DESIGN, SYNTHESIS, AND IMMUNOLOGICAL STUDIES TOWARD THE DEVELOPMENT OF CARBOHYDRATE BASED CANCER VACCINES By Philip Allen Bentley A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Chemistry 2011 ABSTRACT DESIGN, SYNTHESIS, AND IMMUNOLOGICAL STUDIES TOWARD THE DEVELOPMENT OF CARBOHYDRATE BASED CANCER VACCINES By Philip Allen Bentley Cancer is a disease that affects nearly everyone, either directly and/ or indirectly. In 2008 alone, there were nearly 12.7 million new cancer cases reported and close to 7.6 million people lost their battle with this disease worldwide. In the past several years, unique carbohydrate antigens known as tumor-associated carbohydrate antigens (TACAs) were found to be expressed at abnormal levels on cancer cell surfaces differentiating them from normal cells. This makes them viable targets for vaccine design. One such TACA is the Tn antigen. The immunological responses of this B-cell epitope conjugated to T-cell epitopes such as Tobacco Mosaic Virus (TMV), a mutant of TMV, and the Q capsid were evaluated. Additionally, several control vaccines were synthesized and immunologically evaluated to determine the affect of the linkers during an immune response. Analysis of the vaccine constructs by enzyme-linked immunosorbent assay (ELISA) revealed discrepancies between the antibody titers measured for the control and experimental populations. This led to the investigation into the measured background absorbance using each vaccine construct. Changing from an ELISA plate coated with NeutrAvidin to one coated with bovine serum albumin (BSA) resulted in a much lower background. Finally, three glyco-polymers conjugated to the surface of a protein carrier, expressing the Tn antigen in varying degrees, were synthesized in hopes of further enhancing the immune response elicited. ACKNOWLEDGEMENTS I would like to start by thanking my advisor, Professor Xuefei Huang for his patience and guidance during the past three years of my graduate studies. He was instrumental in my growth as a researcher and person and will pave the foundation of my future studies. I also give special thanks to my guidance committee members, Professors William Wulff and Kevin Walker for their critical insight into my research and their valuable time spent serving on my committee. Dr. Daniel Holmes and Mr. Kermit Johnson played vital roles in my NMR training and molecular analysis, Dr. Daniel Jones and Lijun Chen were very helpful in training me how to use the mass spectrometer, and Dr. Jeannie Gaymer shared her expertise in the handling and wellbeing of the mice used during the course of my studies. I also thank those present and past members of the Huang group: Bo, Dino, Gilbert, Gopinath, Herbert, Hongguang, Hovig, Keisuke, Medha, Moe, Qian, Steve, Xiaowei (Vivian), and Zhaojun who, on several occasions, were more than willing to assist me in my day-to-day research. In addition to these group members, undergraduate researchers Sarah McKee and Claire Baniel and high school student David Huynh gave me the opportunity and pleasure to teach and learn from them. Special thanks to Dr. Medha Kamat and her husband Dr. Bala Ramanathan for all the hours they excluded from their own research to help teach me techniques and concepts that will be invaluable in my future research endeavors. Many thanks to Dr. Zhaojun Yin for his dedication and contributions to the vaccine project, specifically the work he contributed to the understanding of the ELISA results, and Dr. Adeline Miermont, whose work in the laboratory prior to my arrival set the foundation for my research project. I thank all the friends I came to know during the course of my graduate studies who helped keep me positive when it seemed as though nothing in the laboratory worked as I hoped. iii All of the soccer games, bowling nights, and just plain hanging out watching Michigan State sports or cheering for our respective countries in the World Cup gave me a much better appreciation for the different cultures; I hope that I too have reciprocated this. I must thank my parents, sister, and grandparents who were always there to support me over the past many years of my education. And last, but certainly not least, I express my deepest thanks to my new wife, Ashley, who was always emotionally in my corner each night I came home either disappointed or perhaps even when I was a bit difficult to be around. iv TABLE OF CONTENTS LIST OF TABLES ......................................................................................................................... vi LIST OF FIGURES ...................................................................................................................... vii LIST OF SCHEMES...................................................................................................................... xi LIST OF ABBREVIATIONS ...................................................................................................... xiii CHAPTER 1 – Design and Synthesis of a Glycopolymer Based Vaccine ..................................... 1 1.1 Introduction ...................................................................................................................... 1 1.2 An Account of On-going Research with TACAs ............................................................ 6 1.3 The Tn Antigen .............................................................................................................. 25 1.4 Synthesis of Tn Analogs ................................................................................................ 26 1.5 Synthesis of the Glycopolymer ...................................................................................... 39 1.6 Conjugation to KLH ...................................................................................................... 48 1.7 Conjugation to TMV and Immunological Studies ......................................................... 53 1.8 Conjugation to Q and Immunological Studies............................................................. 68 1.9 Re-evaluation of the TMV-1Cys Construct Using New ELISA Protocol ..................... 79 1.10 Synthesis and Immunological Studies of a Triazole Linker ........................................ 81 1.11 Conclusions .................................................................................................................. 87 1.12 Experimental Section ................................................................................................... 89 Appendix 1 – Spectral Data ................................................................................................ 107 References ........................................................................................................................... 130 CHAPTER 2 – Synthesis of an Alzheimer’s Disease Imaging Agent........................................ 136 2.1 Introduction .................................................................................................................. 136 2.2 Synthesis of Alzheimer’s Imaging Agent .................................................................... 138 2.3 Experimental Section ................................................................................................... 144 Appendix 2 – Spectral Data ................................................................................................ 151 References ........................................................................................................................... 169 v LIST OF TABLES Table 1.1: Major TACAs Identified in Cancer Tissues ................................................................. 4 Table 1.2: IgM and IgG ELISA titers after vaccination with vaccines 16 and 17 ....................... 23 vi LIST OF FIGURES Figure 1.1: Tumor-Associated Carbohydrate Antigens ................................................................. 7 Figure 1.2: Clustered Tn antigen and conjugated to protein carrier using MBS linker ................. 9 Figure 1.3: Unimolecular trivalent antigen constructs ................................................................. 12 x y Figure 1.4: Dimeric Lewis Lewis antigen construct.................................................................. 14 Figure 1.5: Maleimide and bromoacetamido linkers ................................................................... 15 Figure 1.6: MUC1-STn construct 13 ........................................................................................... 19 Figure 1.7: MUC1-STn construct 14 ........................................................................................... 20 Figure 1.8: Structure of Boons’ first generation three component vaccine (15).......................... 21 Figure 1.9: Boons' second generation three-component vaccine constructs 16 and 17 ............... 22 Figure 1.10: Structure of RAFT construct ................................................................................... 24 Figure 1.11: Native Tn antigen .................................................................................................... 25 Figure 1.12: Tn analogs 19, 20, and 21 ........................................................................................ 27 Figure 1.13: KLH conjugated to glycopolymer based vaccine 38 ............................................... 38 Figure 1.14: Structures of polymers 46A, 46B, and 46C ............................................................ 46 Figure 1.15: Structures of glycopolymers 47A ............................................................................ 47 Figure 1.16: Standard curves for KLH and respective sugars ..................................................... 51 Figure 1.17: Native TMV protein ................................................................................................ 53 Figure 1.18: Structure of TMV-Tn-PEG-N3 conjugate using the native TMV ......................... 56 Figure 1.19: TMV Mutant ............................................................................................................ 57 Figure 1.20: Antibody Titers for Study 1: all mice ...................................................................... 62 Figure 1.21: IgG titers after 35 days using native and mutant TMV constructs .......................... 63 vii Figure 1.22: IgG Antibody Titers for Study 2: all mice............................................................... 64 Figure 1.23: IgG titers at Day 35: Study 1 vs. Study 2 ................................................................ 65 Figure 1.24: Competitive ELISA vs Normal ELISA: IgG Antibodies specific for GalNAc....... 66 Figure 1.25: IgG Antibody Titers for Study 3: all mice............................................................... 72 Figure 1.26: Evaluation of each step in the ELISA protocol ....................................................... 73 Figure 1.27: IgG Antibody Titers against BSA-Tn @ 1: 6400 serum dilutions for Study 3 ....... 75 Figure 1.28: IgG titers for representative mice in Study 3 and native TMV-Tn-PEG-N3 .......... 76 Figure 1.29: IgM titers for representative mice in Study 3 and native TMV-Tn-PEG-N3 .......... 77 Figure 1.30: Antibody Titers for Study 1: all mice using BSA- Tn for ELISA........................... 79 Figure 1.31: BSA conjugated triazole compound for ELISA analysis ........................................ 84 Figure 1.32: IgG antibodies measured against BSA-triazole in TMV study groups ................... 85 Figure 1.33: IgG antibodies measure against BSA- triazole in Q Study groups ....................... 86 1 Figure 1.34: 600 MHz (D2O), H NMR of 19........................................................................... 108 Figure 1.35: 125 MHz (D2O), 13 C NMR of 19 ......................................................................... 109 1 Figure 1.36: 600 MHz (D2O), H NMR of 20........................................................................... 110 1 Figure 1.37: 150 MHz (D2O), H NMR of 20........................................................................... 111 1 Figure 1.38: 500 MHz, (CDCl3), H NMR of 41 ...................................................................... 112 Figure 1.39: 150 MHz, (CDCl3), 13 C NMR of 41 .................................................................... 113 1 Figure 1.40: 600 MHz, (CDCl3), H NMR of 42 ...................................................................... 114 Figure 1.41: 150 MHz, (CDCl3), 13 C NMR of 42 .................................................................... 115 viii 1 Figure 1.42: 600 MHz, (D2O), H NMR of 46A ....................................................................... 116 1 Figure 1.43: 500 MHz, (D2O), H NMR of 47A ....................................................................... 117 1 Figure 1.44: 300 MHz, (D2O), H NMR of 46B ....................................................................... 118 1 Figure 1.45: 600 MHz, (D2O), H NMR of 47B ....................................................................... 119 1 Figure 1.46: 500 MHz, (D2O), H NMR of 46C ....................................................................... 120 1 Figure 1.47: 600 MHz (D2O), H NMR of 52........................................................................... 121 Figure 1.48: 600 MHz (D2O), 13 C NMR of 52 ......................................................................... 122 Figure 1.49: 600 MHz (D2O), gCOSY NMR of 52................................................................... 123 1 Figure 1.50: 600 MHz (D2O), H NMR of 53........................................................................... 124 Figure 1.51: 150 MHz (D2O), 13 C NMR of 53 ......................................................................... 125 Figure 1.52: 600 MHz (D2O), COSY NMR of 53..................................................................... 126 Figure 1.53: 600 MHz (D2O), HMQC NMR of 53 ................................................................... 127 Figure 1.54: 600 MHz (D2O), HMBC NMR of 53 ................................................................... 128 Figure 2.1: Structures of FENE and FDDNP ............................................................................. 137 1 Figure 2.2: 500 MHz, (CD3)2CO, H NMR of 2 ...................................................................... 152 13 Figure 2.3: 150 MHz, (CD3)2CO, C NMR of 2 ..................................................................... 153 1 Figure 2.4: 500 MHz, (CD3CD), H NMR of 10 ...................................................................... 154 Figure 2.5: 125 MHz, (CD3CD), 13 C NMR of 10 ..................................................................... 155 ix 1 Figure 2.6: 300 MHz, (CDCl3), H NMR of 11 ........................................................................ 156 Figure 2.7: 150 MHz, (CDCl3), 13 C NMR of 11 ...................................................................... 157 1 Figure 2.8: 600 MHz, (CDCl3), H NMR of 12 ........................................................................ 158 Figure 2.9: 150 MHz, (CDCl3), 13 C NMR of 12 ...................................................................... 159 1 Figure 2.10: 500 MHz, (CDCl3), H NMR of 13 ...................................................................... 160 Figure 2.11: 125 MHz, (CDCl3), 13 C NMR of 13 .................................................................... 161 Figure 2.12: 500 MHz, (CDCl3), gCOSY of 13 ........................................................................ 162 Figure 2.13: Confirmation of azide stretch using IR of 13 ........................................................ 163 1 Figure 2.14: 500 MHz, (CD3OD), H NMR of 15 .................................................................... 164 Figure 2.15: 150 MHz, (CD3OD), 13 C NMR of 15 .................................................................. 165 Figure 2.16: 600 MHz, (CD3OD), gCOSY of 15 ...................................................................... 166 Figure 2.17: 600 MHz, (CD3OD), HMQC of 15 ....................................................................... 167 x LIST OF SCHEMES Scheme 1.1: Hydrolysis of the Maleimide linker ......................................................................... 17 Scheme 1.2: Synthesis of globally protected 25 ........................................................................... 28 Scheme 1.3: Synthesis of donor 27 .............................................................................................. 28 Scheme 1.4: Synthesis of acceptor 29 .......................................................................................... 29 Scheme 1.5: Synthesis of fully protected Tn building block 31................................................... 31 Scheme 1.6: Synthesis of deprotected common intermediate 34 ................................................. 33 Scheme 1.7: Synthesis of Tn analog 19........................................................................................ 36 Scheme 1.8: Synthesis of azido linker 37 ..................................................................................... 37 Scheme 1.9: Synthesis of Tn analog 20........................................................................................ 38 Scheme 1.10: Retrosynthetic analysis of polymerizable design of 21 ......................................... 40 Scheme 1.11: Generation of TMS protected propargyl acrylamide ............................................. 42 Scheme 1.12: Synthesis of copolymer 46..................................................................................... 44 Scheme 1.13: Synthesis of glycopolymer 47B ............................................................................. 47 Scheme 1.14: Synthesis of KLH conjugated glycopolymer vaccine 38B .................................... 49 Scheme 1.15: Synthesis of TMV-Tn-N3 conjugate using the native TMV ................................ 55 Scheme 1.16: Synthesis of Tn conjugates using mutant TMV .................................................... 59 Scheme 1.17: Synthesis of azido conjugates using Q ................................................................ 69 Scheme 1.18: Synthesis of triazole linker (51) ............................................................................. 82 Scheme 1.19: Synthesis of triazole compounds for ELISA analysis ........................................... 83 Scheme 2.1: Synthesis of 2-6 -substituted aromatic alcohol 2 ................................................... 138 Scheme 2.2: Synthesis of 10 utilizing the Bucherer reaction ..................................................... 139 Scheme 2.3: Predicted mechanism for the synthesis of 10 using the Bucherer reaction ........... 140 xi Scheme 2.4: Synthesis of the desired FDDNP imaging agent (14) ............................................ 142 Scheme 2.5: Synthesis of final imaging agent to be conjugated to magnetic nanoparticles ...... 143 xii LIST OF ABBREVIATIONS AcOH Acetic acid Ac2O Acetic anhydride AD Alzheimer’s Disease AgCl Silver chloride ATRP Atom Transfer Radical Polymerization Bn Benzyl BnBr Benzyl bromide Boc tert-Butyloxycarbonyl BOP Benzotriazole-1-yl-oxy-tris(dimethylamino)-phosphonium hexafluorophosphate BSA Bovine serum albumin cat. Catalytic Cl3CCN Trichloroacetonitrile Ce(NH4)2(NO3)6 Ammonium cerium (IV) nitrate (NH4)6Mo7O24 * 4H2O Ammonium molybdate tetrahydrate CFA Complete Freund's Adjuvant conc. Concentrated COSY Correlation Spectroscopy CPMV Cowpea mosaic virus CuSO4*5H2O Copper (II) Sulfate pentahydrate CuI Copper (I) Iodide xiii d Day DCM Dichloromethane DBU 1,8-Diazabicyclo[5.4.0]undec-7-ene DIPEA Diisopropylethylamine DMAP 4-Dimethylaminopyridine DMF Dimethylformamide DMSO Dimethyl sulfoxide dOSM Desialylated ovine submaxillary mucin EDC. HCl 1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide Hydrochloride ELISA Enzyme-linked immunosorbent assay EtOAc Ethyl acetate FACS Fluorescence-assisted cell sorting FITC Fluorescein isothiocyanate Fmoc 9-Fluorenylmethoxycarbonyl Gal Galactose GalNAc N-Acetyl galactosamine Glc Glucose h Hour H2 Hydrogen gas H2O Water H2SO4 Sulfuric acid HBF4 Fluoroboric acid xiv HCl Hydrochloric acid HMBC Heteronuclear multiple bond correlation HMQC Heteronuclear multiple quantum coherence HOBt Hydroxybenzotriazole HRMS High Resolution Mass Spectrometry HRP Horseradish peroxidase IFA Incomplete Freund's Adjuvant Ig Immunoglobulin K2CO3 Potassium carbonate KHCO3 Potassium bicarbonate KDa Kilo-Dalton KLH Keyhole limpet haemocyanin NHS N-Hydroxysuccinimide ester NH3 Ammonia MeOH Methanol NaOCN Sodium cyanate NaHSO3 Sodium bisulfite NaN3 Sodium azide NaNO2 Sodium nitrite NMP N-Methylpyrrolidone NMR Nuclear Magnetic Resonance OAc Acetate xv OVA Ovalbumin PBS Phosphate Buffered Saline PBST Tween 20 surfactant in Phosphate Buffered Saline py Pyridine RAFT Regioselectivity Addressable Functionalized Template RNA Ribonucleic acid RPM Revolutions per minute rt Room temperature SBA Soybean agglutinin S-CPMV Mutant form of the CPMVcapsid bearing cysteine residues sec Second Ser Serine TACA Tumor-associated carbohydrate antigen TBAI Tetrabutylammonium iodide TEA Triethylamine THF ` Tetrahydrofuran Thr Threonine TLC Thin layer chromatography TMB 3,3’,5,5’-Tetramethylbenzidine TMSOTf Trimethylsilyl trifluoromethanesulfonate TMV Tobacco mosaic virus TMV-Tn-N3 Native TMV Tn conjugate linked to 20 xvi TMV-Tn-PEG-N3 Native TMV- Tn conjugate w/ PEG linker TMV-1Cys Mutant TMV w/ inserted cysteine residue TMV-1Cys-Tn-N3 Mutant TMV- Tn conjugate linked to 20 TMV-1Cys-Tn-bromo Mutant TMV- Tn conjugate linked w/ bromoacetamide linker (48) Tn (C) Tn clustered Tos- Cl 4-methylbenzenesulfonyl chloride TTX Tetanus toxin YAF YAFKYARHANVGRNAFELFL peptide Zn Zinc xvii Chapter 1 – Design and Synthesis of a Glycopolymer Based Vaccine 1.1 Introduction Cancer is a disease that impacts nearly all of the World’s population, either directly and/ 1 or indirectly. According to a report published by the American Cancer Society, cancer is the second leading cause of death in economically developed countries behind only heart disease, and the third leading cause of death in developing countries (following heart diseases and diarrheal diseases). Based on data collected in 2008, the International Agency for Research on Cancer reported nearly 12.7 million new cancer cases and close to 7.6 million people lost their 2 battle with this disease worldwide. This is equal to around 13% of all deaths globally. These numbers are predicted to grow drastically as the world’s population continues to grow and age. It is predicted that by 2050, there could be as many as 27 million new cases and 17.5 million deaths caused by cancer. 1 The current treatment of the various forms of cancer employs such methods as chemotherapy, radiation, and surgery. One major drawback of these current methods, however, is they often do not completely eradicate the tumor cells. Additionally, by the time the cancer is diagnosed, it may be too far progressed to be treatable. As the statistics prove, more effective therapeutic interventions are necessary. One such alternative is immunotherapy because it harnesses the body’s natural defense mechanisms; however, immunotherapy also comes with challenges. For instance, the immune system has developed mechanisms to protect itself from attacking the cells and tissues of the 3, 4 host, which is known as self-tolerance. A few of these examples include negative selection of 1 B cells and T cells in the bone marrow and thymus respectively, expression of tissue-specific proteins in the thymus, suppression of autoimmune responses by regulatory T cells, and the induction of anergy in autoreactive B and T cells. 4 This self tolerance causes tumor associated antigens to be only weakly immunogenic. To develop a vaccine that can avoid self-tolerance, it becomes crucial to fully understand the complexity of the immune system. In brief, when a pathogen, or antigen (in the case of immunotherapy), enters the body, there is an immediate response known as innate immunity. Once the foreign pathogen or microbe has invaded the body, there will be a secondary response, known as adaptive immunity. Both the innate and the adaptive branches of the immune system are mutually important toward the prevention of the disease state. However, the adaptive branch is of most interest toward developing a successful vaccine. Once inside the body, the foreign pathogen will be recognized, up-taken, and digested by circulating cells, know as antigen presenting cells, more specifically macrophages and dendritic cells. These cells will present fragments of these digested antigens on their cell surface known as epitopes. Then, these antigen presenting cells will display the epitopes to naïve B cells (a mature B cell that has not yet been exposed to its specific antigen). Once exposed for the first time, these naïve B cells will become activated B cells that can begin secreting low affinity antibodies of the IgM subtype. This initial response is an important step because it indicates to the researcher that the antigen is being recognized by the host as non-self. This response transient, lasting only a few weeks; therefore, a successful vaccine for application in cancer therapy requires a long-term immunogenic response. For this to occur there must be a class switch from low-affinity IgM isotypes to highaffinity IgG isotypes. This can occur when both a naïve B-cell and naïve T cell recognize the same pathogen or antigen and differentiate into an activated B cell and an effector T cell, 2 respectively. Effector T cells are divided into two classes known as CD4 and CD8 T cells corresponding to the surface protein they present. CD8 T cells, otherwise known as Tc cells or cytotoxic T cells, will circulate and monitor the surface of the host cells for the antigen that they were first activated from. If recognized, the CD8 T cell will bind to the infected cell and cause it to undergo apoptosis, otherwise known as programmed cell death. CD4 T cells, otherwise known as Th cells or T helper cells are designed to help further activate antigen presenting cells. Activated B cells, in addition to secreting low-affinity IgM antibodies, also become antigen presenting cells and can display antigens to these effector CD4 T cells by means of an MHC class II complex. When this happens, the CD4 T cells will bind, and begin releasing small molecule cytokines into the microenvironment, which will cause the B cells to undergo an isotype class switch and begin secreting higher affinity antibodies of the IgG subtype. The complexity of this pathway contributes to the challenges of designing a synthetic strategy toward a successful vaccine. For example, in the study described herein, the desired immune response is developed utilizing a carbohydrate epitope, which are not recognized by T cell receptors; therefore, the activated B cells cannot display antigens to the CD4 T cells along the pathway to IgG isotypes. Because these carbohydrates can only be recognized by B cells, they are considered B cell epitopes. Considering that T cell epitopes can only recognize protein peptides (T cell epitopes), it is believed that a carbohydrate antigen capable of being recognized by a naïve B cell conjugated to a protein, which contains these Th peptide epitopes. should induce activation of B cells and their subsequent interaction with CD4 T cells. When the B cell/ CD4complex is formed, the effector T cell secrete cytokines that induce isotype switching which in turn cause the generation of memory B cells. These memory B cells are desired for long-term immunity, whereby a stronger and faster immune response is induced each time the body is 3 exposed to the same antigen. Knowledge of the mechanism of this complicated communication within the immune system is absolutely necessary when developing an effective immunotherapy. Despite the complexities and difficulties of overcoming intrinsic immunological tolerance in cancer therapy, great interest remains toward treating the disease by targeting Tumor-Associated Carbohydrate Antigens (TACAs). It is known that tumor cells express specific antigens on their cell surface that are be recognized by the immune system, including antigenic carbohydrates. Tumor cells can be distinguished from normal cells by the overexpression of unique oligosaccharides on their cell surfaces. Additionally, different TACAs are shared amongst a variety of different cancer cell lines (Table 1.1). This makes them an intriguing target for cancer immunotherapy. Thus, targeting TACAs on the surfaces of small metastatic cells (cells that have spread from the primary tumor site to other tissues) can potentially improve the efficiency of treatment by eliciting humoral immune responses (one involving B cells) hopefully generating immuno-surveillance. This immuno-surveillance can allow the immune system to recognize cancer cells at an early stage and eliminate them before the cause of disease. This could allow for the treatment of cancer, but more importantly, even preventing it all together. Table 1.1: Major TACAs Identified in Cancer Tissues Cancer Type Melanoma Neuroblastoma Sarcoma B cell lymphoma Small-cell lung Breast Prostate Lung Colon Ovary Stomach Tumor Antigens GM2, GM3, GD2 GM2,GD2,polysialic acid GM2, GD2,GD3 GM2, GD2 GM2, fucosyl-GM1, polysialic acid, Globo-H GM2, Globo-H, TF(c), Tn(c), LeY GM2, Globo-H, Tn(c), TF(c), sTn(c), LeY GM2, Globo-H, LeY GM2,Tn, sTn(c), TF(c), LeY GM2, Globo-H, sTn(c), TF(c), LeY GM2, LeY, LeA, sialyl-LeA 4 However, a challenge of using TACAs in immunotherapy is, that alone, they only weakly activate the antibody secreting B cells. As mentioned above, this weak activation and lack of T cell help prevents isotype switching; thus, only low titers of low affinity IgM antibodies can be generated. To strengthen the antigenicity of TACAs, it therefore is important to link them to an immunogenic carrier that contains epitopes recognizable by CD4 T cells. Once activated, these effector T cells then recognize matched antigens presented by activated B cells and bind to them. This T cell/ B cell complex induces cytokine production by the T cells into the surrounding microenvironment. These cytokines allow additional activation of B cells which undergo isotype class switching from low affinity IgM to high affinity IgG type antibodies. Several studies continue to try and elucidate the complex mechanisms of cancer so an immunotherapy approach can be further honed and optimized. 5 1.2 An Account of On-going Research with TACAs Two protein carriers extensively examined and found to contain helper T-cell epitopes are bovine serum albumin (BSA,) and Keyhole Limpet Haemocyanin (KLH). These macromolecular carriers provide the opportunity for isotype switching and generation of immunological memory which allows the immune system to make a quicker, stronger adaptive immune response to each successive encounter to an antigen. Danishefsky and coworkers y attached several different TACA epitopes, including Globo-H (1), Lewis (2), Fucosyl GM1 (3), 5 TF (4), and Tn (5) antigens, to BSA and KLH (Figure 1.1). 6 Figure 1.1: Tumor-Associated Carbohydrate Antigens Deprotected precursors of the epitopes (Figure 1.1) were designed to be bioconjugated with a protein carrier. These precursors were synthesized to display an aldehyde linker at the reducing end of the sugar moiety which by means of reductive amination 6, 7 were conjugated to the desired protein carrier. Despite the difficulty of controlling the number of antigens attached 7 to the surface of these large protein carriers, encouraging results were observed from in vivo 8 mouse studies. One study by Danishefsky and coworkers involved synthesizing 6 as a control containing three copies of the Tn antigen (Tn-cluster, designated Tn (c)) and vaccines 7 as the experimental vaccine linked to both BSA and KLH (7-BSA and 7-KLH) to demonstrate the advantage of using a protein carrier (Figure 1.2). Both IgG and IgM antibodies, isolated from mouse serum, made against Tn (c) at pre-vaccination isolated for all three of the in vivo vaccines were not detectable by ELISA method, as expected. Ten days following Danishefsky’s two week injection into mice, the ELISA analysis of antibody titers against Tn (c) induced by compound 6 indicated IgM titers of 1,350 (antibody titers is considered a unit-less number) and IgG titers of 150, the 7-BSA complex of 1,350 and 150, and the 7-KLH complex of 12,150 and 450, respectively. These data suggested that KLH functions as a better carrier than BSA. Interestingly, 317 copies of their Tn cluster attached to the surface of KLH as opposed to only 7 copies attached to BSA. Antibody detection by Fluoresence-Activated Cell Sorting (FACS) analysis gave comparable data to the ELISA, except the effects of the two protein carriers were essentially the same. These findings demonstrate that a successful vaccine would require a protein carrier containing a T-cell epitope, to elicit a stronger immune response. 8 Figure 1.2: Clustered Tn antigen and conjugated to protein carrier using MBS linker Considering a protein carrier induced an immune response as a T-cell epitope, Danishefsky and coworkers explored multivalent vaccines, compounds containing more than one antigen, to generate an even stronger immune response. One major shortcoming of previous studies was the lack of significant immune responses mediated by the T-cells likely resulting from the monovalency of the vaccine. They compared immune responses of vaccines, monomer 9 5 (Tn) to trivalent 7 (Tn (c)) (Figure 1.1 and Figure 1.2), in mouse hosts and analysis of the 9 serum by ELISA showed IgG tiers of 30 and 320, respectively, measured against desialylated ovine submaxillary mucin (dOSM; a natural source of Tn), and IgM ELISA titers of 40 and 1,280 for 5 and 7, respectively. It was concluded based on these findings that multivalency has a crucial role in the immune response generated. They also reconfirmed in this study that KLH, elicited a better immune response than BSA. As mentioned above, it remained unknown whether these results were caused by less antigen conjugation to BSA compared to that of KLH; however, based on two different batches of Tn (c)- KLH tested, the authors state, that the “…method and intensity of conjugation plays a major role in increasing the immunogenicity of the antigen.” Danishefsky and coworkers continued the multivalency concept by designing and synthesizing the first multi-antigenic glyco-conjugates 8 and 9 (Figure 1.3). 10 These multi- antigenic constructs were conjugated to KLH, using the same linker as mentioned above, generating the desired vaccines which were then injected into mice. Overall the antibody titers generated in this study were quite low, however, it is interesting to note that antibodies were generated not only against the trivalent constructs but also against each individual antigen portion. 10 They also injected only the trivalent constructs as a control and found that only IgM type antibodies are induced further supporting that a protein carrier is required for the induction of a T-cell response. These findings suggested that combining different antigens, into a multicomponent vaccine, which might be over-expressed on the surface of a certain cancer cell type, could help produce a stronger more efficient immune response than that of a monovalent antigenic vaccine. This earlier work describes multi-antigenic constructs as better mimics of the natural organization of the TACA environment on the surface of cancer cells. They predicted that the organization of the various over-expressed carbohydrate motifs on the cancer cell 10 surfaces is very heterogeneous much like a vaccine presenting multiple antigens. Furthermore, it was also mentioned that vaccine 9 generated a lesser immune response than that of vaccine 8. Danishefsky’s group also found that multiple α-O-linked GalNAc based glycans in the same construct results in an increase in structural rigidity. 11 These findings suggested that vaccine 9, which contained these α-O-GalNAc linkages, should be more rigid than that of 8, causing 9 to more closely match mucin glycoproteins. They suggested that this close similarity between the host mucin glycoproteins and the structural core of vaccine 9, can prevent an immune response from being generated due to the immunological tolerance within the hosts, discussed above. 11 Figure 1.3: Unimolecular trivalent antigen constructs The results from the proof-of-principle study encouraged Danishefsky et a. to synthesize y a first generation unimolecular pentavalent construct displaying the Tn, TF, STn, Lewis , and 12 Globo-H antigens and unimolecular hexavalent construct containing the Tn, TF, STn, GM2, y Lewis , and Globo-H antigens. 12 Modeling the assays used to test the immunogenicity of the trivalent constructs, antibodies were generated against each unimolecular multivalent vaccine, unfortunately the titers analyzed by ELISA were much lower than they had predicted. More recently, the same group synthesized a second generation unimolecular pentavalent construct displaying the Tn, TF, STn, GM2, and Globo-H antigens which is expected to enter phase 1 of 13 clinical trials. The synthetic effort accomplished in the foregoing studies to access a viable cancer vaccine is an amazing step forward in this field. As mentioned above, the IgG antibody titers expressed by the immune response were relatively modest; thus, additional research is needed to further increase the production of the desired IgG antibodies and, more importantly, the generation of long-term immunological memory. Intrigued by the positive results of multi-antigenic vaccinations, Boons and co-workers tested whether the antibodies generated were against the entire construct, each individual y x antigen, or both. They synthesized the dimeric Lewis Lewis heptasaccharide construct (10) 14 (Figure 1.4) and then conjugated it to the protein carrier KLH. 13 x y Figure 1.4: Dimeric Lewis Lewis antigen construct Mice were immunized three times with compound 10 conjugated to KLH (KLH-10) and then, after ~28 days, sera were collected and ELISA was used to measure the antibody titers x y made against Lewis Lewis conjugated to BSA protein. ELISA plates were coated with BSAx y x y Lewis Lewis instead of KLH-Lewis Lewis as to not potentially skew the data by measuring antibodies generated against the protein carrier. They measured titers of IgG (31,645) that were ~30 times higher than those of the IgM subtype (1,060). Next, they evaluated the cross- reactivity of antibodies from their sera samples isolated earlier against BSA-linked separately to x y either Lewis or Lewis antigens. Boon et al. hypothesized that titers against the individual antigens would be similar to those measured for the dimeric compound KLH-10, suggesting that the sera antibodies made against KLH-10 also recognize the monomeric antigens with the same affinity. By contrast, however, IgG antibodies titers measured by ELISA against the BSA-linked y Lewis antigen (3,115) were ~10-fold lower than that for the dimeric form (31,645): IgM titers x were less than 120. By comparison, the titers of IgM for the BSA-linked Lewis antigen were 14 undetectable and those for IgG were at 500. These findings imply that it may be possible to design a specific vaccine to target a certain form of tumor. However, a major limitation to this approach is that to develop an efficient preventative vaccine would be daunting because several different forms of that exist. Even targeted treatment would have its major disadvantages; the design, synthesis, and manufacturing of the several specific compounds needed to address the specific cancer type would be overwhelming. Another important contribution to the field by Boons and coworkers was their work on understanding how different linkers played a role in immunological responses. They wanted to determine how different linkers between protein carrier and TACA affected immune response. 15 y They decided to use two different linkers, conjugating KLH to the Lewis tetrasaccharide. The first linker precursor they employed was 4-(maleimidomethyl)cyclohexane-1-carboxylate (11) because of its widespread use, and when incorporated into the final structure, it looks similar to the MBS linker used by Danishefsky’s group (Figure 1.2). The second linker precursor choice was the smaller and more flexible 3-(bromoacetamido)propionate (12) (Figure 1.5). Figure 1.5: Maleimide and bromoacetamido linkers 15 y In their first study, they injected the KLH-11-Lewis construct into mice, and sera were collected and analyzed by ELISA to determine antibody titers made against KLH-11 alone the y linker alone or the KLH-11-Lewis conjugate. They coated their ELISA plates with BSAy Lewis antigen conjugate, linked separately with each of the two linkers. BSA was used to avoid detection of anti-KLH antibodies. Although high IgG titers (358,928) against BSA-11-Lewis y were induced, they later deduced through control experiments that only a titer of 120 was against y the Lewis antigen. 15 During the conjugation, KLH was derivatized with linker 11 first, and y then Lewis was added to the terminal end of 11. This last reaction step is significant because any unbound linker would likely undergo hydrolysis to form maleamic acid (Scheme 1.1). Boons and coworkers had predicted the principally observed anti-linker immune response was caused by only the terminal hydrolyzed maleimides. To avoid this, they changed the order of y steps in their synthesis by first coupling the linker to Lewis antigen followed by conjugation to KLH. This would ensure that there would be no hydrolyzed linker present. After inoculating this compound in mice and measuring titers by ELISA, they found that the IgG titers against the y Lewis epitope were much better going from 120 to 10,066. Even with improvement in the immune response against the antigen epitope, titers against the linker were still ten times higher (106,894). 16 Scheme 1.1: Hydrolysis of the Maleimide linker y Linker 12 was used to construct KLH-12-Lewis , and this inoculum was injected into y mice. The titers of IgG antibodies against BSA-12-Lewis were much lower (3,957) than when linker 11 was used, however, it is interesting to note that only about 17% (663) of these titers were generated against the linker. These results clearly showed the impact of the linker-type when trying to generate an immune response. If a linker induces a strong immune response, thus producing antibodies against it, then there is a good chance that a desirable immune response may not be generated against the lesser immunogenic TACA. Others, like Kunz and coworkers, took a slightly more synthetically challenging approach to design and synthesize a vaccine that displays the STn antigen on a MUC1 (cell surface 16 associated mucin type) B-cell epitope linked to a T-cell peptide epitope from tetanus toxin (ttx) 17 (13) (Figure 1.6) or ovalbumin (OVA) (14) (Figure 1.7). Based on their findings after inoculating three mice with vaccine (14) (Figure 1.7), one mouse generated a significant immune response as measured by ELISA binding of the serum antibodies to MUC1-BSA coating the plates. More interestingly, the serum antibodies measured, after the third immunization, were significantly higher than those measured after the first and second, indicating that immunological memory was developing. To further verify the specificity of the antibodies generated, they tested the cross- reactivity of the serum samples to that of the tumor associated sialyl-Tn glycopeptide antigen (MUC1), an un-glycosylated MUC1 peptide antigen, sialyl-Tn 17 glycopeptides antigen variant (MUC4), and an un-glycosylated MUC4 peptide antigen. Based on their findings, it was determined that the antibodies were highly specific to the entire vaccine construct containing both carbohydrate and peptide sequence contained in vaccine 14 and not to the individual peptide or carbohydrate motifs. 18 Figure 1.6: MUC1-STn construct 13 19 Figure 1.7: MUC1-STn construct 14 20 Although not previously discussed, an adjuvant was used as a third component with each vaccine discussed above. Adjuvants are “substances used in experimental immunology and in 4 vaccines to enhance the adaptive immune response to an antigen.” This prevents a vaccine 18 from being cleared from the host prior to the immunological response. Boons and coworkers designed and synthesized a vaccine that combined all three of these important parts into one complete construct, represented in construct 15. Included were the Tn antigen as their B-cell epitope, the peptide YAFKYARHANVGRNAFELFL (YAF) as the T-cell epitope, and the lipopeptide Pam3Cys which acts as an immuno-adjuvant (Figure 1.8). Figure 1.8: Structure of Boons’ first generation three component vaccine (15) Due to the low solubility of this compound, it was incorporated into phospholipid based liposomes and then used to inoculate mice. Interestingly, once the antibody titers from the sera were measured by ELISA against BSA conjugated to Tn bound to a 96- well plate, IgM and IgG titers were low, but produced nevertheless. More importantly, however, the IgG titer was higher than the IgM titer indicting that isotype class switch had taken place. With this encouraging 19 proof-of-concept, Boons et al. synthesized a second-generation, three-component vaccine. This vaccine construct was conceived from the first generation construct discussed above, and 21 varied only slightly. First, the YAF T-cell epitope, derived from the outer-membrane protein of Neisseria mengitides, was exchanged with a T-cell epitope with one derived from the polio virus (103-KLFAVWKITYKDT-115). Second, the Tn antigen used as the B-cell epitope was changed to a tumor associated glycopeptide derived from MUC1. And finally, the Pam3Cys T-cell epitope was replaced with either Pam2CysSK4 (16) or Pam3CysSK4 (17) (Figure 1.9). Figure 1.9: Boons' second generation three-component vaccine constructs 16 and 17 After inoculating the vaccine constructs into mice with and without an additional adjuvant (QS-21), antibody titers were measured using ELISA plates coated with BSA-BrAcMUC1. After four immunizations, the IgG titers (without adjuvant added) ranged from 20,900 for vaccine 16 to 169,600 for vaccine 17 (Table 1.2). In addition, it appeared that adding the 22 adjuvant helped to stimulate the generation of IgG titers, thereby indicating isotype switching of the B-cells Table 1.2: IgM and IgG ELISA titers after vaccination with vaccines 16 and 17 Immunization Vaccine 16 Vaccine 16 & QS-21 Vaccine 17 Vaccine 17 & QS-21 IgM 1,400 1,100 7,200 5,000 IgG 20,900 30,200 169,600 322,800 Encouraged by these findings, Boons et al. tested the specificity of the IgG antibodies to the native MUC1 antigen present on cancer cells using flow cytometry. They found that the antisera elicited against vaccines 16 and 17 reacted very strongly to MCF7 human breast cancer 9 cells which are known to express MUC1, whereas they found no binding to SK-MEL-28 cell which do not contain MUC1. These results indicated that their vaccine construct was capable of not only generating high quantities of IgG antibodies but also that they were highly specific. Lo-Man and coworkers have investigated an alternative approach that avoids the requirement of a protein carrier and thus takes advantage regioselectively addressable 20 functionalized templates (RAFTS) (18) as display platforms for the Tn antigen (Figure 1.10). This construct has the advantage over other protein carriers in that it is easier to control of the number of antigen copies being expressed, consequently allowing for better vaccine reproducibility during synthesis. 23 Figure 1.10: Structure of RAFT construct 24 1.3 The Tn Antigen The Tn antigen was first reported in 1957 by Moreau, Dausset, and co-workers 21, 22 at a time when it was associated with an extremely rare disease “Tn syndrome,” which can be characterized by the tendency of certain red blood cells to be agglutinated by all human sera, regardless of the ABO blood group involved. 23 In a study carried out by Dr. Hustings of the Swiss Red Cross in 1994, of 150,000 donated blood samples screened, all came back negative for the disease. 24 It was not until about 20 years after Moreau identified the Tn antigen that 25, 26 Springer and co-workers associated it to carcinoma and determined its structure. native Tn antigen is characterized as GalNAc-α-O-serine/ threonine (Figure 1.11). Figure 1.11: Native Tn antigen 25 The 1.4 Synthesis of Tn Analogs The Tn antigen (Figure 1.1 and Figure 1.11) is expressed on a number of different cancer cell types, primarily breast, colon, and prostate. 27 Notably, the Tn antigen, when administered alone, is not immunogenic. Therefore, to generate a desirable immune response, it is paramount to conjugate the Tn antigen to a protein carrier to elicit a T-cell mediated immune response. The protein carriers principally used in the studies described in this thesis were the tobacco mosaic virus (TMV), enterobacteria phage Q, and Keyhole Limpet Hemocyanin (KLH). With this knowledge and understanding of the ongoing work within the field of immunotherapy, as well as the complexities behind the immune system, our lab developed several carbohydrate based vaccines. Each potential vaccine contained this Tn antigen conjugated to various protein carriers, in the presence of an adjuvant, in hopes of eliciting a strong humoral immune response. In the proceeding sections, these vaccines will be discussed in further detail, including their synthesis and analysis by in vivo immunological methods. To conjugate the Tn antigen to TMV-1Cys, Q, BSA, and my desired glycopolymer, discussed later in Section 1.5, as well as to immobilize Tn on a 96-well plate for ELISA assays, Tn analogs 19 (containing a biotin moiety), 20 (containing a terminal azide for crosslinking), and 21 (containing a terminal acryamide polymerization) were synthesized (Figure 1.12) from a common amine intermediate (34). 26 Figure 1.12: Tn analogs 19, 20, and 21 Amine (34) was synthesized starting from the commercially available D-galactosamine hydrochloride 23. Shelf-stable imidazole-1-sulfonyl azide hydrochloride 22 was used for diazo 28 transfer to make compound 24 (65% yield) (Scheme 1.2). Application of trifluoromethanesulfonyl azide more commonly used in the diazo transfer method was avoided since trifluoromethanesulfonyl azide can react explosively and needs to be freshly made each time. The white crystalline imidazole sulfonyl azide reagent 22 is advantageous since it is suitable for long-term storage and the risk of explosion is drastically reduced. Compound 24 was globally acetylated to afford 25 85% yield. This step was generally quite slow and usually required addition of excess Ac2O throughout the reaction. 27 Scheme 1.2: Synthesis of globally protected 25 Next, selective deacylation of 25 to liberate the anomeric hydroxyl group was accomplished using hydrazine acetate in DMF to yield 26. Then, 26 was reacted under basic conditions with trichloroacetonitrile 29 to afford 27 in 79% yield over two steps (Scheme 1.3). Scheme 1.3: Synthesis of donor 27 28 With donor 27 in hand, acceptor 29 was synthesized (94% yield) from the commercially available Fmoc-serine 28 by means of selective benzylation of the carboxylic acid residue using potassium bicarbonate, tetrabutylammonium iodide and benzyl bromine in DMSO (Scheme 30 1.4). Scheme 1.4: Synthesis of acceptor 29 The native Tn antigen contains only the α-linkage at the anomeric carbon; therefore, the 31 Schmidt’s glycosylimidate synthetic strategy 29 over the unnatural -anomer in high yield. was used to favor the formation of the α-anomer Both donor 27 and acceptor 29 were desiccated under vacuum overnight and then added to a mixture of dichloromethane to ether (1:1, v/v) previously stirred over freshly activated AW-300 molecular sieves for 2 h to remove residual o water which affects glycosylation (Scheme 1.5). The reaction was cooled to -30 C (dry ice in acetonitrile) and then trimethylsilyl trifluoromethanesulfonate was added dropwise. The reaction o was stirred for 30 minutes at -30 C and then quenched with pyridine. The desired α-glycoside 30 was isolated in 71% yield. The Karplus equation 32, 33 was used to predict the 3-bond 3 coupling constants of the proton at the anomeric carbon of the α-isomer ( J predicted to be 29 larger) and -isomer by NMR. The α- and -isomers were identified using this estimation, and 1 the H NMR signals for the anomeric protons were integrated to calculate the ratio of α- to anomer at 4:1, regardless of the scale of the reaction (tested <2.5 g scale). Based on precedents within the Huang group, 34 a one pot reaction containing zinc dust, acetic anhydride, and acetic acid was employed to reduce and acetylate azide 30 to the fully protected Tn building block 31, obtained in 93% yield after ~15 h. 30 Scheme 1.5: Synthesis of fully protected Tn building block 31 31 The next step in this reaction required cleavage of the benzyl ester by catalytic hydrogenolysis: a step which was closely monitored to avoid concomitant removal of the Fmoc protecting group. Initial attempts at hydrogenolysis without acetic acid, used generally to speed up hydrogenation reactions, unfortunately, resulted in significant benzyl and Fmoc removal within 15-20 mins as assessed by TLC analysis. This was surprising since Fmoc is cleaved under basic conditions and the reaction pH was near neutral. Therefore, the pH of the reaction was decreased to 3-4 with acetic acid to prevent Fmoc cleavage. It appeared that acetic acid may have increased the reaction rate only slightly, but, more significantly, prohibited the deprotection of the Fmoc group. Under these conditions and monitoring the reaction progress by TLC, the benzyl ester was removed after 15-30 min to produce the free acid 32. This free acid was purified by silica gel flash column chromatography using a DCM:MeOH:AcOH (94.5:5:0.5, v/v/v) to yield 32 in 92% yield (Scheme 1.6). 32 Scheme 1.6: Synthesis of deprotected common intermediate 34 33 Ethanolamide 33 was made by reacting 32 with ethanolamine. This cap was chosen to help increase the product solubility in the hydrophilic in vivo environment; moreover, it was deemed non-antigenic, and thus the immune response could be localized on the carbohydrate moiety. The acid was pre-activated for 1 h with benzotriazole-1-yl-oxy-tris-(dimethylamino)phosphonium hexafluorophosphate (BOP) in the presence of dichloromethane: THF (1:1, v/v) and DIPEA. Once activated, ethanolamine was added, and the reaction was stirred 2-5 h until completion, as monitored by TLC. This target 33 was purified by silica gel flash column chromatography using either DCM:MeOH (19:1, v/v) or acetone: toluene (3:2, v/v) leading to 69% yield of 33. Although the final global deprotection was rather straight forward, it was essential to use concentrated ammonia in methanol. Early attempts using dilute 2M ammonia in methanol resulted in very low yields, even after 24-48 h. By contrast, the reaction proceeded smoothly when 7N ammonia in methanol solution was used, resulting in 34. Compound 34 needed to be >99% pure prior to its use to make the final Tn analogs so that purification of the latter were simplified. Compound 34 was purified by silica gel column chromatography using a DCM:MeOH:NH4OH (5:3:1, v/v/v), resulting in 92% yield. The overall yield of the 9-step synthesis of 34 starting from the commercially available D-galactosamine hydrochloride was ~17%. With the desired building block in hand, compounds 19 and 20 were synthesized. The biotinated analog 19 was prepared by adding commercially available biotinated amino caproic acid N-hydroxysuccinimide ester (35) to 34 in NMP (Scheme 1.7) and stirred for 15 h. The crude mixture was purified by precipitation using ether followed by silica gel column chromatography using a DCM:MeOH:H2O (60:34:6, v/v/v), and compound 19 was obtained at 57% yield. The relatively low yield was likely caused by a minor impurity (< 5%), undetectable 34 1 by H NMR, with the similar Rf as the starting material (34) on TLC. This impurity was later discovered by TLC analysis of the starting amine (thought to be pure) when 6% (by volume) water was added to the eluent. Compound 19 was used to coat the 96-well plates to quantify antibody titers by ELISA, as discussed later in this chapter. 35 Scheme 1.7: Synthesis of Tn analog 19 36 The next target Tn analog synthesized was 20. This compound required the synthesis of the 6-azido caproic acid linker (37) (Scheme 1.8) from the commercially available 6-amino caproic acid (36). This was accomplished in 66% yield using the imidazole-1-sulfonyl azide 28 hydrochloride (22) method. Scheme 1.8: Synthesis of azido linker 37 Linker 37 was synthesized and coupled to free amine 34 by the classic amidation reaction conditions using BOP, HOBt, and DIPEA in NMP and stirred at room temperature (Scheme 1.9). The desired azido compound (20) was synthesized in 94% yield after reaction overnight. Some of this compound was sent to our collaborators at the Scripps Research Institute, M.G. Finn and coworkers. They conjugated this compound to a mutant strain of the tobacco mosaic virus (TMV) which will be discussed in more detail in a later section of this chapter. 37 Scheme 1.9: Synthesis of Tn analog 20. Our group also wanted to expand on the previous research and knowledge on the multivalency effect, discussed in the introduction. Although an immunological response may be produced when monovalent Tn subunits are attached to the surface of a protein carrier, a multivalent glycopolymer construct might generate and even greater response. This synthesized construct (see target vaccine 38 in Figure 1.13) could then be attached to a carrier protein increasing the number of carbohydrate motifs presented to the immune system. Thus, the multivalent glycopolymer could potentially cause more isotype switching and induce immunological memory than the monovalent cognate. Figure 1.13: KLH conjugated to glycopolymer based vaccine 38 38 1.5 Synthesis of the Glycopolymer Several polymerization methods exist such as the well known atom transfer radical polymerization 35 36, 37 (ATRP), reversible addition-fragmentation chain transfer, 38 mediated polymerization. and nitroxide- Because of the harsh reaction conditions and the challenging synthetic nature of these previous methods, we used a cyanoxyl-mediated free radical 39-42 polymerization method established by Dr. Elliot Chaikof. This method called for milder reaction conditions at reduced temperatures and water as the solvent. The latter was worthwhile since fully deprotected sugars could be dissolved in the reaction medium to facilitate their coupling to the polymer. By contrast, several other techniques needed to employ the protected sugars due to compatibility in organic media. In addition, Chaikof’s research group found that these cyanoxyl persistent radicals react with controlled polymerization, making it an attractive method. Construction of the polymerizable subunit 21. We imagined that compound 21 (Scheme 1.10) could be generated by a simple amidation reaction between glycomonomer 34 and linker 39, which contained a terminal end primed for polymerization. Compound 39 could, in turn, be generated by reaction of 6-amino caproic acid (36) with acrylic acid, acrylic anhydride, or acryloyl chloride. After trying many different conditions and routes, it became clear that the synthetic route to linker 39 was not feasible, due to, primarily, self polymerization of the reagents and product. It was evident that this problem was occurring during the reaction, the work up, and the brief storage periods. A concern was that self-polymerization of 39 would cause low yields from the reaction coupling of Tn compound 34 with 39 in the next step. 39 Scheme 1.10: Retrosynthetic analysis of polymerizable design of 21 Alternative polymerizable design. Thus, an alternative synthetic approach was designed so polymerization occurred prior to addition of the Tn analog. (Trimethylsilyl)propargyl acrylamide (42) (Scheme 1.11) was selected as our polymer precursor, since it could be copolymerized first, after which, click chemistry could be used to conjugate the glyco-monomer to the alkyne moiety. Using procedures previously established by Professor Craig Hawker and 43 coworkers, the synthesis of 42 started by adding to a solution of propargyl amine (40) o (commercially available) in dichloromethane was added DIPEA (at 0 C), and then acryloyl chloride was added (Scheme 1.11). After work up and purification, propargyl acrylamide (41) was obtained in 84% yield. To prevent the terminal alkyne of 41 from participating in the polymerization process, a solution of 41 in dichloromethane was treated with DBU, silver chloride (to activate the alkyne), followed by slow addition of TMS-Cl; the reaction was heated 40 o at 45 C for ~1 day. Although the yield of this protection (Scheme 1.11) was only 57%, it was ~25% higher than that reported in the literature (32.9%). 43 Based on the TLC analysis it appeared as though a large amount of starting material (41) remained in the product mixture. I believe that this step may be reversible due to the reaction conditions, because even when more TMS-Cl was added, the reaction did not proceed further. It is likely that the silver in the reaction is driving this reversibility. 44 41 Scheme 1.11: Generation of TMS protected propargyl acrylamide 42 With (trimethylsilyl)propargyl acrylamide (42) in hand, the next step was to copolymerize it with acrylamide using the cyanoxyl-mediated free radical polymerization technique. Carboxylic acid 43 was selected since it would ultimately conjugate the polymer to the protein carrier KLH, which contains free amines, and the aryl amine functional group was used to generate the arenediazonium salt 44 to initiate polymerization. The radical based reaction was carried out under an argon atmosphere, since it was sensitive to oxygen. Commercially available 4-(4-aminophenyl) butyric acid (43), sodium nitrite, and a 50% flouroboric acid solution were added to a 1:1 mixture of H2O and THF (THF helps with the o solubility of the (TMS) propargyl acrylamide (42)) and cooled to 0 C for 30 min (Scheme 1.12). The reaction flask was opened briefly to add the sodium cyanate, (TMS) propargyl acrylamide (42), and the acrylamide to reduce their exposure to air. The reaction flask was o resealed, evacuated, and filled with argon. The reaction was heated at 50 C for ~16 h. This generated the cyanoxyl radical intermediate 45 in situ, which allowed the polymerization to occur. 43 Scheme 1.12: Synthesis of copolymer 46 To calculate the approximate yield of this reaction, complete incorporation of compound 43, (TMS) propargyl acrylamide (42), and acrylamide, and complete deprotection of the TMS protecting group were assumed. This was an interesting and exciting observation based on NMR analysis of the final polymer 46. Apparently, during extended reaction times, majority of the TMS was deprotected. With those assumptions, yields were generally very low (ranging from 15-25%). One source of the low yield could be atmospheric oxygen dissolved in the reaction solvents. Removal of this residual oxygen by degassing might improve the polymerization yield. Recently, after discussing this issue with a member of Professor Xue-Long Sun’s group at Cleveland State University, the low yields were considered a consequence of the purification method and not from reactive oxygen. Professor Sun’s group informed me that they too had similar low yields following this technique. They were able to drastically enhance their isolated 44 yields by evaporating all THF from the reaction and then dialyzing in pure water. Thus, dialysis over 3-4 days was used to separate the un-polymerized components of the reaction. Sun’s group informed that THF does not appear to affect the integrity of the membrane, based on visualizing by sight, but low recovery from the dialysis step indicated that larger polymers escaped the dialysis tubing. Regardless of the low yields, polymers 46A (50 mg), 46B (24 mg), and 46C (134 mg) (Figure 1.14) were obtained via coupling of the alkyne substituents, added at various ratios. These ratios were determined by integrating the area under the broad peak at ~1.75 ppm (corresponding to CH2 of polymer backbone), the broad peak at ~2.25 ppm (corresponding to CH of polymer backbone), the peak at ~2.75 ppm (corresponding to CH of alkyne), and the peak at ~4.0 ppm (corresponding to CH2 of propargyl amide substituent). [See Figure 1.42 in the Appendix for a visual representation of this.] These chemical shifts values corresponded nicely 39-42 with analogous polymers synthesized by Chaikof et al. The number of propargyl amide substituents were determined based on the integration of the peaks coming around  2.5 and 3.8. By subtracting the number of CH’s from the integration of the broad peak ~2 ppm, the number of acrylamide substituents incorporated into our polymer was determined. We normalized our peak integration to the small aromatic peaks at ~7.25 ppm (4 protons). 45 Figure 1.14: Structures of polymers 46A, 46B, and 46C Addition of Tn antigen 20 to polymer 46A, B, and C. A classic click condition optimized by Professor Morten Meldal and coworkers was used for this polymer coupling reaction with an 45 analog of the Tn antigen (20) (Figure 1.12). To a mixture of the alkyne polymer (46A or 46B) and the azido Tn analog (20) in DMF were added CuI and diisopropylamine, and the reaction was stirred for ~4 days. After dialysis and purification, glycopolymer 47B (12 mg, 79% yield) was obtained (Scheme 1.13); glycopolymer 47A (14 mg) was obtained in similar yields (Figure 1.15). The extent of this click coupling was determined by comparing the integration of the carbohydrate protons to those protons corresponding to the polymer backbone and propargyl amine linker discussed above. A proof of an immunogenic response by 47A and 47B in mice 46 will be evaluated prior to constructing glycopolymer 47C to conserve azido analog 20 because of the amount required for polymer cycloaddition. Scheme 1.13: Synthesis of glycopolymer 47B Figure 1.15: Structures of glycopolymer 47A 47 1.6 Conjugation to KLH The desired glycopolymers 47 were synthetically coupled to KLH protein by standard 146 ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) coupling conditions. A glycopolymer 47 was first activated with EDC under acidic conditions, which generated a highly reactive oacylisourea intermediate. 46 residues on its surface. The KLH protein expresses over 2,000 free amines from lysine Many of these lysine groups are not accessible for reaction with the activated carboxylate due to the sterics of the protein conformation. It was predicted then, that only about 300 to 600 residues are actually exposed for reaction. Assuming complete conjugation to all accessible residues (300-600) of KLH (8,000-9,000 kDa) by the glycopolymer (~13 kDa) results in a massive and intricate glycopolymer complex (38) (Scheme 1.14 for an example). Thus, characterization by standard NMR or mass spectrometry determination techniques is precluded. 48 Scheme 1.14: Synthesis of KLH conjugated glycopolymer vaccine 38B Also since the exact molecular weight of the glycopolymer vaccine is unknown, size separation of the unbound KLH and bound KLH is not practical. However, dialysis was effective to remove any unbound glycopolymer, but it must be noted that during this process, a 1M sodium chloride solution was used. Dialysis in pure water caused the KLH protein and/or KLH conjugated glycopolymer to precipitate; these precipitates were impossible to re-dissolve into solution. A colorimetric assay based on the phenyl sulfuric acid method 47 was employed to quantify KLH-associated complexes in the current study. The same protocol could be used to hydrolyze the sugar groups from the KLH conjugated vaccine, and the quantity of liberated sugar 49 could be calculated by the colorimetric assay by comparison to a standard curve correlating sugar concentration and absorbance. A standard curve for galactose concentration against absorbance (A490) constructed by the Dubois method in our lab were comparable (Figure 1.16) to those previously reported. 47 The limitations of this method quickly became evident when very low concentrations (<30 g/ mL) of KLH gave very high absorbance measurements in comparison to the individual sugars. This was not surprising since KLH is a glycoprotein containing a very large diversity of sugars, including GalNAc, on its surface that can be hydrolyzed off. Different sugar types have been found independently by our group, and earlier by Dubois, to produce significantly different absorbance maxima (e.g., GalNAc and galactose shown in Figure 1.16), even when the same concentrations of carbohydrate were used. This may suggest that the carbohydrates on the surface of KLH produce a higher absorbance response than that of GalNAc. The other possibility is that the KLH protein produces a significant amount of color. The issue these findings present is that even if our Tn antigen was successfully conjugated to KLH, it is extremely likely that the absorbance of KLH would mask that of GalNAc contained within the Tn antigen. 50 Determination of sugar w/ H2SO4 (50 uL Phenol) Absorbance (measured at 490 nm) GalNAc Galactose KLH No sugar 3.5 3 2.5 2 1.5 1 0.5 0 0 0.5 1 1.5 2 2.5 Conc. (mg/mL) Figure 1.16: Standard curves for KLH and respective sugars An alternative colorimetric method was tried that employed 1,3-dihydroxynaphthalene to selectively increase the absorbance of GalNAc over background signal. However, this method resulted in an even less distinguishable response than the phenol sulfuric acid method discussed above. The results prompted a survey of an extremely sensitive method, reported by Jahnel and coworkers, in which high performance anion exchange chromatography with pulsed amperometric detection (HPAE-PAD) of the effluent was used to quantitate carbohydrates. 48 Full analysis of the KLH-conjugates has yet to be determined, but, based on preliminary data, this technique does allow for detection of GalNAc and separation of the different carbohydrates derived from KLH after hydrolysis, even at extremely low concentrations. Once the identities of 38A and 38B are characterized, demonstrating successful conjugation, the two KLH-conjugated glycopolymer vaccines (38A and 38B) and the third KLH conjugated glycopolymer (38C), yet to be synthesized, will be dissolved in PBS buffer mixed with an adjuvant and used to inoculate 51 mice. It is hypothesized that these multivalent glycopolymer constructs can elicit strong immune responses. 52 1.7 Conjugation to TMV and Immunological Studies 49 Based on previous work carried out within our group, it was determined that the tobacco mosaic virus (TMV) could be a promising protein carrier. All past work done within the group, used the native form of TMV (Figure 1.17). a E97 c E106 Y139 b 300 nm 4 nm 18 nm Figure 1.17: Native TMV protein a) reactive sites (tyrosine 139, glutamate 97, glutamate 106) within a monomeric unit of the protein coat b) planar disk aggregate c) fully assembled capsid “For interpretation of the references to color in this and all other figures, the reader is referred to the electronic version of the thesis.” This conjugation to TMV was carried out in Professor Qian Wang’s laboratory at The University of South Carolina. The key step in this reaction used click chemistry to couple the terminal azide of the Tn antigen linker (20) to an alkyne moiety pre-installed on the native TMV platform (Scheme 1.15). This alkyne was attached to residue Tyrosine 139, one of three 53 identified reactive sites on the native TMV protein scaffold. A first generation TMV-Tn-N3 glyco-conjugate (Scheme 1.15), similar to the TMV-1Cys-Tn-N3 constructs (Scheme 1.16) within the current work, was evaluated previously within the Huang group. After immunological analysis of the TMV-Tn-N3, ELISA of the antibodies titers contained in sera had the following subtypes: total IgG + IgM titers were 2,500, IgG titers of 3,200, and IgM titers of 7,000. It was worth noticing that although IgG titers were produced indicating class switching had occurred, the titers of IgM were still much higher. 54 Scheme 1.15: Synthesis of TMV-Tn-N3 conjugate using the native TMV Based on these findings and by other examples presented in the literature, 15 it was predicted that this could be caused by the relatively short length and rigidity of the linker used in the conjugation between protein carrier and Tn antigen construct. To test this theory, a second 55 generation construct TMV-Tn-PEG-N3 was developed which took advantage of a longer PEG linker (Figure 1.18). Figure 1.18: Structure of TMV-Tn-PEG-N3 conjugate using the native TMV Three different TMV-Tn-PEG-N3 constructs were synthesized and immunologically tested, having 1 g, 4 g, and 20 g of Tn loading respectively. After immunological studies using these constructs, it was found that even when a longer linker was used, the responses generated were comparable to those generated against the first generation TMV-TN-N3 construct mentioned above. There were two left over serum samples from the 4 g, and 20 g PEG constructs, referred to from this point forward as AM-4 and AM-20, respectively (AM refers to Dr. Adeline Miermont who carried out these previous studies and the number refers to the amount of Tn antigen used). It would have been more desirable to directly compare serum samples generated against the current constructs to a serum sample of the first generation construct due to the closer similarities between the two. However, serum samples from that study were exhausted. Even so, with knowledge that the second generation constructs gave similar ELISA results to those of the first generation construct, comparisons could still be made. 56 It was thought that if the length of linker alone was not enough to improve immune function than perhaps the location of the carbohydrate presented on the protein coat of TMV could be essential. Professor Kenneth Palmer’s laboratory produced a genetically modified version of the protein coating the TMV RNA in which they inserted a surface reactive lysine at the Nterminus. 50 Modeling this approach, our collaborators Professor Qian Wang, inserted a cysteine residue between residues Serine 2 and Tyrosine 3 in the monomeric subunit of the native TMV protein (designated TMV-1Cys) (Figure 1.19). This modified TMV offered an alternative site to couple with our Tn antigen. This might allow better presentation to the host immune response, thus improving immune function. Glu 106 Newly Inserted Reactive Residue Tyr 139 Glu 97 Cys 3 Figure 1.19: TMV Mutant TMV-1Cys-Tn-N3, like the native TMV-Tn-N3, contains a triazole linker; however, its linker is slightly smaller and less rigid compared to the linker used in the native construct (Scheme 1.16 and Scheme 1.15, respectively). This could play a major role on the immune response generated as Boons and coworkers found with the two linkers they tested (Figure 1.5). 15 Again, our collaborator Professor Qian Wang performed the glyco-conjugation. TMV- 1Cys was reacted directly with a bromoacetamide functional group (48) resulting in TMV-1Cys57 Tn-bromoacetamide vaccine construct (Scheme 1.16). Alternately, TMV-1Cys was first reacted with N-propargylmaleimide, generating an alkyne functionalized TMV particle. This particle can then be subjected to click conditions with the desired azide compound yielding the desired TMV-1Cys-N3 vaccine construct. 58 Scheme 1.16: Synthesis of Tn conjugates using mutant TMV 59 With both TMV-1Cys-Tn-bromo and TMV-1Cys-Tn-N3 in hand, two immunological studies were carried out using C57BL/6 female mice to evaluate the immunogenicity of these conjugates in vivo. In the first study, the mice were divided into four experimental groups. Group 1 (negative control group) included five mice inoculated subcutaneously on day 0 with TMV-1Cys (240 g) (without the Tn antigen tether) as a 1:1 emulsion in Complete Freund’s Adjuvant (CFA). The first experimental group included five mice, each inoculated on day 0 with TMV-1Cys-Tn-N3 (60 g equivalent to 522 ng GalNAc) as a 1:1 emulsion in CFA. The second experimental population of five mice was injected on day 0 with a ~4x concentration of TMV1Cys-Tn-N3 (240 g equivalent to 2.1 g GalNAc) as a 1:1 emulsion in CFA. And the final experimental population, having only three mice, was each inoculated on day 0 with TMV1Cys-Tn-bromo (52 g equivalent to 137 ng GalNAc) as a 1:1 emulsion in CFA. Booster injection were given subcutaneously on days 14 and 21 (Note: day 21 injection was given a week early by mistake should have been day 28) in the same doses as listed for day 0. The only difference was that Incomplete Freund’s Adjuvant (IFA) was used to generate the emulsions instead of CFA. Blood samples were collected from all groups on days 0, 7, 28 and, 35, and serum was isolated and antibody titers were analyzed by Enzyme- Linked Immunosorbent Assay (ELISA). To avoid measuring antibodies against the TMV protein, biotinated 19 (containing a Tn moiety, Figure 1.12) was immobilized on a microtiter plate coated with NeutrAvidin. The serum was then added to each well and the total antibody titers (IgG + IgM) and titers of the IgG and IgM subtypes were measured photometrically using secondary IgG + IgM, IgG, and IgM antibodies conjugated to horseradish peroxidase (HRP). 60 The first few ELISA studies that were undertaken were designed to determine which mice had elicited the best immune responses. This was accomplished by measuring each mouse on day 0, 7, 28, and 35 at one chosen sera dilution (1/1600) point. In addition, all three secondary HRP-coated antibodies were used. Although not pictured, antibody titers at day 0 and 7 and titers measured at days 28 and 35, did not show significant differences. As can be seen in Figure 1.20, there was an obvious increase in antibody titers for all mouse groups over the 35 day time course of the study. Of more importance was that in nearly all mice, the observed IgG titers were higher than those of the IgM subtype at day 35, which was encouraging. However, one puzzling observation was that the control TMV-1Cys group displayed similar trends. The fact that titers were measured against the biotinated Tn antigen, likely indicated that there was a significant amount of non-specific binding of the antibodies produced against the proteins of the TMV coat. 61 Day 0: IgG + IgM Day 35: IgG OD Day 35: IgG + IgM Day 0: IgM Day 0: IgG Day 35: IgM 4 3.5 3 2.5 2 1.5 1 0.5 0 1 2 3 4 5 6 7 9 * 8 10 11 12 13 14 15 16 17 18 Mouse # Figure 1.20: Antibody Titers for Study 1: all mice Based on these findings, it can be seen that there is a significant change in both IgG and IgM titers when comparing Day 0 to Day 35. Study groups are as follows: Mice 1-5: TMV-1Cys Control; Mice 6-10: TMV-1Cys-N3 (137 ng GalNAc); Mice 11-15: TMV1Cys-N3 (2.1 g GalNAc); Mice 16-18: TMV-1Cys-bromo (137 ng GalNAc). *Note: mouse 9 died on Day 0 prior to first immunization Based on the above information, the mouse that elicited the best immune response from each group, toward each respective vaccine, was chosen. Each serum sample, including the two older native TMV serum samples, AM- 4 and AM- 20, were subjected to full analysis by ELISA. It appeared that the native TMV vaccine constructs generated significantly more IgG antibodies than those by the TMV-1Cys carrier (Figure 1.21). However, it must be speculated that perhaps the amount of vaccine administered could be playing a role. Based on past findings within the lab, TMV-Tn-PEG-N3 vaccine (1 g) generated very low levels of IgG compared to the 4 g or 20 g groups. Based on the discrepancies in the present and past findings between the two 62 constructs, the significantly high TMV-1Cys control titer responses, and the fact that the second booster administration had been given a week early out of error, it was decided that it might be worth evaluating the reproducibility of the vaccination process. 450,000 400,000 350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 A B C D Figure 1.21: IgG titers after 35 days using native and mutant TMV constructs Titers generated against the native construct are significantly higher than those against the mutant construct. Each bar represents the best mouse chosen from each experimental group using both native and mutant constructs A) Mouse 7: TMV1Cys-Tn-N3 (137 ng GalNAc); B) Mouse 12: TMV-1Cys- Tn- N3 (2.1 g GalNAc) C) AM-4 D) AM-20 In the second study, the mice were split into three groups. The first group included four mice inoculated subcutaneously on day 0, with TMV-1Cys (240 g) as a 1:1 emulsion in (CFA). The second group included four mice each inoculated on day 0 with TMV-1Cys-Tn-N3 (240 g equivalent to 2.1 g GalNAc) as a 1:1 emulsion in CFA. Booster injections were given subcutaneously on days 14 and 28 in the same doses as listed for day 0. Again the only 63 difference being that IFA was used to generate the emulsions instead of CFA. The third group was simply a no injection control group of two mice. We wanted to ensure that the mice were not generating non-specific antibodies due to environmental or age-related sources which could skew our results. Blood samples were collected from all groups on days 0, 7, and 35, and serum was isolated and antibody titers were analyzed by ELISA as before. As in the first study, to save time and resources, ELISA was run on all mice at 1/1600 dilution (Figure 1.22). This allowed for the mice with the most favorable immune responses to be analyzed further. Day 0 Day 35 2.5 OD 2 1.5 1 0.5 0 1 2 3 4 5 6 7 8 9 10 Mouse # Figure 1.22: IgG Antibody Titers for Study 2: all mice Like in Study 2, it can be seen that significant IgG titers have been generated after the 35 day experiment. Study groups are as follows: Mice 1-4: TMV-1Cys-N3 (2.1 g GalNAc); Mice 5-8; TMV-1Cys Control; Mice 9-10: No Injection Control Just like in study 1, we observed a noticeable difference in the antibody titers between day 0 and Day 35. In addition, we, again, observed unexpectedly high titers in the control groups. Since one of the main reasons for this second study was to determine the reproducibility 64 of the immunological study, it was important to measure antibody titers from both studies at the same time, so as to eliminate the potential for slight differences between experimental analyses. The similar magnitudes of bar A to bar B (Figure 1.23) show that the antibody titers generated in the first study and second study were reproducible. Bar C again suggests that TMV-1Cys control sample must be generating non-specific antibodies. To our surprise, the group of mice that were not injected with anything, bar D, had significant antibodies generated against the Tn antigen. This observation, along with the high titer counts observed for previous control groups, as well as one other experiment, will be discussed below. 12,000 10,000 8,000 6,000 4,000 2,000 0 A B C D Figure 1.23: IgG titers at Day 35: Study 1 vs. Study 2 Based on these results it can be concluded that there is reproducibility between Study 1 and Study 2. In addition, it also shows that the control groups have significant amounts of non specific antibodies A) Study 1, Mouse 15: TMV-1Cys-N3 (2.1 g GalNAc); B) Study 2, Mouse 3: TMV-1Cys-N3 (2.1 g GalNAc) C) Study 2, Mouse 7: TMV-1Cys- Control D) Study 2, Mouse 9: No Injection Having measured noticeable IgG titers in all groups, the next direction was to determine the specificity of these antibodies toward the Tn antigen. The results of a competition ELISA 65 (100 L of GalNAc were added to each well at the same time as the diluted serum samples) were compared to that of a normal ELISA (Figure 1.24). The idea was that if the antibodies are specific to Tn, then they should bind to GalNAc in a competition assay, thus not binding to the antibody stacking complex in the plate, resulting in low optical densities. Bar A shows the results of Mouse 12 that yielded the highest number of IgG antibodies from Study 1. This sample was run alongside the two native TMV serum samples (B: AD-4 and C: AD-20) (Figure 1.24), all three samples, have only slight competitive binding with GalNAc indicating that the antibodies present were not produced in response to the Tn antigen. 450,000 400,000 350,000 300,000 250,000 200,000 150,000 100,000 50,000 0 Figure 1.24: Competitive ELISA vs. Normal ELISA: IgG Antibodies specific for GalNAc Based on these results, it can be deduced that the native TMV constructs elicited a stronger immune response then the mutant. Additionally, it can be determined that none of the samples have generated Tn specific antibodies. If selective antibodies were present, a significant drop in the titer for the competitive cases should have been observed. A) Study 1, Mouse 12: TMV-1 Cys-N3 (2.1 g GalNAc) B) Study 2, Mouse 4: TMV- 1 Cys-N3 (2.1 g GalNAc) C) AM-4 D) AM-20 66 It can be concluded that the native TMV vaccine produced higher quantities of IgG type antibodies than that of the mutant TMV (Figure 1.21); however, few of these titers were specifically generated against Tn (Figure 1.24). Based on these preliminary results described, it was tempting to conclude that the native TMV protein carrier was slightly superior to that of the mutant TMV-1Cys carrier. This conclusion was based on the fact that the two native TMV constructs generated a greater immune response, although few of the titers generated were Tn specific. Finally, these preliminary results also suggested that a significant amount of nonspecific binding in the control populations was occurring. We also thought it might be worth evaluating another promising platform, bacteriophage Q. Previous results published within the Huang group using CPMV, humoral immune response. 34 indicated a strong The protein coating of the wild-type Q has about 120 more subunits than CPMV, and each subunit contains about the same number of surface-accessible lysines as CPMV. 51 Thus, it seemed possible to elicit an even stronger immune response using Q as the Tn carrier. 67 1.8 Conjugation to Q and Immunological Studies In this study, the noninfectious, self-assembled virus-like particle, E. coli bacteriophage 52 Q encapsidating random cellular RNA was used. This icosahedral virion is composed of 180 copies of the coat protein. The capsid protein is composed of 132 amino acids of a mostly antiparallel -sheet structure. 53 A large non-covalent contact area within the capsid allows it to be unusually stable toward extreme temperature, pH, and chemical reagents. There are several benefits to using a virus scaffold toward vaccine development. A few of the many advantages in using virus scaffolds include ease of production and purification in large-scale, being well defined structurally at near atomic resolution, and can have their amino acid sequence engineered 54 by recombinant DNA techniques. The bioconjugation of this Q virion to our antigen constructs was carried out by our collaborators, Professor M. G. Finn and coworkers at the Scripps Research Institute. The method they used for bioconjugation can be seen below in 51 Scheme 1.16. 68 Scheme 1.17: Synthesis of azido conjugates using Q 69 This bioconjugation was accomplished by acylation of the surface lysine chains with alkynyl N-hydroxysuccinimide (NHS) ester, thus installing alkyne substituents on the surface. These alkyne substituents were then subjected to cycloaddition conditions with each of the azide compounds to be immunologically evaluated, affording each respective Q- bioconjugated vaccine construct. These Q-vaccine constructs were then evaluated for their immunological properties. Five groups were assigned, three experimental and two controls. The experimental population, divided into three groups, was designed to test different concentrations of the Q-protein carrier conjugated to the Tn antigen after administration (Q-Tn-N3, 0.45, 1.8, and 9 g respectively). The first of the control groups was Q-propanol. This construct was designed to determine whether an immune response was generated against the carrier, in which antibodies produced, could bind non-specifically to our Tn antigen. The second control group, Q-glucose, was chosen to determine if an immune response could be elicited toward the construct; if so, could these antibodies bind non-specifically to GalNAc as well. One additional experiment that might be interesting to look into, in the future, would be to measure antibodies generated against glucose. Because it has been well established that glucose is a self antigen, it would seem highly improbable that an immune response could be generated against it. If there is a noticeable response, however, this might indicate that the construct was capable of overcoming immunological tolerance. With the five vaccine constructs in hand, immunological studies were carried out using C57BL/6 female mice to evaluate the immunogenicity of these conjugates in vivo. The first, second and third groups: four, five, and four mice respectively were inoculated subcutaneously on day 0 with Q-Tn-N3 (corresponding to 17, 69, and 345 g of the particle or 0.45, 1.8, 9.0 g 70 GalNAc respectively) as a 1:1 emulsion in CFA. The fourth group included four mice each inoculated on day 0 with Q-propanol (220 g equivalent to 2.4 g propanol), as a 1:1 emulsion in CFA. And the fifth and final group, included three mice each inoculated on day 0 with Qglucose (103 g equivalent to 2.9 g GalNAc), as a 1:1 emulsion in CFA. Booster injections were given subcutaneously to all mice on days 14 and 28 in the same doses as listed for day 0. The only difference is that IFA was used to generate the emulsions instead of CFA. Blood samples were collected from all groups on days 0, 7, and, 35 and serum was isolated. Antibody titers were analyzed by ELISA. To avoid measuring antibodies against the Q protein, the biotinated-Tn (19) was immobilized on a microtiter plate coated with NeutrAvidin. The serum was then added to each well and the total antibody titers (IgG + IgM) and titers of the IgG and IgM subtypes were measured photometrically using secondary IgG + IgM, IgG, and IgM antibodies conjugated to horseradish peroxidase (HRP). The first few ELISA studies that were undertaken were designed to determine which mice had elicited the best immune responses. This was accomplished by measuring the IgG antibody titers for each mouse, in all five groups (Figure 1.25). As can be seen, the results from this study were rather alarming. While the titers generated against the Q-Tn-N3 were the same for all groups, more attention must be given to the two control groups. Both the Q-propanol and Q-glucose control groups yielded antibody titers as high as the experimental groups. In fact, the mouse with the highest antibody titer in the entire study, mouse 17, was from the Qpropanol control group. 71 1.2 OD 1 0.8 0.6 0.4 0.2 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Mouse # Figure 1.25: IgG Antibody Titers for Study 3: all mice The results presented here show the alarming findings that both control populations have generated significant non-specific binding. Additionally, mouse 17, from the Qpropanol control group produced the highest titers. Mice 1-4: Q-Tn-N3 (0.45 g GalNAc); Mice 5-9: Q-Tn-N3 (1.80 g GalNAc); Mice 10-13: Q-Tn-N3 (9.00 g GalNAc); Mice 14-17: Q-propanol control; Mice 18-20: Q-glucose control Based on the TMV studies, we had originally assumed that the high background in the control populations were likely due to non- specific binding. However, observing the results of this present study, it was becoming increasingly evident that something else was going on. There were a few key observations that led us to this conclusion. The first of which was alluded to in Figure 1.22 and Figure 1.23 but not discussed. The observation that both mice not injected throughout the course of the experiment (Study 2) had substantial IgG titers was perplexing. One would not expect to see a significant change over the course of the experiment especially considering that they were housed in a well maintained facility where the chance of 72 environmental antigen exposure was minimal. The second realization was that during many of the ELISA experiments that were run, very high background absorbance were detected. The final bit of evidence that clearly brought this issue to the forefront was in this third study in which Q-propanol and specifically the Q-glucose produced titers as high, if not higher, than the experimental groups. It seemed highly improbable that significant quantities of antibodies were being generated against a self antigen like glucose. With this realization, questions began to surface about possible explanations or experimental procedure that could explain this phenomenon. The first thing that we decided to investigate was each individual step of the ELISA protocol, as can be seen in Figure 1.26. OD 4 3.5 3 Normal Elisa No Secondary Ab No Serum No Antigen No BSA No Neutravidin 2.5 2 1.5 1 0.5 0 Qbeta- Glc Qbeta- Prop. Qbeta-Tn (9 ug) AM- 4 Figure 1.26: Evaluation of each step in the ELISA protocol Q-Glc: Study 3, Mouse 20; Q-Prop: Study 3, Mouse 16; Q-Tn: Study 3. Mouse 13; AM-4 73 The idea for this study was that by removing one step from the procedure at a time, we could determine which step was resulting in the high background absorbance. The first step in the protocol was to coat the plate with NeutrAvidin which is capable of binding to our biotinated antigen. In the second step, biotinated antigen was added, which our serum antibodies should be able to recognize, if produced. Third, we used BSA as a small molecule blocker, which was supposed to prevent non- specific binding from occurring. We expected that by eliminating any one of these steps, with the exception of BSA blocking perhaps, there should be little to no antibody binding. This in turn should prevent the secondary HRP-coated antibody from binding. This ultimately meant that no absorbance should be recorded. However, as can be seen from the results, the absorbance measurements were all high. When the serum antibodies and HRPcoated secondary antibodies were omitted, the virtual lack of response in optical density was as expected. This high background absorbance strongly indicates that our serum antibodies were binding non-specifically, which further indicated that a change was required to the ELISA protocol. After many adjustments to conditions, it was ultimately decided that a different means of antigen presentation needed to be tried. Ultimately, with the help of Dr. Zhaojun Yin, bovine serum albumin (BSA) was conjugated to the Tn antigen, This construct was used to replace the NeutrAvidin and biotinated- Tn antigen presenting steps in the ELISA protocol. This BSA system, was capable of carrying between 34-37 copies of the antigen on its surface, which is about 8-9 times more antigen density than NeutrAvidin. NeutrAvidin is known to be able to bind only four copies of biotin. By increasing the antigen density, perhaps the amount of non specific binding would be reduced and more importantly a reduction in the background absorbance would be detected. 74 With this modified system in hand, the results of the titers for all mice were repeated (Figure 1.27). Now when comparing the two control groups, Q-propanol and Q-glucose, the titers measured were negligible as was originally predicted. As for the experimental groups, there was a large discrepancy between individual mice within groups; however, this was the main reason for testing all mice at a given serum dilution (1: 6400), because not all mice should react to the vaccine in the same matter. 3.5 OD 3 2.5 2 1.5 1 0.5 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Mouse # Figure 1.27: IgG Antibody Titers against BSA-Tn @ 1: 6400 serum dilutions for Study 3 All serum samples from Study 3 were analyzed at 1:6400 dilutions against BSA-Tn construct. All control mice gave background level absorbance’s indicating that this new platform can greatly reduce the background absorbances. Mice 1-4: Q-Tn-N3 (0.45 g GalNAc); Mice 5-9: Q-Tn-N3 (1.8 g GalNAc); Mice 10-13: Q-Tn-N3 (9.0 g GalNAc); Mice 14-17: Q-propanol control; Mice 18-20: Q-glucose control From these results, full ELISA determination was performed on the best mouse from each experimental group (Figure 1.28). Mouse 1 was chosen to represent Q-Tn-N3 (0.45 g GalNAc) group, Mice 7 and 8 were chosen to represent Q-Tn-N3 (1.8 g GalNAc), and Mouse 75 13 was chosen to represent Q-Tn-N3 (9.0 g GalNAc). At this time, because no observable titers were detected in the single dilution study for Q-propanol, nor Q-glucose, full ELISA analysis has not been performed. It was interesting to determine how the Q conjugate compared to the native TMV construct that produced such high titers, it was decided to run full ELISAs on those as well. 4,000,000 3,500,000 Titer 3,000,000 2,500,000 2,000,000 1,500,000 1,000,000 500,000 0 1 7 8 13 AM- 4 AM-20 Mouse # Figure 1.28: IgG titers for representative mice in Study 3 and native TMV-Tn-PEG-N3 Based on our findings, it appears as though both the Q conjugates and the native TMV conjugates produces high quantities of IgG titers- Mouse 1: 272,000; Mouse 7: 3,787,000; Mouse 8: 300000; Mouse 13: 226,000; AM- 4: 166,000; AM- 20: 360,000 As it turned out, all the mice tested had very high IgG affinities; Mouse 7 had close to 3 million, whereas the rest all had closer to 300,000, which were still very impressive. With these interesting findings, we decided to determine the IgM antibody titers. Like IgG, all mice were 76 tested and the ones that produced the highest responses where used. We tested Mouse 1, Mouse 13, AM-4 and AM-20 using full dilution analysis. 100,000 90,000 80,000 Titer 70,000 60,000 50,000 40,000 30,000 20,000 10,000 0 1 13 AM- 4 Mouse Number AM-20 Figure 1.29: IgM titers for representative mice in Study 3 and native TMV-Tn-PEG-N3 Respective titers- Mouse 1: 22,400; Mouse 13: 85,800; AM-4: 4,200; AM-20: 9,100 Excitedly, it seemed as though the number of high affinity IgG antibodies generated was significantly higher than the lower IgM affinity antibodies, which indicated that isotype class switching had occurred and the establishment of immunological memory was being produced. Based on these conclusions, it would be interesting to know how specific these antibodies are toward the Tn antigen displayed in its native environment on the tumor cell surface. Preliminary FACS analyses have indicated that there is little specificity of these antibodies toward Tn in its natural environment on cancer cells, NCI-ADR-RES. However, it is now believed that the NCI-ADR-RES ovarian drug resistant cancer cell line that we have been using 77 9 may not express Tn but instead MUC1. This issue is currently being investigated within the lab by testing other cell lines that have been proven to contain Tn on their surface. Additionally, a Q-MUC1-N3 vaccine serum is being tested against this MUC1 containing cell line (NCI-ADRRES). It will be interesting to see what kind of specific binding these vaccine constructs can generate. 78 1.9 Re-evaluation of the TMV-1Cys Construct Using New ELISA Protocol Having proven the effectiveness of the BSA conjugated Tn display for ELISA analysis, we thought it might be important to re-evaluate a few of the past TMV-1Cys samples. First, we re-evaluated all of the mice in study 1. Based on the original results (Figure 1.20), we observed a definitive increase from day 0 to day 35, however, the control groups were all surprising high as well, which we had originally just attributed to non- specific binding. 4 OD 3.5 3 2.5 2 1.5 1 0.5 0 1 2 3 4 5 6 7 8 * 10 11 12 13 14 15 16 17 18 19 20 9 Mouse # Figure 1.30: Antibody Titers for Study 1: all mice using BSA- Tn for ELISA Antibody titers measured against the BSA-Tn construct show much lower titers than when the NeutrAvidin system was used. Mice 1-5: TMV-1Cys Control; Mice 6-10: TMV-1Cys-N3 (137 ng GalNAc); Mice 11-15: TMV-1Cys-N3 (2.1 g GalNAc); Mice 16-18: TMV-1Cys-bromo (137 ng GalNAc). Number 19: AM-4; Number 20: AM-20 *mouse 9 died on Day 0 prior to first immunization 79 Although not shown, IgG titers of Study 2 were analyzed in an analogous fashion. Similar finding were observed. Based on these new realizations, it appeared that the native TMV-Tn construct had resulted in a better immune response than that of the mutant TMV-1Cys control. Additional ELISA analyses are currently underway to quantify the actual titers generated against these constructs, like the Q results already discussed. 80 1.10 Synthesis and Immunological Studies of a Triazole Linker Having determined the antibody titers of the Q and TMV conjugated vaccines, we thought it would be interesting to quantify the amount of antibodies generated against the triazole linker. There has been a lot of debate in the literature about which linkers should and should not be used. However, few of these actually quantify the titers generated against the linker in question. To quantify antibody titers generated against the triazole linkers used in most of the vaccine constructs tested, two target compounds were needed. The first compound that was synthesized was needed to run a competition experiment analogous to those run using GalNAc to quantify titers against Tn. The second compound was needed to coat the ELISA plate to quantify antibody titers. The first compound was synthesized starting from the commercially available 3-chloro-propan-1-ol (49) (Scheme 1.18). Compound 49 was heated to 80 o C overnight for about 18 hours in the presence of sodium azide in water; thus causing an SN2 displacement of chlorine generating 3-azido-propanol (50). Note: Since small molecular weight organic azides have been reported as potential explosives, extreme caution was used in the handing of this compound. Although this compound was concentrated to dryness, it was used immediately in the next step without further analysis due to the nature of the compound. Compound 50 then underwent a classic click reaction using commercially available propargyl o amine, CuI, DIPEA in DMF. The reaction was heated for 16 hours at 50 C resulting in 51 being generated in a 70 % yield over the two steps. 81 Scheme 1.18: Synthesis of triazole linker (51) Free amine, 51, was less than desirable due to its strong nucleophilicity. It was thus decided to acetylate it for use in the competition experiments. The difficulty in acetylation, however, is that there is a primary amine and a primary alcohol. It seemed possible to use normal acetylating conditions to run this reaction as long as the temperature was kept low. We decided to use Ishihara and others 55 method in which they selectively acetylated a variety of primary alcohols in the presence of secondary alcohols. This method involved the reaction of 51 o with 2, 4, 6- collidine, and acetyl chloride in DMF at -60 C for two hours (Scheme 1.19). This reaction resulted in only a 53 % yield with no observable ester formation. The other major spot was starting material. There is a chance that adding more acetyl chloride might have pushed this yield higher however, it was decided against to avoid the risk of ester formation. 82 Scheme 1.19: Synthesis of triazole compounds for ELISA analysis 83 Finally, the last compound synthesized was biotinated triazole 53. This compound was generated by addition of 51 to biotinated amino caproic acid NHS ester (35) and DIPEA to NMP. After 17 hours of stirring at room temperature, the biotinated product, 53, was precipitated out using ice cold ether. Compound 53 was obtained in a 73 % yield. As it turned out, this biotinated- triazole synthesis was completed just prior to the discovery of the background issue discussed above. Consequently, a third construct needed to be synthesized for future ELISA experiments, to quantify the amount of antibodies generated against the triazole linker (Figure 1.31). This synthesis and conjugation, starting from 51 was carried out by Dr. Zhaojun Yin. Figure 1.31: BSA conjugated triazole compound for ELISA analysis Having synthesized our triazole compounds for ELISA analysis, we then tested each of the TMV conjugates to determine how much of an antibody response was generated against the linker. As can be seen in Figure 1.32, a mouse from the TMV-1Cys-bromo and both of the native TMV constructs have significant titers that have an affinity toward triazole. 84 TMV- 1Cys- Tn (522 ng): Day 35 TMV- 1Cys- Tn (2.09 ug): Day 35 OD TMV- 1Cys- Bromo- Tn: Day 0 TMV- 1Cys (control) AM- 20 AM- 4 2.5 TMV- 1Cys- Tn (2.09 ug): Day 35 TMV- 1Cys- Bromo- Tn: Day 35 3 TMV- 1Cys- Tn (522 ng): Day 0 Blank 2 1.5 1 0.5 0 0 1 2 3 4 5 6 Figure 1.32: IgG antibodies measured against BSA-triazole in TMV study groups As for the Q Study groups it appeared as though they all had high amounts of antibodies generated against the triazole linker (Figure 1.33). Q-glucose was the only group that had a lower level of antibodies generated against triazole, however, even one of the mice in that group showed significant antibodies against it. 85 Qbeta- Tn 0.45 ug Qbeta- Tn 9.0 ug 4 Qbeta- Tn 1.8 ug Qbeta- Propanol Qbeta- Glucose 3.5 3 OD 2.5 2 1.5 1 0 1 2 3 4 Figure 1.33: IgG antibodies measure against BSA- triazole in Q Study groups Much work is still ongoing to get a better understanding of the role that linkers have in the immune response. It is believed that if a strong enough immune response is generated against the linker then there is likely a negative consequence to the response generated against the desired antigen. 86 1.11 Conclusions Based on the findings presented herein, it seems as though many factors need to be considered when designing a successful vaccine. For instance, when comparing the native TMV to the mutant TMV-1Cys conjugates, it was predicted that modification of the protein surface would allow better expression of the carbohydrate antigen to the surroundings. However, after quantifying the antibody titers using the new BSA protocol, the native construct was shown to elicit the stronger response. This indicated that the location that the antigen is bound on the subunit of the antigen presenting cell is critical toward developing a successful immune response. Probably one of the most interesting findings in the present work was the major setback caused using the biotinated antigen display for ELISA. The NeutrAvidin/ biotinated-Tn system was used successfully in our group in all past immunological evaluations and very impressive 34 results were obtained. After much thought on this issue, a few possible theories were developed. The first and most convincing, based on the evidence collected thus far, suggest that there may have been a significant difference in the 96-well ELISA plates used in past studies versus those used currently. This theory was suggested based on the findings using the AD-4 and AD-20 samples. Comparing past work carried out by Dr. Adeline Miermont within the 49 group to that of our present work, there was a large discrepancy between the data. She found that, not only was there a strong immune response generated, but the antibodies produced were highly specific toward the Tn antigen based on competitive ELISA. Using the same protocol as well as the serum samples she collected, in the current work little specificity toward the Tn antigen was observed (Figure 1.24). The only unknown variable between the current work and past work is the 96-well plates that were used. There is a chance that the plates used in her case 87 may have been able to bind more NeutrAvidin/ biotinated- Tn thus resulting in a higher density of antigen expression than the plates currently used. A second theory may be that antibodies generated against a certain vaccine have different sensitivities to antigen density. For instance, when using CPMV, excellent results were obtained. 34 Perhaps the antibodies generated, using this construct, were sensitive to the lower density antigen expressed using the NeutrAvidin system. On the other hand, when the TMV and Q constructs were used, potentially the antibodies generated were not as sensitive to the low density antigen presented by the NeutrAvidin system thus could not bind. Whereas, once the higher antigen density system utilizing BSA was applied, these antibodies could bind. This is an extremely interesting idea, because if true, this potentially means that antibodies generated against the TMV or Q constructs may be more selective towards cancer cells, which have high density antigen presentation, and less sensitive to normal cells which express low density antigens. This will be a theory that I am sure will be actively investigated in the near future as soon as the results from the FACS experiments begin to push forward. 88 1.12 Experimental Section General Experimental Procedures. All reactions were carried out under nitrogen or argon with anhydrous solvents in ovendried glassware. All glycosylation reactions were performed in the presence of molecular sieves, which were flame dried immediately prior to the reaction under high vacuum. Glycosylation solvents were dried using a solvent purification system and used directly without further drying. Chemicals used were reagent grade as supplied except where noted. Analytical thin-layer chromatography was performed using silica gel 60 F254 glass plate. Compound spots were visualized by UV light (254 nm) and by staining with a yellow solution containing Ce(NH4)2(NO3)6 (0.5 g) and (NH4)6Mo7O24 * 4H2O (24.0 g) in 6% H2SO4 (500 mL) or a solution of Ninhydrin (1.5g) in n-butanol (100mL) and acetic acid (3mL). Flash column chromatography was performed on silica gel 60 (230–400 Mesh). NMR spectra were referenced 1 using Me4Si (0 ppm), residual CHCl3 (δ H-NMR 7.24 ppm, 1 NMR 4.63ppm), CH3OH (δ H-NMR 3.30 ppm, NMR 2.04 ppm, 1 1 1 13 13 1 C-NMR 77.0 ppm), H2O (δ H1 C-NMR 49.0 ppm), and (CH3)2CO (δ H- 13 C-NMR 29.84 ppm). Peak and coupling constant assignments were based on 1 13 1 13 H-NMR, H– H gCOSY and (or) H– C gHMQC and H– C gHMBC experiments. ESI mass spectra were recorded in positive ion mode. O-(2-Acetamido-2-deoxy-α-D-galactopyranosyl)-L-serine ethanolamide (34) Compound 33 (206.8 mg, 0.296 mmol) was dissolved in a 7N* solution of ammonia in methanol (10 mL) (*Note: using < 7N solutions, reaction will not work) and the reaction was 89 o stirred at 0 C and gradually allowed to warm to room temperature overnight. The reaction mixture was allowed to stir for 15 hours. Completion of the reaction was confirmed using a DCM: MeOH: NH4OH (4: 4: 1) TLC elution mixture. The reaction was then concentrated to dryness and the residue was then purified via column chromatography using a DCM: MeOH: NH4OH (5: 3: 1) elution mixture. The desired product was obtained as a white solid with a 92% yield. 1 H NMR (500 MHz, CD3OD) δ 2.00 (s, 3H), 3.34 (t, J = 5.5 Hz, 2H), 3.56 (dd, J = 5.0, 10.1 Hz, 1H), 3.59 – 3.63 (m, 2H), 3.67- 3.72 (m, 2H), 3.74 (t, J = 5.6 Hz, 1H), 3.76 – 3.80 (m, 1H), 3.80 – 3.85 (m, 1H), 3.87 (d, J = 3.0 Hz, 1H), 3.91 (dd, J = 4.3, 10.1 Hz, 1H), 4.28 (dd, J = 3.7, 10.9 Hz, 1H), 4.78 (d, J = 3.7 Hz, 1H) 13 C-NMR (150 MHz, CD3OD),  22.8, 43.0, 51.4, 55.7, 61.4, 62.9, 69.7, 70.3, 70.4, 72.9, 99.6, 173.8, 174.1. + [HRMS] [M+H] m/z: calc. for C13H26N3O8: 352.1720, found 352.1707. 6-N-Biotinamido-hexanoyl-O-(2-acetamido-2-deoxy-α-D-galactopyranosyl)-L-serine ethanolamide (19) Compound 34 (29 mg, 0.083 mmol), compound 35 (29.7 mg, 0.138 mmol, 1.66 eq) were dissolved in N-methylpyrrolidinone (4 mL). The reaction mixture was stirred at room temperature for 15 hours. Completion of the reaction was confirmed using a DCM: MeOH: H2O (60: 34: 6) TLC elution mixture. A large excess of ether was added to the reaction to precipitate out the product. The solid residue was collected via vacuum filtration and then purified via column chromatography using the same solvent elution mixture used for TLC. The compound was then re-dissolved in water and lyophilized. The desired product was obtained as a white 90 solid with a 57% yield. Comparison with literature data 34 1 confirmed its identity. H NMR (600 MHz, D2O) δ 1.14 – 1.31 (m, 4H), 1.34 – 1.64 (m, 8H), 1.90 (s, 3H), 2.10 (t, J = 7.2 Hz, 2H), 2.20 (t, J = 7.5 Hz, 2H), 2.63 (d, J = 13.2 Hz, 1H), 2.85 (dd, J = 5.1, 12.9 Hz, 1H), 3.99 – 2.08 (m, 2H), 3.16 – 3.26 (m, 3H), 3.47 – 3.54 (m, 2H), 3.57 – 3.63 (m, 2H), 3.65 (dd, J = 5.4, 10.8 Hz, 1H), 3.71 - 3.75 (m, 2H), 3.77 (dd, J = 5.4, 10.8 Hz, 1H), 3.83 (d, J = 3.0 Hz, 1H), 4.02 (dd, J = 3.6, 10.8 Hz, 1H), 4.28 (dd, J = 4.8, 7.8 Hz, 1H), 4.41 – 4.50 (m, 2H), 4.75 (d, J = 3.6 Hz, 1H) 13 C-NMR (125 MHz, CDCl3),  22.2, 25.0, 25.3, 25.8, 27.8, 28.0, 28.2, 35.4, 35.7, 39.2, 39.9, 41.7, 49.9, 53.9, 55.6, 60.0, 60.4, 61.3, 62.3, 67.4, 67.8, 68.6, 71.5, 98.0, 165.5, 171.8, 174.6, 176.8, 177.3. [HRMS] + [M+H] m/z: calc. for C29H51N6O11S: 691.3337, found 691.3347. Azido-caproic-amido-O-(2-acetamido-2-deoxy-α-D-galactopyranosyl)-L-serine ethanolamide (20) Compound 34 (47.5 mg, 0.135 mmol), compound 37 (29.7 mg, 0.189 mmol, 1.4 eq), BOP (72.6 mg, 0.164 mmol, 1.22 eq), HOBT (22.8 mg, 0.168 mmol, 1.24 eq), and DIPEA (45 L, 0.266 mmol, 1.97 eq) were dissolved in N-methylpyrrolidinone (4 mL). The reaction mixture was stirred at room temperature for 15 hours. Completion of the reaction was confirmed using a DCM: MeOH: NH4OH (12: 5: 1) TLC elution mixture. A large excess of ether was added to the reaction to precipitate out the product. The solid residue was collected via vacuum filtration and then purified via column chromatography using the same solvent elution mixture used for TLC. This is to remove any starting material 34 that may have co-precipitated out with the product. The compound was then re-dissolved in water and lyophilized. The desired product 91 was obtained as a white solid with a 94% yield. 1 H NMR (600 MHz, D2O) δ 1.20 – 1.30 (m, 2H), 1.43 – 1.55 (m, 4H), 1.90 (s, 3H), 2.22 (t, J = 7.4 Hz, 2H), 3.19 (t, J = 6.8 Hz, 2H), 3.22 (t, J = 4.8 Hz, 2H), 3.49 – 3.53 (m, 2H), 3.59 – 3.63 (m, 2H), 3.63 – 3.68 (dd, J= 5.4, 10.8, 1H), 3.72 – 3.76 (m, 2H), 3.78 (dd, J = 5.4, 10.8 Hz, 1H), 3.84 (d, J = 3.6 Hz, 1H), 4.02 (dd, J = 3.6, 10.8 Hz, 1H), 4.45 (t, J = 5.4 Hz, 1H), 4.76 (d, J = 3.6 Hz, 1H) 13 C-NMR (150 MHz, CDCl3),  24.6, 27.3, 28.1, 30.3, 37.8, 44.2, 52.4, 53.5, 56.3, 62.4, 63.7, 69.9, 70.3, 71.0, 73.9, 100.4, 174.2, + 177.0, 179.7. [HRMS] [M+H] m/z: calc. for C19H35N6O9: 491.2466, found 491.2485. N-(Prop-2-yn-1-yl)acrylamide (41) Propargyl amine (40) (3.98 g, 72.26 mmol) and DIPEA (12.5 mL, 71.99 mmol, 0.996 o eq), were dissolved in dichloromethane (132 mL). The reaction mixture was cooled to 0 C. Once chilled, acryloyl chloride (4.65 mL, 57.49 mmol, 0.80 eq) was added dropwise slowly (~30 mins). The reaction was then allowed to slowly warm to room temperature overnight. Stirring was continued for 20 hours after which the reaction mixture was washed with brine and water, dried over anhydrous sodium sulfate, and concentrated under vacuum. The product was obtained as an amorphous light yellow solid in an 84% yield after purification by column chromatography using an EtOAc/ Hex (2:1) mixture. Comparison with literature data 43 1 confirms its identity. H NMR (500 MHz, CDCl3) δ 2.22 (t, J = 2.5 Hz, 1H), 4.09 – 4.13 (dd, J = 2.5, 5.0 Hz, 2H), 5.63 – 5.69 (dd, J = 3.0, 10.5 Hz, 1H), 6.00 (broad s, 1H), 6.10 (dd, J = 10.5, 17.0 Hz, 1H), 6.29 (dd, J = 3.0, 17.0 Hz, 1H) 13 C-NMR (150 MHz, CDCl3),  29.2, 71.7, 79.3, 127.3, 130.1, 165.2. + [HRMS] [M+H] m/z: calc. for C6H8NO: 110.0606, found 110.0611. 92 N-(3-(Trimethylsilyl)prop-2-yn-1-yl)acrylamide (42) To a suspension of silver chloride (332.8 mg, 2.32 mmol, 0.25 eq) in dichloromethane (65 mL), compound (41) (1.01 g, 9.26 mmol), and DBU (6.85 mL, 45.80 mmol, 4.87 eq) were added. The reaction was allowed to stir at room temperature for 20 minutes, after which time, chlorotrimethylsilane (12.5 mL, 98.49 mmol, 10.64 eq) was added. After stirring the reaction at o room temperature for 1.5 hours the reaction was heated to 45 C for 22 hours. The reaction was then diluted with 200 mL of dichloromethane. The organic phase was washed with saturated sodium bicarbonate (2x 100mL), followed by 1% hydrochloric acid solution (2x 100mL), and water (1x 100mL). The organic phase was then dried over anhydrous sodium sulfate, filtered, and concentrated under vacuum. The product was obtained as an amorphous white solid in a 57% yield after purification by column chromatography using an EtOAc/ Hex (1:3) mixture. Comparison with literature data 43 confirms its identity. 1 H NMR (600 MHz, CDCl3) δ 0.14 (s, 9H), 4.13 (d, J = 5.1 Hz, 2H), 5.65 (dd, J = 1.2, 10.3 Hz, 1H), 5.85 ( broad s, 1H), 6.09 (dd, J = 10.3, 17.0 Hz, 1H), 6.28 (dd, J = 1.2, 17.0 Hz, 1H), 13 C-NMR (150 MHz, CDCl3),  -0.2, + 30.3, 88.7, 100.9, 127.1, 130.2, 164.9. [HRMS] [M+H] m/z: calc. for C9H16NOSi: 182.1001, found 182.0998. 3-(4-(Aminomethyl)-1H-1,2,3-triazol-1-yl)propan-1-ol (51) Step 1: 3-azido-propan-1-ol (50) needed to be freshly synthesized. Due to the low molecular weight of this azide, caution must be taken during the concentration step for this compound. This reaction was accomplished by addition of commercially available 3-chloropropan-1-ol (49) (1.5 g, 15.87 mmol, 1.0 eq.) to sodium azide (2.06g, 31.69 mmol, 2.0 eq.) and water (40 mL). 93 o This reaction was then heated to 80 C and left to react for 18 hours. After which time, the aqueous phase was extracted with EtOAc 3 to 4 times (40 mL). The organic layers were washed with brine (30 mL) and then dried over anhydrous sodium sulfate, filtered, and concentrated under close supervision under reduced pressure. This colorless oil (50) was used directly in the next step. Step 2: 3-Azido-propan-1-ol (50) (1.6 g, 15.82 mmol, 2.5 eq) synthesized in the previous step, propargyl amine (0.4 mL, 6.25 mmol, 1 eq), copper (I) iodide (1.00 g, 5.25 mmol, 0.84 eq.), and DIPEA (2.0 mL, 13.31 mmol, 2.13 eq.) were mixed together in a reaction flask o containing DMF (50 mL). The reaction was then heated to 50 C and continued overnight. After 16 hours, the product was isolated in a 70 % yield after both steps. Comparison with 55 literature data confirms its identity. 1 H NMR (600 MHz, D2O) δ 1.92 – 2.03 (m, 2H), 3.42 (t, J = 6.3 Hz, 2H), 3.74 (t, J = 7.1 Hz, 2H), 4.35 (t, J = 7.0 Hz, 2H), 7.75 (s, 1H), MHz, D2O) δ 31.93, 35.68, 47.29, 58.29, 123.39, 148.30. 13 C NMR (150 + ESI-MS [M+H] m/z: calc. for C6H12NaN4O0: 57.11 found 157.11. N-((1-(3-Hydroxypropyl)-1H-1,2,3-triazol-4-yl)methyl)acetamide (52) Compound (51) (202.3mg, 1.30 mmol, 1eq.) was first dissolved in DMF (10 mL). The o reaction mixture was then cooled to -60 C at which point 2, 4, 6- collidine (511 L, 4.63 mmol, 3.56 eq) was added. This mixture was allowed to stir for 20 mins to ensure the reaction mixture o was at -60 C. Then, acetyl chloride (91 L, 1.05 mmol, 0.8 eq.) was added dropwise to the chilled solution. Once added, the reaction was allowed to stir for 2 hours. At this point, it was o warmed to -30 C, diluted with EtOAc (10 mL) and washed with 1M HCl (10 mL), followed by 94 saturated sodium bicarbonate (10mL), and brine (10mL). The organic phase was then dried over anhydrous sodium sulfate, filtered, and concentrated under vacuum. The product was obtained in a 53 % yield, as a light yellow solid, after purification by column chromatography using a 56 DCM/ MeOH/ H20 (33: 15: 2) mixture. Comparison with literature data confirms its identity. 1 H NMR (600 MHz, D2O) δ 1.87 (s, 3H), 1.97 (p, J= 6.6 Hz, 2H), 3.42 (t, J = 6.6 Hz, 2H), 4.29 (s, 2H), 4.35 (t, J= 6.6 Hz, 2H), 7.75 (s, 1H), 13 C-NMR (150 MHz, CDCl3),  21.9, 31.9, 34.7, + 47.4, 58.3, 123.9, 144.7, 174.3. [HRMS] [M+H] m/z: calc. for C8H15N4O2: 199.1195, found 199.1202. N-((1-(3-Hydroxypropyl)-1H-1,2,3-triazol-4-yl)methyl)-6-(5-((3aS,4S,6aR)-2-oxohexahydro1H-thieno[3,4-d]imidazol-4-yl)pentanamido)hexanamide (53) Compound (51) (25.5 mg, 0.16 mmol, 1 eq.) and biotinated amino caproic acid Nhydroxysuccinimide ester (35) (80.2 mg, 0.18 mmol, 1.1 eq.) were dissolved in NMP (2 mL). Then DIPEA (26.4 L, 0.16 mmol, 1 eq.) was added and the reaction allowed to stir at room temperature for 17 hours. Once reaction had gone to completion as monitored my TLC using a o DCM/ MeOH/ H2O (33: 15: 2) solvent mixture, reaction was cooled to 0 C and then the product was precipitated out using ice cold ether. Product was then filtered and then re-dissolved in H2O. Product (53) was then lyophilized under high vacuum producing a pure white product in a 73 % yield. 1 H NMR (600 MHz, D2O) δ 1.31 – 1.38 (m, 2H), 1.39 – 1.50 (m, 3H), 1.52 – 1.59 (m, 2H), 1.59 – 1.83 (m, 7H), 2.15 – 2.22 (m, 2H), 2.30 (t, J = 7.2 Hz, 2H), 2.33 – 2.39 (m, 2H), 2.79 – 2.87 (m, 1H), 3.04 (dd, J = 13.1, 5.0 Hz, 1H), 3.16 – 3.28 (m, 2H), 3.35 – 3.42 (m, 1H), 95 3.63 (t, J = 6.2 Hz, 2H), 4.48 (dd, J = 7.9, 4.5 Hz, 1H), 4.51 (s, 2H), 4.57 (t, J = 7.0 Hz, 2H), 4.64 – 4.69 (m, 1H), 7.96 (s, 1H), 13 C-NMR (150 MHz, CDCl3),  25.0, 25.3, 25.6, 27.8, 28.0, 28.1, 32.0, 34.5, 35.59, 35.64, 39.2, 39.8, 47.4, 55.5, 58.3, 60.4, 62.2, 124.0, 144.8, 165.5, 176.8, + 177.0. [HRMS] [M+H] m/z: calc. for C22H38N7O4S: 496.2706, found 496.2719. General procedure for colorimetric determination of carbohydrate concentration. Serial dilutions of known carbohydrate (i.e., GalNAc, Gal) concentrations in PBS were made, ranging from 2 mg/ mL down to 4 g/ mL. Then, 200 L of each solution were pipetted into either a 96-well plate or Eppendorf tubes depending on the number of samples to be measured. Next, 50 L of an 80 % phenol solution in H2O was added to each tube. After this point, 1 mL of concentrated H2SO4 was added rapidly to each to allow good mixing. The mixtures were then allowed to stand at r.t. for 10 minutes before being placed for an additional o 15 minutes in a water bath set at 30 C. After this time, the solutions should develop a yellow color which can be measured at 480 nm. General procedure for vaccine emulsion preparation. Samples were combined with Freunds adjuvant (complete for first inject, and incomplete for all booster injections) in a 1: 1 ratio by volume. Samples were then vortexed on high overnight (>15-20 hours prior to immunizations). Note: all samples were calculated and prepared assuming one extra mouse more than what was used during the study to account for volume loss due to the formation of this emulsion and for loss in the syringe. After stirring overnight, emulsions were thick, white, and homogeneous. 96 Study 1: Immunization of mice with TMV-1 Cys conjugates Pathogen-free C57BL/6 female mice age 6-10 weeks were obtained from Jackson Laboratory (Bar Harbor, ME, USA) and maintained in the Animal Care Facility at the biochemistry building of Michigan State University. All animal care procedures and experimental protocols have been approved by the Institutional Animal Care and Use Committee (IACUC) Michigan State University. Vaccine Injections: Group 1- Control: Five C57BL/6 mice were injected subcutaneously (90 L) under the scruff on day 0 with 240 g of TMV-1Cys as an emulsion in CFA (0.1 mL, Fisher). Booster injections (90 L) were given subcutaneously under the scruff on days 14 and 21 with 240 g of TMV1Cys as emulsions in IFA (0.1 mL, Fisher). Group 2: Five C57BL/6 mice were injected subcutaneously (90 L) under the scruff on day 0 with 60 g of TMV-1Cys-Tn-N3 (containing 522 ng of GalNAc) as an emulsion in CFA. Booster (90 L) injections were given subcutaneously under the scruff on days 14 and 21 with 60 g of TMV-1Cys-Tn-N3 (containing 522 ng of GalNAc) as emulsions in IFA. Group 3: Five C57BL/6 mice were injected subcutaneously (90 L) under the scruff on day 0 with 240 g of TMV-1Cys-Tn-N3 (containing 2.1 µg of GalNAc) as an emulsion in CFA. Booster injections (90 L) were given subcutaneously under the scruff on days 14 and 21 with 240 g of TMV-1Cys-Tn-N3 (containing 2.1 µg of the GalNAc) as emulsions in IFA. Group 4: Three C57BL/6 mice were injected subcutaneously (90 L) under the scruff on day 0 with 53 g of TMV-1Cys-Tn-bromoacetamide (containing 137 ng of GalNAc) as an emulsion in CFA. Booster injections (90 L) were given subcutaneously under the scruff on days 14 and 21 97 with 53 g of TMV-1Cys-Tn-bromoacetamide (containing 137 ng of GalNAc) as emulsions in IFA. Blood Collections: Blood (~ 0.2 mL/mouse) was collected from all groups of mice on days 0, 7, 28 and 35. Sera o from each group of mice were allowed to clot at 0 C overnight and then centrifuged at 3,000 RPM for 10 minutes (longer time may be required, but it is recommended not to increase centrifugal rate as red blood cells will lyse). Sera was then extracted from the red blood cell clot o and stored at -20 C until needed for ELISA analysis Study 2: Immunization of mice with TMV-1 Cys conjugates reproduced Pathogen-free C57BL/6 female mice age 6-10 weeks were obtained from Jackson Laboratory (Bar Harbor, ME, USA) and maintained in the Animal Care Facility at the biochemistry building of Michigan State University. All animal care procedures and experimental protocols have been approved by the Institutional Animal Care and Use Committee (IACUC) Michigan State University. Vaccine Injections: Group 1: Four C57BL/6 mice were injected subcutaneously (90 L) under the scruff on day 0 with 240 g of TMV-1Cys-Tn-N3 (containing 2.1 µg of GalNAc) as an emulsion in CFA (0.1 mL, Fisher). Booster injections (90 L) were given subcutaneously under the scruff on days 14 and 28 with 240 g of TMV-1Cys-Tn-N3 (containing 2.1 µg of GalNAc) as emulsions in IFA (0.1 mL, Fisher). 98 Group 2- Control: Four C57BL/6 mice were injected subcutaneously (90 L) under the scruff on day 0 with 240 g of TMV-1Cys as an emulsion in CFA. Booster injections (90 L) were given subcutaneously under the scruff on days 14 and 28 with 240 g of TMV-1Cys as emulsions in IFA. Group 3- Control: Two C57BL/6 mice were NOT injected during the course of this study to test ensure that the mice antibody titers remain low throughout the course of the experiment. Blood Collections: Blood (~ 0.2 mL/mouse) was collected from all groups of mice on days 0, 7, and 35. Sera from o each group of mice were allowed to clot at 0 C overnight and then centrifuged at 3,000 RPM for 10 minutes (longer time may be required, but it is recommended not to increase centrifugal rate as red blood cells will lysis can occur). Sera was then extracted from the red blood cell clot and o stored at -20 C until needed for ELISA analysis Study 3: Immunization of mice with Q conjugates Pathogen-free C57BL/6 female mice age 6-10 weeks were obtained from Jackson Laboratory (Bar Harbor, ME, USA) and maintained in the Animal Care Facility at the biochemistry building of Michigan State University. All animal care procedures and experimental protocols have been approved by the Institutional Animal Care and Use Committee (IACUC) Michigan State University. Vaccine Injections: Group 1: Four C57BL/6 mice were injected subcutaneously (90 L) under the scruff on day 0 with 17 g of Q-Tn-N3 (containing 450 ng of GalNAc) as an emulsion in CFA (0.1 mL, 99 Fisher). Booster injections (90 L) were given subcutaneously under the scruff on days 14 and 28 with 17 g of Q-Tn-N3 (containing 450 ng of GalNAc) as emulsions in IFA (0.1 mL, Fisher). Group 2: Five C57BL/6 mice were injected subcutaneously (90 L) under the scruff on day 0 with 69 g of Q-Tn-N3 (containing 1.8 µg of GalNAc) as an emulsion in CFA. Booster injections (90 L) were given subcutaneously under the scruff on days 14 and 28 with 68.97 g of Q-Tn-N3 (containing 1.8 µg of GalNAc) as emulsions in IFA. Group 3: Four C57BL/6 mice were injected subcutaneously (90 L) under the scruff on day 0 with 345 g of Q-Tn-N3 (containing 9 µg of GalNAc) as an emulsion in CFA. Booster injections (90 L) were given subcutaneously under the scruff on days 14 and 28 with 345 g of Q-Tn-N3 (containing 9 µg of GalNAc) as emulsions in IFA. Group 4- Control: Four C57BL/6 mice were injected subcutaneously (90 L) under the scruff on day 0 with 220 g of Q-propanol (containing 2.4 µg of propanol) as an emulsion in CFA. Booster injections (90 L) were given subcutaneously under the scruff on days 14 and 28 with 220 g of Q-propanol (containing 2.4 µg of propanol) as emulsions in IFA. Group 5- Control: Three C57BL/6 mice were injected subcutaneously (90 L) under the scruff on day 0 with 103 g of Q-glucose (containing 2.9 µg of Glc) as an emulsion in CFA. Booster injections (90 L) were given subcutaneously under the scruff on days 14 and 28 with 103 g of Q-glucose (containing 2.9 µg of Glc) as emulsions in IFA. 100 Blood Collections: Blood (~ 0.2 mL/mouse) was collected from all groups of mice on days 0, 7, and 35. Sera from o each group of mice were allowed to clot at 0 C overnight and then centrifuged at 3,000 RPM for 10 minutes (longer time may be required, but it is recommended not to increase centrifugal rate as red blood cells will lyse). Sera was then extracted from the red blood cell clot and stored at o 20 C until needed for ELISA analysis General procedure for ELISA with NeutrAvidin To each well of a 96-well microtiter plate (Thermo scientific Nunc Immuno 96-well solid plates (clear, passive surfaces) flat-bottomed; Maxisorp: pinchbar; 400 L/ well), 100 L of a NeutrAvidin solution (8 g / mL in PBS) was added and the plate was allowed to incubate o overnight at 4 C (cover with aluminum foil). The next morning (~15 hours), the NeutrAvidin solution was thrown away and the plate washed with 200 L PBST (0.5 % tween 20 in PBS) three times. Wells were dried by tapping over absorbent pads and residual air bubbles were carefully removed. Next, 200 L of 1% BSA/PBS was added as a blocking agent and the plate was allowed to incubate for 1 hour at room temperature. After the hour had passed, the blocking solution was thrown away and the plate washed again three times with 200 L PBST like above. Wells were dried by tapping over absorbent pads and residual air bubbles were carefully removed. Then, 200 L of biotinated antigen (5 g/ mL in 0.1 % BSA/ PBS) was added to each o well. It was then incubated at 37 C for 1 hour. During this wait time, 1:4 serial dilutions of the diluted serum samples (serum to 0.1 % BSA/PBS 1:10 ratio) were prepared. After the incubation was complete, the plate was washed three times with PBST and the wells dried 101 thoroughly. 100 L of the freshly prepared serum samples was added to each well and incubated o at 37 C for 2 hours. The plate was again washed three times with PBST and dried. Next, 100 L of a 1: 2000 dilution (secondary antibody (Ab) to 0.1 % BSA/PBS) of horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG + IgM, IgG, or IgM (Jackson ImmunoResearch Laboratory IgG + IgM catalog # 115-035-068, IgG catalog # 115-035-071, or IgM catalog # 115o 035-075) antibody was added to each well. This secondary antibody was incubated at 37 C for 1 hour. During the last 10 minutes of this wait period, a solution of 3, 3’, 5, 5’- tetramethylbenzidine (TMB) was prepared (5 mg of TMB was dissolved in 2 mL of DMSO and then diluted to 20 mL using citrate phosphate buffer). Immediately prior to adding the TMB solution to the plate, 20 L of H2O2 was added. After incubation the plate was washed three times with PBST and dried. The freshly prepared TMB solution was then added (200 L/ well). Color (blue) was allowed to develop for 10-20 minutes. At this time, 50 L of a 0.5 M H2SO4 solution was added to quench the reaction (blue color changes to yellow). Note: if the color became too intense, the reaction should be quenched sooner. The optical density (OD) was then measured using a plate reader at 450 nm. Each experiment was repeated three to four times and the average was used to calculate the titers. Antibody titers were defined as the maximum folds of dilution resulting in a 0.1 OD unit higher than the average absorbance after greater than one hundred thousand folds of dilution. General procedure for ELISA with BSA To each well of a 96-well microtiter plate (Thermo scientific Nunc Immuno 96-well solid plates (clear, passive surfaces) flat-bottomed; Maxisorp; pinchbar; 400 L/ well), 100 L of a 102 BSA- Tn (10 g / mL in PBS) was added and the plate was allowed to incubate overnight at 4 o C. The next morning (~15 hours), the BSA- Tn solution was thrown away and the plate washed with 200 L PBST (0.5 % tween 20 in PBS) three times. Wells were dried by tapping over absorbent pads and residual air bubbles were carefully removed. Next, 200 L of 1% BSA/PBS was added as a blocking agent and the plate was allowed to incubate for 1 hour at room temperature. During this wait time, 1:4 serial dilutions of the diluted serum samples (serum to 0.1 % BSA/PBS 1:10 ratio) were prepared. After the hour had passed, the blocking solution was thrown away and the plate washed again three times with 200 L PBST like above. 100 L of o the freshly prepared serum samples was added to each well and incubated at 37 C for 2 hours. The plate was again washed three times with PBST and dried. Next, 100 L of a 1: 2000 dilution (secondary antibody (Ab) to 0.1 % BSA/PBS) of horseradish peroxidase (HRP)conjugated goat anti-mouse IgG + IgM, IgG, or IgM (Jackson ImmunoResearch Laboratory IgG + IgM catalog # 115-035-068, IgG catalog # 115-035-071, or IgM catalog # 115-035-075) o antibody was added to each well. This secondary antibody was incubated at 37 C for 1 hour. During the last 10 minutes of this wait period, a solution of 3, 3’, 5, 5’- tetramethylbenzidine (TMB) was prepared (5 mg of TMB was dissolved in 2 mL of DMSO and then diluted to 20 mL using citrate phosphate buffer). Immediately prior to adding the TMB solution to the plate, 20 L of H2O2 was added. After incubation the plate was washed three times with PBST and dried. The freshly prepared TMB solution was then added (200 L/ well). Color (blue) was allowed to develop for 10-20 minutes. At this time, 50 L of a 0.5 M H2SO4 solution was added to quench the reaction (blue color changes to yellow). Note: if the color became too intense, the reaction 103 should be quenched sooner. The optical density (OD) was then measured using a plate reader at 450 nm. Each experiment was repeated three to four times and the average was used to calculate the titers. Antibody titers were defined as the maximum folds of dilution resulting in a 0.1 OD unit higher than the average absorbance after greater than one hundred thousand folds of dilution. General procedure for competitive ELISA The protocol for competitive ELISA was almost identical to that of the normal ELISA protocol mentioned previously. The only difference occurred after incubation with the biotinated antigen and wash. A 0.1 M GalNAc solution (dissolved in 0.1 % BSA/PBS) was used to dilute the serum samples (instead of 0.1 % BSA/ PBS like before). Note: The serum was allowed to incubate for 1 hour before addition to the plate as well as being directly added to the plate. No significant difference was noticed between the two methods. General Procedure for FACS NCI/ ADR- RES ovarian cancer cells were obtained from Professor Paul Erhardt, the University of Toledo, and were grown in serum containing RPMI-1640 containing 2.0 mM Lo glutamine supplemented with 10% FBS, and incubated at 37 C and 5% CO2. On the day of the experiment, 100,000 cells/ mL were transferred to eppendorf tubes, and the tubes were labeled corresponding to the serum sample being tested. The cells were than washed with FACS buffer (1% BSA, 0.1% NaN3/PBS) using centrifugation. The cells were then re-suspended in 1:10 o diluted test sera (75 L), and incubated at 4 C for 30 minutes. The cells were then washed 104 twice with FACS buffer (1 mL). All tubes, including the control tube which does not contain serum, received 100 L of a 1:100 diluted goat antimouse IgG labeled with FITC (Jackson ImmunoResearch Laboratory, catalog #115-095-164). Tubes were incubated with this secondary o antibody at 4 C for 30 minutes. The cells were washed again with FACS buffer (1 mL) and then resuspended in FACS buffer (200 L). Fluorescence was then accessed immediately on a Vantage FACS to determine the percentage of positive cells. 105 APPENDICES 106 Appendix 1 – Spectral Data 107 HDO 9.5 8.5 7.5 6.5 5.5 4.5 δ (ppm) 1 Figure 1.34: 600 MHz (D2O), H NMR of 19 108 3.5 2.5 1.5 0.5 200 180 160 Figure 1.35: 125 MHz (D2O), 140 120 100 80 δ (ppm) δ 13 C NMR of 19 109 60 40 20 0 HDO 9.5 8.5 7.5 6.5 5.5 4.5 δ δ (ppm) 1 Figure 1.36: 600 MHz (D2O), H NMR of 20 110 3.5 2.5 1.5 0.5 200 180 160 140 120 100 80 δ (ppm) 1 Figure 1.37: 150 MHz (D2O), H NMR of 20 111 60 40 20 0 CHCl3 9.5 8.5 7.5 6.5 5.5 4.5 δ δ (ppm) 1 Figure 1.38: 500 MHz, (CDCl3), H NMR of 41 112 3.5 2.5 1.5 0.5 CHCl3 200 180 160 Figure 1.39: 150 MHz, (CDCl3), 140 120 100 80 δδ (ppm) 13 C NMR of 41 113 60 40 20 0 CHCl3 9.0 8.0 7.0 6.0 5.0 4.0 δ (ppm) 1 Figure 1.40: 600 MHz, (CDCl3), H NMR of 42 114 3.0 2.0 1.0 0.0 CHCl3 200 180 160 Figure 1.41: 150 MHz, (CDCl3), 140 120 100 80 δ (ppm) 13 C NMR of 42 115 60 40 20 0 HDO a a b b a d b c d c 9.5 8.5 7.5 6.5 5.5 4.5 δ δ (ppm) 1 Figure 1.42: 600 MHz, (D2O), H NMR of 46A 116 3.5 2.5 1.5 0.5 HDO 9.5 8.5 7.5 6.5 5.5 4.5 δ δ (ppm) 1 Figure 1.43: 500 MHz, (D2O), H NMR of 47A 117 3.5 2.5 1.5 0.5 HDO 9.5 8.5 7.5 6.5 5.5 4.5 δδ (ppm) 1 Figure 1.44: 300 MHz, (D2O), H NMR of 46B 118 3.5 2.5 1.5 0.5 HDO 9.5 8.5 7.5 6.5 5.5 4.5 δδ(ppm) 1 Figure 1.45: 600 MHz, (D2O), H NMR of 47B 119 3.5 2.5 1.5 0.5 HDO X X= X X 9.5 8.5 7.5 6.5 5.5 4.5 δ δ (ppm) 1 Figure 1.46: 500 MHz, (D2O), H NMR of 46C 120 3.5 2.5 1.5 0.5 HDO 9.5 8.5 7.5 6.5 5.5 4.5 δδ (ppm) 1 Figure 1.47: 600 MHz (D2O), H NMR of 52 121 3.5 2.5 1.5 0.5 200 180 160 Figure 1.48: 600 MHz (D2O), 140 120 100 80 δ (ppm) 13 C NMR of 52 122 60 40 20 0 1 2 3 5 6 7 8 8.5 7.5 6.5 5.5 4.5 δ (ppm) Figure 1.49: 600 MHz (D2O), gCOSY NMR of 52 123 3.5 2.5 1.5 δ (ppm) 4 HDO 9.5 8.5 7.5 6.5 5.5 4.5 δ δ (ppm) 1 Figure 1.50: 600 MHz (D2O), H NMR of 53 124 3.5 2.5 1.5 0.5 200 180 160 Figure 1.51: 150 MHz (D2O), 140 120 100 80 δδ (ppm) 13 C NMR of 53 125 60 40 20 0 1 2 5 6 7 8 9 9.0 7.5 6.0 4.5 δ δ (ppm) Figure 1.52: 600 MHz (D2O), COSY NMR of 53 126 3.0 1.5 δ 4 δ (ppm) 3 20 30 60 70 80 90 100 110 120 130 140 9.0 7.5 6.0 4.5 δ δ (ppm) Figure 1.53: 600 MHz (D2O), HMQC NMR of 53 127 3.0 1.5 δ (ppm) δ 40 50 9.0 7.5 6.0 4.5 δ δ (ppm) Figure 1.54: 600 MHz (D2O), HMBC NMR of 53 128 3.0 1.5 δ (ppm) δ 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 REFERENCES 129 References 1. In Cancer Facts & Figures 2008, American Cancer Society, Ed., Atlanta: American Cancer Society: 2008, 52. 2. Globocan 2008. International Agency for Research on Cancer. 2010. 3. Gajewski, T. F.; Meng, Y.; Blank, C.; Brown, I.; Kacha, A.; Kline, J.; Harlin, H., Immune Resistance Orchestrated by the Tumor Microenvironment. Immun. Rev. 2006, 213, 131145. 4. Parham, P., The Immune System. Third Ed. ed.; Garland Science, Taylor & Francis Group, LLC: 2009; p 506. 5. Danishefsky, S. J.; Allen, J. R., From the Laboratory to the Clinic: A Retrospective on Fully Synthetic Carbohydrate-Based Anticancer Vaccines. Angew. Chem. Int. Ed. 2000, 39, 836863. 6. Bernstein, M. A.; Hall, L. D., A General Synthesis of Model Glycoproteins: Coupling of Alkenyl Glycosides to Proteins, Using Reductive Ozonolysis Followed by Reductive Amination with Sodium cyanoborohydride. Carbohydr. Res. 1980, 78, C1-C3. 7. Lemieux, R. U., Haworth Memorial Lecture. Human Blood Groups and Carbohydrate Chemistry. Chem. Soc. Rev. 1978, 7, 423-452. 8. Kuduk, S. D.; Schwarz, J. B.; Chen, X.-T.; Glunz, P. W.; Sames, D.; Ragupathi, G.; Livingston, P. O.; Danishefsky, S. J., Synthetic and Immunological Studies on Clustered Modes of Mucin-Related Tn and TF O-Linked Antigens: The Preparation of a Glycopeptide-Based Vaccine for Clinical Trials against Prostate Cancer J. Am. Chem. Soc. 1998, 120, 12474-12485. 9. Kagan, E.; Ragupathi, G.; Yi, S. S.; Reis, C. A.; Gildersleeve, J.; Kahne, D.; Clausen, H.; Danishefsky, S. J.; Livingston, P. O., Comparison of Antigen Constructs and Carrier Molecules for Augmenting the Immunogenicity of the Monosaccharide Epithelial Cancer Antigen Tn. Cancer Immunol. Immunother. 2005, 54, 424-430. 10. Ragupathi, G.; Coltart, D. M.; Williams, L. J.; Koide, F.; Kagan, E.; Allen, J.; Harris, C.; Glunz, P. W.; Livingston, P. O.; Danishefsky, S. J., On the Power of Chemical Synthesis: Immunological Evaluation of Models for Multiantigenic Carbohydrate-Based Cancer Vaccines. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 13699. 130 11. Coltart, D. M.; Royyuru, A. K.; Williams, L. J.; Glunz, P. W.; Sames, D.; Kuduk, S. D.; Schwarz, J. B.; Chen, X.-T.; Danishefsky, S. J.; Live, D. H., Principles of Mucin Architecture: Structural Studies on Synthetic Glycopeptides Bearing Clustered Mono-, Di-, Tri-, and Hexasaccharide Glycodomains. J. Am. Chem. Soc. 2002, 124, 9833-9844. 12. Ragupathi, G.; Koide, F.; Livingston, P. O.; Cho, Y. S.; Endo, A.; Wan, Q.; Spassova, M. K.; Keding, S. J.; Allen, J.; Ouerfelli, O.; Wilson, R. M.; Danishefsky, S. J., Preparation and Evaluation of Unimolecular Pentavalent and Hexavalent Antigenic Constructs Targeting Prostate and Breast Cancer: A Synthetic Route to Anticancer Vaccine Candidates. J. Am. Chem. Soc. 2006, 128, 2715-2725. 13. Zhu, J.; Wan, Q.; Lee, D.; Yang, G.; Spassova, M. K.; Ouerfelli, O.; Ragupathi, G.; Damani, P.; Livingston, P. O.; Danishefsky, S. J., From Synthesis to Biologics: Preclinical Data on a Chemistry Derived Anticancer Vaccine. J. Am. Chem. Soc. 2009, 131, 9298-9303. 14. Buskas, T.; Li, Y.; Boons, G.-J., Synthesis of a Dimeric Lewis Antigen and the Evaluation of the Epitope Specificity of Antibodies Elicited in Mice. Chem. Eur. J. 2005, 11, 5457-5467. 15. Buskas, T.; Li, Y.; Boons, G.-J., The Immunogenicity of the Tumor-Associated Antigen y Lewis May Be Suppressed by a Bifunctional Cross-Linker Required for Coupling to a Carrier Protein. Chem. Eur. J. 2004, 10, 3517-3524. 16. Keil, S.; Claus, C.; Dippold, W.; Kunz, H., Towards the Development of Antitumor Vaccines: A Synthetic Conjugate of a Tumor-Associated MUC1 Glycopeptide Antigen and a Tetanus Toxin Epitope. Angew. Chem. Int. Ed. 2001, 40, 366-369. 17. Dziadek, S.; Hobel, A.; Schmitt, E.; Kunz, H., A Fully Synthetic Vaccine Consisting of a Tumor-Associated Glycopeptide Antigen and a T-Cell Epitope for the Induction of a Highly Specific Humoral Immune Response. Angew. Chem. Int. Ed. 2005, 44, 7630-7635. 18. Buskas, T.; Ingale, S.; Boons, G.-J., Towards a Fully Synthetic Carbohydrate-Based Anticancer Vaccine: Synthesis and Immunological Evaluation of a Lipidated Glycopeptide Containing the Tumor-Associated Tn Antigen. Angew. Chem. Int. Ed. 2005, 44, 5985-5988. 19. Ingale, S.; Wolfert, M. A.; Gaekwad, J.; Buskas, T.; Boons, G.-J., Robust Immune Responses Elicited by a Fully Synthetic Three-Component Vaccine. Nat. Chem. Biol. 2007, 3, 663-667. 131 20. Grigalevicius, S.; Chierici, S.; Renaudet, O.; Lo-Man, R.; Dériaud, E.; Leclerc, C.; Dumy, P., Chemoselective Assembly and Immunological Evaluation of Multiepitopic Glycoconjugates Bearing Clustered Tn Antigen as Synthetic Anticancer Vaccines. Bioconjugate Chem. 2005, 16, 1149-1159. 21. Moreau, R.; Dausset, J.; Bernard, J.; Moullec, J., Anémie Hémolytique Acquise Avec Polyagglutinabilité des Hématies par un Nouveau Facteur Présent dans le Sérum Humain Normal (Anti-Tn). Bull. Soc. méd. HÔp., paris 1957, 73, 569-587. 22. Dausset, J.; Moullec, J.; Bernard, J., Acquired Hemolytic Anemia with Polyagglutinability of Red Blood Cells Due to a New Factor Present in Normal Human Serum (Anti-Tn). Blood 1959, 14, 1079-1093. 23. Eric G, B., Tn-syndrome. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease 1999, 1455, 255-268. 24. Berger, E. G.; Thurnher, M.; Dinter, A., The cell-specific expression of glycan antigens exemplified by the deficiency of 1, 3 galactosyltransferase in the Tn-syndrome. Transfus. Clin. Biol. 1994, 1, 103-108. 25. Springer, G. F.; Desai, P. R.; Banatwala, I., Blood Group MN Specific Substances and Precursors in Normal and Malignant Human Breast Tissues. Naturwissenschaften 1974, 61, 457458. 26. Springer, G. F., T and Tn, General Carcinoma Autoantigens Science 1984, 224, 11981206. 27. Dabelsteen, E., Cell Surface Carbohydrates as Prognostic Markers in Human Carcinomas. J. Pathol. 1996, 179, 358-369. 28. Goddard-Borger, E. D.; Stick, R. V., An Efficient, Inexpensive, and Shelf-Stable Diazotransfer Reagent: Imidazole-1-sulfonyl Azide Hydrochloride. Org. Lett. 2007, 9, 37973800. 29. Koeller, K. M.; Smith, M. E. B.; Wong, C.-H., Chemoenzymatic Synthesis of PSGL-1 Glycopeptides: Sulfation on Tyrosine Affects Glycosyltransferase-Catalyzed Synthesis of the Oglycan. Bioorg. Med. Chem. 2000, 8, 1017-1025. 132 30. Huang, Y.; Dey, S.; Zhang, X.; Sönnichsen, F.; Garner, P., The α-Helical Peptide Nucleic Acid Concept: Merger of Peptide Secondary Structure and Codified Nucleic Acid Recognition. J. Am. Chem. Soc. 2004, 126, 4626-4640. 31. Grundler, G.; Schmidt, R. R., Glycosylimidate, 13. Anwendung des TrichloracetimidatVerfahrens auf 2-Azidoglucose- und 2-Azidogalactose-Derivate. Liebigs Ann. Chem. 1984, 1984, 1826-1847. 32. Karplus, M., Vicinal Proton Coupling in Nuclear Magnetic Resonance. J. Am. Chem. Soc. 1963, 85, 2870-2871. 33. Karplus, M., Contact Electron- Spin Coupling of Nuclear Magnetic Moments. J. Chem. Phys. 1959, 30, 11-15. 34. Miermont, A.; Barnhill, H.; Strable, E.; Lu, X.; Wall, K. A.; Wang, Q.; Finn, M. G.; Huang, X., Cowpea Mosaic Virus Capsid: A Promising Carrier for the Development of Carbohydrate Based Antitumor Vaccines. Chem. Eur. J. 2008, 14, 4939-4947. 35. Matyjaszewski, K.; Xia, J., Atom Transfer Radical Polymerization. Chemical Reviews 2001, 101, 2921-2990. 36. Bernard, J.; Hao, X.; Davis, T. P.; Barner-Kowollik, C.; Stenzel, M. H., Synthesis of Various Glycopolymer Architectures via RAFT Polymerization: From Block Copolymers to Stars. Biomacromolecules 2006, 7, 232-238. 37. Chiefari, J.; Chong, Y. K.; Ercole, F.; Krstina, J.; Jeffery, J.; Le, T. P. T.; Mayadunne, R. T. A.; Meijs, G. F.; Moad, C. L.; Moad, G.; Rizzardo, E.; Thang, S. H., Living Free- Radical Polymerization by Reversible Addition- Fragmentation Chain Transfer: The RAFT Process. Macromolecules 1998, 31, 5559-5562. 38. Hawker, C. J.; Bosman, A. W.; Harth, E., New Polymer Synthesis by Nitroxide Mediated Living Radical Polymerizations. Chem. Rev. 2001, 101, 3661-3688. 39. Baskaran, S.; Grande, D.; Sun, X.-L.; Yayon, A.; Chaikof, E. L., GlycosaminoglycanMimetic Biomaterials. 3. Glycopolymers Prepared from Alkene-Derivatized Mono- and Disaccharide-Based Glycomonomers. Bioconjugate Chem. 2002, 13, 1309-1313. 133 40. Grande, D.; Baskaran, S.; Baskaran, C.; Gnanou, Y.; Chaikof, E. L., GlycosaminoglycanMimetic Biomaterials. 1. Nonsulfated and Sulfated Glycopolymers by Cyanoxyl-Mediated FreeRadical Polymerization. Macromolecules 2000, 33, 1123-1125. 41. Grande, D.; Baskaran, S.; Chaikof, E. L., Glycosaminoglycan Mimetic Biomaterials. 2. Alkene- and Acrylate-Derivatized Glycopolymers via Cyanoxyl-Mediated Free-Radical Polymerization. Macromolecules 2001, 34, 1640-1646. 42. Sun, X.-L.; Grande, D.; Baskaran, S.; Hanson, S. R.; Chaikof, E. L., Glycosaminoglycan Mimetic Biomaterials. 4. Synthesis of Sulfated Lactose- Based Glycopolymers That Exhibit Anticoagulant Activity. Biomacromolecules 2002, 3, 1065-1070. 43. Malkoch, M.; Thibault, R. J.; Drockenmuller, E.; Messerschmidt, M.; Voit, B.; Russell, T. P.; Hawker, C. J., Orthogonal Approaches to the Simultaneous and Cascade Functionalization of Macromolecules Using Click Chemistry. J. Am. Chem. Soc. 2005, 127, 14942-14949. 44. Carpita, A.; Mannocci, L.; Rossi, R., Silver(I)- Catalysed Protiodesilylation of 1(Trimethylsilyl)-1-alkynes. Eur. J. Org. Chem. 2005, 2005, 1859-1864. 45. Tornøe, C. W.; Christensen, C.; Meldal, M., Peptidotriazoles on Solid Phase: [1,2,3]Triazoles by Regiospecific Copper(I)-Catalyzed 1,3-Dipolar Cycloadditions of Terminal Alkynes to Azides. J. Org. Chem. 2002, 67, 3057-3064. 46. Hermanson, G. T., Bioconjugate Techniques. second ed.; Academic Press, Inc: 2008; p 1202. 47. DuBois, M.; Gilles, K. A.; Hamilton, J. K.; Rebers, P. A.; Smith, F., Colorimetric Method for Determination of Sugars and Related Substances. Anal. Chem. 1956, 28, 350-356. 48. Jahnel, J. B.; Ilieva, P.; Frimmel, F. H., HPAE-PAD- A Sensitive Method for the Determination of Carbohydrates. Fresenius J. of Anal. Chem. 1998, 360, 827-829. 49. Miermont, A. Syntheses and Immunological Studies of Tumor Associated Carbohydrate Antigens. Ph.D. Dissertation, University of Toledo. Toledo, OH, 2008. 50. McCormick, A. A.; Corbo, T. A.; Wykoff-Clary, S.; Nguyen, L. V.; Smith, M. L.; Palmer, K. E.; Pogue, G. P., TMV- Peptide Fusion Vaccines Induce Cell- Mediated Immune Responses and Tumor Protection in Two Murine Models. Vaccine 2006, 24, 6414-6423. 134 51. Kaltgrad, E.; O'Reilly, M. K.; Liao, L.; Han, S.; Paulson, J. C.; Finn, M. G., On-Virus Construction of Polyvalent Glycan Ligands for Cell-Surface Receptors. J. Am. Chem. Soc. 2008, 130, 4578-4579. 52. Overby, L. R.; Barlow, G. H.; Doi, R. H.; Jacob, M.; Spiegelman, S., Comparison of Two Serologically Distinct Ribonucleic Acid Bacteriophages I. Properties of the Viral Particles. J. Bacteriol. 1966, 91, 442-448. 53. Strable, E.; Prasuhn, D. E.; Udit, A. K.; Brown, S.; Link, A. J.; Ngo, J. T.; Lander, G.; Quispe, J.; Potter, C. S.; Carragher, B.; Tirrell, D. A.; Finn, M. G., Unnatural Amino Acid Incorporation into Virus-Like Particles. Bioconjugate Chem. 2008, 19, 866-875. 54. Prasuhn, D. E.; Singh, P.; Strable, E.; Brown, S.; Manchester, M.; Finn, M. G., Plasma Clearance of Bacteriophage Q Particles as a Function of Surface Charge. J. Am. Chem. Soc. 2008, 130, 1328-1334. 55. Ishihara, K.; Kurihara, H.; Yamamoto, H., An Extremely Simple, Convenient, and Selective Method for Acetylating Primary Alcohols in the Presence of Secondary Alcohols. J. Org. Chem. 1993, 58, 3791-3793. 56. Jlalia, I.; Beauvineau, C.; Beauvière, S.; Önen, E.; Aufort, M.; Beauvineau, A.; Khaba, E.; Herscovici, J.; Meganem, F.; Girard, C., Automated Synthesis of a 96 Product- Sized Library of Triazole Derivatives Using a Solid Phase Supported Copper Catalyst. Molecules 2010, 15, 3087-3120. 135 Chapter 2 – Synthesis of an Alzheimer’s Disease Imaging Agent 2.1 Introduction Currently, Alzheimer’s Disease (AD) is attributed to the sixth leading cause of death in 1 the United States and the fifth leading cause of death in Americans over the age of 65. As the average age of the current population increases, the number of people affected by this disease is projected to soar. The cost surrounding the disease is staggering. It is estimated that $148 billion is spent annually in direct and indirect costs. This disease can be characterized by a progressive loss of cognitive function, memory decline, and altered behavior. In the later stages of this disease, motor function deteriorates and global amnesia sets in. What makes this disease so difficult to detect and treat, is that the only absolute definitive diagnosis for the disease is during postmortem neuropathological examinations. Because of this, it is absolutely essential that a new method for detection be developed allowing earlier detection. Earlier detection would also be a great assisting tool in attempts directed toward treatment of the disease. One such option was developed by Barrio and coworkers in which they developed the hydrophobic, fluorescent molecular-imaging naphthyl)ethylidene malononitrile 18 probe 2-(1-(6-[(2-[ F]fluoroethyl)(methyl)amino]-2- 18 ([ F] 18 FDDNP) and its analog, 1-(6-[(2- 18 [ F]fluoroethyl)(methyl)amino]naphthalene-2-yl)ethanone ([ F] FFENE) both of which were found to be capable of crossing the blood- brain barrier (Figure 2.1). These two compounds were found to have Kd values of 0.12 and 0.16 nM for high- affinity binding sites for FDDNP and FENE respectively, and Kd values of 1.86 and 71.2 nM for low-affinity sites. 136 Figure 2.1: Structures of FENE and FDDNP These two compounds have previously been used to label senile plaques and neurofibrillary tangles (two hallmark pathologies accompanying the neurodegeneration involved in the diseased state) in the living brains of patients using positron emission tomography (PET). One of the biggest limitations to these current compounds is that they require the use of radiolabeled fluorine. Many locations (i.e., hospitals/ clinics), do not have the ability nor the expertise to work with such compounds. Design of a more universal analog would therefore be highly advantageous. Realizing that these current compounds have demonstrated success, the goal of our project was to design and synthesize an analog that would be capable of being attached to a nanoparticle. Once conjugated to the nanoparticle scaffold, these compounds could then be observed using magnetic resonance imaging (MRI). The original direction of our design was to synthesize a compound in which we installed an amine linker in place of the radio- labeled 18 18 fluorine in [ F] FENE. Once demonstrated as proof of concept in MRI, the [ F] FDDNP compound would also be synthesized in an analogous fashion. 137 2.2 Synthesis of Alzheimer’s Imaging Agent A paper published in a 1999 symposium abstract by J.R. Barrio and coworkers made 2 mention of the Bucherer reaction in the reversible conversion of aromatic alcohols with 3 secondary amines. After looking into this named reaction, literature precedence indicated that this reaction might be an excellent starting point in our synthesis. To perform this reaction, a 2naphthol compound needed to be synthesized that would allow the synthesis of the 2, 6 substitution pattern that we desired. It was decided to start from commercially available 1-(6methoxynaphthalen-2-yl)ethanone (1) which could easily be converted into 2 in the presence of concentrated hydrochloric acid (Scheme 2.1). This reaction was completed after two hours and the product was isolated in an 85-90% yield. Scheme 2.1: Synthesis of 2-6 -substituted aromatic alcohol 2 With the desired aromatic alcohol in hand, it was time to determine if the Bucherer reaction could couple the secondary amine to the aromatic alcohol. Compound 2 was placed in a Schlenk tube, along with 2-methylaminoethanol, sodium bisulfite, and water. The flask was o sealed and then heated to approximately 140 C for four days (Scheme 2.2). After work-up and purification, it was determined that the reaction had proceeded in a 70 %. 138 Scheme 2.2: Synthesis of 10 utilizing the Bucherer reaction 4 This reaction mechanism had been elucidated by Seeboth in 1967 (Scheme 2.3). It is believed that the aromatic alcohol will first pick up a proton from its surroundings forming intermediate 3. Then, when the resonance form intermediate 4 interacts with sodium bisulfite, intermediate 5 could be generated. Intermediate 5 can then undergo tautermerization to form intermediate 6. This cyclic ketone can interact with the secondary amine generating intermediate 7. Next, intermediate 7 would undergo a proton transfer from the amine to the alcohol, which could subsequently undergo dehydration forming intermediate 8. Next a molecule of sodium bisulfite could eliminate a proton from the 5- position generating intermediate 9. And finally, syn-elimination of sodium sulfite from intermediate 9 would produce the desired compound 10. 139 Scheme 2.3: Predicted mechanism for the synthesis of 10 using the Bucherer reaction 140 With this compound in hand, the rest of the reaction steps are quite straight forward (Scheme 2.4). Based on literature precedence in which they converted a primary alcohol to a 5 primary amine, primary alcohol (10) was reacted with p- toluenesulfonyl chloride in dichloromethane, with DMAP and pyridine. This reaction was allowed to stir at room temperature overnight at which point, the reaction was found to contain two major products. As it turned out, beside the desired tosylate compound (11), a significant amount of the chlorosubstituted compound was observed (12). The ratio of 11 to 12 varied anywhere from 1:1 to 5:1 respectively. This by-product is likely a result of the chloride anion generated during the reaction of the primary alcohol with Tos-Cl acting as a nucleophile and displacing the tosylate leaving group via an SN2 mechanism. Although these compounds could be separated via column chromatography, it was decided to use them both directly in the next step. The only disadvantage to having the chloro- compound in the reaction mixture was that it is, of course, not as good a leaving group as the tosylate compound. Thus, it caused the reaction time to be increased. Both 11 and 12 were combined with sodium azide, DMF, and a few drops of water to o help with the solubility of the sodium azide. After 15 hours of stirring at 55 C azide compound 13 was isolated in a 68 % yield over the past two steps. The final step was a straight forward hydrogenation of azide compound 13 to the desired primary amine target (14) using standard conditions. This reaction resulted in a 71 % yield of an extremely fluorescent yellow solid. Unfortunately however, it appears as though this compound is extremely unstable because even after very short periods of storage or use (i.e., NMR analysis, mass measurements) decomposition appears to occur. 141 Scheme 2.4: Synthesis of the desired FDDNP imaging agent (14) 142 With this unfortunate occurrence, it was decided that it would be more advantageous to synthesis a free acid instead of this free amine. This might prevent the compound from being quite so sensitive to decomposition. Since the nanoparticles can be synthesized to express either acid or amine functionality, it does not make a difference which our compound includes. It was decided to perform a one pot reduction of azide (13) which subsequently reacted with succinic anhydride generating the free acid compound (15) (Scheme 2.5). Scheme 2.5: Synthesis of final imaging agent to be conjugated to magnetic nanoparticles 143 2.3 Experimental Section General Experimental Procedures. All reactions were carried out under nitrogen or argon with anhydrous solvents in ovendried glassware. Chemicals used were reagent grade as supplied except where noted. Analytical thin-layer chromatography was performed using silica gel 60 F254 glass plate. Compound spots were visualized by UV light (254 nm) and by staining with a yellow solution containing Ce(NH4)2(NO3)6 (0.5 g) and (NH4)6Mo7O24 * 4H2O (24.0 g) in 6% H2SO4 (500 mL) or a solution of Ninhydrin (1.5g) in n-butanol (100mL) and acetic acid (3mL). Flash column chromatography was performed on silica gel 60 (230–400 Mesh). NMR spectra were referenced 1 using Me4Si (0 ppm), residual CHCl3 (δ H-NMR 7.24 ppm, 1 NMR 4.63 ppm), CH3OH (δ H-NMR 3.30 ppm, NMR 2.04 ppm, 1 1 1 13 13 1 C-NMR 77.0 ppm), H2O (δ H1 C-NMR 49.0 ppm), and (CH3)2CO (δ H- 13 C-NMR 29.84 ppm). Peak and coupling constant assignments were based on 1 13 1 13 H-NMR, H– H gCOSY and (or) H– C gHMQC and H– C gHMBC experiments. ESI mass spectra were recorded in positive ion mode. 1-(6-Hydroxynaphthalen-2-yl)ethanone (2) In a 3L three neck round bottom flask, equipped with a reflux condenser, dropping funnel and glass stopper, concentrated HCl (2L) was stirred and heated to boiling. Meanwhile, commercially available 1-(6-methoxynaphthalen-2-yl)ethanone (1) (5.92g, 29.57 mmol) was dissolved in the minimal amount of dichloromethane and added to the drop funnel. This solution was added drop-wise to the HCl solution. Once completely added, the reaction was allowed to stir at reflux for 3 hours. The reaction was then allowed to cool to r.t. at which point a solid 144 precipitated out which was then filtered and then dissolved in EtOAc (130mL). EtOAc was extracted 3x with 100 mL of water and then washed once with 100 mL of brine. The organic phase was dried over anhydrous sodium sulfate, and concentrated under vacuum, and used without further purification. Product was obtained as a white solid in 86-89% yield. 3 Comparison with literature data confirms its identity. 1 H NMR (500 MHz, (CD3)2CO) δ 2.64 (s, 3H), 7.23 (dd, J = 2.4, 8.8 Hz, 1H), 7.26 (d, J = 2.4 Hz, 1H), 7.74 (d, J = 8.7 Hz, 1H), 7.93 (dd, J = 1.8, 8.7 Hz, 1H), 7.97 (d, J = 8.8 Hz, 1H), 8.52 (d, J=1.8 Hz, 1H), 9.01 (broad s, 1H), 13 C NMR (150 MHz, (CD3)2CO) δ 26.5, 109.9, 120.1, 123.3 125.0, 127.3, 128.3, 131.2, 132.4, + 133.2, 138.5, 197.5. [HRMS] [M+H] m/z: calc. for C12H11O2: 187.0759, found 187.0755. 1-(6-((2-Hydroxyethyl)(methyl)amino)naphthalene-2-yl)ethanone (10) Compound 2 (2g, 10.74mmol), sodium bisulfite (4.47g, 42.96mmol), and 2(methylamino) ethanol (4.03g, 53.7 mmol), and water (15mL) were added to a schlenk tube. The atmosphere within the schlenk tube was evacuated under high vacuum and then filled with nitrogen. This process was repeated three times before sealing the schlenk tube filled with o nitrogen atmosphere. The reaction vessel was then placed in an oil bath and heated at 140 C for 4 days. The reaction was allowed to cool to r.t. and was then extracted with EtOAc and water. The organic phase was then washed once with brine and then dried over anhydrous sodium sulfate, and concentrated under vacuum The desired compound was obtained in a 70% yield 3 after column purification using EtOAc/ Et2O (1:2). Comparison with literature data confirms its identity. 1 H NMR (500 MHz, CD3OD) δ 2.69 (s, 3H), 3.18 (s, 3H), 3.67 (t, J = 5.8 Hz, 2H), 3.83 (dd, J = 3.1, 5.8 Hz, 2H), 6.99 (d, J = 2.6 Hz, 1H), 7.31 (dd, J = 2.6, 9.2 Hz, 1H), 7.66 145 (d, J = 8.8 Hz, 1H), 7.84 – 7.92 (d, J = 9.2 Hz, 1H), 7.84 – 7.92 (dd, J = 8.8, 1.7 Hz, 1H), 8.42 (d, J = 1.7 Hz, 1H), 13 C NMR (125 MHz, CD3OD) δ 26.4, 39.4, 55.5, 60.4, 105.9, 117.4, + 125.1, 126.4, 127.2, 131.4, 131.9, 132.0, 139.5, 151.1, 200.3 [HRMS] [M+H] m/z: calc. for C15H18NO2: 244.1338, found 244.1339. 2-((6-Acetylnaphthalen-2-yl)(methyl)amino)ethyl 4-methylbenzenesulfonate (11) Compound 10 (495 mg, 2 mmol), p- toluenesulfonyl chloride (796.2 mg, 4.2 mmol), and dichloromethane (5 mL) were added to a 10 mL round bottom flask. This mixture was allowed o to cool to 0 C. Once chilled, DMAP (179.6 mg, 0.72 mmol) and pyridine (520 L, 2.82 mmol) were added to the reaction, which was then allowed to warm to r.t. and stirring continued overnight. Once starting material was consumed as monitored by TLC using EtOAc/Hex (2:3), the reaction was quenched using water. The aqueous phase was extracted using DCM. The combined organic layers were then washed with 1N HCl and then dried over anhydrous sodium sulfate, and concentrated under vacuum. Compound was purified as a yellow solid by column chromatography using EtOAc/Hex (2:3); however, the yield of compound 11 was highly variable due to the formation of byproduct 12. The ratio of 11 to 12 ranges from 1:1 to 5:1. Once other minor impurities are removed, 11 and 12 can be used together directly in the next step. Reaction 5 was designed based on the following literature precedence for the conversion of compound 10 1 to 11. H NMR (600 MHz, CDCl3) δ 2.28 (s, 3H), 2.65 (s, 3H) 3.02 (s, 3H), 3.75 (t, J = 5.8 Hz, 2H), 4.23 (t, J = 5.8 Hz, 2H), 6.73 (d, J = 2.6 Hz, 1H), 6.97 – 7.02 (dd, J = 2.6, 9.2 Hz, 1H), 7.09 -7.12 (m, 2H), 7.57 (d, J = 8.7 Hz, 1H), 7.61 – 7.66 (m, 2H), 7.72 (d, J = 9.2 Hz, 1H), 7.91 (dd, J 146 = 1.8, 8.7 Hz, 1H), 8.28 (d, J = 1.8 Hz, 1H), 13 C NMR (150 MHz, CDCl3) 21.5, 26.4, 39.3, 51.2, 66.8, 102.4, 105.7, 115.7, 124.7, 125.4, 126.3, 127.7, 129.7, 130.2, 130.9, 131.2, 132.5, 137.5, 144.9, 148.0, 197.6. + [HRMS] [M+H] m/z: calc. for C22H24NO4S: 398.1426, found 398.1419. 1-(6-((2-Chloroethyl)(methyl)amino)naphthalen-2-yl)ethanone (12) Compound 10 (495 mg, 2mmol), p- toluenesulfonyl chloride (796.2 mg, 4.2 mmol), and dichloromethane (5mL) were added to a 10 mL round bottom flask. This mixture was allowed to o cool to 0 C. Once chilled, DMAP (179.6mg, 0.72mmol) and pyridine (520 L, 2.82mmol) were added to the reaction which was then allowed to warm to r.t. and stirring continued overnight. Once starting material was consumed as monitored by TLC using EtOAc/Hex (2:3), reaction was quenched using water. The aqueous phase was extracted using DCM. The combined organic layers were then washed with 1N HCl and then dried over anhydrous sodium sulfate, and concentrated under vacuum. Compound was purified as a brownish yellow solid after column chromatography using EtOAc/Hex (2:3), however, the yield of compound 12 is highly variable due to the formation of byproduct 11. The ratio of 12: 11 ranges from 1:1 to 1:5. Once other minor impurities are removed, 11 and 12 can be used together directly in the next 5 step. Reaction was designed based on the following literature precedence for the conversion of compound 10 to 11. 1 H NMR (600 MHz, CDCl3) δ 2.65 (s, 3H), 3.15 (s, 3H), 3.68 (t, J = 7.1 Hz, 2H), 3.81 (t, J = 7.1 Hz, 2H), 6.91 (d, J = 2.6 Hz, 1H), 7.15 (dd, J = 2.6, 9.1 Hz, 1H), 7.63 (d, J = 8.8 Hz, 1H), 7.81 (d, J = 9.1 Hz, 1H), 7.92 (dd, J = 1.8, 8.8 Hz, 1H), 8.31 (d, J=1.8 Hz, 1H), 13 C NMR (150 MHz, CDCl3) δ.26.4, 39.2, 40.4, 54.3, 105.6, 115.7, 124.8, 125.4, 126.3, 147 130.3, 131.1, 131.3, 137.6, 148.1, 197.7. + [HRMS] [M+H] m/z: calc. for C15H17ClNO: 262.0999, found 262.1003. 1-(6-((2-Azidoethyl)(methyl)amino)naphthalen-2-yl)ethanone (13) To a mixture of compounds 11 and 12 (938 mg, 3.6 mmol, assumes 100% lower molecular weight compound 12), sodium azide (3.49 g, 53.7 mmol), and DMF (5 mL) were added. A few drops of water were added to help with the solubility of the sodium azide. The o reaction was stirred at r.t. for 1 hour before being heated to 55 C overnight (15-20 hours). Solvent was then removed under reduced pressure. The product was obtained as a bright yellow solid in a 68% yield over two steps after purification by flash column chromatography using either a EtOAc/ Hex (1:3) or an acetone/ toluene (1:9) system. 1 H NMR (500 MHz, CDCl3) δ 2.65 (s, 3H), 3.14 (s, 3H), 3.52 (t, J = 6.1 Hz, 2H), 3.68 (t, J = 6.1 Hz, 2H), 6.89 (d, J = 2.6 Hz, 1H), 7.15 (dd, J = 2.6, 9.1 Hz, 1H), 7.63 (d, J = 8.8 Hz, 1H), 7.81 (d, J = 9.1 Hz, 1H), 7.92 (dd, J = 1.8, 8.8 Hz, 1H), 8.28 – 8.33 (d, J=1.8 Hz, 1H), 13 C-NMR (125 MHz, CDCl3), δ 26.4, 39.0, 49.0, 51.8, 105.6, 115.8, 124.8, 125.4, 126.3, 130.3, 131.0, 131.1, 137.6, 148.4, 197.7 [HRMS] + [M+H] m/z: calc. for C15H17N4O: 269.1402, found 269.1404. 4-((2-((6-Acetylnaphthalen-2-yl)(methyl)amino)ethyl)amino)-4-oxobutanoic acid (15) Compound 13 (92.6 mg, 0.35 mmol) was dissolved in dry THF (5 mL). To this solution, succinic anhydride (69 mg, 0.69 mmol, 1.97 eq) was added and allowed to dissolve. Once dissolved, 10% Palladium on carbon (10.8 mg, 0.10 mmol, 0.29 eq) was added and allowed to stir at room temperature, under a hydrogen atmosphere for about 5 hours. Upon completion of 148 the reaction, as monitored by TLC, the reaction mixture was filtered through Celite and concentrated under reduced pressure. The product, 15, was obtained in an 87 % yield after purification by flash column chromatography using a MeOH/ DCM (1: 19) solvent mixture. 1 H NMR (600 MHz, CD3OD) δ 2.36 (t, J = 7.5 Hz, 2H), 2.44 (t, J = 7.5 Hz, 2H), 2.63 (s, 3H), 3.10 (s, 3H), 3.34 (broad s, 1H), 3.42 (t, J = 6.6 Hz, 2H), 3.61 (t, J = 6.6 Hz, 2H), 6.96 (d, J=2.5 Hz, 1H), 7.28 (dd, J = 2.5, 9.0 Hz, 1H), 7.63 (d, J = 8.8 Hz, 1H), 7.82 – 7.85 (m, 2H), 8.37 (d, J = 1.8 + Hz, 1H), [HRMS] [M+H] m/z: calc. for C19H23N2O4: 343.1658, found 343.1649. 149 APPENDICES 150 Appendix 2 – Spectral Data 151 (CD3)2CO 9.5 8.5 7.5 6.5 5.5 4.5 δ δ (ppm) 1 Figure 2.2: 500 MHz, (CD3)2CO, H NMR of 2 152 3.5 2.5 1.5 0.5 (CD3)2CO (CD3)2CO 230 200 Figure 2.3: 150 MHz, (CD3)2CO, 170 140 100 δδ (ppm) 13 C NMR of 2 153 80 60 40 20 0 CD3OD 9.5 8.5 7.5 6.5 5.5 4.5 δ δ (ppm) 1 Figure 2.4: 500 MHz, (CD3CD), H NMR of 10 154 3.5 2.5 1.5 0.5 CD3OD 200 170 Figure 2.5: 125 MHz, (CD3CD), 140 110 δδ (ppm) 13 C NMR of 10 155 80 60 40 20 0 CHCl3 H2O 9.5 8.5 7.5 1 6.5 5.5 4.5 δ (ppm) δ Figure 2.6: 300 MHz, (CDCl3), H NMR of 11 156 3.5 2.5 1.5 0.5 CHCl3 200 170 Figure 2.7: 150 MHz, (CDCl3), 140 110 δδ(ppm) 13 C NMR of 11 157 80 60 40 20 0 CHCl3 H2O 8.5 7.5 6.5 5.5 4.5 3.5 δ (ppm) 1 Figure 2.8: 600 MHz, (CDCl3), H NMR of 12 158 2.5 1.5 0.5 CHCl3 200 170 Figure 2.9: 150 MHz, (CDCl3), 140 110 δδ(ppm) 13 C NMR of 12 159 80 60 40 20 0 CHCl3 H2O 9.5 8.5 7.5 6.5 5.5 4.5 δ δ (ppm) 1 Figure 2.10: 500 MHz, (CDCl3), H NMR of 13 160 3.5 2.5 1.5 0.5 CHCl3 250 210 Figure 2.11: 125 MHz, (CDCl3), 170 130 δδ (ppm) 13 C NMR of 13 161 90 60 30 0 0 1 2 4 5 6 7 8 8.5 7.0 5.5 4.0 δδ (ppm) Figure 2.12: 500 MHz, (CDCl3), gCOSY of 13 162 2.5 1.0 δ (ppm) δ 3 philazid Tue Jan 19 17:49:21:46 2010 70 70 65 65 % Transmittance % T r a n 60 60 s m i t t a 55 55 n c e 50 50 N3 45 45 4000 4000 3500 3500 3000 3000 2500 2500 2000 2000 1500 1500 Wavenumbers Wavenumber (cm Figure 2.13: Confirmation of azide stretch using IR of 13 163 -1 ) 1000 1000 CD3OD 9.5 8.5 7.5 6.5 5.5 4.5 δ δ (ppm) 1 Figure 2.14: 500 MHz, (CD3OD), H NMR of 15 164 3.5 2.5 1.5 0.5 CD3OD 200 170 Figure 2.15: 150 MHz, (CD3OD), 140 110 δ δ (ppm) 13 C NMR of 15 165 80 60 40 20 0 6.5 7.0 7.5 8.0 8.5 8.0 7.0 6.0 5.0 δ (ppm) ` Figure 2.16: 600 MHz, (CD3OD), gCOSY of 15 166 4.0 3.0 δ (ppm) 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 9.0 8.0 7.0 6.0 5.0 δ (ppm) Figure 2.17: 600 MHz, (CD3OD), HMQC of 15 167 4.0 3.0 2.0 δ (ppm) δ 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 REFERENCES 168 References 1. 2009 Alzheimer's Disease Facts and Figures. Alzheimer's & dementia: the journal of the Alzheimer's Association 2009, 5, 234-270. 2. Barrio JR; Huang SC; Cole G; Satyamurthy N; Petric A; Phelps ME; G, S., PET Imaging of Tangles and Plaques in Alzheimer Disease with a Highly Hydrophobic Probe. J. Label Compd. Radiopharm 1999, 42, S194- S195. 3. Barrio, J. R.; Petric, A.; Satyamurthy, N. Methods for Binding Agents to - Amyloid Plaques. WO/2005/040337, 2005. 4. Seeboth, H., The Bucherer Reaction and the Preparative Use of its Intermediate Products. Angew. Chem. Int. Ed. 1967, 6, 307-317. 5. Ghosh, A. K.; Leshchenko-Yashchuk, S.; Anderson, D. D.; Baldridge, A.; Noetzel, M.; Miller, H. B.; Tie, Y.; Wang, Y.-F.; Koh, Y.; Weber, I. T.; Mitsuya, H., Design of HIV-1 Protease Inhibitors with Pyrrolidinones and Oxazolidinones as Novel P1'-Ligands To Enhance Backbone-Binding Interactions with Protease: Synthesis, Biological Evaluation, and ProteinLigand X-ray Studies. J. Med. Chem. 2009, 52, 3902-3914. 169