Clayton10 in 1966, two years before it was isolated as a metabolite from Streptomyces citricolor.11 NpcA on the other hand was isolated from Ampulariella regularis in 1981.12 Both compounds possess excellent chemical and metabolic stability of the glycosidic bond as well as broad-spectrum antiviral activity. In addition, both demonstrate potent inhibitory activity of S-adenosyl-L-homocysteine hydrolase (AdoHcy hydrolase). Cools and De Clerq13 have shown a correlation between their antiviral potency and their inhibitory effects towards AdoHcy hydrolase. AdoHcy hydrolase is an enzyme involved in the regulation of the biomethylation reactions dependent on S-adenosylmethionine, which, in turn, is crucial to viral replication.14,15 Severe toxicity has banned these two compounds from clinical use; however, it opened the door for a huge research area into the reasons behind their toxicity and how to modulate it for the sake of better antiviral analogues. 5 Figure 5. Aristeromycin and Neplanocin A The cytotoxicity of Ari16 and NpcA17 is attributed to the fact that they are both substrates to adenosine kinase (Scheme 2). Interaction with kinases in the cell leads to the formation of the 5?- triphosphate analogues of the naturally occurring adenosine thus interfering with biological processes in which ATP is involved.18 Another aspect concerning NpcA9 toxicity is that it is converted to neplanocylmethionine 19 (the AdoMet analogue) via interaction with methionine which is believed to be the mechanism by which NpcA exerts its cytotoxicity. Scheme 2. Toxicity of Ari and NpcA due to phosphorylation. 6 What is SAM/AdoMet? S-adenosyl-L-methionine (SAM, AdoMet) is one of the most versatile molecules in nature, is found in almost all phyla of living things, and is crucial for the existence of living organisms (Figure 6). Figure 6. Structure of S-adenosyl-L-methionine (AdoMet). First, identified by Cantoni20 as a product formed from L-methionine and adenosine triphosphate (ATP) under catalysis of SAM synthetase, AdoMet is involved in a number of metabolic activities.21 One of the most important processes is that AdoMet serves as a cofactor methyl donor for numerous methyltransferases that catalyze the methylation of various molecules. These include macromolecules like proteins, RNA, DNA and polysaccharides as well as small molecules like phospholipids, histamines and catecholamines.19 The methylation of mRNA, has drawn a lot of attention among other methylation reactions, could happen either at the 2?-hydroxyl group of the ribose moiety or at the N-7 of the guanosine residue. Figure 7 shows the involvement of AdoMet in various biological processes. 7 M e t hi oni ne S - A de nos y l - m e t hi oni ne S - A de nos y l - hom oc y s t e i ne H om oc y s t e i ne T H F 5 - Me - T H F M e t hi oni ne S y nt ha s e Di e t B e t a i ne D i m e t hy l g l y c i ne Me t h y l A c c e p t o r Me t h y l a t e d A c c e p t o r A d e n o s i n e In o s i n e A T P AD AK Cy s t a t h i o n i n e Cy s t e i n e Cy s t a t h i o n i n e ? - S y n t h a s e V i t a m i n B6 T ra n ss u l f u ra t i o n p a t h w a y S A H H Figure 7 AdoMet Metabolism. One of the most well-defined methylation reactions in RNA is the mRNA methylation. A variety of eukaryotic cells and viruses require enzymatic methylation of the 5?-terminal residue of their mRNA to form the 5?-methylated cap structure essential for proper translation to proteins and, subsequently, viral replication.22 In this 5?-cap (Scheme 3) a guanosine residue, transferred from guanosine triphosphate (GTP), blocks the 5?-terminal of the penultimate base of mRNA through a unique 5?-5? triphosphate linkage. Subsequently, the N-7 of the guanosine residue as well as the ribose moiety of one or more of the penultimate bases are methylated via AdoMet-dependent methylation enzymatic reactions. 8 Scheme 3. Structure of 5?-cap of mRNA. The enzymes involved in the cap formation include23: (i) RNA triphosphatase: cleavage of the 5? triphosphate terminus in a transcribed RNA to a diphosphate-terminated RNA. (ii) mRNA guanylyltransferase: capping of the diphosphate terminus with GMP. (iii) mRNA (guanine-7-)methyltransferases: methylation at the N-7 position of guanine (methyl group from AdoMet). (iv) 2?-O-Methyltransferase: methylation of the 2?-O-ribose moiety of the penultimate nucleoside.22 9 The 5?-cap structure of mRNA offers stability against phosphatases and ribonucleases hence imparting stability to the mRNA in the cell.22 It has also been recognized as playing a significant role in the interaction between mRNA and ribosomal RNA or its protein components during initiation of protein synthesis and for the promotion of splicing.14 Why has S-Adenosylhomocysteine (AdoHcy) Hydrolase (AdoHcy hydrolase; EC 3.3.1.1) become an intriguing target for antiviral drug design?14 After AdoMet participates in a methylation reaction in which an activated methyl group is being transferred to a recipient biological molecule in the cell, the product is the formation of S- adenosylhomocysteine (AdoHcy). (Scheme 4) Scheme 4. Formation of AdoHcy as a byproduct of methylation reactions AdoHcy hydrolase is the only known enzyme that catalyzes the reversible conversion of AdoHcy to adenosine (Ado) and homocysteine (Hcy). In vitro, the equilibrium of this reaction lies towards the synthetic pathway, although the reaction is pulled towards the hydrolytic direction due to efficient enzymatic removal of adenosine (Ado) and homocysteine (Hcy) under the physiological conditions.19 AdoHcy is a potent competitive feedback inhibitor of all AdoMet- 10 dependent methyltransferases. On the hydrolytic side, Ado is being converted via the purine degradation pathway into inosine by adenosine deaminase (ADA) or to adenosine triphosphate (ATP) by adenosine kinases (AK). On the other hand, Hcy enters the remethylation and transsulfuration pathways in a 50:50 ratio.24 Remethylation of Hcy occurs via two routes: irreversible methylation via betaine (trimethyl glycine) in the liver and kidney25 and the other involves a transfer of a methyl group to Hcy from 5-methyltetrahydrofolate via methionine synthase. In the transsulfuration pathway Hcy is being converted to cystathionine via cystathionine-?-synthase. Figure 6 shows an overall outlay for the position of AdoHcy hydrolase in the methionine, transmethylation and transsulfuration metabolic pathways. The reaction mechanism of AdoHcy hydrolase has been reported by Palmer and Abeles26,27 (Scheme 5). The first step is the oxidation of the 3?-hydroxyl group in the enzyme-bound adenosine complex to a ketone with the concomitant reduction of the tightly bound NAD+ to NADH. This step results in the activation of 4?-H that is now more acidic and removed via a basic residue in the active site. This is followed by loss of homocysteine. Water adds in a Michael type fashion to the ?, ?-unsaturated ketone intermediate. NADH then reduces the 3?- ketone back to a hydroxyl group generating Ado, which then dissociates from the enzyme. Scheme 5 shows both the hydrolytic and synthetic pathways involved in the catalytic cycle of AdoHcy hydrolase with B being the residue responsible for accepting and returning the proton at the 4? position. There are two distinct conformations of AdoHcy hydrolase which are the open and closed forms.28 In the resting state, the catalytic and the cofactor binding domains are opened to allow exposure of the active site to the environment. Upon ligand binding, the enzyme transforms to the closed conformation to allow for catalysis. Both conformations are responsible for regulation of the steps involved in the catalytic cycle. 11 Scheme 5. Mechanism of action of AdoHcy hydrolase 29 12 AdoHcy inhibitors can be divided into three groups based on the history of their discovery. Across the evolution of these inhibitors, the common goal was to develop more potent and more specific inhibitors.30 (A) First generation inhibitors: include both naturally occurring carbocyclic analogues and synthetic cyclic or acyclic Ado analogues. Figure 8. First Generation AdoHcy hydrolase inhibitors Some inhibitors inactivate AdoHcy Hydrolase irreversibly and become tightly bound to the enzyme like Ado dialdehyde and NpcA, while others are reversible competitive inhibitors of the enzyme like Ari. NpcA inactivation belongs to Type-I-mechanism-based inhibitors14 where neplanocin inhibits AdoHcy hydrolase by a cofactor-depletion mechanism. After binding to the active site, the inhibitor is oxidized at the C3? position to give the 3?-keto along with the concomitant reduction of the enzyme- bound NAD+ leading to consumption of NAD+ by the 13 inhibitor and this leads to inactivation of the enzyme. These are exemplified by NpcA. A common feature of this generation is their cellular toxicity (Scheme 6).30 Scheme 6. Mechanism of Type I mechanism-based AdoHcy hydrolase inhibitors (B) Second Generation Inhibitors: These were designed with an attempt to be more specific along with fewer capabilities to serve as substrates for adenosine kinase and adenosine deaminase; which is the source of their toxicity. The design was based on two directions: the first was to replace the adenine moiety in NpcA and Ari where with 3-deazaadenine (C3-NpcA, C3-Ari) and the second approach used to remove the 4?-hydroxymethyl group thus prevent 5?-phosphorylation via Ado kinase and deamination via Ado deaminase and these efforts resulted in DHCA and C3-DHCA and their saturated counterparts ( Figure 9).31,32 These compounds exhibit broad-spectrum antiviral activity while their toxicity is lower than the first generation and they are not substrates for either Ado deaminase or Ado kinase. The deaza nucleosides are reversible competitive inhibitors of AdoHcy hydrolase while their truncated counterparts irreversibly inactivate the enzyme via a type I mechanism. 14 Figure 9. Second generation AdoHcy hydrolase inhibitors. Another set of compounds was designed as part of the second generation as type II- mechanism based inhibitors of AdoHcy Hydrolase. These inhibitors are able to bind covalently at the active site of the enzyme and cause permanent irreversible inhibition in addition to causing NAD+ depletion. An example is (Z) 4?, 5?-didehydro-5?-deoxy-5?-fluoroAdo. The type ? II mechanism of inhibition is shown in Scheme 7. 14,33 Scheme. 7. Mechanism of Type II mechanism-based AdoHcy hydrolase inhibitors14 (C) Third Generation Inhibitors: This generation has been designed as a prodrug that would be converted within the active site into a potent drug33, and is based on the hydrolytic activity of AdoHcy hydrolase. 15 In conclusion, inhibition of AdoHcy hydrolase results in the accumulation of AdoHcy, which is both a product and a feedback inhibitor of AdoMet-dependent methylation reactions. As these methylations are important in securing the formation of the 5?-cap in viral mRNA, this capping process is being disrupted and, consequently, the viral replication is inhibited.14,34 AdoHcy hydrolase inhibitors are potent broad spectrum antiviral agents inhibiting an array of (-) RNA (measles, vesicular stomatitis, and parainfluenza), double stranded (?) RNA (i.e., reo and rotavirus) viruses and DNA poxviruses (vaccinia and African swine fever).14 They don?t exhibit activity towards (+) RNA or HIV. 16 Computational Background: To allow for us to understand the biological function of nucleotides and nucleic acids, we need to study their conformational characteristics.35 This can be done with X-ray crystallography, spectroscopic techniques, mainly NMR, and theoretical studies. All of these methods will allow for a better understanding and characterization of the genetic materials. In addition, a ligand?s activity might be correlated with its preferred structure in solution.36 Research directed towards the determination of the stereochemistry of molecules in solution plays a crucial role in the understanding of interactions in both the chemical and biochemical reactions.37 Defining the terms used in describing the geometry of nucleotides and nucleic acids is fundamental in understanding and explaining the results obtained via the various techniques used in their structural elucidation and giving way to a more comprehensive understanding of their properties. The nucleotide is the basic motif in DNA and RNA. It is composed of the heterocyclic base, the sugar and the phosphate group (Figure 1).35,38 The conformational details of their structure has been well defined by the torsion angles ?,?,?,?,? and ? in the phosphate backbone, ?0-?4 in the furanose ring and ? for the glycosylic bond. However, due to the interdependency of those torsional angles, the shape of nucleotides (sides) can be described in terms of four parameters: (1) The sugar pucker. (2) The syn-anti conformation of the glycosylic bond (?). (3) The orientation of the C4?-C5? (?). (4) The shape of the phosphate ester bond. 17 Figure 10. Purine Nucleoside numbering The focus will be on the first three parameters because of their relevance to our study of purine nucleosides (Figure 10). (1) The Sugar Puckering Modes: The Pseudorotation Cycle: Planar ribose is energetically unfavorable since in such arrangement all the torsion angles would be 0o which means substituents on carbons are all eclipsed. The furanose ring twists out of plane to minimize these non-bonded interactions between substituents. Furanose can be puckered into an envelope (E) form with one atom out of plane and four atoms in a plane; or in a twist form (T) with two adjacent atoms displaced on opposite sides from the other in-plane three atoms (Figure 11). If the atom displaced from the plane is on the same side as the C5?, it is referred to as endo, and if it is on the other side it is referred to as exo. Endo and exo notations 18 are translated into superscripts and subscripts, respectively. If the endo displacement of C2? is greater than that of C3?, it is called C2?-endo. This corresponds to a South conformer along the pseudorotation cycle, the analytical representation initially introduced for cyclopentane,39 extended to substituted furanose rings40 and applied towards the study of carbocyclic nucleosides later on.41 Altona has introduced two useful parameters to describe the concept of pseudorotation in naturally occurring nucleosides42,43 the phase angle of pseudorotation ?P? (0- 360o) and the degree of pucker ??max?. No r t h , C 2' - e x o - C 3' - e n do S o ut h , C 2' - e n do - C 3' - e x o ? max 10 20 30 40 50 Figure 11. Pseudorotation cycle 19 By convention a phase angle P=0o corresponds to an absolute North (N) conformation having a symmetrical twist (3?T2?), while a phase angle P=180o corresponds to the South (S) antipode with a symmetrical twist (3?T2?). The value of P depends on five torsion angles (?0-?4) of the furanose ring (Figure 10). The relationship between these five torsion angles can be described by a simple cosine function35: ?j = ?max cos (P + j ?) where j = 0-4, ? = 144o (eq. 1) And from this equation, a useful formula can be derived for the calculation of P: tan P = [(?4+ ?1)- (?3+ ?0) ]/2 .?2 .( sin 36+sin 72) (eq. 2) A close inspection of the available crystallographic data for individual nucleosides reveals that the puckering modes of the furanose ring cluster in two antipodal domains around a C3?- endo (3?E, N) and a C2?-endo (2?E, S) envelope conformations. However, in solution, the sugar puckering fluctuates rapidly between these two conformational extremes. In the N conformation, the P values are found in a range between 342o and 18o [2?E?3?T2??3?E (C3?-endo)], whereas in the antipodal S conformation the range is between 162o and 198o [2?E?3?T2??3?E (C2?-endo)].42,44 The average magnitude of equilibrium can be estimated via the 3J NMR coupling constants 3J1?2? and 3J3?4? (Equation 3, 4, 545)42,46 which gives the relative population of N/S conformations. The equilibrium is influenced mainly by the preference of the substituents at the C2? and C3? for axial orientation as well as the orientation of the base (usually C3?-endo is preferred for anti and C2?-endo prefers syn).44 J1?2?=9.3(1-XN) = 9.3XS (eq. 3) J3?4? = 9.3XN (eq. 4) % C2?-endo = [J1?2?/ (J1?2?+J3?4?)] x100 (eq. 5) 20 (2) The syn-anti conformation of the glycosyl bond (?): Although the sugar pucker is the main conformational determinant in nucleosides, an important associated parameter is the glycosyl torsion angle ? (defined by X-C1?-N9-C4) (Figure 10). The value of ? determines the syn and anti disposition of the base relative to the sugar moiety.44 In the syn conformer, the N-3 of the purine lies above the plane of the sugar moiety, while in the anti conformer the H-8 of purine lies in the same position (Figure 12). According to the currently accepted IUPAC definition of the angle ?, the torsion angle range of 180o ? 90o is generally referred to as anti while the range 0o ? 90o is referred to as syn.44,47 The syn-anti conformation of nucleosides is one of the most important conformational aspects used in the study of nucleoside-enzyme interactions.48 Figure 12. Adenosine showing Syn and Anti conformations. In solution, the energy barrier between different conformations is around (25 kJ/mol)36,49and normally the different conformations can be described by two-state models (syn-anti; N/S-type sugar pucker).44 Stolarski et al 50 demonstrated that the chemical shifts (?) of H2? in purine nucleosides can be used as an indicator of the syn/anti orientation. Typical values of ? (H2?) in DMSO are 5.2 ppm for syn and 4.2 ppm for anti conformations. The big difference between the 21 two values can be explained in terms of the influence exerted by N-3 on the H2? in the syn conformation. Adenosine has a reported value for H2? of 4.62 ppm which demonstrates that it exists in equilibrium between the syn and anti conformations.50 (3) The orientation of the C4?-C5? (?): The conformation of the C4?-C5? determines the orientation of the hydroxyl group in nucleosides or the 5?-phosphate in nucleotides relative to the sugar ring. There are three favored conformers shown in Figure 13. In purines +sc and ap are equally populated. Figure 13. Classical staggered rotational isomers around C4?-C5?.46,51 Rosemeyer et al 52 used equations based on coupling between H4? and H5? as well as H5?? to calculate the population of each conformer as shown in Equations (7,8 and 9). % ? g+ = {1.46-(3J4?5? +3J4?5??) /8.9} x 100 (eq. 7) % ? t = {(3J4?5?? /8.9) ? 0.23} x 100 (eq. 8) % ? g- = {(3J4?5? /8.9) ? 0.23} x 100 (eq. 9) 22 Rational for Research (A) 8-Alkylaristeromycin The search for useful therapeutic agents is controlled by various factors. One aspect that makes carbocyclic nucleosides appealing is that they are highly resistant to phosphorylases, which cleaves the glycosidic bond, while still good substrates for cellular kinases.53 The broad spectrum activity for both Ari and NpcA coupled with their toxicity, resulted into modification of those two important nucleosides to retain antiviral activity and lower toxicity has been done. An important consideration in nucleoside chemistry is the structural orientation of the base and sugar and its effect on the activity of nucleoside as well as the information it can provide for a better design of structural analogues. Purine nucleosides with no substitutions at the 8 position preferentially exist in the anti conformation.44 Analogues with 8-substitution with bromo, fluoro, or methyl substituents54 have been studied via circular dichroism and NMR spectroscopy55 and have been assigned as syn conformations. It was confirmed specroscopically that due to the rapid equilibrium between syn-anti that those C8-substituted nucleosides still afford access to the anti conformer in solution,50 which would allow binding at the enzyme active site.56 For carbocyclic nucleosides, especially aristeromycin, the 8-alkyl substitution has not been well explored. However, various studies regarding adenosine and its analogues have been published,57,58 with a variety of substituents in the 8 position. Alkylation of the 2-, 6, and 8- positions of adenosine and its deoxy analogues was accomplished and their biological activity was reported. Of the 8-substituted adenosine analogues, 8-methyladenosine exhibits a 50-60 % syn50 conformations and it shows high selective inhibition against vaccinia virus.57 They have also been studied theoretically for structural conformation.59 23 So in an attempt to gain more insight into the effect of 8-alkylation on the structural and chemical aspects of carbocyclic nucleoside, mainly aristeromycin, two targets (Figure 14) were synthesized, 8-methylaristeromycin (1) and 8-ethylaristeromycin (2) within this part of the research. Figure 14. Targets 1, 2. 24 (B) Neplanocin Analogues As mentioned before NpcA is an antiviral agent effective against a broad range of viruses, in particular (-) RNA strand viruses (vesicular stomatitis, parainfluenza, and measles), 34,60 and whose effect is mediated by its inhibition of AdoHcy hydrolase, whilst its toxicity is mediated by its phosphorylation via adenosine kinase and subsequent transformation into S- neplanocylmethionine.61 Borchardt et al.,32,60 designed truncated analogues of NpcA that retained the inhibitory activity towards AdoHcy hydrolase (antiviral activity) while lacking substrate properties towards cellular adenosine kinase and deaminase (reduced cellular toxicity) via removal of the 4? hydroxymethyl group hence the term 4?-norneplanocin. Figure 15. DHCA analogues As part of the synthetic research, five target compounds were designed and synthesized based on the DHCA framework and are shown in Figure 15. 25 Chapter 1. Theoretical Investigation of Carbocyclic Nucleosides The computational study involved five compounds: (A) Adenosine (B) Aristeromycin (C) 8-methylaristeromycin (Target 1) (D) 3-deazaaristeromycin Figure 16. Compounds involved in the study Methods: All density functional calculations were carried out with the Gaussian program62 using the computational resources available through the Alabama Supercomputer Authority. Full geometry 26 optimizations were carried out with the Becke three-parameter exchange functional63 with the nonlocal correlation functional of Lee, Yang and Parr and the 6-31G(d) basis set.64 The PCM continuum solvation model65 was utilized to approximate the solvent effect (water) via single point energy evaluations on the optimized gas phase structures. Further optimizations in solution using different basis set namely the augmented correlation-consistent double-zeta66 (aug-cc- pVDZ) level of theory, was followed by NMR calculation using the mixed method at the same level. The study aims at investigating a group of carbocyclic nucleosides (A, B, C, D) with the objective of studying the relationship between pseudorotation and the glycosyl torsion angle as well as identifying the conformations that contribute to the overall structure of the nucleoside through NMR coupling constants. This will involve optimization of molecules in the gas phase followed by stepwise constraints on the P angle and ?max and further optimization in the gas phase as well as single point energy calculations in the gas phase and in solution {using both B3LYP/3-21G and B3LYP/6-31G(d)}. The final step will involve optimization in solution and NMR calculations both with the same level of theory (B3LYP/aug-cc-pVDZ, B3LYP/aug-cc- pVDZ nmr = mixed). The search for the lowest energy conformations for a compounds A, B, C and D (Figure 16) was based on work done by Akdag67 as a starting point for geometry optimization. The sampling was based on variation of the torsion angle ? (X-C1?-N9-C4) (see Figure 2 for definition) as well as the orientation of the substituents on the five-membered ring. This included various combinations of North/South and syn/anti conformations. This led to the generation of six starting conformations (Figure 17, compound A is shown as example with AA1, AA2, AA3, AA4, AA5 and AA6 in order). These conformations were optimized using the Density 27 Functional Theory (DFT). They were initially optimized at the B3LYP/3-21G level63,64 followed by B3LYP/6-31G (d) and the results are shown in (Table 1). The conformers were checked and their ?, P and ?max were tabulated as well. Figure 17. Starting six geometries for all 4 compounds, example shown is compound A. 28 Table 1. Results of initial geometry optimization of 4 compounds: Conformer ? P ?max ?E* 3-21G ?E* 6-31G(d) AA1 168.29 160.46 40.98 0.00 0.00 AA2 67.53 174.11 38.64 10.23 5.97 AA3 165.82 14.40 41.62 4.20 4.67 AA4 49.03 19.75 38.48 1.93 2.61 AA5 69.40 181.22 33.11 8.16 4.63 AA6 168.39 160.43 40.98 0.00 0.00 ? AA1 BB1 162.63 153.61 45.78 0.92 0.00 BB2 63.29 178.92 43.51 9.14 4.55 BB3 250.49 37.49 46.14 3.82 2.16 BB4 55.42 37.09 46.10 0.00 0.24 BB5 61.98 181.40 39.24 7.76 3.59 BB6 162.63 153.63 45.77 0.92 0.00 ?BB1 CC1 145.82 154.05 46.78 4.09 2.69 CC2 63.35 178.65 43.76 9.13 4.41 CC3 249.95 44.51 46.62 5.52 3.72 CC4 54.86 38.04 46.18 0.00 0.00 CC5 38.71 184.18 38.68 5.44 4.28 CC6 145.88 154.10 46.77 4.09 2.69 ?CC1 DD1 191.50 151.39 47.03 7.58 3.40 DD2 59.20 160.10 45.53 8.53 3.16 DD3 250.55 38.22 46.05 0.35 0.29 DD4 60.88 45.07 46.27 0.00 0.00 DD5 58.32 161.29 41.15 8.66 3.01 DD6 127.39 156.49 46.73 10.12 6.22 *?E (kcal/mol) After the 3-21G optimization, the relative energy (kcal/mol) for each set was calculated from the returned SCF of optimized molecules. Conformers within relative energy of 4 kcal/mol were chosen to go to the next level of the pseudorotation study. Some of the conformers collapsed after this initial optimization (See Table 1). At this point and further on, PROSIT68 the online 29 interactive tool for pseudorotation was used and molecules were uploaded and analyzed for various data including the glycosyl torsion angle, P , ?max , and endocyclic torsion angles. This gave way to the data presented in Table 1. Table 2. Summary of the molecules from B3LYP/3-21G. Conformer P ? ?E* 3-21G AA1 S ANTI 0.0 AA4 N SYN 1.9 BB1 S ANTI 0.9 BB3 N ANTI 3.8 BB4 N SYN 0.0 CC1 S ANTI 4.1 CC4 N SYN 0.0 DD4 N ANTI 0.4 DD4 N SYN 0.0 After the initial optimization via the B3LYP/3-21G level of theory, the conformers returned are considered our starting geometries for the pseudorotational study and they are summarized in Table 2. It is worth mentioning here that, at a later step, another level of theory 6-31G(d) had been applied to the original 6 conformations in addition to the B3LYP/3-21G. The goal was to confirm the accuracy of choice of conformers within more refined limitations. Based on equations 1 and 2, restriction of the pseudorotation phase angle ?P? across the five- membered ring was constrained in 30o increments with ??max? taken as average of the initial six conformations from Table 1. This resulted in 12 new molecules with restricted phase angles. These restricted molecules were subjected to geometry optimization in the gas phase at the B3LYP/3-21G level followed by single point energy calculation at the B3LYP/6-31G (d) both in the gas phase and in solution. 30 Using SCF and information from PROSIT, data were tabulated for each compound showing P, ?, and relative energies (kcal/mol). A typical output of this online tool is displayed in (Table 3) confirming our 30o stepwise constraint for the phase angle as well as giving the corresponding constrained values for the endocyclic torsion angles (?0-?4). It also lists ? thus determining the syn or anti disposition of the base and the sugar as well as the torsion angle ?, which defines the orientation of the 5?-OH. And the last column describes the type of the molecule submitted in terms of the characteristic north, east, west and south designation, which reflects the value of ?P? on the pseudorotation cycle (See Figure 11). Table 3. PROSIT output for (A) after single point energy in solution. (A) ?0 ?1 ?2 ?3 ?4 P ?max ? ? TYPE AA1_0 12.51 -32.30 39.94 -32.41 12.43 0.01 39.94 164.52 -69.62 ade, C3?-endo AA1_30 -8.11 -16.22 34.53 -39.82 29.88 30.09 39.91 174.73 -71.82 ade, C3?-endo AA1_60 -26.69 4.05 19.77 -36.62 39.23 60.29 39.88 178.83 -65.70 ade, C4?-exo AA1_90 -38.16 23.39 0.08 -23.33 38.11 89.88 39.96 171.90 -57.19 ade, O4?-endo AA1_120 -39.21 36.54 -19.75 -3.87 26.88 119.72 39.84 165.46 -57.15 ade, C1?-exo AA1_150 -29.78 39.78 -34.50 16.43 8.36 149.93 39.87 165.54 -61.60 ade, C2?-endo AA1_180 -12.46 32.35 -39.89 32.44 -12.41 180.02 39.89 170.68 -62.62 ade, C3?-exo AA1_210 8.48 16.35 -34.72 39.70 -29.66 209.92 40.06 175.89 -62.28 ade, C3?-exo AA1_240 26.93 -4.04 -20.02 36.46 -39.09 239.96 39.98 181.20 -64.00 ade, C4?-endo AA1_270 38.11 -23.51 -0.13 23.36 -38.07 269.82 39.98 147.71 -65.17 ade, O4?-exo AA1_300 39.13 -36.52 19.86 3.93 -26.92 299.84 39.90 147.77 -61.17 ade, C1?-endo AA1_330 29.78 -39.73 34.56 -16.43 -8.39 329.98 39.92 154.65 -62.20 ade, C2?-exo The relative energy for each series based upon SCF from the single point energy calculation in solution (all arranged with respect to the most stable in the series; highlighted in Tables (4-7). 31 This arrangement led to construction of a graph for each compound to pick the molecules that will undergo total optimization in solution and NMR calculations. The criteria for choosing were molecules that are minima on the energy surface and that their relative energy was 4 kcal/mol or less. The results for each compound are listed in Tables 4-7 and Graphs 1-4 with the chosen molecules being pinpointed (single point energy optimization in solution). 32 Table 4. Results shown for AA1 and AA2 based on single point energy in solution. Adenosine AA1 ? ?E kcal/mol AA4 ? ?E kcal/mol 0 164.52 A 0.00 0 50.34 S 8.20 30 174.73 A 1.73 30 50.42 S 8.05 60 178.83 A 8.26 60 64.03 S 8.19 90 171.9 A 7.86 90 71.77 S 7.24 120 165.46 A 7.70 120 67.36 S 3.19 150 165.54 A 7.02 150 58.81 S 2.41 180 170.68 A 5.46 180 67.24 S 0.60 210 175.89 A 5.73 210 84.27 S 1.99 240 181.2 A 8.76 240 105.30 A 5.47 270 147.71 A 5.99 270 142.85 A 6.45 300 147.77 A 2.88 300 103.29 A 8.19 330 154.65 A 0.10 330 70.08 S 10.72 Graph 1. Relative energy in solution for A 0.00 2.00 4.00 6.00 8.00 10.00 12.00 0 30 60 90 120 150 180 210 240 270 300 330 360 ?E (kcal/m ol) Phase Angle AA1 and AA4 AA1 AA4 33 Table 5. Results shown for BB1, BB3 and BB4 based on single point energy in solution. BB1 ? ?E kcal/mol BB3 ? ?E kcal/mol BB4 ? ?E kcal/mol 0 209.63 A 3.64 0 222.35 A 8.29 0 36.36 S 12.66 30 185.25 A 5.57 30 249.01 A 5.23 30 52.81 S 9.60 60 178.86 A 8.29 60 251.17 A 4.23 60 56.52 S 8.04 90 169.68 A 5.80 90 247.27 A 4.99 90 52.88 S 8.44 120 163.29 A 5.73 120 247.55 A 2.61 120 53.59 S 5.68 150 161.82 A 6.70 150 243.52 A 0.00 150 54.04 S 4.05 180 166.66 A 8.49 180 241.17 A 1.43 180 60.80 S 5.85 210 177.07 A 10.19 210 245.92 A 4.53 210 81.63 S 8.41 240 180.67 A 14.67 240 346.25 S 12.17 240 107.97 A 12.22 270 143.71 A 11.60 270 329.12 S 14.64 270 117.60 A 13.41 300 134.42 A 7.67 300 308.09 S 11.71 300 116.40 A 13.60 330 161.68 A 5.09 330 225.32 A 13.00 330 55.40 S 18.60 Graph 2. Relative energy in solution for B. 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 0 30 60 90 120 150 180 210 240 270 300 330 360 ?E (kcal/m ol) Phase Angle BB1,BB3 and BB4 BB1 BB3 BB4 34 Table 6. Results shown for CC1 and CC4 based on single point energy in solution. CC1 ? ?E kcal/mol CC4 ? ?E kcal/mol 0 200.41 A 1.94 0 35.22 S 7.70 30 191.33 A 6.25 30 52.20 S 5.09 60 180.86 A 4.03 60 30.10 S 3.99 90 169.91 A 2.40 90 31.05 S 4.60 120 154.69 A 3.08 120 29.59 S 1.46 150 145.59 A 3.39 150 30.78 S 0.00 180 151.85 A 5.02 180 38.14 S 1.81 210 182.91 A 8.02 210 25.10 S 3.54 240 178.94 A 11.83 240 106.29 A 6.67 270 113.51 A 6.34 270 110.41 A 8.50 300 116.83 A 2.38 300 106.93 A 9.17 330 120.62 A 1.08 330 52.37 S 15.14 Graph 3. Relative energy in solution for C 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 0 30 60 90 120 150 180 210 240 270 300 330 360 ?E (kcal/m ol) Phase Angle CC1 and CC4 CC1 CC4 35 Table 7. Results shown for DD3 and DD4 based on single point energy in solution. DD3 ? ?E kcal/mol DD4 ? ?E kcal/mol 0 205.3 A 7.95 0 37.05 S 10.04 30 250 A 5.37 30 52.43 S 6.75 60 245 A 4.24 60 62.38 S 3.91 90 240.3 A 4.89 90 57.92 S 4.67 120 258.3 A 2.46 120 56.22 S 1.58 150 262.84 A 0.00 150 57.24 S 0.00 180 238.9 A 1.57 180 56.58 S 1.59 210 177.75 A 4.88 210 329.1 S 7.27 240 297.16 A 8.56 240 99.47 A 11.21 270 294.85 A 11.25 270 104.99 A 15.26 300 291.91 A 10.22 300 118.72 A 12.55 330 182.4 A 11.76 330 57.67 S 15.86 Graph 4. Relative energy in solution for D 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 0 30 60 90 120 150 180 210 240 270 300 330 360 ?E (kcal/m ol) Phase Angle DD3 and DD4 DD3 DD4 36 From the above graphs, these are the molecules chosen for the full optimization in solution followed by NMR calculations (also showing their final designation based on lowest in energy): I- AA1_0 (A1), AA4_180 (A2). II- BB1_0 (B3), BB1_90 (B1), BB3_150 (B2), BB4_150 (B4). III- CC1_90 (C1), CC1_330 (C2), CC4_60 (C3), CC4_150 (C4). IV- DD3_60 (D3), DD3_150 (D1), DD4_60 (D2), DD4_150 (D4). For the selected conformers, geometry optimizations in solution using B3LYP/aug-cc-pVDZ level of theory with conductor-like polarizable continuum model (CPCM) with universal force field (UFF) cavity set is applied and the spin-spin coupling constants for each conformer are determined by the same level of theory. Spin-spin coupling constants were collected, and data for final molecules were tabulated and compared with experimental data in Tables 8-11. The best fit between experimental and theoretical data was obtained via linear regression as shown in Appendix. 37 Table 8. A Comparison of experimental and theoretical spin-spin coupling constants for Adenosine (A) Adenosine (A) 3J1?-2? 3J2?-3? 3J3?-4? P ?max ? ?Ec C2?- endod ?H2?e Exp.69(D2O) 6.1 5.3 3.4 64 4.6 Calc.a A1 7.8 6.4 -0.2 163.8 36.3 S 0.0 3.9 A2 0.0 6.1 7.8 0.5 33.5 A 0.8 5.0 Linear Regressionf 5.1 (1.1)b 6.3 (1.2) 2.8 (0.6) 64.6 4.3 aThe spin-spin coupling constant by ab initio calculation was done in aqueous solution. bThe value in parentheses is the difference between experimental data and linear regression results. c?E is the energy difference from the most stable conformer (kcal/mol). dC2?-endo is the relative population of the South conformer and is given by [3J1?-2? /(3J1?-2? +3J3?-4?)]*100. e?H2?is the chemical shift of the in ppm.70 fLinear regression equation = 0.66*(A1)+0.34*(A2) Adenosine (A) prefers an anti conformation with 65% in DMSO and 75% in water.50 While according to other studies based on 1D 1H NOE, it exists as a syn conformation with 60% preference.36 However, most of the studies refer to Ado as possessing roughly equal equilibrium between both conformations.59 Also, the population of the C2?-endo (South conformation) was determined, from the coupling constants with the aid of the two-state model of Altona,43 calculated to be 67% in DMSO71 and 63% in water.69 By comparing these experimental data with our theoretical ones, we find that we have a syn?anti conformation of 66%?34% shifted towards the syn conformer which conforms to literature findings. In terms of pseudorotation, we find that the percentage of population of the C2?-endo is within 0.6% difference which is in good 38 agreement with the reported values. The ? H2? chemical shift is averaged at 4.3 ppm, which is within 0.3 ppm difference from the literature. This calculated chemical reflects a more preference to the anti conformations50 even though our best fit comes mainly from the syn conformation. In addition the 3J spin-spin coupling constants agree well within ? 1.2 Hz. Table 9. A Comparison of experimental and theoretical spin-spin coupling constants for Aristeromycin (B) Aristeromycin (B) 3J1?-2? 3J2?-3? 3J3?-4? P ?max ? ?Ec C2?- endod ?H2?e Exp72.(D2O) 9.3 5.9 3.4 A 73.2 4.5 Calc.a B1 10.8 7.4 0.6 136.5 43.4 A 0.0 3.9 B2 10.9 6.6 0.1 140.4 42.9 A 0.3 3.9 B3 1.1 7.5 10.0 47.0 42.8 A 1.5 4.0 B4 10.3 6.6 0.0 147.6 42.4 S 2.8 5.0 Linear Regressionf 8.5 (0.8)b 6.8 (0.9) 2.6 (0.6) 76.6 4.0 aThe spin-spin coupling constant by ab initio calculation was done in aqueous solution. bThe value in parentheses is the difference between experimental results and linear regression results. c?E is the relative energy difference from the most stable conformer (kcal/mol). dC2?-endo is the relative population of the South conformer and is given by [3J1?-2? /(3J1?-2?+3J3?-4?)]*100. e?H2?is the chemical shift in ppm. fLinear regression equation = 0.75*(B2)+0.25*(B3) For aristeromycin (B), the 3J spin-spin coupling constants73 compare well within only a ? 0.9 Hz. The % C2?-endo differs from experiment by ? 3.4% which is still in good agreement 39 signifying that the major conformer is the C2?-endo. ? H2? is averaged at 4.0 ppm which is 0.5 ppm different from the experimental and which indicates the preference of aristeromycin for anti conformation. Also, based on the linear regression of the coupling constants it shows major contributions from B2 and B3 which coincides with the experimental preference for anti. Table 10. A Comparison of experimental and theoretical spin-spin coupling constants for 8- methylaristeromycin (C) 8-methyl Aristeromycin (C) 3J1?-2? 3J2?-3? 3J3?-4? P ?max ? ?Ec C2?- endod ?H2?e Exp.(D2O) 9.1 5.6 2.8 S 76.5 4.9 Calc.a C1 2.6 7.4 9.9 53.6 42.8 A 0.0 4.5 C2 1.2 7.7 10.3 49.7 42.4 S 0.7 4.3 C3 1.2 7.2 9.8 50.1 43.4 S 2.1 4.4 C4 10.3 6.9 0.2 143.5 42.8 S 2.1 5.2 Linear Regressionf 8.6 (0.5)b 7.0 (1.4) 2.3 (0.5) 78.9 5.0 aThe spin-spin coupling constant by ab initio calculation was done in aqueous solution. bThe value in parentheses is the difference between experimental results and linear regression results. c?E is the energy difference from the most stable conformers (kcal/mol). dC2?-endo is the relative population of the South conformer and is given by [3J1?-2?/(3J1?-2?+3J3?-4?)]*100. e?H2? is the chemical shift in ppm. fLinear regression equation = 0.22*(C1)+0.78*(C4) 8-methylaristeromycin (C): By looking at Table 10, we find that the difference between experimental and calculated spin-spin coupling constants for 3J1?-2? and 3J3?-4? is 0.5 Hz whereas 40 the difference is higher for 3J2?-3? (1.4 Hz). The C2?-endo population agrees with experiment to a ? of 2.4%. The theoretical ? H2? is 5.0 ppm which is only 0.1 ppm difference from the experimental value obtained from the NMR experiments (4.9 ppm). This indicates that C exits as the syn conformer. Table 11. A Comparison of experimental and theoretical spin-spin coupling constants for 3- deazaaristeromycin (D) 3-deaza Aristeromycin(D) 3J1?-2? 3J2?-3? 3J3?-4? P ?max ? ?Ec C2?- endod ?H2?e Exp74,75.(DMSO) 9.3 5.4 2.8 76.9 4.2 Calc.a D1 10.5 6.9 0.3 150.0 46.2 A 0.0 3.9 D2 1.8 7.7 9.7 60.2 46.1 S 0.4 4.4 D3 1.3 7.3 9.6 60.2 46.1 A 0.9 4.1 D4 10.6 6.3 0.0 150.0 46.2 S 1.2 4.5 Linear Regressionf 8.6 6.5 2.0 81.1 4.4 0.7 1.1 0.8 aThe spin-spin coupling constant by ab initio calculation was done in aqueous solution while experimental data was derived from DMSO data. bThe value in parentheses is the difference between experimental results and linear regression results. c?E is the energy difference from the most stable conformer (kcal/mol). dC2?-endo is the relative population of the South conformer and is given by [3J1?-2?/(3J1?-2?+3J3?-4?)]*100. e?H2? is the chemical shift in ppm. fLinear regression equation = 0.21*(D3)+0.79*(D4) 41 For 3-deazaaristeromycin (D): The coupling constants agree well to a difference of 1.1 Hz with experimental. ? H2? chemical shift at 4.4 ppm is within 0.2 ppm difference from experiment and it coincides with a preference to the anti conformation. The population of the C2-endo is within 4.2 % difference from experiment but still reflects the preference of (D) for the South conformation. Again here we see, like (A) that the major contributor is D4 which is syn while the second contributor is D3 which is anti which constitutes a controversy for this compound. Results The orientation of the base relative to the sugar moiety has been analyzed in different ways including but not limited to: spin relaxation times, NOE effects, chemical shifts perturbation, Karplus-like dependence of the 3J(C-8, H1?) coupling constant on ?.36 All methods have been aiming at gaining reliable information regarding the dynamic equilibrium between syn and anti conformation as well as the conformation around the five-membered ring(North/South). This theoretical study started with three known nucleosides that represent three nucleosides classes and for which information is available for comparison and validation of the methodology adopted. These nucleosides are: Adenosine (A) represents the ribofuranoside nucleoside, Aristeromycin (B) as a representative of the carbocyclic nucleosides with the oxygen of the tetrahydrofuran of adenosine being replaced with a methylene group and finally 3- deazaaristeromycin (D) where it lacks the N3 nitrogen. Our targets for the study were the 8- methyl and 8-ethyl Aristeromycin (C) and (F). The developed methodology has good agreements in terms of the C2?-endo population in cases of A, B, C and D. 42 *In terms of prediction of the syn-anti conformer, it has been inconsistent with the predictions of which is the lowest energy conformer. *In terms of predicting the spin-spin coupling constants, with the exception of E, it was in good agreement with the experimental data. More refinement into the conditions of the calculations may be needed to increase the precision of the methodology and address the issues of description of the conformation around the glycosidic torsion angle. 43 Chapter 2. Synthesis of Required Precursors Synthesis of the Carbocycle: The synthesis of carbocyclic nucleosides depends mainly on the condensation of the heterocyclic nucleobase and the pseudo cyclopentyl sugar.76 This can be done via one of two methods: a convergent or a linear route. In the former, an intact nucleobase is coupled directly to an activated carbocycle, while the latter depends on the construction of the heterocyclic nucleobase onto a suitable chiral cyclopentylamine.8 Both routes have been studied extensively for their advantages and disadvantages. The convergent route is generally less tedious than the linear one and hence was utilized throughout this research. In this regard, whether the target is A or B (Scheme 8), the convergent synthesis will follow a pathway outlined by attaching a functionalized heterocyclic base with an appropriately constructed version of intermediate 15 (Scheme 8). The direct substitution can be accomplished via several methods3,53 including: (a) palladium catalyzed displacement of an allylic ester or carbonate, (b) Mitsunobu coupling with a cycloalkanol, (c) nucleophilic displacement of a halide ion or activated hydroxyl such as mesylate or triflate, (d) ring opening of an epoxide, (e) Michael addition to an olefin activated by a carbonyl or other electron withdrawing group. 44 Scheme 8. Convergent retrosynthesis towards target compounds. One of the most useful common methods is the Mitsunobu coupling. It depends on activation of a hydroxyl group by a complex formed from an azodicarboxylate and triphenylphosphine which allows for direct substitution of the alcohol. The relative high acidity of the NH- on the aromatic base makes it a useful partner in the coupling.53 A major drawback of this reaction is the poor atom economy displayed by the use of triphenylphosphine (PPh3) and diisopropyl azodicarboxylate (DIAD) only for hydroxyl group activation and resulting inevitably 45 in the formation of by-products insoluble in aqueous media and need to be removed via chromatography.76 And although the yield for the Mitsunobu coupling was low, it was still a more desirable practical step that avoided the lengthy linear route of building the purine base in a stepwise fashion.77 The requisite cyclopentenone 15 has been synthesized previously via various pathways.32,78-80 Adaptation of a concise and efficient method developed in the Schneller group81 was called on for this research. The synthesis of cyclopentenone 15 is outlined in Scheme 9. The preparation of this important intermediate started with commercially available D-ribose by its reaction with 2, 2- dimethoxypropane, methanol and hydrochloric acid to achieve 2, 3- ketal protection and methylation of the anomeric hydroxyl group in 11. This was followed by transformation of the primary hydroxyl group of 11 into the iodide site via reaction with triphenylphosphine and iodine to give compound 12.78 Reductive cleavage of 12 using activated zinc in methanol gave the ring opened aldehyde 13.80 This aldehyde underwent a Grignard 1,2-addition process using vinylmagnesium bromide to yield diene 14. The compound 14 was subjected to a ring closing metathesis using Grubbs first generation catalyst (Figure 18) resulting in the formation of the allylic alcohol, which was oxidized directly, and without purification, using pyridinium chlorochromate to give 15 (in 30 % overall yield over five steps from D-ribose). 46 Scheme 9. The synthesis of cyclopentenone 15 from D-Ribose Olefin metathesis has emerged as a powerful tool in C-C bond formation 82 and was called upon here for the necessary ring-closing metathesis (RCM) whereby medium or large sized rings are constructed from acyclic diene precursors. It has been an efficient approved tool in the total 47 synthesis of natural products83,84 and nucleoside chemistry.85 The ruthenium olefin metathesis catalysts (Grubbs catalysts) offer high activity, functional group tolerance and adequate air and moisture stability for this purpose.86 In 2005, Yves Chauvin, Robert Grubbs and Richard Schrock received the Nobel Prize for their contributions and the development of metatheses reactions. 87 Figure 18. Grubbs catalysts The mechanism of action88,89 as shown in Scheme 10 was proposed by Chauvin where a metal alkylidene reacts with an olefin leading to the formation of the metallocyclobutane intermediate. The newly formed metal alkylidene reacts with the other olefinic group resulting in another metallocyclobutane intermediate which upon decomposition yields the product and the metal alkylidene which is ready for another catalytic cycle. 48 Scheme 10. Reaction mechanism of RCM Having the enone 15 in hand allowed for its functionalization to yield all the required targets. For target A (Scheme 8), the enone was transformed into compound 16 through a 1, 2-Michael addition of vinylmagnesium bromide to 15. Reduction of the vinyl ketone using lithium aluminum hydride gave the vinyl alcohol 17 in 92% yield. (Scheme 11) 49 Scheme 11. Synthesis of precursor 17 Cyclopentenone 15 was also employed to build up the needed precursor 18, which also illustrated the versatility of 15 for the synthesis of carbocyclic nucleosides. Thus cyclopentenone 15 was reduced to allylic alcohol 18 under the stereoselective Luche reduction conditions6 employing sodium borohydride and cerium(III)chloride heptahydrate (Scheme 12). Scheme 12. Synthesis of precursor 18 50 Synthesis of the Heterocyclic Nucleobase: Attention turned to the synthesis of the C-8 substituted heterocyclic nucleobase. The literature shows that a large number of purines are synthesized from pyrimidine precursors. For example the Traube synthesis of 2,4,5-triamino-1,6-dihydro-6-oxopyrimidine with formic acid and a wide variety of cyclising agents provide a versatile and efficient route to numerous substituted purines.90 This approach is based on introduction of a one-carbon fragment to bridge the nitrogens of the amino groups at the 4 and 5 positions of the pyrimidine. Some of the cyclizing agents used are carboxylic acids, acid chlorides, acid anhydride, orthoesters, ureas and amidines. Almost all chloropurines that appear in the literature come from chlorination of purinones with phosphorus oxychloride (POCl3) and with or without a tertiary ?-amine resulting in variable yields and unpredictability of formation of the chloropurine.91 Other methods involve the use of chloro-4, 5-diaminopyrimidines with conventional reagents as formic acid which leads to hydrolysis of the chlorine atom. The use of chloro-4, 5-diaminopyrimidines with orthoesters and anhydride mixtures leads to the formation of chloropurines and N-acetyl purines. The N-acetyl purines are hydrolyzed to the corresponding purines upon dissolution in 10% NaOH solution for 5-10 minutes. 51 Scheme 13. Purine synthesis based on various reagents agents In Scheme 13, four routes to the synthesis of purines via the Traube method were attempted based on compatibility, availability and efficiency. The first starting material tried was the commercially available 5, 6-diaminopyrimidin-4-ol that is supplied as a hemisulfate salt (A, Scheme 14). Treatment of the hemisulfate salt with acetic anhydride at refluxing conditions followed literature conditions,92 failed to produce any products. Alternatively the salt was treated with acetic acid and sodium acetate, and was refluxed overnight. The excess acetic acid was distilled off under reduced pressure (Scheme 14), and the residue treated with acetic anhydride 52 under refluxing conditions to produce 20.92 Refluxing 20 in phosphorus oxychloride in the presence of N, N-diethylaniline afforded 21. Scheme 14. Synthesis of 6-chloro-8-methylpurine 21. Failure of the hemisulfate to give ring closed products without treatment with acetic acid and sodium acetate may be attributed to the stability of the salt and the reduced nucleophilicity of the amino nitrogens. Hence, this procedure was eliminated for optimization towards other alkyl substituents at the purine-8-position. Looking back at Scheme 13, 4, 6-dichloropyrimidin-5-amine (B) arose as a promising possibility for the desired syntheses of the 8-alkyl-6-chloropurines. Thus the less expensive 53 reagent available 4, 6-dichloropyrimidin-5-amine was transformed with methanolic ammonia into 19 in high yields (Scheme 15). Scheme 15. Amination of 4, 6-dichloropyrimidin-5-amine Using 19 as a starting material, an optimization study using different routes as described in Scheme 12 was conducted towards the synthesis of 6-chloro-8-ethyl-9H-purine. Synthesis of 6-chloro-8-ethyl-9H-purine: (A) Synthesis based on use of carboxylic acids: This route was based on using carboxylic acids with 19 under refluxing conditions (Scheme 16).93 Use of propionic acid led to the formation of the monoacylated product with concomitant displacement of the chlorine atom. Cyclization and chlorination were achieved with phosphorus oxychloride in the presence of N,N-diethylaniline.92 This reaction has been inconsistent and at times the chlorination step didn?t occur simultaneously with cyclization which led to exploring other routes for cyclization and alkylation at the 8-position of the purine. 54 Scheme 16. Synthesis of 28 using carboxylic acid (B) Synthesis based on use of acid anhydrides: Reaction of 19 with propionic anhydride under reflux gave 28 in very low yield along with the formation of unidentified byproducts. (C) Synthesis based on use of acid chlorides: Scheme 17. Synthesis of 28 using acid chloride. 55 Heating 19 with propionyl chloride in toluene led to acylation of the amino group at the 5 position of 19 and the cyclized product 28. These conditions also caused displacement of the chlorine atom. However, maintaining anhydrous conditions for the reaction and using a low concentration of sodium hydroxide in the second step and refluxing for a short time produced a good yield of 28 (Scheme 17). (D) Synthesis based on use of orthoesters: Scheme 18. Synthesis of 28 using orthoesters. It is known that direct cyclization of diamino pyrimidines is possible with various combinations of orthoesters and acid mixtures or acid anhydrides with desired purine functionalization guiding reagent choice.90,94 Thus reacting 19 with triethyl orthopropionate in the presence of a catalytic amount of formic acid gave a crude mixture of mono- 26 and diimino 27 substituted intermediates as well as 28 (Scheme 18). After the reaction was done, as evident by disappearance of starting material as shown via thin layer chromatography and LC/MS, both compounds were separated and refluxed in NaOH separately. Compound 26 was cyclized in less than 30 min of reflux. On the other hand, compound 27 took longer to cyclize into 28. Upon repetition of this synthesis it was apparent that separation of 26 and 27 and cyclization in NaOH 56 didn?t improve the yields. Hence the crude mixture, without separation, was refluxed in sodium hydroxide for short periods of time to furnish 21 in 58% yield. Scheme 19. Intermediates 26 and 27 from the orthoester reaction. A proposal for the reaction mechanism is outlined in Scheme 20. For compound 26, NaOH abstracts a proton from the amino group at the 4-position of the pyrimidine ring leading to a nucleophilic attack on the imine carbon and subsequent loss of ethanol leading to the cyclized product 28. Cyclization of compound 27 follows a similar mechanism with the subsequent loss of ethanol and ethyl propionate. 57 Scheme 20. Proposed reaction mechanism for the formation of 28 from intermediates 26, 27. 58 Synthesis of 6-chloro-8-isopropyl-9H-purine using orthoesters: Scheme 21. Synthesis of 29 Due to the availability of the orthoester that furnishes the isopropyl substitution at the 8- position of the purine ring, 19 was reacted with trimethyl orthoisobutyrate [(CH3)2CHC (OCH3)3] in the presence of formic acid (Scheme 21). The crude mixture then refluxed in sodium hydroxide to provide 29 which occurred in low yield with the appearance of a side-product 45 which is believed to be a cyclization product following displacement of the chlorine atom. The proposed mechanism is outlined in Scheme 22. A nucleophilic attack of the pyrimidine amino group at the 5-position happened at the orthoester with concomitant loss of methanol. This was followed with imine formation and loss of another methanol. Nucleophilic attack at the imine carbon coupled with loss of another methanol furnished the cyclized product 29, which under these reaction conditions caused a nucleophilic aromatic substitution of the chlorine resulting in 45 resulting in low yields of 29. 59 Scheme 22. Proposed mechanism for formation of 45 Synthesis of 8-tert-butyl-6-chloro-9H-purine: Scheme 23. Synthesis of 6-chloro-8-t-butylpurine using acid chloride 60 Due to the fact that longer chain orthoesters are either unavailable or very expensive, to realize compound 30, trimethylacetyl chloride was the reagent of choice (Scheme 23). Refluxing 19 with this acid chloride gave the monoacylated product 32 as a hydrochloride salt. Figure 19 shows structure of 32 from analysis of a single crystal by X-ray spectroscopy and it serves as a confirmation as to which amino group (at the C5 of the pyrimidine, ortho to chlorine atom) was involved in the acylation reaction. Upon refluxing of 32 with sodium hydroxide for 2 h, the cyclized product 30 was obtained as a hydrochloride salt in 55% yield. Figure 19. Structure of 32 from X-ray analysis. 61 Chapter 3. Synthesis of 8-Alkylaristeromycin Based on the computational analysis described in chapter 1, two compounds were chosen for preparation as prototypes for the experimental of 8-substituted aristeromycin. Once obtained, these two compounds were analyzed and the data collected to measure the precision of the methodology developed from the theoretical calculations (spin-spin coupling constants, syn/anti, N/S, and relevant chemical shifts). The synthesis will depend on the convergent route to carbocyclic nucleosides. Alkylation of the 8-position of purine nucleosides has been extensively studied.57 One of the routes used to accomplish this, is C-8 bromination followed by nucleophilic displacement.95,96 However, this pathway is time consuming and offers low yields. Another route to this alkylation is lithiation at C-8 followed by reaction with a desired electrophile.97,98 Based on some unpublished results by Dr. Wei Ye from our labs, the latter linear route was adopted as shown in Scheme 24 as a way of synthesis. 62 Scheme 24. Synthesis of 1 based on unpublished results using the linear route. 63 The synthesis was based on obtaining the protected aristeromycin then protecting the primary hydroxyl group with a t-butyldimethyl silyl group and subjecting it to lithiation using 5 equivalents of lithium diisopropylamide (LDA) and finally quenching with iodomethane. This step resulted in the formation of the desired product in addition to two other side products as well as recovering some staring material. The side products obtained B and C were studied, and a mechanism was proposed for their formation as shown in (Scheme 25). This is based on the abstraction of H1? via LDA and the formation of the C1?-C2? double bond, followed by opening of the isopropylidene and subsequent loss of acetone. Scheme 25. Proposed mechanism for side products. Based on these results, optimization of this procedure was undertaken. The convergent route provides advantages for this purpose as being shorter and more efficient. The proposed retro synthesis of 8-alkylaristeromycin is presented in (Scheme 26) where R = methyl or ethyl. 64 Scheme 26. Retro synthesis of 8-alkyl substituted Aristeromycin The synthesis (Scheme 27) starts with precursor 17 (Scheme 11) and 6-chloro-8-methyl- purine 21 in (Scheme 14) with a Mitsunobu coupling in the presence of triphenylphosphine and diisopropyl azodicarboxylate yielded 22. Manipulation of the vinyl group in 22 to the requisite hydroxymethyl side chain was accomplished in two steps81: (i) oxidative cleavage with osmium tetraoxide and sodium periodate gave an intermediate aldehyde, (ii) direct reduction of the aldehyde, without separation, with sodium borohydride into the alcohol giving 23 in 49% yield over two steps. 65 Scheme 27. Synthesis of compound 1 by the convergent route. Treatment of 23 with methanolic ammonia at 110 oC for two days gave 24 in good yield. Finally, deprotection of the isopropylidene group under acidic conditions followed by chromatographic purification afforded target 1 which was confirmed with X-ray crystallography. Target 1 was characterized via NMR spectroscopy and LC/MS. Both the NMR data and the X- ray crystallography show a syn conformation of compound 1. 66 Figure 20. Structure of 1 from X-ray analysis 67 The preparation of target 2 followed the retro synthesis outlined in Scheme 25. Thus, a Mitsunobu coupling between 17 and 28 (Scheme 28) was carried out. The crude mixture of 34, contaminated with diisopropylhydrazine dicarboxylate, was subjected to oxidative cleavage with osmium tetraoxide followed by reduction with sodium borohydride to yield 35. Ammonolysis of 35 with methanolic ammonia furnished 36 in 39 % over two steps. Deprotection of 36 with 1 N HCl gave target 2 in 46% yield. Scheme 28. Synthesis of compound 2 by the convergent route. 68 Chapter 4. Synthesis of 8-Substituted 4?-Norneplanocin Analogues This chapter will describe the investigations into alkyl substitution at the 8 position of purine bases within the framework of 4?-norneplanocin. The synthetic approach made use of the convergent pathway of synthesis of carbocyclic nucleosides stressing the versatility of this route. The efficiency of this route became more evident in instances where the series of nucleosides built on a specific carbocycle or specific heterocycle is required for building libraries of nucleosides for biological screenings.76 In this research the synthetic investigation of 8-alkyl-4?- norneplanocin depended on the common allylic alcohol 18 with the preformed heterocyles: 21, 28, 29, 30 and 46. Target compounds 3, 4, 5, 6 and 7 pursued are shown in Scheme 29. 69 Scheme 29. Retrosynthesis of targets 3, 4, 5, 6 and 7 using the convergent route. 70 The synthesis of target nucleosides made use of the allylic alcohol 18 described earlier in Scheme 12. One of the most common and useful methods for joining the heterocycle and the carbocycle is the afore described Mitsunobu coupling reaction. However, one of the drawbacks of this process is the formation of side products, namely, triphenylphosphine oxide and the diisopropyl hydrazine dicarboxylate (DIAD by-product). In this case, the closeness of the retention factor Rf between diisopropyl hydrazine dicarboxylate and the coupled product led to problematic separations of the coupled adducts i.e. 37, 39, 41, 43 and 47. Fortunately, the next step in the synthesis was not affected by the use of the crude material in the ammonolysis reaction. For 43, the reaction required longer heating during the Mitsunobu coupling step, up to 5 days, and only a minor portion of 30 reacted as demonstrated by its 90 % recovery. This suggests that the bulky t-butyl is interfering with the SN2 reaction of the Mitsunobu coupling. Having prepared the heterocyclic bases 21, 28, 29 and 30 allowed coupling under regular Mitsunobu conditions with 18 and the adduct products (37, 39, 41 and 43) were confirmed via mass spectrometry as well as NMR spectroscopy (Scheme 29). However, due to contamination with the diisopropyl hydrazine dicarboxylate (DIAD by-product), the NMR spectra were sometimes difficult to reconcile. Consequently, the crude material was subjected to amination under standard methanolic ammonia conditions to furnish the protected nucleoside (38, 40, 42 and 44). A final deprotection under acidic conditions followed with purification with column chromatography gave target compounds 3, 4, 5 and 6 (Scheme 30) for which verification via NMR spectroscopy and mass spectrometry was accomplished. 71 Scheme 30. Synthesis of targets 3, 4, 5, and 6 72 In order to summarize the conditions used for both the synthesis of the heterocycle and the convergent route used in this research, a final target compound 7 was pursued showing the consequence of synthetic optimization in Scheme 31. Scheme 31. Convergent Synthesis of target compound 7 The synthesis of 46 proceeded via the use of valeroyl chloride under refluxing conditions. This reaction required anhydrous conditions to circumvent displacement of the chlorine atom. Upon reflux, a mixture of acylated products (mono- and di- acylation at the 4 and 5 positions of 73 the pyrimidine ring) resulted which upon treatment with NaOH under reflux yielded 46 in 31 % yield. The Mitsunobu coupling proceeded smoothly in 47 % yield. This was followed by standard amination with methanolic ammonia giving 48 in 53% yield then deprotection under acidic conditions furnished target 7 in quantitative yield. X-ray crystallography, NMR spectroscopy and LC/MS were used for identification of target 7 and its x-ray crystal structure is show in Figure 21 as a syn conformer. Figure 21. Structure of 7 from X-ray analysis 74 Chapter 5. Analysis of 8-Ethylaristeromycin (2) Target 2 (referred to as F for the theoretical study) was the second probe to test the theoretical methodology developed in this research. For this target, the molecule was optimized according to the method described in Chapter 2 and all data generated for this molecule were collected. The NMR calculations were done in water and the coupling constants were gathered. The optimization started with six geometries for 2 that were optimized using both the 3-21G and the 6-31G (d) level of theory. This resulted in three conformations that were being chosen based on our choice of 4 kcal/mol in relative energy (Table 12). Table 12. Initial conformers chosen for pseudorotation study of (F). Conformer ?E 3-21G kcal/mol ?E 631G(d) kcal/mol P ?max ? FF1 4.2 2.9 152.9 47.2 144.2 FF2 9.1 4.2 177.1 43.9 61.4 FF3 5.4 3.8 43.9 46.6 248.6 FF4 0.0 0.0 37.7 46.0 55.7 FF5 5.4 4.3 183.6 38.5 38.7 FF6 4.2 2.9 152.9 47.2 144.2 75 Figure 22. Initial optimization of 8-ethylaristeromycin (F). For each of FF1, FF3 and FF4 shown in Figure 22, the phase angle has been constrained in 30o increments to generate 12 structures that were fully optimized under the 3-21G in the gas phase followed by single point energy calculations in the gas phase and in solution via the 6-31G (d) level of theory. A graph was constructed to choose the final conformations for total solvation optimization and NMR calculations. 76 Table 13. Conformations based on single point energy calculations in solution of (F). FF1 ? ?E (kcal/mol) FF3 ? ?E (kcal/mol) FF4 ? ?E (kcal/mol) 0 200.62 A 5.7 0 226.1 A 9.8 0 34.26 S 10.1 30 241.32 A 2.9 30 246.9 A 5.0 30 52.11 S 7.0 60 180.75 A 6.6 60 243.9 A 2.1 60 30.25 S 6.3 90 170.2 A 5.2 90 248 A 4.1 90 31.14 S 6.8 120 146.58 A 5.8 120 247.8 A 2.2 120 30.37 S 4.0 150 143.29 A 6.3 150 250.5 A 0.0 150 32.08 S 2.8 180 150.02 A 7.7 180 246.4 A 0.7 180 38.22 S 4.6 210 184.28 A 11.3 210 247.2 A 4.8 210 25.23 S 6.4 240 179.34 A 15.1 240 352.7 S 10.8 240 106.4 A 9.2 270 115.24 A 9.0 270 289.5 S 13.3 270 111.4 A 11.2 300 117.66 A 4.9 300 309.1 S 14.9 300 109.5 A 11.6 330 123.77 A 3.9 330 3.533 A 15.3 330 52.25 S 17.8 Graph 5. Relative energy in solution for F. 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 16.0 18.0 20.0 0 30 60 90 120 150 180 210 240 270 300 330 360 ?E (kcal/m ol) Phase Angle FF1, FF3 and FF4 FF1 FF3 FF4 77 Based on graph 5, five final geometries were chosen for total solvation optimization and NMR calculations and the coupling constants were collected and tabulated in Table 14. The final geometries obtained from the NMR calculations are shown in Figure 23. Figure 23. Final geometries for F after NMR calculations. 78 In addition to the theoretical calculated data, following the synthesis of 8-ethyl aristeromycin 2, 1D and 2D NMR data were collected: 1HNMR, 13CNMR, COSY, HMBC and ROESY for the proper assignment of all the protons and carbons in the molecule. This was used to correlate experimental with theoretical as shown in Table 14. For the calculation of the coupling constant involving H2? i.e. 1?-2? and 2?-3?, the 1HNMR spectrum had an overlapping water peak (Figure 24), which led to acquiring the NMR at higher temperature to move the water peak and be able to calculate the coupling constants. Figure 24. Portion of 1HNMR spectrum of F in D2O showing overlapping H2?. 79 The NMR of F was taken in D2O to allow for minimum variations as possible from the theoretical study and it was run at 327 oK and the result is shown in Figure 25. Figure 25. Portion of 1HNMR spectrum of F in D2O (327 oK) showing resolved peaks. 80 Table 14. A Comparison of experimental and theoretical spin-spin coupling constants for 8-ethylaristeromycin (F) 3J1?-2? JH2?-H3? JH3?-H4? P ?max ? ?Ec C2?- endod ?H2?e 1-Exp.(D2O)(327oK) 8.4 5.8 3.6 S 70.0 5.2 2-Exp(D2O)(294oK) 9.2 N/A 3.2 4.7 Calc.a F1 5.8 9.7 8.9 73.5 41.7 A 0.0 4.2 F2 10.3 6.1 0.2 143.9 42.3 S 1.6 5.3 F3 1.9 7.4 9.7 52.3 43.5 A 0.0 4.2 F4 11.6 6.6 0.0 148.4 42.7 A 1.6 4.4 F5 10.3 7.3 0.4 139.1 43.1 S 2.9 5.2 Linear Regressionf 8.0 6.6 3.1 72.0 4.3 0.4 0.8 0.5 aThe spin-spin coupling constant by ab initio calculation was done in aqueous solution. bThe value in parentheses is the difference between experimental data and linear regression results. c?E is the energy difference from the most stable conformer (kcal/mol). dC2?-endo is the relative population of the South conformer and is given by [JH1?-2?/ (J H1?-2?+J H3?-4?)]*100. e?H2? is the chemical shift of the in ppm. fLinear regression equation for 1 = 0.37 F2 + 0.31 F3 + 0.24 F4 The data show agreement between the theoretical and the experimental NMR spin-spin coupling constants with a ? 0.8 Hz as the highest. Also, from experiment it shows that F prefers the C2?-endo with a population of 70 % and theoretical data shows preference for the C2?-endo as well with a population of 72 %. 81 Finally, the experimental chemical shift shows 5.2 ppm for H2? (at 327 oK) which according to the literature36,99 is a syn conformer, while it is 4.7 ppm (at 294 oK) where the theoretical one is 4.3 ppm which is closer to the anti conformation. Overall, there is satisfactory agreement between the theoretical study and the experimental study in terms of the coupling constants and the conformation of the five-membered ring. Figure 26. Structure of 2 from X-ray analysis. The structure of target 2 has also been shown as the syn conformer via X-ray crystallography as shown in Figure 26. 82 Summary of results: (i) In terms of the relationship between the syn/anti and North/South conformation: The experimental coupling constants observed have been explained by the existence of two (for A, B, C and D) or more (three conformations in case of E) conformations in solution as shown by theoretical calculations. There is a correlation between syn/anti and pseudorotation as has been documented in the literature.44 A syn conformation prefers a C2?-endo (South) and this was shown to be true through the studied compounds. For the anti conformers, both the C2?- endo and the C3?-endo are equally probable. This has been more evident with compound (F) where it showed three conformations, two of which are anti/north (31%) and anti/south (24%) which is represents a nearly equal probability of N/S with the anti conformer. That is in addition to the syn/south conformer which is the major contributor to the observed experimental values, Figure 27. Figure 27. Summary of correlation study between theoretical and experimental data. Ado (A) A1(Syn/S), 66% A2 (Anti/N), 34% Ari (B) B2 (Anti/S), 75% B3(Anti/N), 25% 8-meari(C) C1(Anti/N), 22% C4(Syn/S), 78% 3-deazaari(D) D3(Anti/N), 21% D4(Syn/S), 79% 8-ethylari(F) F2(Syn/S), 37% F3(Anti/N), 31% F4(Anti/S), 24% 83 (ii) In terms of syn/anti conformations the 8-substitution: Alkyl substitution at the 8-position produces a pronounced shift in equilibrium to the syn conformation yet the structure is explained by contributions from both conformations. Future biological activity studies could corroborate these findings for target molecules as well as give insight into the possibility of extrapolating the results to other carbocyclic nucleosides. 84 Conclusion A theoretical methodology for explaining the conformational structural parameters of 8- substituted carbocyclic nucleosides have been investigated. In that direction, a series of optimizations have been carried out using the DFT level of theory to produce coupling constants for the optimized lowest energy conformations obtained. Tables 8-12 provide the calculated and experimental spin coupling constants of 3J1?2?, 3J 2?3? and 3J 3?4? for 5 compounds. A theoretical calculation describes each compound in terms of two or more conformations as being the most probable conformations in solution. To obtain the best fit to experiment, linear regression statistics were then applied and contributions for each conformation were measured. The five compounds demonstrated a preference for the south conformation. A correlation between syn conformations and C2?-endo was found as well as an equal preference of the anti conformation for both the north/south. This suggests that antiviral activity could reside in the south conformation. Further biological studies could corroborate this finding. In terms of the syn/anti relationship each compound is represented by an equilibrium that is shifted towards one conformation. For the 8-alkyl compound it was found that a preference for the syn conformation existed whereas unsubstituted aristeromycin showed a shift toward the anti conformation. Yet, in the end, the results for both the substituted and unsubstituted illustrated that syn and anti conformations were accessible. Thus, in regard to syn/anti conformational preference the results here are equivocal. 85 To avail authentic sample for the aforementioned NMR analyses, it was necessary to establish successful synthetic procedures for those molecules central to the investigation. For that purpose procedures were designed and accomplished via convergent methods involving (i) the construction of the purine bases from pyrimidine units, which required synthesis, and (ii) coupling the purines with appropriately designed and prepared cyclopentyl derivatives. The final products were obtained in overall reasonable yields and their structures verified by X-ray crystallography and a thorough NMR analysis using contemporary, multi-dimensional methods. 86 Experimental Details General Melting points were recorded on a Meltemp II melting point apparatus and the values are uncorrected. 1H and 13C NMR spectra were recorded on either a Bruker AV 250 spectrometer or a Bruker AV 400 spectrometer. All 1H chemical shifts are reported in ? relative to the internal standard tetramethylsilane (TMS, ? 0.00). 13C chemical shift are reported in ? relative to CDCl3 (center of triplet, ? 77.23) or relative to DMSO-d (center of septet, ? 39.51). The spin multiplicities are indicated by the following symbols: s (singlet), d (doublet), t (triplet), q (quartet), quin (quintet), sex (sextet), sep (septet), m (multiplet) and br (broad). The reactions were monitored by thin-layer chromatography (TLC) using 0.25 mm Whatman Diamond silica gel 60-F254 precoated plated with visualization by irradiation with a Mineralight UVGL-25 lamp or exposure to potassium permanganate. Column chromatography was performed using Whatman silica, 230-400 mesh and 60 Ao using elution with the indicated solvent systems. Yields refer to chromatographically and spectroscopically (1H and 13C NMR) homogeneous materials. All mass spectroscopic data (LC/MS) were collected using Waters Aquity UPLC and Q-Tof Premier Mass Spectometer. Methyl 2, 3-O-isopropylidene-?-D-ribofuranoside (11). Concentrated hydrochloric acid (10.0 mL) was added to a suspension of D-ribose (100 g, 0.67 mol) in acetone (420 mL), 87 methanol (420 mL), and 2, 2-dimethoxypropane (200 mL). The reaction was stirred at room temperature overnight and neutralized with pyridine. Water (1000 mL) and ether (300 mL) were added and the separated aqueous layer was washed with ether and ethyl acetate. The combined organic layers were washed with water, brine and dried over sodium sulfate. The solvent was evaporated by rotary evaporator and the residue was distilled in vacuum to give 104.5 g (77 %) of 11 as colorless oil. The NMR spectral data agreed with literature.78 Methyl-5-deoxy-5-iodo-2,3-O-isopropylidene-?-D-ribofuranoside (12): A solution of 11 (90.92 g, 0.45 mol), imidazole (45.5 g, 0.67 mol) and triphenylphosphine (140.3 g, 0.5 mol) in toluene (500 mL) and acetonitrile (100 mL) was treated with iodine (135.8 g, 0.5 mol) portionwise. The reaction mixture was refluxed for 30 minutes and cooled to room temperature. The white precipitate was decanted and the remaining solution was diluted by ether. The organic phase was washed with 10% sodium thiosulfate solution, water and brine and dried over sodium sulfate. The residue after concentration was loaded on silica gel and eluted with hexane/ethyl acetate (15:1) to give 119.04 g (85.1 %) of 12 as colorless oil. The NMR spectral data agreed with literature.78 (4R,5R)-2,2-dimethyl-5-vinyl-1,3-dioxolane-4-carbaldehyde (13). Activated powdered zinc (52.0 g, 0.8 mol) was added to 12 (50 g, 0.2 mol) in methanol (200 mL). The mixture was left at room temperature for 2 hours. After filtration, the filtrate was concentrated in vacuum around 20oC and the residue was loaded on a silica gel column and eluted with hexane /ethyl acetate (4:1) to afford 21.34 g (86%) of 13 as a colorless liquid. The NMR spectral data agreed with literature.80 88 1-((4S,5R)-2,2-dimethyl-5-vinyl-1,3-dioxolan-4-yl)prop-2-en-1-ol (14). To a well stirred solution of 13 (4.1 g, 26.3 mmol) in THF (100 mL), was added vinyl magnesium bromide (1.0 M in THF, 31.5 mL, 31.5 mmol) dropwise between -20o and -30oC. The mixture was kept at that temperature for 2 hours and allowed to warm to room temperature. The mixture was quenched with saturated ammonium chloride solution and extracted with ethyl acetate. The combined organic layers were washed with brine and dried over sodium sulfate. Filtration, evaporation of the filtrate followed by column chromatography (4:1 hexane: ethyl acetate) gave 3.8 g (76%) of 14 as a colorless oil. The NMR spectral data agreed with literature.81 (3aR,6aR)-2,2-dimethyl-3aH-cyclopenta[d][1,3]dioxol-4(6aH)-one; ((4R,5R)-4,5-O-iso- propylidene-2-cyclopentenone) (15). To a solution of the diene 14 (25 g, 135.7 mmol) in anhydrous methylene chloride (300 mL) was added Grubbs catalyst benzylidene bis (tricyclohexylphosphine) dichlororuthenium (1.2 g, 1.458 mmol) after the solution was flushed with nitrogen for 20 minutes. After stirring at room temperature for 12 hours, pyridinium chlorochromate (23.5 g, 271.4 mmol), and 4 Ao molecular sieves (20 g) were added to the dark brown solution. The reaction mixture was stirred at room temperature overnight and filtered over silica gel pad with ethyl acetate. The filtrate was concentrated in vacuum and the residue was purified with column chromatography at afford 11.3 g (54%) of 15 as white crystals. The NMR spectral data agreed with literature.79 (3aR,6R,6aR)-2,2-dimethyl-6-vinyldihydro-3aH-cyclopenta[d][1,3]dioxol-4(5H)-one (16). To a suspension of CuBr.Me2S (166.7 mg, 0.8 mmol) in THF (120 mL) at -78 oC was added vinylmagnesium bromide (24.3 mL, 24.3 mmol) dropwise. The reaction mixture was 89 stirred for 20 minutes and HMPA (6.8 mL, 38.9 mmol) was added followed by a solution of cyclopentenone 15 (2.5 g, 16.2 mmol) and TMSCl (4.1 mL, 32.4 mmol) in THF (20 mL) dropwise. After the reaction was stirred for 3 hours at -78 oC, the mixture was warmed to 0 oC and quenched with a saturated solution of ammonium chloride. The reaction mixture was then stirred for 30 minutes and ethyl acetate was added (300 mL). The organic phase was separated and washed with water and brine, dried with MgSO4. Solvent was evaporated in vacuo and the residue was purified by column chromatography (8:1 hexane: ethyl acetate) to give 2.18 g (74%) of 16 as a colorless liquid. The NMR spectral data agreed with literature.100 (3aS,4S,6R,6aR)-2,2-dimethyl-6-vinyltetrahydro-3aH-cyclopenta[d][1,3]dioxol-4-ol (17). To a suspension of lithium aluminum hydride (0.7 g, 17.8 mmol) in THF (50 mL) was added a solution of 16 (1.6 g, 8.9 mmol) in THF (20 mL) drop wise at 0 oC. The reaction mixture was stirred at room temperature for 4 hours and was quenched sequentially with water (0.6 mL), aqueous NaOH 15% (0.6 mL) and water (1.5 mL). After filtration, the filtrate was evaporated to give 1.5 g (92%) of 17 as a colorless liquid. The NMR spectral data agreed with literature.81 (3aS,4S,6aR)-2,2-dimethyl-4,6a-dihydro-3aH-cyclopenta[d][1,3]dioxol-4-ol (18). To a stirred solution of cyclopentenone 15 (1.0 g, 6.5 mmol) and CeCl3.H2O (2.4 g, 6.5mmol) in MeOH (50 mL) at 0 oC was added NaBH4 (0.5 g, 13.0 mmol) portionwise. After stirring at room temperature for 1 h the mixture was neutralized with NH4Cl, and extracted with CH2Cl2. The organic layers were combined, dried (anhydrous Na2SO4) and concentrated to give 18 as a colorless syrup (0.7 g, 68%). 1H NMR (400 MHz, CDCl3) ? 5.81 (s, 2H), ? 4.94 (d, 1H), ? 4.67 90 (t, 1H), ? 4.48 (m, 1H), ? 2.74 (d, 1H), ? 1.35 (s, 3H), ? 1.32 (s, 3H); 13C NMR (100 MHz, CDCl3) ? 136.5, 132.1, 112.5, 83.7, 77.3, 74.3, 27.8, 26.7. 6-chloropyrimidine-4,5-diamine (19). 1.5 g of 4,6-dichloropyrimidin-5-amine was dissolved in 25 mL MeOH and cooled to 0 oC for 30 min before being saturated with ammonia gas at the same temperature for 1 h. The solution was heated at 110 oC for 48 h in a sealed stainless steel Parr. The solvent was evaporated under reduced pressure and the residue was washed with EtOAc to give 1.2 g (90 %) of 19 as white crystals. 1H NMR (400 MHz, DMSO): ? 7.637 (s, 1H, H-2), ? 6.757 (s, 2H, NH2), ? 4.975 (s, 2H, NH2), 13C NMR (100 MHz, DMSO) ? 153.6, 145.8, 137.5, 123.2. ESI-MS calcd for C4H5N4: [(M + H) +]: 145.0281, found: 145.0278 N-(4-amino-6-hydroxypyrimidin-5-yl) acetamide (20). 4, 5-Diamino-6-hydroxypyrimidine hemisulfate (1.9 g, 5.3 mmol) was added to a mixture of acetic acid (30 mL) and sodium acetate (0.8 g, 10.0 mmol) and left to reflux for 12 h. Excess acetic acid was evaporated under reduced pressure. 40 mL of acetic anhydride was added to the residue and refluxed for 4 h. The excess acetic anhydride was azeotropically evaporated with toluene. The residue was dissolved in 40 mL of 1.5 N NaOH and boiled for 20 min. The solution was acidified while hot with glacial acetic acid and cooled. The cooled solution gave 0.8 g (quantitative yield) of 20 as white crystals. 1H NMR (400 MHz, DMSO): ? 11.75 (br s, 1H), ? 8.608 (s, 1H), ? 7.742 (s, 1H), ? 6.203 (s, 2H), ? 3.322 (br s, 1H), ? 1.93 (s, 3H); 13C NMR (100 MHz, DMSO): ? 168.8, 159.2, 159.1, 146.9, 98.5, 22.8. ESI-MS calcd for C6H6N4O: [(M + H) +]: 169.0731, found: 169.0714. 91 8-Methyl-6-chloropurine (21). To a solution of phosphors oxychloride (35 mL), containing N, N-diethyl aniline (3 mL), was added compound 20 (2.3 g, 15.1 mmol) slowly. The mixture was refluxed overnight and the excess phosphorus oxychloride was distilled under reduced pressure. The residue was poured on cracked ice and the solution was made strongly basic with 10N potassium hydroxide and allowed to stand for 20 minutes. The solution was then extracted with ether (2x100 mL). The solution was then acidified to pH 4 with concentrated aqueous HCl and was continuously extracted with ether for 48 h. The organic layer was collected and evaporated under reduced pressure to yield 21 as yellow crystals. 1H NMR (400 MHz, DMSO): ? 8.63 (s, 1H), 2.57 (s, 3H); 13C NMR (100 MHz, DMSO) ? 156.5, 152.6, 150.8, 149.0, 113.9, 15.1. ESI-MS calcd for C6H5N4Cl: [(M + H) +]: 169.0281 found: 169.0270. 6-chloro-9-(2,2-dimethyl-6-vinyltetrahydro-3aH-cyclopenta[d][1,3]dioxol-4-yl)-8- methyl-9H-purine (22). To a solution of 17 (92.5 g, 13.6 mmol), triphenylphosphine (7.1 g, 27.3 mmol) and 6-chloropurine (3.0 g., 19.2 mmol) in THF (100 mL) was added DIAD (diisopropyl azodicarboxylate) (5.5 g, 27.3 mmol) dropwise at 0 oC. The reaction mixture was kept at 0 oC for 2 hours followed by stirring at 50 oC for 8 hours. The solvent was removed under reduced pressure and the residue was purified with column chromatography (hexane: ethyl acetate 4:1) to give (0.4 g, 55%) of 22 as yellow solid. 1H NMR (400 MHz, CDCl3): ? 8.51 (s, 1H), 5.86 (m, 1H), 5.11-5.01 (m, 3H), 4.63-4.57 (m, 2H), 2.83-2.74 (m, 2H), 2.64 (s, 3H), 2.19 (m, 1H), 1.46 (s, 3H), 1.14 (s, 3H). 13C NMR (100 MHz, CDCl3) ? 155.2, 152.6, 150.6, 149.0, 137.4, 131.1, 116.1, 113.9, 83.7, 82.6, 61.5, 48.2, 35.2, 27.5, 25.0, 14.9. ESI-MS calcd for C16H19N4O2Cl: [(M + H) +]: 335.1275 found: 335.1271. 92 ((3aR,4R,6R,6aS)-6-(6-chloro-8-methyl-9H-purin-9-yl)-2,2-dimethyltetrahydro-3aH- cyclopenta[d][1,3]dioxol-4-yl)methanol (23). To a suspension of 22 from above and sodium periodate (4.3 g, 20.2 mmol) in methanol (35 mL) and water (20 mL) was added OsO4 (30 mg, 0.1 mmol) at 0o C. The suspension was stirred for 2 hours at 0 oC and then for 3 h at room temperature. The suspension was filtered and the solid was washed with EtOAc. The combined filtrates were concentrated and the residue was diluted with water (20 mL) and extracted with EtOAc (3x40 mL). The combined organic layers were dried over MgSO4. After filtration, the filtrate was evaporated at ambient temperature and dissolved in MeOH (40 mL). This solution was cooled to 0 oC and NaBH4 (1.2 g, 30.8 mmol) was added portion wise. After 10 min, the solvent was removed by reduced pressure and the residue was neutralized by saturated ammonium chloride solution followed by extraction with EtOAc. The combined organic layers were washed with brine and dried over MgSO4. After filtration, the solvent was removed under reduced pressure and the residue was purified via column chromatography using (hexane: EtOAc 1:2) to afford (0.4 g, 49%) of 23 as white solid. 1H NMR (250 MHz, CDCl3): ? 8.64 (s, 1H), 5.23 (t, J=10.4, 10 Hz, 1H), 4.80-4.70 (m, 2H), 3.89 (d, J=8.8 Hz, 2H), 3.49 (br s, 1H), 2.85 (m, 1H), 2.76 (s, 3H), 2.57-2.31 (m, 2H), 1.59 (s, 3H), 1.31 (s, 3H); 13CNMR (60 MHz, CDCl3) ? 155.6, 152.6, 150.6, 149.3, 131.3, 113.6, 83.2, 81.9, 63.8, 62.9, 45.7, 32.1, 27.8, 25.2, 15.0. ESI- MS calcd for C15H19N4O3Cl: [(M + H) +]: 339.1224 found: 339.1222. ((4R,6R)-6-(6-amino-8-methyl-9H-purin-9-yl)-2,2-dimethyltetrahydro-3aH- cyclopenta[d][1,3]dioxol-4-yl)methanol (24): A solution of 23 (130 mg, 0.4 mmol) was dissolved in MeOH (20 mL) and was cooled to 0 oC for 30 min before being saturated with ammonia gas at the same temperature for 1 h. The solution was heated to 100 oC for 48 h in a 93 sealed stainless steel Parr. The solvent was removed under reduced pressure, and the residue purified by column chromatography (EtOAc/MeOH = 8:1) to give (100 mg, 82 %) as white solid. 1H NMR (400 MHz, CDCl3) 8.16 (s, 1H), 6.53 (s, 2H), 5.12 (t, J = 6.4, 6 Hz, 1H), 4.63 (dd, J = 4, 7.6 Hz, 1H), 4.64-4.58 (dddd, 1H), 3.81 (d, J = 5.2 Hz, 2H), 2.70 (q, 1H), 2.54 (s, 3H), 2.50-2.42 (m, 1H), 2.39-2.31 (m, 1H), 1.53 (s, 3H), 1.25 (s, 3H). 13C NMR (100 MHz, CDCl3) ? 155.01, 151.58, 150.34, 149.53, 118.73, 112.95, 83.89, 82.56, 64.03, 62.80, 45.38, 32.18, 27.90, 25.30, 14.59. (3R,5R)-3-(6-amino-8-methyl-9H-purin-9-yl)-5-(hydroxymethyl)cyclopentane-1,2-diol (1). Compound 23 (110 mg, 0.3 mmol) was dissolved in a mixture of 1N HCl (6mL) and MeOH (6 mL). The mixture was stirred at room temperature for 3 h. The solution was then neutralized with weakly basic exchange resin (Amberlite IRA-67). Filtration followed with evaporation of solvent, the crude product was purified by chromatography (EtOAc/MeOH/NH3:H2O = 8:1:1) to give (70 mg, 73 %) of 1 as white crystals. 1H NMR (400 MHz, DMSO-d6) 8.04 (s, 1H), 7.03 (s, 2H), 4.85 (d, 1H), 4.76 (t, 1H), 4.67 (d, 1H), 4.60 (q, 1H), 4.53 (q, 1H), 3.84 9dd, 1H), 3.56 (quintet, 1H), 3.45 (quintet, 1H), 2.51 (s, 3H), 2.13-2.01 (m, 3H); 13C NMR (100 MHz, DMSO- d6) ? 155.09, 151.06, 150.39, 149.38, 149.37, 149.36, 118.16, 73.16, 71.49, 63.36, 60.33, 45.68, 27.08, 14.4. ESI-MS calcd for C12H17N5O3: [(M + H) +]: 280.1410, found: 280.1401. 6-chloro-8-ethyl-9H-purine (28). To 16 ml of trimethyl orthopropionate was added 19 (0.98 g, 6.81 mmol) and the mixture was heated at 100 oC until clear. Formic acid (1 ml) was added dropwise and the solution refluxed overnight until the disappearance of starting material. The solution was evaporated and refluxed with 1.5 N NaOH for 1 h. The crude mixture was 94 evaporated and purified via column chromatography (EtOAc: MeOH = 15:1) to give (0.65 g, 58 %) of 28 as white solid. 1H NMR (250 MHz, CDCl3): ? 13.6 (br s, 1H, NH), ? 8.61 (s, 1H, H-2), 2.912 (q, 2H, CH2), ? 1.321 (t, 3H, CH3). 13C NMR (60 MHz, CDCl3): ? 161.0, 155.2, 150.4, 145.4, 129.5, 22.2, 11.3. ESI-MS calcd for C7H8N4Cl: [(M + H) +]: 183.0437, found: 183.0439. 6-chloro-8-isopropyl-9H-purine (29). 19 (0.5 g, 3.5 mmol) was added to 1, 1, 1-trimethoxy- 2-methylpropane (5.6 mL, 35 mmol) and the mixture was heated to 100 oC until it became almost clear. Formic acid (1 mL) was added dropwise and the solution was refluxed for 8 h. The solution was refluxed as such in 1.5 N NaOH for 1 h. The crude mixture was purified with column chromatography (EtOAc: MeOH = 15:1) to give (0.2 g, 32%) of 29 as a white solid and (0.5 g, 67%) of 31 as yellow oil. 6-chloro-8-isopropyl-9H-purine (29). 1H NMR (400 MHz, CDCl3) ? 13.59 (br s, 1H), ? 8.65 (s, 1H), ? 3.22 (sep, 1H), ? 1.36 (d, 6H); 13C NMR (60 MHz, CDCl3). ESI-MS calcd for C8H10N4Cl: [(M + H) +]: 197.0594 found: 197.0587 8-isopropyl-6-methoxy-9H-purine (31). 1H NMR (400 MHz, CDCl3): ? 11.60 (br s, 1H), ? 8.37 (s, 1H), ? 4.08 (s, 3H), ? 2.49 (sep, 1H), ? 1.38 (d, 6H). N-(4-amino-6-chloropyrimidin-5-yl) pivalamide (32). To a solution of 19 (1.0 g, 7.2 mmol) in dry toluene, pivaloyl chloride (1.2 mL, 8.6 mmol) was added dropwise. The solution was refluxed for 4 h. The solution was filtered and washed several times with cold toluene to yield (1.3 g, 80 %) of 32 as white solid. 1H NMR (400 MHz, MeOD): ? 8.29 (s, 1H), 1.34 (s, 9H); 13C NMR (100 MHz, MeOD): ? 181.5, 163.1, 156.4, 154.5, 15.6, 40.8, and 27.8. ESI-MS calcd for C9H14N4OCl: [(M + H) +]: 229.0862, found: 229.0868. 95 8-tert-butyl-6-chloro-9H-purine (30). 0.3 g of 32 was refluxed in 1 N NaOH for 2 h. The solution was evaporated and the residue was purified with column chromatography EtOAc/MeOH = 15:1 to give 0.24 g (55%) of 30 as white solid. 1H NMR (400 MHz, DMSO-d6) ? 13.61 (br s, 1H), 8.66 (s, 1H), 1.42 (s, 9H); 13C NMR (100 MHz, DMSO-d6) ? 166.6, 154.8, 151.4, 147.5, 130.7, 34.4, 29.0. ESI-MS calcd for C9H11N4Cl: [(M + H) +]: 211.0756, found: 211.0748. 6-chloro-9-((3aS,6aR)-2,2-dimethyl-6-vinyltetrahydro-3aH-cyclopenta[d][1,3]dioxol-4- yl)-8-ethyl-9H-purine (34). To a solution of 17 (0.4 g, 2.2 mmol), triphenylphosphine (1.4 g, 5.2 mmol) and 28 (0.5 g, 2.6 mmol) in THF (100 mL) was added DIAD (diisopropyl azodicarboxylate) (1.1 g, 5.2 mmol) dropwise at 0 oC. The reaction mixture was kept at 0 oC for 2 hours followed by stirring at 50 oC for 8 hours. The solvent was removed under reduced pressure and the residue was purified with column chromatography (hexane: ethyl acetate 4:1) to give 34 (1.3 g) as yellow solid contaminated with by-product of diisopropyl azodicarboxylate as confirmed by mass spectrometry and was used as such in the next step. ESI-MS calcd for C17H21N4O2Cl: [(M + H) +]: 349.1431, found: 349.1429. ((3aR,6aS)-6-(6-chloro-8-ethyl-9H-purin-9-yl)-2,2-dimethyltetrahydro-3aH- cyclopenta[d][1,3]dioxol-4-yl)methanol (35). To a suspension of 34 from above and sodium periodate (1.3 g 3.9 mmol) in methanol (35 mL) and water (20 mL) was added OsO4 (30 mg, 0.1 mmol) at 0 oC. The suspension was stirred for 2 hours at 0 oC and then for 3 h at room temperature. The suspension was filtered and the solid was washed with EtOAc. The combined filtrates were concentrated and the residue was diluted with water (20 mL) and extracted with 96 EtOAc (3 x 40 mL). The combined organic layers were dried over MgSO4. After filtration, the filtrate was evaporated at ambient temperature and dissolved in MeOH (40 mL). This solution was cooled to 0 oC and NaBH4 (0.4 g, 5.9 mmol) was added portion wise. After 10 min, the solvent was removed by reduced pressure and the residue was neutralized by saturated ammonium chloride solution followed by extraction with EtOAc. The combined organic layers were washed with brine and dried over MgSO4. After filtration, the solvent was removed under reduced pressure and the residue was purified via column chromatography using (hexane: EtOAc 1:2) to afford (0.4 g, 49%) of 35 as white solid. 1H NMR (400 MHz, CDCl3): ? 8.58 (s, 1H), 5.19 (t, J = 6.4, 6.4 Hz, 1H), 4.72-4.61 (m, 2H), 4.05 (q, 1H), 3.81 (d, J = 5.6 Hz, 2H), 2.97 (q, 2H), 2.68 (q, 1H), 2.42 (m, 1H), 2.29 (m, 1H), 1.51 (s, 3H), 1.42 (s, 3H), 1.24 (s, 3H); 13C NMR (100 MHz, CDCl3) ? 159.9, 152.7, 150.7, 149.8, 113.7, 83.4, 82.4, 64.5, 62.8, 45.7, 32.4, 27.9, 25.3, 22.1, 12.2. ESI-MS calcd for C16H21N4O3Cl: [(M + H) +]: 353.1380 found: 353.1372. ((3aR,6aS)-6-(6-amino-8-ethyl-9H-purin-9-yl)-2,2-dimethyltetrahydro-3aH- cyclopenta[d][1,3]dioxol-4-yl)methanol (36). A solution of 35 (0.2 g, 0.5 mmol) was dissolved in MeOH (30 mL) and was cooled to 0 oC for 30 min before being saturated with NH3 at the same temperature for 1 h. The solution was heated to 100 oC for 48 h in a sealed stainless steel Parr. The solvent was removed under reduced pressure, and the residue purified by column chromatography (EtOAc/MeOH = 8:1) to give 36 as a white solid. 1H NMR (400 MHz, MeOD) ? 8.13 (s, 1H), 5.26 (t, 1H), 4.80-4.73 (m, 1H), 4.70-4.67 (q, 1H), 3.76 (dd, 1H), 3.71 (dd, 1H), 2.98 (q, 2H, CH2), 2.69 (q, 1H), 2.41-2.29 (m, 2H), 1.55 (s, 3H), 1.43 (t, 3H), 1.29 (s, 3H); 13C NMR (100 MHz, MeOD) ? 156.3, 155.9, 152.6, 151.7, 119.3, 114.8, 84.0, 82.9, 64.3, 63.0, 47.4, 97 33.8, 27.9, 25.4, 22.0, 12.1. ESI-MS calcd for C16H23N5O3: [(M + H) +]: 334.1879, found: 334.1864. (1R, 2S)-3-(6-amino-8-ethyl-9H-purin-9-yl)-5-(hydroxymethyl) cyclopentane-1,2-diol (2). Compound 36 was dissolved in a mixture of 1N HCl (6mL) and MeOH (6 mL). The mixture was stirred at room temperature for 3 h. The solution was then neutralized with weakly basic exchange resin (Amberlite IRA-67). Filtration followed with evaporation of solvent, the crude product was purified by chromatography (EtOAc/MeOH/NH3:H2O = 8:1:1) to give (90 mg, 64 %) of 2 as white crystals. 1H NMR (400 MHz, DMSO-d6) ? 8.05 (s, 1H), ? 6.98 (s, 2H), 4.84 (d, J = 6.8, 1H), 4.75 (t, 1H), 4.69-4.63 (m, 2H), 4.53 (q, 1H), 3.87-3.84 (dd, 1H), 3.60-3.55 (quintet, 1H), 3.48-3.40 (quintet, 1H), 2.87 (q, 2H), 2.11-2.03 (m, 3H), 1.30 (t, 3H); 13C NMR (100 MHz, DMSO) ? 155.2, 153.6, 151.0, 150.4, 118.2, 72.9, 71.5, 63.3, 60.1, 45.6, 27.3, 20.7, 12.0. ESI- MS calcd for C13H19N5O3: [(M + H) +]: 294.1566, found: 294.1564. 6-Chloro-9-(2,2-dimethyl-4,6a-dihydro-3aH-cyclopenta[1,3]dioxol-4-yl)-8-methyl-9H- purine (37). To a solution of 18 (0.2 g, 1.4 mmol), triphenylphosphine (0.9 g, 3.4 mmol) and 21 (0.3 g., 1.7 mmol) in THF (100 mL) was added DIAD (diisopropyl azodicarboxylate) (0.7 mL, 3.4 mmol) dropwise at 0 oC. The reaction mixture was kept at 0 oC for 2 hours followed by stirring at 50 oC for 48 h. The solvent was removed under reduced pressure and the residue was purified with column chromatography (hexane: ethyl acetate 1:1) to give (0.5 g, 69%) of 37 as yellow oil. 1H NMR (400 MHz, CDCl3) ? 8.60 (s, 1H), ? 6.31-6.28 (dt, 1H), ? 5.84-5.82 (dd, 1H), ? 5.75-5.73 (dd, 1H), ? 5.58 (br s, 1H), ? 5.01 (d, J = 5.6 Hz, 1H), ? 2.77 (s, 3H), ? 1.52 (s, 3H), ? 1.39 (s, 3H); 13C NMR (100 MHz, CDCl3) ? 154.8, 152.6, 150.9, 148.9, 137.3, 130.9, 98 128.9, 112.2, 85.6, 83.0, 66.8, 27.3, 25.6, 15.2. ESI-MS calcd for C14H15N4O2C1: [(M + H) +]: 307.0962 found: 307.0967. 9-(2,2-Dimethyl-4,6a-dihydro-3aH-cyclopenta[1,3]dioxol-4-yl)-8-methyl-9H-purin-6- ylamine (38). A solution of 37 (0.4 g, 1.3 mmol) was dissolved in MeOH (30 mL) and was cooled to 0 oC for 30 min before being saturated with ammonia gas at the same temperature for 1 h. The solution was heated to 100 oC for 48 h in a sealed stainless steel Parr. The solvent was removed under reduced pressure, and the residue purified by column chromatography (EtOAc/MeOH = 8:1) to give (0.3 g, 77%) of 38 as a white solid. 1H NMR (400 MHz, CDCl3) ? 8.09 (s, 1H), ? 6.59 (s, 2H), ? 6.13-6.11 (dt, 1H), ? 5.69-5.67 (dd, 1H), ? 5.60-5.58 (dd, 1H), ? 5.38 (br s, 1H), ? 4.73 (d, 1H), ? 2.47 (s, 3H), ? 1.40 (s, 3H), ? 1.27 (s, 3H). 13C NMR (60 MHz, CDCl3) ? 154.7, 151.8, 150.6, 148.5, 136.1, 129.6, 118.1, 111.8, 85.3, 83.03, 66.0, 27.1, 25.3, 13.9. ESI-MS calcd for C14H17N5O2: [(M + H) +]: 288.1461 found: 288.1476. 5-(6-Amino-8-methyl-purin-9-yl)-cyclopent-3-ene-1,2-diol (3). Compound 38 (0.2 g, 0.6 mmol) was dissolved in a mixture of 1N HCl (6mL) and MeOH (6 mL). The mixture was stirred at room temperature for 3 h. The solution was then neutralized with weakly basic exchange resin (Amberlite IRA-67). Filtration followed with evaporation of solvent, the crude product was purified by chromatography (EtOAc/MeOH/NH3:H2O = 8:2:1) to give (69.8 mg, 54 %) of 3 as white crystals. 1H NMR (400 MHz, DMSO-d6) ? 8.01 (s, 1H), ? 6.98 (br s, 2H), ? 6.10-6.17 (m, 1H), ? 6.02 (dd, 1H), ? 5.35-5.32 (m, 1H), ? 5.05 (d, J = 7.6Hz, 1H), ? 4.92 (d, J = 5.6 Hz, 1H), ? 4.50-4.46 (m, 1H), ? 4.43-4.38 (m, 1H), ? 2.51 (s, 3H); 13C NMR (100 MHz, DMSO-d6) ? 155.0, 99 151.37, 150.7, 149.0, 133.8, 133.6, 117.8, 75.5, 72.1, 64.6, 14.6. ESI-MS calcd for C11H13N5O2: [(M + H) +]: 248.1147 found: 248.1148. 6-Chloro-9-(2,2-dimethyl-4,6a-dihydro-3aH-cyclopenta[1,3]dioxol-4-yl)-8-ethyl-9H- purine (39). To a solution of 18 (0.2 g, 1.3 mmol), triphenylphosphine (1.9 g, 3.1 mmol) and 28 (0.3 g., 1.5 mmol) in THF (100 mL) was added DIAD (diisopropyl azodicarboxylate) (1.0 mL, 3.1 mmol) dropwise at 0 oC. The reaction mixture was kept at 0 oC for 2 hours followed by stirring at 50 oC for 48 h. The solvent was removed under reduced pressure and the residue was purified with column chromatography (hexane: ethyl acetate 1:1) to give 0.7 g of 39 as yellow oil, contaminated with diisopropyl azodicarboxylate by-product, that was used without further purification in the next reaction. ESI-MS calcd for C15H17N4O2Cl: [(M + H) +]: 321.1118 found: 321.1120. 9-(2,2-Dimethyl-4,6a-dihydro-3aH-cyclopenta[1,3]dioxol-4-yl)-8-ethyl-9H-purin-6- ylamine (40). A solution of 39 from the above step was dissolved in MeOH (30 mL) and was cooled to 0 oC for 30 min before being saturated with ammonia gas at the same temperature for 1 h. The solution was heated to 100 oC for 48 h in a sealed stainless steel Parr. The solvent was removed under reduced pressure, and the residue purified by column chromatography (EtOAc/MeOH = 8:1) to give (200 mg, 45 %) of 40 as a yellow oil. 1H NMR (400 MHz, MeOD) ? 8.01 (s, 1H), ? 6.06-6.04 (dt, 1H), ? 5.62 (d, 1H), 5.52 (d, 1H), ? 5.29 (br s, 1H), ? 4.08 (d, 1H), 2.92 (g, 2H), ? 1.43 (s, 3H), ? 1.38 (t, 3H), ? 1.31 (s, 3H). 13C NMR (100 MHz, MeOD) ? 156.3, 155.7, 153.0, 152.4, 151.9, 147.5, 137.5, 131.4, 113.4, 87.2, 84.8, 67.7, 27.8, 25.9, 11.9. ESI-MS calcd for C15H19N5O2: [(M + H) +]: 302.1617 found: 302.1615. 100 5-(6-Amino-8-ethyl-purin-9-yl)-cyclopent-3-ene-1,2-diol (4). Compound 40 was dissolved in a mixture of 1N HCl (6mL) and MeOH (6 mL). The mixture was stirred at room temperature for 3 h. The solution was then neutralized with weakly basic exchange resin (Amberlite IRA-67). Filtration followed with evaporation of solvent, the crude product was purified by chromatography (EtOAc/MeOH/NH3:H2O = 8:2:1) to give (80 mg, 46 %) of 4 as yellow crystals. 1H NMR (400 MHz, CDCl3) ? 8.05 (, 1H), 6.21-6.17 (dt, 1H), 6.11-6.08 (dd, 1H), 5.52- 5.47 (m, 1H), 4.67-4.65 (m, 2H), 4.48 (q, 2H), 1.44 (t, 3H); 13C NMR (100 MHz, MeOD) ? 156.7, 156.4, 152.8, 152.1, 135.4, 134.9, 77.2, 74.3, 66.5, 58.5, 18.5, 11.9. ESI-MS calcd for C12H15N5O2: [(M + H) +]: 262.1304 found: 262.1309. 6-Chloro-9-(2,2-dimethyl-4,6a-dihydro-3aH-cyclopenta[1,3]dioxol-4-yl)-8-isopropyl-9H- purine (41). To a solution of 18 (0.1 g, 0.4 mmol), triphenylphosphine (0.2 g, 0.9 mmol) and 29 (90.7 mg, 0.5 mmol) in THF (100 mL) was added DIAD (diisopropyl azodicarboxylate) (0.2 mL, 0.9 mmol) dropwise at 0 oC. The reaction mixture was kept at 0 oC for 2 hours followed by stirring at 50 oC for 48 h. The solvent was removed under reduced pressure and the residue was purified with column chromatography (hexane: ethyl acetate 1:1) to give 0.4 g of 41 as yellow oil. 1H NMR (400 MHz, CDCl3) ? 8.50 (s, 1H), ? 6.22-6.20 (dt, 1H), ? 5.72-5.67 (m, 2H), ? 5.45 (br s, 1H), ? 4.93 (d, 1H), ? 3.34-3.27 (sep, 1H), ? 1.45 (t, 9H), ? 1.32 (s, 3H); 13C NMR (100 MHz, CDCl3) ? 162.9, 152.9, 151.0, 149.6, 137.6, 131.4, 129.1, 112.5, 86.1, 83.4, 66.8, 27.6, 25.8, 22.0, 21.5. ESI-MS calcd for C16H20N4O2Cl: [(M + H) +]: 335.1275 found: 335.1272. 9-(2,2-Dimethyl-4,6a-dihydro-3aH-cyclopenta[1,3]dioxol-4-yl)-8-isopropyl-9H-purin-6- ylamine (42). A solution of 41 (440 mg) was dissolved in MeOH (30 mL) and was cooled to 0 101 oC for 30 min before being saturated with ammonia gas at the same temperature for 1 h. The solution was heated to 100 oC for 48 h in a sealed stainless steel Parr. The solvent was removed under reduced pressure, and the residue purified by column chromatography (EtOAc/MeOH = 8:1) to give (120 mg, 82 % over 2 steps) of 42 as a white solid. 1H NMR (400 MHz, CDCl3) ? 8.14 (s, 1H), ? 6.19-6.17 (dt, 1H), ? 6.09 (s, 2H), ? 5.72-5.69 (m, 2H), ? 5.39 (br s, 1H), ? 4.96 (d, 1H), ? 3.21 (sep, 1H), ? 1.46 (s, 3H), ? 1.39 (d, 3H), ? 1.37 (d, 3H), ? 1.33 (s, 3H); 13C NMR (100 MHz, CDCl3) ? 157.3, 154.9, 152.0, 151.0, 136.6, 130.0, 118.8, 112.1, 86.0, 83.5, 66.2, 27.5, 27.0, 25.7, 21.7, 21.6. 5-(6-Amino-8-isopropyl-purin-9-yl)-cyclopent-3-ene-1,2-diol (5). Compound 42 (120 mg, 0.4 mmol) was dissolved in a mixture of 1N HCl (6mL) and MeOH (6 mL). The mixture was stirred at room temperature for 3 h. The solution was then neutralized with weakly basic exchange resin (Amberlite IRA-67). Filtration followed with evaporation of solvent, the crude product was purified by chromatography (EtOAc/EtOH/NH3:H2O = 8:1:1) to give 5 as white crystals in quantitative yield. 1H NMR (400 MHz, MeOD) ? 8.06 (s, 1H), ? 6.21-6.18 (dt, 1H), ? 6.08-6.06 (dd, 1H), ? 5.53 (d, 1H), ? 4.75 (t, 1H), ? 4.70-4.67 (dddd, 1H), ? 3.38 (sep, 1H), ? 1.46 (d, 3H), ? 1.40 (d, 3H); 13C NMR (100 MHz, MeOD) ? 158.8, 154.9, 151.2, 150.4, 134.0, 133.4, 117.8, 75.4, 72.8, 64.9, 26.4, 20.7, 20.6. ESI-MS calcd for C13H18N5O2: [(M + H) +]: 276.1461 found: 276.1461. 8-tert-Butyl-6-chloro-9-(2,2-dimethyl-4,6a-dihydro-3aH-cyclopenta[1,3]dioxol-4-yl)-9H- purine (43). To a solution of 18 (0.4 g, 2.2 mmol), triphenylphosphine (1.4 g, 6.5 mmol) and 30 (0.6 g., 2.7 mmol) in THF (100 mL) was added DIAD (diisopropyl azodicarboxylate) (1.1 mL, 102 6.5 mmol) dropwise at 0o C. The reaction mixture was kept at 0oC for 2 hours followed by stirring at 50 oC for 72 h. The solvent was removed under reduced pressure and the residue was purified with column chromatography (hexane: ethyl acetate 1:1) to give 0.1 g of crude 43 as yellow oil that was used as such in the next step. ESI-MS calcd for C17H22N4O2Cl: [(M + H) +]: 349.1431, found: 349.1422. 8-tert-Butyl-9-(2,2-dimethyl-4,6a-dihydro-3aH-cyclopenta[1,3]dioxol-4-yl)-9H-purin-6- ylamine (44). A solution of 43 from the above step 0.1 g was dissolved in MeOH (30 mL) and was cooled to 0 oC for 30 min before being saturated with ammonia gas at the same temperature for 1 h. The solution was heated to 100 oC for 48 h in a sealed stainless steel Parr. The solvent was removed under reduced pressure, and the residue purified by column chromatography (EtOAc/MeOH = 8:1) to give 40 mg of 44 with a yield of 51% over two steps as a white solid. 1H NMR (400 MHz, CDCl3): ? 8.18 (s, 1H), 6.24-6.20 (dt, 1H), 5.80-5.66 (m, 3H), 5.08 (d, 1H), 2.07 (s, 2H), 1.58 (s, 9H), 1.51 (s, 3H), 1.38 (s, 3H); 13C NMR (100 MHz, CDCl3) ? 158.8, 154.9, 152.0, 147.9, 129.8, 112.1, 86.6, 83.5, 68.3, 60.7, 34.6, 30.0, 27.6, 25.8. ESI-MS calcd for C17H24N5O2: [(M + H) +]: 330.1930, found: 330.1906. 5-(6-Amino-8-tert-butyl-purin-9-yl)-cyclopent-3-ene-1, 2-diol (6). Compound 44 was dissolved in a mixture of 1N HCl (6mL) and MeOH (6 mL). The mixture was stirred at room temperature for 3 h. The solution was then neutralized with weakly basic exchange resin (Amberlite IRA-67). Filtration followed with evaporation of solvent, the crude product was purified by chromatography (EtOAc/EtOH/NH3:H2O = 8:1:1) to give 6 as white crystals. 1H NMR (400 MHz, CDCl3): ? 8.07 (s, 1H), 5.82 (dt, 1H), 5.73 (m, 1H), 5.03 (m, 2H), 4.92 (m, 103 1H), 3.96 (s br, 2H), 1.49 (s, 9H); 13C NMR (100 MHz, MeOD) ? 159.6, 155.2, 152.4, 151.3, 135.1, 133.0, 73.8, 73.7, 67.8, 34.4, 29.7. 22.7. ESI-MS calcd for C14H20N5O2: [(M + H) +]: 290.1617, found: 290.1600. 8-butyl-6-chloro-9H-purine (46). To a solution of 19 (0.5 g, 3.5 mmol) in dry toluene, valeroyl chloride (1.4 mL, 3.5 mmol) was added dropwise. The solution was refluxed for 4 h. The solvent was evaporated under reduced pressure. The residue was tested via mass spectroscopy and contained a mixture of both the monoacylated product 45 (major) as well as the bisacylated product. ESI-MS calcd for C9H14N4OCl: [(M + H) +]: 229.0856, found: 229.0855. The residue was refluxed in 1 N NaOH for 2 h. The solution was evaporated and the residue was purified with column chromatography EtOAc/MeOH = 15:1 to give (0.23 g, 31 %) of 46 as white solid. 1H NMR (400 MHz, CDCl3): ? 13.29 (s, 1H), ? 8.69 (s, 1H), ? 3.07 (t, 2H), ? 1.87 (quintet, 2H), ? 1.39 (m, 2H), ? 0.89 (t, 3H); 13C NMR (100 MHz, MeOD) ? 154.8, 152.9, 152.1, 150.8, 130.2, 30.1, 29.7, 22.5, 13.8. ESI-MS calcd for C9H12N4Cl: [(M + H) +]: 211.0750, found: 211.0778. 8-butyl-6-chloro-9-((3aS,6aR)-2,2-dimethyl-4,6a-dihydro-3aH-cyclopenta[d][1,3]dioxol- 4-yl)-9H-purine (47). To a solution of 18 (0.1 g, 0.7 mmol), triphenylphosphine (0.4 g, 1.7 mmol) and 46 (0.2 g., 0.9 mmol) in THF (100 mL) was added DIAD (diisopropyl azodicarboxylate) (0.2 mL, 1.7 mmol) dropwise at 0 oC. The reaction mixture was kept at 0 oC for 2 hours followed by stirring at 50 oC for 36 h. The solvent was removed under reduced pressure and the residue was purified with column chromatography (hexane: ethyl acetate 1:1) to give 0.1 g of crude 47 as yellow oil that was used as such in the next step 1H NMR (400 MHz, 104 CDCl3): ? 8.52 (s, 1H), 6.22-6.20 (dt, 1H), 5.82 (s, 1H), 5.82-5.67 (m, 2H), 5.41 (s, 1H), 2.95 (t, 2H), 1.83 (quin, 2H), 1.45 (s, 3H), 136 (q, 2H), 1.32 (s, 3H), 0.94 (t, 3H) ; 13C NMR (100 MHz, CDCl3) ? 158.2, 152.4, 150.6, 149.0, 137.2, 128.7, 112.1, 85.6, 82.9, 74.0, 66.6, 29.7, 28.0, 27.1, 25.4, 22.4, 13.6. ESI-MS calcd for C17H22N4O2Cl: [(M + H)+] : 349.1431, found : 349.1422. 8-butyl-9-((3aS,6aR)-2,2-dimethyl-4,6a-dihydro-3aH-cyclopenta[d][1,3]dioxol-4-yl)-9H- purin-6-amine (48). A solution of 47 from the above step 0.1 g was dissolved in MeOH (30 mL) and was cooled to 0 oC for 30 min before being saturated with ammonia gas at the same temperature for 1 h. The solution was heated to 100 oC for 48 h in a sealed stainless steel Parr. The solvent was removed under reduced pressure, and the residue purified by column chromatography (EtOAc/MeOH = 8:1) to give (0.13 g, 25 % yield over two steps) of 48 as a white solid. 1H NMR (400 MHz, CDCl3): ? 8.15 (s, 1H), 6.20 (dt, J = 2.8, 1.8 Hz, 1H), 6.05 (br s, 2H), 5.72-5.69 (m, 2H), 5.37 (br s, 1H), 4.95 (d, J = 5.6 Hz, 1H), 2.84 (t, J = 8, 7.6 Hz, 2H), 1.77 (quin, 2H), 1.46 (s, 3H), 1.43 (q, 2H), 1.33 (s, 3H), 0.93 (t, J = 7.6, 7.2 Hz, 3H); 13C NMR (100 MHz, CDCl3): ? 154.8, 152.9, 152.1, 151.0, 136.5, 130.0, 118.8, 112.1, 83.5, 86.0, 66.4, 30.2, 28.1, 27.5, 25.7, 22.9, 13.94. ESI-MS calcd for C17H24N5O2: [(M + H) +]: 330.1930, found: 330.1939. (1S,2R)-5-(6-amino-8-butyl-9H-purin-9-yl)cyclopent-3-ene-1,2-diol (7). Compound 48 (70 mg) was dissolved in a mixture of 1N HCl (6mL) and MeOH (6 mL). The mixture was stirred at room temperature for 3 h. The solution was then neutralized with weakly basic exchange resin (Amberlite IRA-67). Filtration followed with evaporation of solvent, the crude product was purified by chromatography (EtOAc/EtOH/NH3:H2O = 8:1:1) to give 7 (120 mg) as white 105 crystals. 1H NMR (400 MHz, CDCl3): ? 8.08 (s, 1H), 6.19-5.97 (m, 4H), 5.42 (dd, J = 2, 1.6 Hz 1H), 4.81-4.78 (m, 1H), 4.66 (t, J = 6, 5.6 Hz, 1H), 2.80 (br s, 1H), 2.78 (t, J = 8, 5.6 Hz, 2H), 2.58 (s, 1H), 1.71 (quintet, 2H), 1.35 (sextet, 2H), 0.88 (t, J = 7.2, 7.2 Hz, 3H); 13C NMR (100 MHz, CDCl3): ? 154.81, 152.89, 152.1, 151.0, 136.5, 130.0, 118.8, 86.1, 83.5, 65.3, 30.1, 27.8, 22.5, 13.8. ESI-MS calcd for C14H20N5O2: [(M + H) +]: 290.1617, found: 290.1616. 106 References (1) Garrett, R. H.; Grisham, C. M. Principles of biochemistry: with a human focus; Brooks/Cole Pub Co, 2001. (2) Neidle, S.; ScienceDirect Principles of nucleic acid structure; Elsevier, 2008. (3) Agrofoglio, L.; Suhas, E.; Farese, A.; Condom, R.; Richard Challand, S.; A. Earl, R.; Guedj, R. Tetrahedron 1994, 50, 10611. (4) Volpini, R.; Dal Ben, D.; Lambertucci, C.; Marucci, G.; Mishra, R.; Ramadori, A.; Klotz, K. N.; Trincavelli, M.; Martini, C.; Cristalli, G. ChemMedChem 2009, 4, 1010. (5) Cristalli, G.; Eleuteri, A.; Vittori, S.; Volpini, R.; Lohse, M. J.; Klotz, K. N. Journal Of Medicinal Chemistry 1992, 35, 2363. (6) Seley, K. L.; Mosley, S. L.; Zeng, F. Organic Letters 2003, 5, 4401. (7) Rodriguez, J. B.; Comin, M. J. Mini Reviews in Medicinal Chemistry 2003, 3, 95. (8) Borthwick, A. D.; Biggadike, k. Tetrahedron 1992, 48, 571. (9) Yuan, C.-S.; Liu, S.; Wnuk, S. F.; Robins, M. J.; Borchardt, R. T.; Clercq, E. D. In Advances in Antiviral Drug Design; Elsevier: 1996; Vol. Volume 2, p 41. (10) Shealy, Y. F.; Clayton, J. D. Journal of the American Chemical Society 1966, 88, 3885. (11) Kusaka, T.; Yamamoto, H.; Shibata, M.; Muroi, M.; Kishi, T. The Journal of Antibiotics,1968,21,255. 107 (12) Hayashi, M.; Yaginuma, S.; Yoshioka, H.; Nakatsu, K. The Journal of Antibiotics 1981, 34, 675. (13) Cools, M.; De Clercq, E. Biochemical Pharmacology 1990, 40, 2259. (14) Wolfe, M. S.; Borchardt, R. T. Journal of Medicinal Chemistry 1991, 34, 1521. (15) Borcherding, D. R.; Narayanan, S.; Hasobe, M.; McKee, J. G.; Keller, B. T.; Borchardt, R. T. Journal Of Medicinal Chemistry 1988, 31, 1729. (16) Guranowski, A.; Montgomery, J. A.; Cantoni, G. L.; Chiang, P. K. Biochemistry 1981, 20, 110. (17) De Clercq, E.; Cools, M.; Balzarini, J.; Marquez, V. E.; Borcherding, D. R.; Borchardt, R. T.; Drach, J. C.; Kitaoka, S.; Konno, T. Antimicrobial Agents and Chemotherapy 1989, 33, 1291. (18) Glazer, R. I.; Knode, M. C. Journal of Biological Chemistry 1984, 259, 12964. (19) Turner, M.; Yang, X.; Yin, D.; Kuczera, K.; Borchardt, R.; Howell, P. Cell Biochemistry and Biophysics 2000, 33, 101. (20) Cantoni, G. L. Journal of Biological Chemistry 1953, 204, 403. (21) Borchardt, R. T. Journal of Medicinal Chemistry 1980, 23, 347. (22) Banerjee, A. K. Microbiology and Molecular Biology Reviews 1980, 44, 175. (23) Mao, X.; Schwer, B.; Shuman, S. Molecular and Cellular Biology 1995, 15, 4167. (24) Lentz, S. R. Life Sciences 1997, 61, 1205. (25) McKeever, M. P.; Weir, D. G.; Molloy, A.; Scott, J. M. Clinical Science 1991, 81, 551. (26) Abeles, J. L. P. R. H. Journal of Biological Chemistry 1976, 251, 5817. (27) Palmer, J. L.; Abeles, R. H. Journal of Biological Chemistry 1979, 254, 1217. 108 (28) Lee, K. M.; Choi, W. J.; Lee, Y.; Lee, H. J.; Zhao, L. X.; Lee, H. W.; Park, J. G.; Kim, H. O.; Hwang, K. Y.; Heo, Y.-S.; Choi, S.; Jeong, L. S. Journal of Medicinal Chemistry 2011, 54, 930. (29) Sumin Cai, Q.-S. L., Jianwen Fang,Ronald T. Borchardt,; Krzysztof Kuczera, C. R. M., and Richard L. Schowen Nucleosides, Nucleotides and Nucleic Acids 2009, 28, 485. (30) Liu, S.; Wolfe, M. S.; Borchardt, R. T. Antiviral Research 1992, 19, 247. (31) Wolfe, M. S.; Lee, Y.; Bartlett, W. J.; Borcherding, D. R.; Borchardt, R. T. Journal of Medicinal Chemistry 1992, 35, 1782. (32) Borcherding, D. R.; Scholtz, S. A.; Borchardt, R. T. Journal of Organic Chemistry 1987, 52, 5457. (33) Chiang, P. K., Pharmacology and Therapeutics 1998, 77, 115. (34) De Clercq, E. Antimicrobial Agents and Chemotherapy 1985, 28, 84. (35) Saenger, W.; Springer Advanced Ttexts in Chemistry. Springer-Verlag, New York, NY: 1984. (36) Rosemeyer, H.; Toth, G.; Golankiewicz, B.; Kazimierczuk, Z.; Bourgeois, W.; Kretschmer, U.; Muth, H. P.; Seela, F. Journal of Organic Chemistry 1990, 55, 5784. (37) Tvaroska, I.; Hricov?ni, M.; Petr?kov?, E. Carbohydrate Research 1989, 189, 359. (38) Blackburn, G. M. Nucleic acids in chemistry and biology; Royal Society of Chemistry, 2006. (39) Kilpatrick, J. E.; Pitzer, K. S.; Spitzer, R. Journal of the American Chemical Society 1947, 69, 2483. (40) Hall, L. D.; Steiner, P. R.; Pedersen, C. Canadian Journal of Chemistry 1970, 48, 1155. 109 (41) Thibaudeau, C.; Chattopadhyaya, J. Stereoelectronic effects in nucleosides and nucleotides and their structural implications; Uppsala University Press Uppsala, 1999. (42) Altona, C.; Sundaralingam, M. Journal of the American Chemical Society 1972, 94, 8205. (43) Altona, C.; Sundaralingam, M. Journal of the American Chemical Society 1973, 95, 2333. (44) Saenger, W. Principles of Nucleic Acid Structure, C.R. Cantor (Ed.); Springer Advanced Texts in Chemistry. Springer-Verlag, New York, NY, 1984. (45) Westhof, E.; R?der, O.; Croneiss, I.; L?demann, H. D. Zeitschrift f?r Naturforschung. Section C: Biosciences 1975, 30, 131. (46) Davies, D. B.; Danyluk, S. S. Biochemistry 1974, 13, 4417. (47) Sundaralingam, M. Annals of the New York Academy of Sciences 1975, 255, 3. (48) Aoyagi, M.; Minakawa, N.; Matsuda, A. Tetrahedron Letters 1993, 34, 103. (49) Rhodes, L. M.; Schimmel, P. R. Biochemistry 1971, 10, 4426. (50) Stolarski, R.; Dudycz, L.; Shugar, D. European Journal of Biochemistry 1980, 108, 111. (51) Blackburn, B. J.; Grey, A. A.; Smith, I. C. P.; Hruska, F. E. Canadian Journal of Chemistry 1970, 48, 2866. (52) Rosemeyer, H.; Zulauf, M.; Ramzaeva, N.; Becher, G.; Feiling, E.; M?hlegger, K.; M?nster, I.; Lohmann, A.; Seela, F. Nucleosides, Nucleotides and Nucleic Acids 1997, 16, 821. (53) Crimmins, M. T. Tetrahedron 1998, 54, 9229. (54) Ikehara, M.; Uesugi, S.; Yoshida, K. Biochemistry 1972, 11, 830. 110 (55) Sarma, R. H.; Lee, C. H.; Evans, F. E.; Yathindra, N.; Sundaralingam, M. Journal of the American Chemical Society 1974, 96, 7337. (56) Bookser, B. C.; Matelich, M. C.; Ollis, K.; Ugarkar, B. G. Journal Of Medicinal Chemistry 2005, 48, 3389. (57) Van Aerschot, A. A.; Mamos, P.; Weyns, N. J.; Ikeda, S.; De Clercq, E.; Herdewijn, P. A. Journal Of Medicinal Chemistry 1993, 36, 2938. (58) Van der Werten, E. M.; Hartog-Witte, H. R.; Roelen, H. C. P. F.; von Frijtag Drabbe K?nzel, J. K.; Pirovano, I. M.; Math?t, R. A. A.; Danhof, M.; Van Aerschot, A.; Lidaks, M. J.; Ijzerman, A. P.; Soudijn, W. European Journal of Pharmacology: Molecular Pharmacology 1995, 290, 189. (59) Palczewski, K.; Kahn, N.; Hargrave, P. A. Biochemistry 1990, 29, 6276. (60) Hasobe, M.; McKee, J. G.; Borcherding, D. R.; Borchardt, R. T. Antimicrobial agents and chemotherapy 1987, 31, 1849. (61) Keller, B. T.; Clark, R. S.; Pegg, A. E.; Borchardt, R. T. Molecular Pharmacology 1985, 28, 364. (62) M. J. Frisch, G. W. T., H. B. Schlegel, G. E. Scuseria,; M. A. Robb, J. R. C., J. A. Montgomery, Jr., T. Vreven,; K. N. Kudin, J. C. B., J. M. Millam, S. S. Iyengar, J. Tomasi,; V. Barone, B. M., M. Cossi, G. Scalmani, N. Rega,; G. A. Petersson, H. N., M. Hada, M. Ehara, K. Toyota,; R. Fukuda, J. H., M. Ishida, T. Nakajima, Y. Honda, O. Kitao,; H. Nakai, M. K., X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross,; V. Bakken, C. A., J. Jaramillo, R. Gomperts, R. E. Stratmann,; O. Yazyev, A. J. A., R. Cammi, C. Pomelli, J. W. Ochterski,; P. Y. Ayala, K. M., G. A. Voth, P. Salvador, J. J. Dannenberg,; V. G. Zakrzewski, S. D., A. D. Daniels, M. C. Strain,; O. Farkas, D. K. M., A. D. 111 Rabuck, K. Raghavachari,; J. B. Foresman, J. V. O., Q. Cui, A. G. Baboul, S. Clifford,; J. Cioslowski, B. B. S., G. Liu, A. Liashenko, P. Piskorz,; I. Komaromi, R. L. M., D. J. Fox, T. Keith, M. A. Al-Laham,; C. Y. Peng, A. N., M. Challacombe, P. M. W. Gill,; B. Johnson, W. C., M. W. Wong, C. Gonzalez, and J. A. Pople,; Gaussian, I., Wallingford CT, 2004. 2004. (63) Becke, A. D. Physical Review A 1988, 38, 3098. (64) Lee, C.; Yang, W.; Parr, R. G. Physical Review B 1988, 37, 785. (65) Tomasi, J.; Persico, M. Chemical Reviews 1994, 94, 2027. (66) Dunning, T. H. Journal of Chemical Physics 1989, 90, 1007. (67) Akdag, A.; Carver, C. M.; McKee, M. L.; Schneller, S. W. Journal of Physical Chemistry A 2002, 106, 11254. (68) Sun, G.; Voigt, J. H.; Filippov, I. V.; Marquez, V. E.; Nicklaus, M. C. Journal of chemical information and computer sciences 2004, 44, 1752. (69) Arpalahti, J.; D. Klika, K.; Sillanpaa, R.; Kivekas, R. Journal of the Chemical Society, Dalton Transactions 1998, 1397. (70) Ciuffreda, P.; Casati, S.; Manzocchi, A. Magnetic Resonance in Chemistry 2007, 45, 781. (71) Stolarski, R.; Pohorille, A.; Dudycz, L.; Shugar, D. Biochimica et Biophysica Acta (BBA) - Nucleic Acids and Protein Synthesis 1980, 610, 1. (72) Miyashita, O.; Kasahara, F.; Kusaka, T.; Marumoto, R. The Journal of Antibiotics 1985, 38, 981. (73) Madhavan, G. V. B.; Martin, J. C. Journal of Organic Chemistry 1986, 51, 1287. 112 (74) Montgomery, J. A.; Clayton, S. J.; Thomas, H. J.; Shannon, W. M.; Arnett, G.; Bodner, A. J.; Kion, I. K.; Cantoni, G. L.; Chiang, P. K. Journal Of Medicinal Chemistry 1982, 25, 626. (75) Yang, M.; Zhou, J.; Schneller, S. W. Tetrahedron Letters 2004, 45, 8981. (76) Ludek, O. R.; Marquez, V. E. Synthesis 2007, 2007, 3451. (77) Rodriguez, J. B.; Marquez, V. E.; Nicklaus, M. C.; Mitsuya, H.; Barchi, J. J. Journal Of Medicinal Chemistry 1994, 37, 3389. (78) Palmer, Andreas M.; J?ger, V. European Journal Of Organic Chemistry 2001, 2001, 1293. (79) Jin, Y. H.; Liu, P.; Wang, J.; Baker, R.; Huggins, J.; Chu, C. K. Journal of organic chemistry 2003, 68, 9012. (80) Paquette, L. A.; Bailey, S. Journal of Organic Chemistry 1995, 60, 7849. (81) Yang, M.; Ye, W.; Schneller, S. W. Journal of organic chemistry 2004, 69, 3993. (82) Grubbs, R. H.; Chang, S. Tetrahedron 1998, 54, 4413. (83) Poeylaut-Palena, A. A.; Testero, S. A.; Mata, E. G. Journal of Organic Chemistry 2008, 73, 2024. (84) Hale, K. J.; Domostoj, M. M.; Tocher, D. A.; Irving, E.; Scheinmann, F. Organic Letters 2003, 5, 2927. (85) Kurteva, V. B.; Afonso, C. A. M. Chemical Reviews 2009, 109, 6809. (86) Hong, S. H.; Grubbs, R. H. Journal of the American Chemical Society 2006, 128, 3508. (87) Casey, C. P. Journal of Chemical Education 2006, 83, 192. (88) Schmalz, H.-G. Angewandte Chemie International Edition in English 1995, 34, 1833. (89) Trnka, T. M.; Grubbs, R. H. Accounts of Chemical Research 2000, 34, 18. 113 (90) Lister, J. H. Syntheses from Pyrimidines; John Wiley & Sons, Inc., 2008; Vol. 24. (91) Montgomery, J. A. Journal of the American Chemical Society 1956, 78, 1928. (92) Koppel, H. C.; Robins, R. K. Journal of Organic Chemistry 1958, 23, 1457. (93) Mornet, R.; Leonard, N. J.; Theiler, J. B.; Doree, M. Journal of the Chemical Society, Perkin Transactions 1 1984, 879. (94) Lister, J. H.; Fenn, M. D. Synthesis from Pyrimidines; John Wiley & Sons, Inc., 2008; Vol. 54. (95) Ueda, T.; Nomoto, Y.; Matsuda, A. Chemical & Pharmaceutical Bulletin 1985, 33, 3263. (96) Kitade, Y.; Kozaki, A.; Miwa, T.; Nakanishi, M. Tetrahedron 2002, 58, 1271. (97) Barton, D. H. R.; Hedgecock, C. J. R.; Lederer, E.; Motherwell, W. B. Tetrahedron Letters 1979, 20, 279. (98) Hayakawa, H.; Haraguchi, K.; Tanaka, H.; Miyasaka, T. Chemical & Pharmaceutical Bulletin 1987, 35, 72. (99) Rosemeyer, H.; Seela, F. Journal of the Chemical Society, Perkin Transactions 2 1997, 2341. (100) Johnson, C. R.; Chen, Y. F. Journal of organic chemistry 1991, 56, 3344. 114 Appendix Data for linear regression calculation of AA: #/bin/ksh # touch EE.txt for ((a=0.0; a<=1.0; a=a+0.01)) do b=$((1.0-$a)) min1=$((7.8*$a+0.0*$b-6.1)) echo $(( $min1*$min1 )) > MIN1 min2=$((6.4*$a+6.1*$b-5.3)) echo $(( $min2*$min2 )) > MIN2 min3=$((-0.2*$a+7.8*$b-3.4)) echo $(( $min3*$min3 )) > MIN3 if (( $a+$b <= 1.0 )) then mint=$(( $min1*$min1 + $min2*$min2 + $min3*$min3 )) 115 echo $mint > MINT echo $a > A echo $b > B paste A B MINT|awk '{printf ("%5.2f%5.2f%8.2f\n",$1,$2,$3)}' >> EE.txt fi done /bin/rm MIN1 MIN2 MIN3 MINT A B