u a w . . . ... .w _... my. .“5. #Mmmv a Wig... V V . V 5.. its: . ”I fiJA .u‘ I V3.3 .. ”a? h . ..V . . 1:. “MM...“ 1. ... z :52, aw. , H — 'Ar ‘ . . wflfinn s fiwsrvf. , fl... V , V 2%. .T u V. ., "map. 5 . . “flaw. .n V. . “my; 5.. . . . k... V A a». .. Ewan” as.” . . V .mfinjv . 3 3.: I .. ,I. .a... I ; .h urtvflou: . n Y ran“ 3... 1,? 1...». Tu? {Flf ... .1. .v. . V . 3V nfifiw‘mm. In}. L % V. 9 .31.. .5 A ; firinuio :- THESIS 2501 This is to certify that the dissertation entitled PART I: DETECTION AND SEQUENCING OF t OLIGOSACCHARIDES AND PEPTIDES. PART II: CI—lROMnTOGhAI‘h‘Y AND DETECTION OF SULI’IlOQUINOVOSE nND SUSPECTED METABOLITES leOlVi ltHIZOBIUl‘Vi MELILOTI 0" - 1021 presented by James J. Bradford has been accepted towards fulfillment of the requirements for Pho D. degreeinChemiStrL Date 19 play , 2000 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE moo Wipes-p.14 Part Part I: Detection and Sequencing of Oligosaccharides and Peptides. Part II: Chromatography and Detection of Sulphoquinavose and Suspected Metabolites From Rhizobium meliloti bv. 1021. By James J. Bradford A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 2000 Dc glycosyle SOphistic These 5 glycofon required ABSTRACT By James J. Bradford Detailed analysis of carbohydrate structures associated with glycosylation remains a major analytical challenge that requires sophisticated instrumentation and a variety of sample preparation strategies. These samples are often available only in limited quantities of each glycoform. Therefore, sensitive techniques that are highly reproducible are required for the structure determination of oligosaccharides. All sugars are polyalcohols. Modifications to these sugars include the presence of an amine group, a carboxylic acid group, and/or an acetal group. All reducing sugars also are in equilibrium either as a ketone or aldehyde. The chemistry of each of these groups can be taken advantage of to improve detection and structural determination through derivatization. These reactions provide for analysis specificity and, through the formation of derivatives, detectability and structural information can be obtained. The techniques used for the separation and detection of phosphorilated and sulfated saccharides are similar to the techniques used for saccharides. The attempts to resolve the biosynthesis of Sulphoquinovosediacylglycerol so far have proven fruitless. Only recently has it become feasible to rigorously bathe biOSITIil andthe test the different chemical and genetic approaches. Work on determining the biosynthetic pathway has fallen into two areas: the sulfoglycolytic pathway, and the sugar-nucleotide pathway. For Sean and Kaitlyn Live Long and Prosperl I l none of research strongly approac'r and hisi M discussi chemist} I hadn‘t theoretj1 Biocheni and Ca experien djTECtioi them. I deSEn'e DEOple 5 that CO. th‘JSe S ACKNOWLEDGIVIENTS I have had many teachers and advisors that have influenced me, but none of them compare to my mentor Dr. Hollingsworth. His approach to research is similar to his approach to mentoring, very hands on. He is strongly focused and very intent on his approach to science. I know that my approach to mentoring will be derived from my experiences within his lab and his interactions with me. My fellow lab mates past and present, with which I have had many discussions, certainly deserve mention. Jie and Gang opened my eyes to the chemistry occurring within the laboratory. Our discussions revealed a world I hadn’t envisioned. Guanfei and Hussein showed ‘me the practical and theoretical chemistry that can be achieved through personal diligence. Biochemical discussions with Carol and Atima, and conversations with them and Gabriela nearly always resulted in some fun and a good learning experience. Ben and Rob gave me a View of biochemistry from a different direction than the one I had been trained in, it was a pleasure working with them. I have had many teachers that influenced me. The two that most deserve mention are Mr. Michael Crowl and Dr. Richard Lees. These two people showed me how to teach and gave me an appreciation for instructors that could treat students as people. Thanks to Mr. Crowl, I was able to find those students that didn’t know how to interact with instructors and get them to. teaching: student i think ah F;I support me to SI little ha instance helped i fmstratt possible them to. Thanks to Dr. Lees I knew how to treat them as equals while still teaching. To them I owe the most rewarding complaint I ever received from a student “He was a tough grader. He never gave direct answers, he made us think about the answers.” Finally, I must thank my parents, friends, wife, and children. The support from them has been overwhelming. My parents have always wanted me to succeed at any project that I had started, and knew when to push a little harder for me to complete it. My friends have kept me going, in some instances literally, from Bob Hazekamp, to Preston Davis and Ken Roth, they helped me retain some form of sanity by listening to me when I was frustrated. My two beautiful children have affected me more than I thought possible, they bring joy to my life. vi Carboh; Chapter Introdu Materig ProcedL Results COUCluE Chapte Utilizir IntrOdu Xpem ESUlts COHCI U: Chapte nirodL actefi to...” Benefit Epara and SL1 TABLE OF CONTENTS List of Tables and Graphs ................................................................................. ix List of Figures ..................................................................................................... x List of Abbreviations .......................................................................................... xv Chapter 1: Carbohydrate Analysis Literature Review Introduction ......................................................................................................... 1 Carbohydrate Chemistry .................................................................................... 3 Carbohydrate Analysis: Derivatization Techniques ........................................................................... 7 Polyacrylamide Gel Electrophoresis ........................................................... 9 Column Chromatography .......................................................................... 11 Ion Exchange Chromatography ................................................................. 11 Reductive Amination ................................................................................. 13 Microbore High Pressure Liquid Chromatography .................................. 15 Peptide Analysis ............................................................................................... 17 Peptide Nomenclature ...................................................................................... 20 Carbohydrate Nomenclature ............................................................................ 22 Chapter 2: Structure Determination of Reducing Saccharides Introduction ....................................................................................................... 26 Materials ........................................................................................................... 29 Procedures ......................................................................................................... 30 Results and Discussion ..................................................................................... 36 Conclusion ......................................................................................................... 69 Chapter 3: Primary and Secondary Sequence Determination for Peptides Utilizing a Single Quadrapole Mass Spectrometer Introduction ....................................................................................................... 72 Experimental ..................................................................................................... 79 Results and Discussion ..................................................................................... 80 Conclusion ......................................................................................................... 93 Chapter 4: Sulphoquinovose Literature Review Introduction ....................................................................................................... 96 Bacterial Symbiosis ........................................................................................ 103 Biosynthesis and Bacterial Factors ............................................................... 105 Genetic Studies ............................................................................................... 110 Separation and Detection of Sulphoquinovose and Sulphoquinovose Diacylglycerol .............................................................. 111 vii "4 Isolation of Sulphoquinovose and Metabolites .............................................. 113 Chapter 5: Synthesis of Synthetic Standards of Sulphoquinovose and Suspected Metabolites Introduction ..................................................................................................... 115 Materials and Methods Synthesis .................................................................................................. 120 Separations .............................................................................................. 124 Results and Discussion Organic Reactions .................................................................................... 129 Separations .............................................................................................. 140 Conclusion ....................................................................................................... 146 Acknowledgements ......................................................................................... 147 Chapter 6: Extraction and Structure Determination of Sulphoquinovose Introduction ..................................................................................................... 148 Materials and Methods ................................................................................... 149 Results and Discussion ................................................................................... 153 Conclusion ....................................................................................................... 155 Acknowledgement ........................................................................................... 157 Appendices ....................................................................... , ............................... 158 Bibliography .................................................................................................... 165 viii LIST of TABLES and GRAPHS Graph 2.1 ........................................................................................................... 39 Relationship between the average retention time over three consecutive injections on an ODS column verses the degree of polymerization of the saccharide. Error bars are included to indicate the range of retention times. Table 2.1 ............................................................................................................ 40 Analysis of the retention times of saccharides of varying degrees of polymerization. Table 3.1 ............................................................................................................ 88 Data from the in-source fragmentation of twenty one peptides. The type of fragment and the peptide bond that fragmented are indicated. Table 3.2 ............................................................................................................ 90 Continuation of the data from the in-source fragmentation of twenty one peptides. The type of fragment and the peptide bond that fragmented are indicated. LIST OF FIGURES 1.1 3 Isomerization of linear and pyranose form of glucose. For derivatization purposes, the linear form, containing the aldehyde, is required. 1.2 20 Peptide fragmentation nomenclature, typically for low energy CID only b and y type fragments are observed. 1.3 22 Fragmentation nomenclature for linear saccharides as developed by Domon and Costello. 1.4 24 Nomenclature for branched saccharides as developed by Domon and Costello. 2.1 33 Scheme for reductive amination of saccharides with a derivatizing reagent. This example utilizes a-aminopyfidine. For most experiments pyrenemethylamine was used. 2.2 35 Maltodextrin separated on an ODS column after derivatization with PMA. The degree of polymerization of the maltodextrin decreases with increasing retention time. 2.3 38 Alternative separation for maltodextrins derivatized with PMA (DP. 5 15) on an ODS column using a methanol gradient. 2.4 41 Electrospray mass spectrum of pyrenemethylamine derivatized maltodextrin (DP. 3 15). 2.5 42 Two pmol 11.1:1 of disaccharide directly injected into the mass spectrometer. Total volume infused, 10 111. 2.6 43 Schematic of the intermediate pressure region in an electrospray mass spectrometer . 2.7 44 Configuration of Fisons Platform electrospray mass spectrometer. 2.8 51 Fragmentation of PMA derivatized maltose. The first set of data utilizes increased nebulization gas flow rate to fragment (Figure A). The second set of data utilize conventional in source fragmentation, increased potential on the extraction cone, to fragment derivatized maltose (Figure B). increa at voltage ‘ 2.11 Spect1 a! men . " extrat ti of37“. respet t' 2.12 Resul s cone. Spect: u Voltag e increz 5 2.13 SPGCt. '1. deriv; t applit d' TESpeI i 2.14 Spect; '3. Won 3;. eXirm ti Spect; '1.- 2.9 53 Spectra from experiments in which gentiobiose, derivatized with pyrenemethylamine, is subjected to increasing voltage applied to the extraction cone. Spectrum A is table with the normal operating voltage of 37 V. In spectrum B the voltage applied to the cone is increased to 45 V, while in spectrum C the voltage applied is further increased to 50 V. 2.10 55 Spectra from experiments in which gentiobiose derivatized with pyrenemethylamine is subjected to increasing voltages applied to the extraction cone. Spectrum A has the operating voltage at 55 V. Spectrum B increases the voltage applied to the cone to 60 V, and spectrum C has the voltage applied increased to 65 V. 2.11 57 Spectra depicting experimental results where lactose, derivatized with pyrenemethylamine, is subject to increasing potentials applied to the extraction cone. The results shown here are for the normal operating voltage of 37 V, 45 V, and 50 V applied to the cone for spectra A, B, and C, respectively. 2.12 59 Results from experiments in which increasing voltages are applied to the cone. The saccharide derivatized with PMA for these spectra is lactose. Spectrum A has the operating voltage set to 55 V. Spectrum B increases the voltage applied to the cone to 60 V, and spectrum C has the voltage applied increased to 65 V. 2.13 61 Spectrum A, B, and C show the experimental results where melibiose, derivatized with pyrenemethylamine, is subjected to increasing voltages applied to the extraction cone. The voltages applied are 37, 45, and 50 V, respectively. 2.14 63 Spectra from experiments in which melibiose, derivatized with pyrenemethylamine, is subjected to increasing potentials applied to the . extraction cone. Spectrum A has the operating potential set to 55 V. Spectrum B increases the voltage applied to the cone to 60 V, and spectrum C has the potential applied increased to 65 V. 2.15 65 Spectra from experiments in which sophorose, derivatized with pyrenemethylamine, is subjected to increasing voltage applied to the extraction cone. Spectrum A is table with the normal operating voltage of 37 V. In spectrum B the voltage applied to the cone is increased to 45 V, while in spectrum C the voltage applied is further increased to 50 V. xi 2.16 67 Spectra depicting experimental results where sophorose, derivatized with pyrenemethylamine, is subjected to increasing voltages applied to the extraction cone. Spectrum A has the operating potential set to 55 V. Spectrum B increases the voltage applied to the cone to 60 V, and spectrum C has the potential applied increased to 65 V. 2.17 69 Separation of maltodextrin (n S 15) by microbore HPLC. There is overlap of the various degrees of polymerization. The first eluting band at 37.42 minutes contains degrees of polymerization from ten to eight, the second eluting band at 40 minutes contains degrees of polymerization from nine to seven. 3.1 74 Two hundred pmol [11'1 of fetuin characterizing the poorer sensitivity of a glycosylated peptide due to a lack of single, distinct charge series. The presence of a mixture, due to differences in glycosylation, decreases sensitivity. For these materials, ES-MS may provide a spectrum that appears as a large unresolved “hump” as opposed to nonglycosylated, or deglycosylated, species of similar molecular weight. 3.2 76 Two hundred fmol ul'1 of horse heart myoglobin, HHM, showing an envelope of multiple charge states. 3.3 82 Example of a peptide (QKSQRSQAENPV) analyzed by the Fisons Platform mass spectrometer with the potential set to 37 V to obtain molecular weight information. 3.4 84 The same peptide as figure 3.3 fragmented using in-source fragmentation (cone potential set to 55 V). Nearly the entire sequence of the peptide is obtained, and the linkages not cleaved can be inferred from the data. 3.5 86 Ion current obtained for two peptides (QKSQRAQAENPV, solid line, and QKSQRSQDENPV, dashed line) obtained under conditions of increasing cone potential. 4.1 99 The carbon (A) and nitrogen (B) cycles: the chemical forms and key processes in the biogeochemical cycling of carbon and nitrogen, respectively. Bacteria mediate all the critical steps of nitrogen fixation, nitrification, and denitrification. 4.2 101 The sulfur cycle, showing various transformations of organic and inorganic compounds. xii 4.3 105 The general structure of nodulation factors. 4.4 107 Schematic for the sugar-nucleotide pathway of sulfoquinovose synthesis. 4.5 109 Schematic of sulfoquinovose with the nod genes indicated that are responsible for the formation of the various bonds. 5.1 116 One proposed pathway for a nucleotide sugar to form sulfolipids starting with either UDP-glucose or UDP-galactose. 5.2 118 An alternative route to the formation of sulfolipids using the gluconeogenesis pathway. 5.3 121 A-B. Formation of the sulfonic acid of acetaldehyde. C. Formation of the sulfonic acid of lactic acid. 5.4 123 The formation of the methylglycoside of sulfoquinovose from methyl- glucopyranoside. B. Formation of the sulfonic acid of propanal-2-01-3- sulfonic acid from the methylglycoside of sulfoquinovose by periodate oxidation. , 5.5 126 NMR first reaction product, the sulfonic acid of acetaldehyde dimethylacetal from the starting material of bromo-acetaldehyde-dimethylacetal. 5.6 128 The final reaction product of the first series, the sulfonic acid of lactic acid, one of the suspected metabolites of sulfoquinovose. Formation of the sulfonic acid of lactic acid after the reaction of the sulfonic acid of acetaldehyde with - CN and Ca(COa)2. 5.7 131 The 13C spectrum from the sulfonic acid of acetaldehyde dimethylacetal. 5.8 132 NMR of the synthesized SQ-Me from the starting material of the glycopyranose methylglycoside. 5.9 134 NMR of the reaction product of SQ-Me after periodate oxidation, the formation of the sulfonic acid of propanal-2-ol-3-sulfonic acid. 5.10 137 Reverse phase HPLC separation with absorbance detection at 240 nm of or- aminopyridine, p-nitrobenzylamine, and pyrenemethylamine after reductive amination using glucose as the test molecule. 5.11 139 Anion exchange separation of SQ and the sulfonic acid of acetaldehyde. xiii 5.12 145 Separation of SQ derivatized with a—aminopyridine, p-nitrobenzylamine, and pyrenemethylamine. 6.1 151 First eluting fraction after sample was purified from R. 1021. 6.2 153 NMR of SQDG from the published work of Cedergren et a1. xiv LIST OF ABBREVIATIONS 2-AA ................................................................................................ 2-Aminobenzoic acid 2-AP or a—AP ........................................................................................ a-Aminopyridine 2-D ....................................................................................................... Two Dimensional ABEE ..................................................................................... Aminobenzoic Ethyl Ester CAD ........................................................................ Collisionally Activated Dissociation CCD ........................................................................................... Charge Coupled Device CE ........................................................................................... Capillary Electrophoresis CID ................................................................................. Collision Induced Dissociation D.P. ......................................................................................... Degree of Polymerization ECD ...................................................................................... Electron Capture Detector EI ...................................................................................................... Electron Ionization ESI ............................................................................................. Electrospray Ionization FAB ......................................................................................... Fast Atom Bombardment FACE ........................................... Fluorophore-Assisted Carbohydrate Electrophoresis FID ......................................................................................... Flame Ionization Detector GAG .............................................................................................. Glucoseaminoglycans GC ................................................................................................. Gas Chromatography HPAEC ...................................................... High pH Anion Exchange Chromatography HPLC ............................................................... High Pressure Liquid Chromatography ISAA ....................................................................... Ion-Suppression Amine-Absorption LIF ................................................................................... LASER Induced Fluorescence m/z .................................................................................................. mass to charge ratio MS .................................................................................................... Mass Spectrometry MS/MS ................................................................................ Tandem Mass Spectrometry NMR ................................................................................. Nuclear Magnetic Resonance ODS ........................................................................................................... Octadecylsilyl PA ........................................................................................................ Pyridylamination PAD ............................................................................. Pulsed Amperometric Detection PAGE ..................................................................... Polyacrylamide Gel Electrophoresis PAGEFS .......................................................... PAGE Fluorophore-labeled Saccharides PGPR ................................................................ Plant Growth Promoting Rhizobacteria PMA ................................................................................................ Pyrenemethylamine PMP ............................................................................. 1-Phenyl-3-Methyl-5-Pyrazolone PNGase F .................................................................................... Protein N-glyconase F RIC ............................................................................ Reconstructed Ion Chromatogram RP-HPLC ...................................................................................... Reverse Phase HPLC SFD ......................................................................... Saccharide Fluorophore Derivative SQDG ............................................................................ Sulphoquinovosyldiacylglycerol TIC ...................................................................................................... Total Ion Current UV .................................................................................................................. UltraViolet vis ......................................................................................................................... Visible PART I: ‘ DETECTION AND SEQUENCING OF OLIGOSACCHARIDES AND PEPTIDES. Chapter 1 CARBOHYDRATE ANALYSIS LITERATURE REVIEW Introduction: Carbohydrates may comprise as much as 10% of the weight of some membranes, usually in the form of glycolipid or glycoprotein.1 They are widely distributed in nature and the determination of their exact structure is often an important feature in the study of their functional or biological properties. Glycosylation is therefore a prevalent post- translational modification of proteins and these glycoproteins constitute a major segment of protein profiles in most eukaryotes.2 It is also known that the extent of glycosylation is not static, the carbohydrate components on both glycolipids and glycoproteins are known to change during cell development and differentiation. These alterations in the carbohydrate moieties of glycoproteins affect their solubility, stability, and most of all their functionality. The glycan compliment, of glycoconjugates, has been shown to play important biological roles these include: protection from proteases, modulation of activity, molecular recognition, immunogenicity, neuronal cell adhesion, sperm and egg fertilization, homing of lymphocytes, and bacterial and viral interactions.1 A precise description of the role played by carbohydrates in cellular processes is frequently compromised by their structural complexity and, as in many areas of biological science, understanding will require the pursuit of greater molecular detail. This -1- requiril charac oroeti are ad sacchz any 0 any ‘ Struc hydr: {ram linke lhyd, Rece the I ada; N311 Whil Ge} com] refiu requires more information than other biopolymers. Typically, in order to characterize a polymer it is required to know the monomer composition, the order of appearance and the size of the polymer. For oligosaccharides there are additional requirements such as: the configuration and ring form of the saccharide residues, the position and character of any substituent groups on any of the monomers, the positions of the inter-residue linkages including any branching points, and the configuration of the glycosidic linkages.3 Structural elucidation of the carbohydrate moieties typically involves specific hydrolysis of the oligosaccharide followed by detailed analysis of the resulting fragments. Methods that have been previously used for characterizing the N- linked oligosaccharides attached to glycoproteins employ chemical (hydrazinolysis) or enzymatic (PNGase-F) release from the polypeptide chain. Recently, the increasing use of instrumental approaches has been used for the determination of composition and sequence of oligosaccharides. Some are adaptations of generally applicable analytical methods (e.g., two-dimensional NMR spectroscopy, reverse phase HPLC, GC-MS, FAB-MS, and ESI-MS) while others have been developed specifically for oligosaccharides (e.g., Bio- Gel P4 column chromatography, HPAEC, lectin blotting). The most comprehensive structural information is currently provided by NMR, but this requires large pure quantities (> 500 nmol) of sample.4 Carbohydrate Chemistry: All sugars are polyalcohols. Modifications to these sugars include the presence of an amine group (glucosamine, galactosamine, etc.), a carboxylic acid group (e.g., sialic acid), and/or an acetal group (N -acetylglucosamine, N-acetylgalactosamine, etc). All reducing sugars also are in equilibrium either as a ketone or aldehyde (Figure 1 and Appendix A and B). The chemistry of each of these groups can be taken advantage of to improve detection and structural determination through derivatization. These reactions provide for analysis specificity and, through the formation of derivatives, detectability can be improved. RO Figure 1.1: Isomerization of linear and pyranose form of glucose. For derivatization purposes, the linear form, containing the aldehyde, is required. HPLC for GA GC an detecti and si deriva improi chromi applici in the derivat than .l could l 20-30 1 With 7i 211,7 an UTOnic the FEA ind“e only or Which Derivatization of carboxylic acids has been employed for GC, MS, and HPLC analysis. Although monocarboxylic acids may be sufficiently volatile for GC, multifunctional carboxylic acids require derivatization for successful GC analysis. Various derivatives have been prepared which yield specific detectability in GO. Derivatization is also of value in improving volatility and stability of carboxylic acid compounds for MS analysis. For HLPC, derivatives have been used to confer UV absorption or fluorescence to improve detectability. The determination of carboxylic acids by chromatography continues to be a very widely used technique and is applicable to some modified sugars. Several reports deal with improvements in the sensitivity of the detection primarily by the used of fluorescent derivatives.5 Recently, 1-pyrenyldiazomethane was found to be more stable than other diazomethane derivatizing agents and resulted in derivatives that could be separated by reversed-phase chromatography. Detection limits were 20-30 fmols. In addition carboxylated carbohydrates have been derivatized with 7 -aminonaphthalene-1,3-disulfonic acid as a fluorescent tag. El Rassi et al.7 analyzed glucosaminoglycans (GAGs) by first digesting the bipolymer of uronic acid and hexosamine residues into disaccharides and then derivatizing the resulting fractions for separation by capillary electrophoresis and laser induced fluorescence (LIF). In this instance, the resulting disaccharide has only one acid group. For the purpose of quantitating large polysaccharides, which may have varying numbers of carboxylic acid groups on labile residues, this approach would not give the necessary information to determine the degree of polymerization and therefore the concentration of each component. Since there may be a varying number of carboxylic acid groups, the concentration of the components in each chromatographic band would be uninterpretable since the number of derivatizing reagents present may vary from peak to peak; therefore, quantitation becomes impossible. Derivatizing alcohols to improve detectability is more difficult than derivatizing carboxylic acids. Alcohols do not undergo SN2 reactions very efficiently, since the -OH functionality is a poor leaving group. In addition, most alcohols are insufficiently acidic or basic to form stable ions by modification of solution pH. Therefore the most useful and general routes to produce derivatives of alcohols make use of reactions in which the oxygen acts as a nucleophile8. The derivatives of alcohols fall primarily into the three classes of alkyl, acyl, and silyl. Alcohols are insufficiently acidic (pKa 16-19) to undergo alkylation by the common alkylation reagents but may be alkylated by use of very strong bases and alkyl halides or by diazomethane with suitable catalysis. Acyl derivatives of alcohols are most commonly prepared with acid anhydrides, acid chlorides, or reactive amide reagents. The classical technique for preparing acyl derivatives involves acylation with acetic anhydride, which may be base (frequently pyridine) catalyzed or acid catalyzed. The perfluoroacyl derivatives are more volatile than the corresponding nonfluorinated derivatives and may be used to impart n3- raj 3T8 Ia: Hf Stll me electron-capturing ability.9 As with carboxylic acids, the use of fluorescent derivatizing agents enhances the sensitivity of the detection of alcohols in HPLC analyses. Anthracene—9-carbonyl chloride has been used as a derivatizing agent in the determination of short-chain alcohols.lo For the purpose of quantitation, in this instance, it would be necessary to have a first hand knowledge of the number of -OH groups present on the saccharide since varying numbers of alcohols can be present on oligosaccharides. For this reason, the best approach to quantitation and detection utilizes the chemistry of a functional group that is represented only once on a reducing sugar. Many of the derivatives that have been used for GC, GC-MS, MS, and HPLC analysis of aldehydes and ketones stem from the classical carbonyl derivatives such as hydrazones and oximes.11 Whereas classically these derivatives had to be isolated, purified, and analyzed, modern chromatographic and spectrometric techniques can be applied to the derivatives without prior isolation. In the past, G0 has been used for phenylhydrazones and dinitrophenyl hydrazones. The dinitrophenyl- hydrazones can be detected with ECD to enhance detection sensitivity by 2-3 orders of magnitude over that obtained with FID. The dinitrophenyl- hydrazones and the 5-dimethylamino-1-naphtha1enesolfonyl hydrazones have been found useful for chromogenic derivatization for HPLC analysis. Other substituted hydrazones are also used as analytical derivatives, and the basic theme still offers potential for development of new reagents. Carbohydrate Analysis: Derivatization Techniques: The determination of carbohydrates begins with the knowledge of the monosaccharide composition, and the linkages between the individual residues. There are numerous derivatizing reagents that have been utilized for this purpose. The first application of GC to carbohydrates was reported in 1958 and described efforts to separate the fully methylated methyl glycosides of some monosaccharides.12 The turning point toward widespread application of GC to carbohydrates was the publication in 1963 by Sweeley et al.13 of the use of trimethylsilyl derivatives for GO analysis of carbohydrates including mono- through tetrasaccharides. In addition, alkyl derivatives of carbohydrates have been utilized in conjunction with acid hydrolysis of glycoside bonds and further derivatization.” Utilizing GC/MS it is then possible to determine the linkage positions between the residues by interpreting the fragmentation patterns associated with intramolecular bond cleavages. The fragmentation patterns associated with many of the derivatizing agents for GC/EIMS have been reviewed.14 More recently, HPLC methods have been utilized to determine the monosaccharide composition. Kim15 utilized reductive amination to determine the monosaccharide content of glycoproteins. His derivatization used p-aminobenzoic ethyl ester (ABEE) as the chromophore and RP-HPLC for separation. Fu and O’Neill16 utilized a procedure developed by Honda and cowoq I pyraz metht. amint separi llldllCt oligos simila have I SeDar: differ definel associ glycos CarbO} micm] Show: \‘an'at coworkers17 to derivatize monosaccharides with 1-phenyl-3-methyl-5- pyrazolone (PMP) and RP-HPLC to separate the resulting material. Another method developed by Sato et al.18 used reductive amination with 2- aminobenzoic acid (2-AA) to derivatize monosaccharides that were then separated using capillary electrophoresis (CE) and detected using laser induced fluorescence (LIF). The monosaccharide compositions of oligosaccharides and glycoproteins were found to be highly accurate with similar or improved sensitivities to those assays previously developed. These high resolution separation techniques, coupled to detectors that have extremely high sensitivity are required. This is due to the diversity in the oligosaccharide structures, combine this with the multiple sites of glycosylation that may occur, results in an impressive array of glycoforms, separating these glycoforms requires high resolution separations. These different structures can originate from glycoprotein heterogeneity that can be defined as the varying carbohydrate core branching patterns and composition associated with a unique protein/peptide glycosylation position. Each core glycostructure can exhibit further structural modifications to a given carbohydrate moiety, thus greatly increasing the structural diversity or microheterogeneity. The heterogeneity and microheterogeneity have been shown to vary between cell types and stages of development. Due to these variations in the structures from cell to cell and time to time, it is not posut ghcai protei that prepa weigh these unflth and : Adda each that 2 oligos dfieq data. neces Deak5 “Sedi 335i? K'- possible for a monosaccharide analysis to give an accurate picture of the glycan.19 Detailed analysis of carbohydrate structures associated with unique protein/peptide glycosylation position(s) remains a major analytical challenge that requires sophisticated instrumentation and a variety of sample preparation strategies.20 For some glycans, a determination of molecular weight alone would be adequate to identify structure and ES profiling of these samples is direct and informative. However, it is common to find multiple structures (glycomers) that are consistent with an established mass, and alternative strategies are necessary to resolve these glycoforms.21 Additionally, these samples are often available only in limited quantities of each glycoform obtained from the sample. Therefore, sensitive techniques that are highly reproducible are required for the structure determination of oligosaccharides. The goals of these derivatizations are to improve detectability of the saccharides and ease the interpretation of the resulting data. Many approaches have been taken to improve the separations necessary to remove ambiguities inherent in overlapping chromatographic peaks. Polyacrylamide Gel Electrophoresis: One technique that is being used to determine saccharide structures is polyacrylamide gel electrophoresis of fluorophore-labeled saccharides22 (PAGEFS) also known as fluorophore- assisted carbohydrate electrophoresis (FACE). In this method of glycan 'tec llll - ec Que De lOTr analysis, the saccharides with hemiacetal functions are covalently labeled with a fluorophore and the derivatives are separated by polyacrylamide gel electrophoresis (PAGE). The saccharide-fluorophore derivatives can be subjected to PAGE in high-density gels yielding high-resolution separations. Detection of the bands on the gels was obtained in some instances by illuminating the slab gel on a UV transilluminator and photographing the resulting fluorescence to give high-quality negatives which could be scanned densitometrically. Over the range of 10 to 500 pmol, of saccharide per band, the response was nonlinear, but repeatable. The lower limit of detection using this technique was about 1 pmol.3 Alternatively, gels are imaged digitally using a camera system based on a cooled charge coupled device (CCD). The quantitative potential of the system was shown to be good in that a linear response was obtained over the range 12 to 500 pmol per band. Repeatability was similar to that obtained with the photographic system. The sensitivity of the CCD camera was about 0.2 pmol per band. Digestions are carried out using glycosidases that will give an indication as to the sequence, linkage type, and, ultimately, structure of the saccharide, and can be performed either before or after the saccharides have been labeled with a fluorophore. In the enzymatic structural analyses described above, the glycan structural features are deduced by interpretation of the changes that occur after specific glycosidase treatments. -10- npe igos *par ét-Er' Column Chromatography: Column chromatography has long been one of the preferred techniques for separating mono and oligosaccharides. High-pH anion-exchange chromatography (HPAEC) with a pulsed amperometric detector (PAD) is a powerful tool for analysis of oligosaccharides from glycoproteins. There are also many protocols for the separation and analysis of neutral and acidic oligosaccharides”:26 and monosaccharides.23,27'29 Among the more recent variations on these protocols, Hayase et al.30 used HPAEC-PAD for O-linked saccharides. They found that reduced desialylated O-linked oligosaccharides, especially those with short chains, are barely retained on a HPAE column because they lack the negative charge of either a sialic acid or an acidic anomeric hydroxyl group. In their work, the desialylated oligosaccharide-alditols were de—N -acety1ated and then N -succinylated before HPAEC separation. The succinylation introduces a negative charge on an amino group of each amino sugar residue and causes the oligosaccharides to be retained much longer on a HPAE column. HPAEC was also used by Clarke?’1 to separate peptidoglycan hydrolysates from several species of Proteeae. The single analysis afforded base line separation of glucosamine, the basic, and neutral, amino acids in an isocratic period and the peptide portion during a gradient segment. Ion Exchange Chromatography: HPAEC-PAD is very sensitive, capable of very high resolution, and can separate very similar anionic oligosaccharide structures. However, while excellent for analytical purposes -11- HPAEC-PAD is less suitable for preparative work since sodium hydroxide can cause epimerization and degradation of sugars.32 Also, the use of nonvolatile buffers requires an additional step for their removal. For buffers such as sodium acetate the nonvolatile cations can be removed using cation- exchange resin and then the volatile anions can be evaporated. However, this would require the use of large amounts of cation-exchange resin. Buffers with nonvolatile anions such as phosphate can be removed by gel filtration but this is time consuming.33v34 The use of anion exchange columns is not applicable to mass spectrometry without extensive desalting procedures. Also, gradients cannot be used with refractive index detection which limits the applicability. Several studies have shown that bonded-phase columns bearing amine functional groups can be used to separate oligosaccharides by HPLC. Ion- exchange HPLC35 fractionates oligosaccharides into groups with different charges, but these remain heterogeneous with respect to carbohydrate content and linkage. Additionally, amine adsorption HPLC36 cannot proceed unless the anionic fraction is removed. A different analytical procedure, ion- suppression amine-adsorption HPLC (ISAA-HPLC)37 can fractionate neutral and anionic oligosaccharides, but does not adequately separate large neutral oligosaccharides or highly sialylated small oligosaccharides. By modifying the conditions used for amine adsorption and ion-exchange HPLC, Kondo et 31.35 established methods for the separation of complex mixtures of neutral -12- and carbo- using derit' with are amin redui chroxJ apph' at pH Come a UV been polyr and sialylated PA-oligosaccharides based upon differences in net carbohydrate content (size) and the number of sialic acid moieties (charge), using an amine-bearing silica column. Reductive Amination: To study closely related oligosaccharides derived from glycoconjugates, it is essential to be able to fractionate them with good resolution. A pyridylamination method in which oligosaccharides are fluorescently labeled with a—aminopyridine (or-AP), by reductive amination, has been developed by Hase et a1."8 Since its introduction reductive amination has been an important approach to placing a chromophore onto reducing sugars, and the approach has seen many applications. Initially, the specimens were analyzed by paper electrophoresis at pH 5.0. Each sugar derivative giving a single spot in addition to one corresponding to excess oc-aminopyridine when the paper was scanned under a UV lamp. Since then, many primary amines with large chromophores have been utilized to obtain information on oligosaccharide degree of polymerization, and to quantitate the amount of saccharide present. Kondo et al.39 reported an improved pyridylamation procedure in which sialic acid residues are not released during the reaction. The combination of pyridylamination with HPLC allows sensitive analysis of oligosaccharides, with a detection limit at the femtomole level of pyridylamino (PA)- oligosaccharides. PA-oligosaccharides have been analyzed by HPLC on reversed-phase and normal phase columns, and the elution positions of many -13- PA-oligosaccharides have been plotted two-dimensionally by Hase et a1.“0 They separated pyridylaminated oligosaccharides by two-dimensional sugar mapping, a combination of size-fractionation and reversed-phase HPLC. This technique allows each sugar residue to make its own contribution to the elution time (partial elution time), and therefore the elution time of a PA- oligosaccharide is calculated as the sum of all contributing partial elution times. Utilizing this approach mono, di, and trisaccharides were obtained by partial acid hydrolysis and compared to authentic pyridylaminated standards. Under these conditions, mannose and glucose derivatives could not be separated in this system, although they were well separated by TSK gel sugar AXI column.41 Hase et a1. and Tomiya, et al.“2 have compiled an extensive 2-D map for neutral PA-oligosaccharides. This same group has also extended the use of this 2-D map to include GalNAc-containing N-linked oligosaccharides.43 To utilize this map, the structure of unknown oligosaccharides are estimated by comparing the elution position (expressed in glucose units) of an unknown sample with those of standard oligosaccharides analyzed by both reversed- and normal-phase chromatography. The resulting coordinates are plotted as a two-dimensional map. The structure of unknowns is then characterized by comparing the elution position with those of the standard oligosaccharides on the map. In this study, the 2-D positions of the GalNAc-containing oligosaccharides did not correspond to any of those of the known GalNAc-free oligosaccharides. -14- .ff"f"" ante] Chan to th degrt used inclu tn al octylz has ( afiac 2liCH deny anthl 31.46 Stmc' Capfl cofllh Cunt mix:- Cons; After B-N-acetylhexosaminidase digestion, however, their elution positions changed and coincided with known positions on the original map. Derivatization with a chromophore of high, known, molar absorptivity to the reducing saccharide is key to quantitating saccharides, based on the degree of polymerization. Reverse phase HPLC is one of the most widely used techniques for separating oligosaccharides. Samples of this technique include the assays developed by Hase, as already mentioned. Hollingsworth, et al.,44 also utilized reverse phase HPLC to separate N,N-(2,4-dinitrophenyl) octylamine derivatized oligosaccharides for FAB-MS analysis. Anumula“5 has developed an HPLC assay for sialic acids where a label is specifically attached, the fluorescent tag utilized in this instance is o-phenylenediamine 2HCl, and this yields a stable fluorescent quinoxaline derivative. Other derivatizing reagents investigated include 2-amino benzamide and 2- anthranilic acid, both were utilized to analyze N- and O-glycans by Bigge et a1.“6 Microbore HPLC: The characterization of the gross covalent structure of glycoproteins has been revolutionized with the advent of capillary and microbore high performance liquid chromatography (HPLC) combined with electrospray ionization mass spectrometry (ESI-MS). Using a combination of proteolytic enzymatic digestion and separation of the digest mixture by HPLC/ESI—MS, a glycoprotein mass map can be obtained consisting of both peptides and N- and O- linked glycopeptidesfl":50 This -15- netl ialy tear; methodology allows the nature and extent of site heterogeneity, including sialylation, of intact native glycopeptides. The location of those components bearing glycosylation in such digests is usually difficult because of inadequate chromatographic resolution. However, as recently reported by Carr et al.49, collisional activation of electrospray generated ions, fragmentation occurring either in the source or within the second quadrupole of a triple quadrupole mass analyzer, induces dissociation reactions preferentially within the carbohydrate moieties themselves. The fragment ions are specific to the oligosaccharide structure attached to the peptide backbone.“52 Hence, monitoring of these, lower mass, carbohydrate ions throughout the chromatogram permits detection of glycosylated species. Burlingame53 reported the use of this technique utilizing an ethanol: propanol: water: formic acid solvent system for the reversed-phase separation and electrospray mass spectrometric detection of tryptic peptides and glycopeptides from bovine fetuin. However, because of the complexity of carbohydrate heterogeneity, which is often present within and among sites in a glycoprotein, it is not possible in most cases to completely resolve the individual glycoforms using only one method. Depending on the degree of glycosylation, several glycoforms from the same amino acid sequence usually coelute, making interpretation of such mass spectra difficult. In addition, when several components (whether peptides and glycopeptides or glycopeptide glycoforms) coelute, possible problems with suppression and/or -16- usable dynamic range may arise, especially when attempting to detect and measure nonstoichiometric quantities of glycopeptides by LC/ESIMS. For some glycans, a determination of molecular weight alone is sufficient to identify structure and ES profiling of these samples is direct, rapid, and informative. However, it is common to find multiple structures (glycomers) that are consistent with an established mass, and alternative strategies are necessary to resolve these glycoforms. Peptide Analysis: Traditionally, short stretches of amino acid sequences have been used to search databases to locate protein or gene sequences, or to identify proteins of similar sequence. This constitutes a fairly unique address for a protein and can lead to successful searches. It is generally accepted that the molecular structure of a globular protein is ultimately determined by its amino acid sequence, within the constraints imposed by its chemical environment, and that the molecular structure determines its biological function.54 The determination of the specificity and activity of a protein depends on obtaining the correct sequence of amino acids (i.e., primary structure) that comprise the polypeptide chain. This includes the knowledge of the presence of specific post-translational modifications needed for the proper folding of the oligopeptide chain to a precise three- dimensional structure. The amino acid sequence information is usually obtained by N-terminal analysis using Edman degradation. Conventional -17- sequ amirl anal.i the deter of d modi assii mea: Isola met}- my; and mal 1 Sir; CID Dec: Epf sequencing strategies typically employ chemical reagents to remove one amino acid at a time from the amino terminus followed by isolation and analysis of the released amino acid derivative.55t56 Mass spectrometry offers the advantage of speed and sensitivity in protein primary structure determination and also has a unique and important role in the determination of structural changes introduced into proteins by post-translational modifications. Both chemical degradation and peptide mapping approaches require the use of fairly homogeneous samples to avoid ambiguity in assigning the amino acid sequence and establishing the origin of the set of measured peptide ions, respectively. These approaches are not well suited to the study of biological problems where complex mixtures of peptides may be isolated. Innovations such as electrospray ionization have led to improved methods for interfacing liquid separation techniques to mass spectrometers. Combining microcolumn liquid chromatography (micro-bore HPLC) with tandem mass spectrometry improves the ability to manipulate, for sequence analysis, small quantities of peptides contained in complex mixtures. Ions of a single mass-to-charge ratio can be selected for collision-induced dissociation (CID) even though that peptide may be introduced into the mass spectrometer as part of a mixture of peptides.57 Under CID conditions, peptides fragment to create patterns characteristic of a specific amino acid sequence. These patterns are reproducible and, in general, predicable. But -18- T rid ". VEH equ hiie tom text n/z orec )rod 3815 men Dion HQa Lhe lTQ; URn 00f even still, interpretation of a low energy CID spectrum for an unknown sequence proceeds by identifying a consecutive series of fragment ions whose differences correspond to residue masses for amino acids. This ion series can correspond to either the type —b or -y ion series or a combination of both (see next section for a description of the nomenclature). Interpretation of limited m/z range, low-resolution tandem mass spectra from multiply charged precursors can present difficulties due to the possibility of multiply charged product ions and the lack of unambiguous charge-state information. Even so, ESI/MS with instruments usable over a relatively low m/z range (<2000) have been shown to be attractive for molecular weight measurements of large biomolecules.53,59 Electrospray interfaces serve as the basis for a near ideal interface for liquid chromatography and a range of separation formats. The potential of these methods is further enhanced by tandem mass spectrometry (MS/MS), which, ideally, permits efficient dissociation of the precursor molecular ion into product ions that are structurally distinctive to be useful for fingerprinting. Electrospray techniques have found application in accurate measurements of molecular weights of peptides and proteins by exploiting the phenomenon of multiple protonation, which yields a coherent series of progressively more highly protonated species (M + nH)“. Smith and coworkers have published an excellent review.60 For larger molecules, the more highly charged ions are more susceptible to dissociation.61 Because -19- ‘1. :oHi reas< hu'a diar. relat protc COLA 5aCci Slmp 30nd collision energy is proportional to the number of charges at a given m/z, it is reasonable to assume that the greater number of charges is a primary reason for an increased CAD efficiency. In addition, efficiency for MS/MS of multiply charged ions may be enhanced due to electrostatic repulsive forces.62 The relatively facile fragmentation of extremely large proteins in their multiply- protonated forms doubtless reflects, in part, the destabilizing influence of the Coulombic repulsions within the multiply charged species.63 2 zn-Z n-l Yn-l Yn-Z Xn-1 xn_2 xii-3 Rt 2 Rs HzN-CH CO NH H CO NH CH CO—--- 31 32 a3 b1 b2 C2 Cn-2 cl Figure 1.2: Peptide fragmentation nomenclature, typically for low energy CID only b and y type fragments are observed. Peptide Nomenclature: Compared to the complexity of the saccharide fragmentation nomenclature, the system used for peptides is simple, particularly for low energy CID. The full series of fragments is indicated in Figure 1.4. Not only are fragments that account for backbone bonds breaking included, but also for side chains, just as in saccharide -20- mm the : addi the s grou bone for l the a fragmentation. Fortunately, for low energy CID, the only fragments that are typically present are the -b and -y series. These are the ions that account for the sequence of the peptide and form from breaking the peptide bonds. Any additional fragmentation that may occur typically only relays information on the side chains. The fragmentations that arise from the side chains, the “R” groups, typically do not occur, nor do the fragmentations that arise from bonds breaking other than the amide linkage. These fragments are reserved for high energy CID, which may occur on magnetic sector instruments due the additional energy imparted during acceleration. -21- fOr 8 here, app}; Clean Instr 30m]. CH2 OH Figure 1.3: Fragmentation nomenclature for linear saccharides as developed by Domon and Costello.“ Carbohydrate Nomenclature: A modification of the nomenclature for saccharides that was developed by Domon and Costello64 will be used here. Their system, developed for FAB ionization and high energy CID is applicable here since the fragmentations are sufficiently similar. Some ring cleavages observed at the high potentials utilized with magnetic sector instruments (high energy CID) are not observed, but are included for completeness. -22- comm nucl strut dete liert The types of fragmentations that occur in glycoconjugates are more complicated than that of other large biomolecules, such as peptides and nucleic acids. Whereas with these other biomolecules the polymer of interest is linear, saccharides have numerous possible branch, or connection, points and all of these must be taken into account when revealing and reporting the structure. Therefore, it is necessary to define all connectivities, however determined. This system expands upon the nomenclature used for peptides. Here, Ai, Bi, and Ci labels are used to designate fragments containing a terminal (non—reducing end) sugar unit, whereas Xj, Yj, and Z,- represent ions still containing the aglycone (or the reducing sugar unit). Subscripts indicate the position relative to the termini analogous to the system used in peptides, and "superscripts indicate cleavages within carbohydrate rings (Figure 1.2). -23- tha: bra: bee. lntr 0rd. t‘fi‘r‘. CHQOH HO 0 Ybs OH CHQOH o 0 Yea HO OH CHQOH O O 851 YaS HO OH H CHQOH 832 O 0 Y“ B 0 Y2 /Y03 63 A OH CHQOH H O CH? Y1 O 0 O HO OH CHZOH OH O CHQOH Bai / HO OH Ho OH 0 CHon yo 0 Baz BaS/ '34 HO OH 35 36 Figure 1.4: Nomenclature for branched saccharides as developed by Domon and Costello.“ As previously stated, oligosaccharides are unique for biomolecules in that their three dimensional structure includes branching. To describe this branching in the fragmentation nomenclature the convention introduced has been extended. The “core” of the oligosaccharide is labeled as previously introduced, the branches are labeled with the Greek letters or, B, y, ..., in order of decreasing molecular weight (Figure 1.3). In the instances where two fragments are obtained from the same sugar, the Greek letters used to -24- rep] indi bra: represent each branch are given together (e.g., Yag). Primes are used to indicate branch points on saccharides, other than the core region, the number of primes increases in the order of decreasing molecular weight of each sub- branch. -25- I incl teas four the biol det; Cari fEa: are l l 1 mol and Chapter 2 STRUCTURE DETERMINATION OF REDUCING SACCHARIDES Introduction: The rapid expansion of knowledge in the field of molecular biology during the third quarter of the twentieth century did not include carbohydrates. This exclusion is due to a preoccupation with proteins and nucleic acids. But, as contemporary researchers try to bring chemical reasoning to physiological function, carbohydrates and their conjugates are found as crucial participants in cellular processes.65 A precise description of the role played by carbohydrates in cellular processes is frequently compromised by their structural complexity, and as in many areas of biological science, understanding will require the pursuit of greater molecular detail. The pursuit of this detail has its cost though; the analysis of carbohydrates is impeded by the lack of outstanding structurally detectable features. There are no groups that fluoresce or absorb well, nor groups that are easily ionized. From a structural standpoint, oligosaccharides are ideally suited for events requiring molecular specificity. Their polyhedric character provides a platform with numerous functional groups for modification and interaction, while monomers within an oligomer can participate in multiple linkage and branching patterns, creating an array of structural possibilities from a -26- lin an: 1110 det pur pro rele prer to l bioe limited set of monosaccharides.66 Traditional methods of carbohydrate analysis rely on chromatographic and electrophoretic techniques.67'69 Since most sugars do not contain chromophoric moieties, the methods of direct detection of sugars are thus limited. This problem is compounded since the purification of N-linked, or O-linked, oligosaccharides from glycoproteins is problematic, a single glycoprotein often contains numerous types of closely related oligosaccharide structures.70 Furthermore, it may be necessary to prepare milligram quantities of individual oligosaccharide glycomers in order to perform complete structural characterization, by NMR, and to assess bioactivity. Although sugars do absorb at 185 nm,71 the detection sensitivity is poor due to their low molar absorptivities.72 Underivatized oligosaccharides have been purified using high-pH anion exchange chromatography (HPAEC).73 The major advantage of this approach is that the isolated oligosaccharides retain their reducing end and thereby can be derivatized for further analysis. In this light it has been suggested that no single technique can be used to determine a complete saccharide structure. That is, the problems inherent in determining the degree of polymerization and the linkage positions cannot be determined by only one technique. Typically multiple stages of separation, and/or detection are involved in the analysis of oligosaccharides. -27.. me bee imp acct lim Bor reli rep but Der flUI inc: Wit 1110' (lg. HP HPLC has played an important role in carbohydrate analysis. Several methods for quantification of monosaccharides without derivatization have been described.”77 Refractive index detectors are used for detecting underivatized sugars. However they are insensitive, nonspecific, and impossible to employ under gradient conditions. HPAEC-PAD has been accomplished with high sensitivity and selectivity,78-82 but PAD has the limitation of possible electrochemical modification of the detected species. Borate anion-exchange chromatography with suitable postcolumn detection is reliable and easily implemented, but requires long analysis times and is relatively insensitive due to peak broadening.83 Utilizing a combination of techniques thus appears required for structure determination of saccharides. To this end a combination of HPLC with absorbance detection and ESI-MS has been employed in the work reported here to obtain not only information on the degree of polymerization, but also information on the linkage positions for numerous saccharides. Derivatization melds carbohydrate structural analysis with UV/V is, or fluorescence, and is required for three reasons: First, detection sensitivity is increased, second, it enables quantitation, based on Beer’s law, and third, with charge localization on the derivatizing reagent, structural features are more evident. Therefore, our procedure calls for the saccharides to be derivatized first, for ease in quantitating the individual species present. The HPLC separation is used to separate the individual saccharides to obtain -23- IL- ' infc det: qua sep. pa frag incl inf mat util Sin 0011 Wei this this DEE. The information on the degree of polymerization. Then, once separated, a UV/Vis detector set at the absorption maximum of the derivatizing reagent allows quantitation based on Beer’s Law. The fractions obtained from the separation are then introduced into an electrospray mass spectrometer and, partially, fragmented. Typically, tandem analyzers are used for these fragmentations so that ions may be separated in complex mixtures and individually studied by CID. This dramatically increases the structural information available and introduces another level of separation, based on mass and charge. But, low energy CID is also available in electrospray by utilizing in-source fragmentation and a single quadruple mass spectrometer. Since the bonds that are fragmented by low-energy collisions, typically connect the monosaccharide residues, the resulting fragments provide monomer sequence and branching information. This allows the linkage position of the species to be determined and also confirms the molecular weight and, therefore, the degree of polymerization. The data obtained with this procedure are partially redundant, in order to decrease the chances of misinterpreting the degree of polymerization. Materials: Pyrenemethylamine was purchased from Sigma, benzylamine, analine, and 2-aminopyridined were purchased from Aldrich. The maltodextrin (n S 15) was obtained from Synthon. Most sugars (arabinose, fructose, fucose, gentiobiose, glucose, lyxose, mannose, melibiose, -29- rh: rhz W8 to eqt mic Th: ha. L! .IIIIII ll“— . I: . .vEF .. N-e Wat the det oft to an 5 ill rhamnose, ribose turanose, and xylose) came from Sigma. Sophorose and rhamnose were also from Sigma. Galactose came from Aldrich, while lactose was obtained from Kodak. Procedures: The reducing saccharide(s) to be derivatized are added to a one-dram vial, with two equivalents of derivatizing reagent and two equivalents of sodium cyanoborohydride. To this mixture, two hundred microlitres of water: methanol solution (1:1) with 5% acetic acid is added. The vial is capped with a screw cap, having a Teflon liner, and heated at 65 °C for one hour. Subsequently the solution is dried by evaporation using an N-evap. The sample is then ready for further analysis. Most saccharides were purified by passing them through a column (6 x 1.5 inches) of silica with a mobile phase of isopropanol: ammonium hydroxide: water (6:4:1). These fractions were collected for further analysis. Aliquots of the samples were then spotted onto TLC plates, and the chromatograms were developed in a solvent of isopropanol: methanol: ammonium hydroxide: water (1:1:121). The resulting zones were visualized by observing the fluorescence of the derivatizing reagent under a UV lamp. Additionally, orcinol was used to visualize the saccharides. The fractions that were found to be fluorescent and to contain saccharides were then analyzed by mass spectrometry. Reversed-phase HPLC analysis was accomplished using a Waters 600 (Milford, MA) multisolvent delivery system equipped with a Kratos (Ramsey, -30- \ ah se; ma of lint Alt mo inc the Tet qu: Opg NJ) Spectroflow 783 programmable absorbance detector set to monitor the absorbance maximum for each derivatizing reagent (240 nm for PMA). The separation of derivatized saccharides is achieved in either eighty or forty minutes on a Beckman (San Ramon, CA) Ultrasphere ODS column (4.6 mm x 25 cm with a mean particle diameter of 5 pm). The first separation has the elution performed at a flow rate of 0.4 ml min'1 at 30° C using a gradient. For maximum resolution of all saccharides, a mobile phase that initially consisted of 10% methanol: water, with 0.1% TFA was utilized, this was increased linearly to 100% methanol over seventy minutes after a ten minute hold. Alternatively, for increased resolution of mono through tetrasaccharides, the mobile phase was instead modified with 0.1 % HCl initially not only did this increase resolution for the lower degrees of polymerization, but also moved the bands further away from the derivatizing reagent and decreased the retention times. Mass spectrometry was performed on a Platform (Fisons, MA) single quadrupole mass spectrometer equipped with an electrospray interface and operated in the positive ion mode. Solutions of analytes were obtained by dissolving approximately 1 ug of sample in 1 ml of methanol: water (1:1), this was diluted further as needed. Samples were introduced into the ESI source with a Sage 361 syringe infusion pump at a flow rate of 20 pl min:1. The filter electrode was heated to 100 °C, the nebulization gas flow rate was 18 l hrl, -31- and optin and the drying gas flow rate was 350 l hrl, all other parameters were optimized for maximum sensitivity while maintaining unit resolution. -32- Figure 2.1: Scheme for reductive amination of saccharides with a derivatizing reagent. This example utilizes (x-aminopyridine. For most experiments pyrenemethylamine was used. -33- CHZC {Q OR CH20H CHQOH CHZOH CHZOH o o HO—————H OH HO H OH 5% CH3C02H O H ’ O H on OR OH OH H-—-——-OH H—————oH HO H H0-———H CH A CH N NH / CJ .5 'HgO CH CH CH CH 2 CH20H 2 CHZOH o 0 OH H0———H 0H HO———H 0 OR H H NaBH CN H OH H 3 H—————OH ‘ H OH HO————-—-H H0—-—-H CH2 CH NH N ill IV III -34- 0. 00 Figure 2 With PM. Increasi o.' 00 30.00 60:00 90700 Figure 2.2: Maltodextrin separated on an ODS column after derivatization with PMA. The degree of polymerization of the maltodextrin decreases with increasing retention time. -35- Ri performe principle aldose 0 amino water fro . is revers reductior cyanobor Th HPLC usl obtained from one detected ChrOmatO Under-ta}. PresenteC experime disacchar In rEtenti fTOm tab monosacc Results and Discussion: Reductive amination (Figure 2.1) was performed using a modification of the procedure outlined by Hase et al.35 The principle of the method is as follows: Saccharides (I) which are either an aldose or a ketose are reacted with a fluorophore which contains a primary amino group to yield a carbinolamine, or Schiff base (II). With the loss of water from the carbinolamine, an imonium ion is formed (III). This reaction is reversible and to obtain the final product, the imine (III) is trapped by reduction to a secondary amino derivative (IV) using sodium cyanoborohydride, a mild reducing agent. The saccharide-fluorophore derivatives (SFDs) were separated by HPLC using a reversed phase, ODS column, where baseline separation is obtained for Oil-4 linked maltodextrins having a degree of polymerization from one to fifteen (Figure 2.2). The fluorophore enables the derivatives to be detected with femtomole sensitivity using absorbance detection.84 The chromatographic analysis of saccharides other than Glua1—>4Glu was also undertaken with this separation. The results for various saccharides are presented in graph 2.1 and table 2.1. The first obvious result from these experiments is that the retention times overlap for the mono and disaccharides while, individually, the saccharides maintain a small deviation in retention time (graph 2.1, error bars and the relative standard deviations from table 2.1). For the purpose of this study, it was found that the monosaccharide retention times overlap with the disaccharides, and their -35- configure. deoxy-mt typically to be exp the colun the HPL piece of i plays on] Hence, tl I POlmeri: seen wh. POb'merii Fol attempt Separatio these we This may content. have bee, Utilizin g othSacC} apply to configuration is much more critical than for the disaccharides. When a deoxy-monosaccharide is chromatographed, the retention time is longer, typically a barely resolved shoulder on the derivatizing reagent band. This is to be expected due to the increased interaction with the stationary phase of the column. Therefore retention time presents a problem for utilizing only the HPLC separation to determine the degree of polymerization. Another piece of information that is shown in this chart is that the linkage position plays only a small role in determining the retention time for disaccharides. Hence, the retention time is useful for approximating the degree of polymerization determination, but not the linkage position. It remains to be seen whether this finding applies to saccharides with larger degrees of polymerization when the PMA derivatizing reagent is used. For lower degrees of polymerization, a separate assay was developed to attempt to improve resolution. By modifying the mobile phase, the separation between the mono through tetra saccharides was increased, and these were further separated from excess derivatizing reagent (Figure 2.3). This may allow for more of an unambiguous determination of monosaccharide content. But, other obstacles to strictly chromatographic determinations have been encountered. Hase et a1.“1 and Tomiya et al.42 suggest a method of utilizing glucose standards to set a “glucose unit” designation for oligosaccharides. Although, according to Tomiya, this does not successfully apply to saccharides containing sialic acid residues, their response to this -37- problem labile se branchec Since it ‘ the infor‘ detector i Figure 1 (DP. 5 1 problem is to remove the sialic acid residues before analysis since these are labile saccharide units. This approach allowed them to categorize the branched saccharides correctly based upon Hase’s glucose unit designation. Since it is obvious that a single method of detection could not provide all of the information required, the approach we have taken is to add a second detector that is mass sensitive. ”/ Iir‘ ' ’ ' Li Figure 2.3: Alternative separation for maltodextrins derivatized with PMA (DR S 15) on an ODS column using a methanol gradient. -33- 90- 'r o Arabinose 80 I * h a: a: 3: I Deoxygalactose ‘ Deoxymannose 70 j; x Fructose 1 I . _ x Fucose 50 4- . Galactose i + Glucose 50 i - Lyxose -- Mannose 4O » Rhamnose Ribose 30 ‘ Xylose >< Gentiobiose 20 x Lactose a Maltose 10 e Melibiose - Maltohexaose O r I . I f t t t I r I 0 2 4 6 8 1 1 1 1 1 0 2 4 6 8 Graph 2.1: Relationship between the average retention time over three consecutive injections on an ODS column verses the degree of polymerization of the saccharide. Error bars are included to indicate the range of retention flaws -39- Sugar D.P. Mean S.D.mean R.S.D. Arabinose 1 64.8 0.4 0.01 Deoxygalactose 1 78.6 0.2 0.00 Deoxymannose 1 81.4 1.3 0.02 Fructose 1 81.4 0.1 0.00 Fucose 1 77.2 0.3 0.00 Galactose 1 54.2 2.0 0.04 Gentiobiose 2 77.4 0.4 0.01 Glucose 1 62.6 0.2 0.00 Lactose 2 67.5 1.8 0.03 Lyxose 1 72.0 0.1 0.00 Maltohexaose 4 62.6 1.2 0.02 Maltose 2 78.4 0.2 0.00 Mannose 1 50.1 1.6 0.03 Rhamnose 1 76.3 0.3 0.00 Ribose 1 55.4 0.3 0.00 Xylose 1 60.7 0.9 0.02 Melibiose 2 77 .5 0.3 0.00 Gentiobiose 2 80.5 0.6 0.0 1 Table 2.1: Analysis of the retention times of saccharides of varying degrees of polymerization. -40- 32.5 “’1’ 12.06 ‘ 132 %.I 1368 720 . 1044 1530 H l 0 J' i ——"u- —- in i4 1- ,_v ..J ‘_{ _'__n‘_5___ .-_i__“____ _._ J J m/z 200 400 600 800 1000 1200 1400 Figure 2.4: Electrospray mass spectrum of pyrenemethylamine derivatized maltodextrin (DP. 5 15). By utilizing mass spectrometry to obtain mass information fiom the bands obtained from the HPLC separation the degree of polymerization can be determined with pmol sensitivity up to a degree of polymerization of ten. Figure 2.4 is two hundred pmol of the derivatized maltodextrin injected, shown are DPN .<_ 8. Sensitivity for this stage of structure determination is in -41- the tens of picomole range (e.g., figure 2.5 for the derivatized disaccharide maltose). By obtaining confirming results by mass spectrometry, not only can the degree of polymerization be confirmed, but also any modifications to the saccharide(s) can be detected due to changes in mass. Therefore, even if, for example, a deoxy-sugar or acetylated sugar were present, the mass of the resulting ion would indicate its presence. But, this will not give us all of the information that we seek. Information on branching, linkage positions, or where modifications to the sugars occur is not present, only indications that they exist. It is therefore necessary to fragment the saccharides to obtain ions that indicate this information. 558 559 / In ll." llllllllllllillll lunltllndlllll j J III'II |fiTlTl 540 560 580 Figure 2.5: Two pmol ill" of disaccharide directly injected into the mass spectrometer. Total volume infused, 10 pl. -42- Utilizing a “soft” ionization method typically yields only molecular weight information. In order to obtain structural information, fragmentation of the ions is required and collisionally induced dissociation (CID) is used. This technique, for low energy collisions, typically is performed in the center quadrupole of a triple quadrupole mass spectrometer where a collision gas has been introduced. However, for the purposes of these experiments, a single quadrupole mass spectrometer was used. In-source fragmentation, a.k.a. cone fragmentation, relies on imparting additional energy to the ionized molecules to fragment them prior to any mass analysis. There are two ways in which this can be done. Conventional in-source fragmentation relies on manipulating the potential between the extraction cone and skimmer cone (Figure 2.6). Several groups have observed this.85'37 Additionally, in-source fragmentation can be obtained, with similar results, by increasing the flow rate of the nebulization gas. ...,5_,.,-5-.h;~_.ys,oq; , I. .V_ t}"".‘x_y‘~ 3)? flirt“: -. 53W '0‘“- '5"«. r ' ,,"'"" ' '5“; 1"”);92'45' f’o,‘ ”caveat-wee" .. , 5 ‘25-: saw:- treati- Lara. Figure 2.6: Schematic of the intermediate pressure region in an electrospray mass spectrometer . -43- Figure 2.7: Configuration of Fisons Platform electrospray mass spectrometer. -44- 5330 980 nos—5.3. m HOHmHHQHH- emafiflmgflfiudw mflmv msmNmfin-Awoz -45- By adjusting the potential between the extraction cone and the skimmer cone in the region of intermediate vacuum (Figures 2.6 and 2.7), 88 where the gas expands through the sampling cone, or capillary, into the lower pressure region, forming a supersonic jet, fragmentation can occur. At the pressures typically encountered in this region, a shock wave is formed due to collisions with background gas, and there is considerable scattering of the expanding beam. Here, the potentials applied between the skimmer and the cone/capillary serve both to focus the ion beam through the skimmer orifice and to accelerate ions through this intermediate pressure region. Raising the cone potential has been shown to improve signal-to-‘noise ratios for some species in ESI which has been attributed to improved ion focusing.89 Acceleration of ions through the intermediate-pressure region has also been found to cause collision-induced dissociation by increasing the energy of collisions with background gas, by imparting excess energy to the ions to induce fragmentations. This phenomenon has been attributed to the effects of collisional activation of the ions as they are accelerated through the intermediate pressure region of the ion source . The second method of inducing fragmentation involves increasing the flow rate of the nebulization gas. In the ion-spray configuration, an additional flow of gas concurrent with the mobile phase flow is present (Figure 2.7). For the configuration of the Fisons Platform mass spectrometer, -45- the flow rate for obtaining pseudomolecular ions is 18 l hrl. Increasing this flow increases the rate of evaporation of the droplets and also increases the fragmentation present in the resulting mass spectrum. The data obtained are similar to those for conventional in source fragmentation (Figure 2.8 {maltose}). Since the “curtain” gas flow rate is more than ten times higher than the flow rate of the nebulization gas, the fragmentations are likely due to the increased evaporation of the droplets. This allows for more collisions prior to the ions passing through the extraction cone into the region of intermediate vacuum. The fragmentations described here are sufficient to obtain linkage position as well as to indicate where modifications are present in the saccharide. From observing the differences in fragmentation as the potential is increased, differences in linkage position become evident, although there are similarities among all of the data. In particular, as the potential is increased, the intensity of adduct ions (Na+ and K+) also increases (Figures 2.8 - 2.16). These ions offer no increased structural information, they only decrease the intensity of the ions that do contain structural information. A potential needs to be chosen that is a trade off between the intensity of adduct ion(s) and the intensity of the fragmentation ions. This can provide a generalized procedure for sample analysis, even though increased sensitivity for the fragment ions could be obtained by optimizing the potential for each saccharide present. Optimization assumes prior knowledge of the sample -47- composition, which in instances of unknowns is not present. The potential that was chosen was 55 V. Once the amount of energy required to induce fragmentations was investigated, another series of experiments was carried out that focused on disaccharides with varying linkage positions and configurations, and monosaccharides with varying configurations. This was pursued in order to determine differences in the fragmentation that would indicate the presence, or absence, of specific linkage configurations (Figures 2.9 — 2.16). In all of the fragmentations in this data set, the linkage position of the two saccharide systems was observed by a unique fragment ion. In the case of gentiobiose (GluBl-96Glu), the linkage position was indicated by a cleavage on the derivatized saccharide between carbons five and six, with the charge being retained on the derivative end of the molecule. In addition, a fragment ion is observed corresponding to 3’5X1 (two bonds being broken within one ring) with the charge again being retained on the derivative side of the molecule. The fragment ions obtained for lactose (GalBl—>4Glu) (Figure 2.11 and 2.12) and maltose (GluBl—>4Glu) are nearly identical, just as is the situation for EIMS.9° Therefore, strictly utilizing any one form of mass spectrometry for the purpose of monosaccharide determination is likely impossible. The linkage position is well indicated with a characteristic cleavage, an ion indicating that the bond between carbons three and four was broken. In this instance the species that imparts the charge is sodium (Nat), and the charge -43- is retained on the saccharide portion of the SFD (m/z 278). As with gentiobiose (GluBl-96Glu), a ring fragmentation is also present yielding a fragment ion, again with Na+ imparting the charge. Only, in this instance, the derivative part of the molecule retains the charge, and yields a 1:5X1 fragment. The similarities expected between the spectra of gentiobiose and melibiose (Galoc1—96Glu) are obvious upon inspection of their structures (Figures 9 and 10 and 13 and 14). Both disaccharides have a one to six linkage: gentiobiose has a beta linkage with glucose as both monomers, while melibiose has an alpha linkage with galactose linked to glucose. Both structures fragment between carbons five and six, leaving the charge on the derivatized saccharide, and both have a ring cleavage at 3’5X1. The expanded region in figure 14 C gives an indication of the only difference. An additional bond is broken in melibiose, Zr. This bond is more labile, possibly due to a decreased pi overlap in the glycosidic bond, from the differences in linkage type (i.e., alpha vs. beta linkage). The final disaccharide observed was sophorose. This disaccharide consists of two glucose monomers with a beta linkage at the two position (i.e., GluBl—)2G1u). In this instance, the bonds are very labile and, with increased potential applied to the extraction cone, the amount of secondary fragmentation increases. These additional fragmentations only lead to more difficult interpretation, and do not readily give more structural information. From the spectrum with the potential on the extraction cone at forty-five -49- volts, the relevant structural information is obtained. Again the fragmentation of interest occurs near the glycosidic linkage, and gives an ion characteristic of this position, m/z 246. Also a fragment can be seen in this spectrum involving the saccharide ring, which is a 1’5X1 fragment. -50- Figure 2.8: Fragmentation of PMA derivatized maltose. The first set of data utilizes increased nebulization gas flow rate to fragment (Figure A). The second set of data utilize conventional in source fragmentation, increased potential on the extraction cone, to fragment derivatized maltose (Figure B). -51- 50 voflag° InienslIY Relative -52' ml! value Figure 2.9: Spectra from experiments in which gentiobiose, derivatized with pyrenemethylamine, is subjected to increasing voltage applied to the extraction cone. Spectrum A is table with the normal operating voltage of 37 V. In spectrum B the voltage applied to the cone is increased to 45 V, while in spectrum C the voltage applied is further increased to 50 V. -53- 367 Rom H \/ H OH H0——-—H H0 HO OH H——OH 0 OH I E——0H H——o 485 H u L l i . .V L l -54- Figure 2.10: Spectra from experiments in which gentiobiose derivatized with pyrenemethylamine is subjected to increasing voltages applied to the extraction cone. Spectrum A has the operating voltage at 55 V. Spectrum B increases the voltage applied to the cone to 60 V, and spectrum C has the voltage applied increased to 65 V. -55- ' 558.2 Rdm H :T H e—OH Ho—-—H HO HO OH H 03 OH %. 5mm 337 H————OH H flflfl 0 . .lV - Vie-t", L .4. A . emaz or 5i30.1 . :515 0. .l l I I..L B ~ amz mmz °/o. . :mrt o . I1 r'rllr -, va . .-- 1 I . . . r . .. ‘nl “hi-L C -55- Figure 2.11: Spectra depicting experimental results where lactose, derivatized with pyrenemethylamine, is subject to increasing potentials applied to the extraction cone. The results shown here are for the normal operating voltage of 37 V, 45 V, and 50 V applied to the cone for spectra A, B, and C, respectively. -57- 558.2 580.0 I v .I. 461— 343 11an H 215 V H——OH OH ,5 Ho__H 30% I Hfl—OWCF OH 255 + Na H——OH H——OH 95 H I Gill. .1III‘LJJ. . . -. .1‘. J. .l I! i .' A - 558.4 %. o .tvdr'r' . . 't‘lliv .‘l'.L ‘L 171' J- .“HLv .‘ l5“ :1 B . 558.2 °/o. -53- Figure 2.12: Results from experiments in which increasing voltages are applied to the cone. The saccharide derivatized with PMA for these spectra is lactose. Spectrum A has the operating voltage set to 55 V. Spectrum B increases the voltage applied to the cone to 60 V, and spectrum C has the voltage applied increased to 65 V. w C) -59- 0. °/o . %- °/o . 461— 343 558.1 H——- OH OH HO——H 50M I—‘HT—o 0.. OH 255 + Na H——ON 215 H——OH 95 580.2 5 I hA'L n r' in Jul l'l L l' 515' 558.1 5 9.7 ‘51|‘.'11'Lt.'lv 1115' ‘ L. Y xtl't '55'1‘ It. IL: la 558.2 , 580.2 ‘ "rv'lnllrnq ll 1 1 1 I ll A Al dipll Lia‘l‘. -60- Figure 2.13: Spectrum A, B, and C show the experimental results where melibiose, derivatized with pyrenemethylamine, is subjected to increasing voltages applied to the extraction cone. The voltages applied are 37 , 45, and 50 V, respectively. -51- .OJ-L—w.‘ O/o. 558.1 “ R-NII H \/ H———OH i HO—I—H 367 H——OH I H——OH %T a——o (m H O ‘ 395+Na OH 485 OH HO 5f0.3 O 0 I l I y~ v 1“} A * 558.2 I 0 /°“ 530.2 485.5 1 l. :1. B 5 558.2 °/.. 580.2 0 -- I .111 -52- Figure 2.14: Spectra from experiments in which melibiose, derivatized with pyrenemethylamine, is subjected to increasing potentials applied to the extraction cone. Spectrum A has the operating potential set to 55 V. Spectrum B increases the voltage applied to the cone to 60 V, and spectrum C has the potential applied increased to 65 V. -53- AO 558.2 RmH H \/ H OH Ho H 367 H 0H [3 H OH % - H'fi—O OH 580 2 H o 395 + Na 0H ms 0H , Ho 0, . 36.6.4 . ll , p,“ A - 558.3 0 /° “ 580.3 0 . 490:0. ' . II. B " 558.2 °/o .. 580.1 _ 367.1 0 h t' 1‘ . Illlfl|l 1' I'll pllr .‘l.|. l l Yllt -54- Figure 2.15: Spectra from experiments in which sophorose, derivatized with pyrenemethylamine, is subjected to increasing voltage applied to the extraction cone. Spectrum A is table with the normal operating voltage of 37 V. In spectrum B the voltage applied to the cone is increased to 45 V, while in spectrum C the voltage applied is further increased to 50 V. -55- 246 5- l - . it - . .-. Lt. 15L: 1 ‘. vv. , . . . . . . .— 1 . 4.111.! JAI.‘ fllvlLll £41111: Ll J +41 n.1hrn. L} L -‘l J lav-L1. .ALJAW' vh‘vlzv l!1 I 1 l l' r v r l v V‘ V 1' V l " ‘ ' l C . In, It I _ IV III...1luruudnid.llIlliittiltthIltll -66- 1111151111111de Figure 2.16: Spectra depicting experimental results where sophorose, derivatized with pyrenemethylamine, is subjected to increasing voltages applied to the extraction cone. Spectrum A has the operating potential set to 55 V. Spectrum B increases the voltage applied to the cone to i l: 60 V, and spectrum C has the potential applied increased to 65 V. j -57- 246 1'1.Y.IV.1‘1KIVII 1.11. J 1111‘. vr'v " . l I J - '.l . 5' ".51 , ..1 5-1111“ -..u L,- :1011..~L1.1...L5.LA ‘5 l 1 All! 1 I"x11 L 1 1 tva Jun-“-14.... ..JJ+LJ», v! IVY -53- {Vvv ' .S F.l m0 ni th b) 01' C0 41.77 1. 40.49 | I 87.42 ‘ 7. If \. Figure 2.17: Separation of maltodextrin (n S 15) by microbore HPLC. There is overlap of the various degrees of polymerization. The first eluting band at 37.42 minutes contains degrees of polymerization from ten to eight, the second eluting band at 40 minutes contains degrees of polymerization fi'om nine to seven. Conclusion: This assay has many strong points, but there is always room for improvement. The separation of the maltodextrin (n S 15) is baseline, and gives a good indication of the degree of polymerization. The absorbance properties of the PMA allow for detection of the saccharides on the femtomole level, and the lower limit of detection can be improved further by using fluorescence detection for the HPLC separation in the first stage. Even though for monosaccharides the retention appears to be very dependent on the saccharide, this in itself can be used as a diagnostic tool to obtain more information from the separation. Since, the degree of polymerization is confirmed using MS, the retention time is also a diagnostic tool. -59- 'Unfintt 2.5, the througl nflzz:l the ser assay,t 1 thing ; microb 0f free | infiial the re conver Suflici first 9 ten h POlyU OVErlz ions i 1 reSuI “Um‘t Unfortunately, the MS limits the sensitivity of the assay. As shown in figure 2.5, the detection limit is two pmol [114, this appies to the disaccharide through the decasaccharide. This instrument’s sensitivity begins to fall off at m/z 2 1000 and declines even more rapidly beyond m/z 1500. In this instance, the sensitivity is only limited by technology. This does not invalidate the assay, or the techniques. Improvements to the assay are needed. The separation was developed using a conventional 25 cm x 4 mm ODS HPLC column. By utilizing microbore HPLC (ti-bore HPLC), the assay can be improved, and the problem of fraction collection, with the concomitant sample loss, is removed. The initial steps toward a separation have already begun (Figure 2.17). Although the resolution is poor, for the same degrees of polymerization as the conventional separation, many of the oligosaccharides are separated, and it is sufficient to indicate another instrumental shortcoming. The chromatogram (Figure 2.17) is the reconstructed ion chromatogram (RIC) from the MS. The first eluting band at 37 .42 minutes contains degrees of polymerization fi'om ten to eight, the second eluting band at 40 minutes contains degrees of polymerization from nine to seven. This could be an indication that there is overlap between the two bands, but the intensities of the pseudomolecular ions increase for all of the ions in the second eluting band. Included with this result are the following: first, that there are apparent differences in the number average degree of polymerization obtained from the chromatographic -70- separa mixtur eleven picture| throu these with I gToups higher increa Althow (n < 1 ~ larger separations (Figures 2.3 and 2.4) and the mass spectrum of the unseparated mixture (Figure 2.4). Second, that the higher degrees of polymerization, eleven through fifteen, are not detected. With this information a more clear picture is obtained. All indications are that the linear polymer, upon passing through the extraction cone, is being sheared into smaller pieces and that these fragments are being detected. This does not appear to be associated with the skimmer cone potential since it has been observed by other groups9‘4tli’2’105 that increasing the potential on the extraction cone can reveal higher molecular weight species. But what has been shown here, is that increasing the cone potential on oligosaccharides increases fragmentation. Although this is a benefit for structure determination of smaller saccharides (n S 10), it is not beneficial for the detection, or structure determination, of larger saccharides. -71- PR glyCOPi Hence (e.g., a 0f glyr to an isolatt 3ch” Chapter 3 PRIMARY AND SECONDARY SEQUENCE DETERMINATION FOR PEPTIDES UTILIZING A SINGLE QUADRAPOLE MASS SPECTROMETER Introduction: Understanding the glycobiology (structure-function) of glycoproteins requires the structural determination of the entire molecule. Hence structure of the protein as well as any post-translational modifications (e.g., any attached sugar chains, which often occur at any of the possible sites of glycosylation) must be determined. These glycoforms invariably give rise to an impressive array of structures. Individual glycopeptide forms may be isolated using modern methods of separation, although, profiling the structure in the absence of workup may best represent a natural distribution of species. Attempts at isolation and derivatization introduce opportunities for sample disproportionation, especially for samples with contrasting structural types. However, for protein glycoconjugates the assumption is that variations in the glycan do not affect the ionization of the glycopeptide, an postulation that is more plausible as the peptide size increases relative to the size of the carbohydrate.91 Even so, higher molecular weight glycoproteins are often difficult to directly ionize by electrospray. This reflects a number of factors, such as high molecular weights leading to high-charge states and the accumulation of adducts that may challenge the resolution limits of the mass -72- spectrometer. For these materials, ES-MS may provide a spectrum that appears as a large unresolved “hump” (Figure 3.1) as opposed to nonglycosylated, or deglycosylated, species of similar molecular weight (Figure 3.2). The presence of adducts (e.g., multiple glycoforms) can limit the effectiveness of HPLC/MS interfacing by decreasing the number of glycopeptides that can be directly ionized. Alternatively, the peptide sequence in addition to the glycan distribution may be assessed separately after deglycosylation. Additional possibilities, in the study of large biological polymers, are offered by collisional activation of multiply charged ions created by electrospray ionization. Multiple charging allows higher molecular weight precursors to be collisionally activated and enhances fragmentation in these heavier species. To date, most reports of collisional activation of selected multiply charged peptides have involved triple quadrupole instruments and multiple low-energy collisions, typically with an inert collision gas in the second quadrupole.”94 Peptide precursors with mixed charge states have been fragmented by low-energy collisions in the ESI source.“96 High-energy collisional activation has been examined with doubly charged ions produced by fast-atom bombardment (FAB).97'99 -73- Figure 3.1: Two hundred pmol pl1 of fetuin characterizing the poorer sensitivity of a glycosylated peptide due to a lack of single, distinct charge series. The presence of a mixture, due to differences in glycosylation, decreases sensitivity. For these materials, ES-MS may provide a spectrum that appears as a large unresolved “hump” as opposed to nonglycosylated, or deglycosylated, species of similar molecular weight. -74- oowH coma cox oomH oooH cow NE liiillil iiIlI i I L 4- I .— H as CO or 00 CD N55 . mm «mm 455 / Is oomfi NHH moofi 555 458 wwmfi msofi . mmhfi Hmmfi osvfl Hmefl . wwvfi mmmfi NHHNwOH eemfi weHH . mam“ . saws $2 IOOH -75- Figure 3.2: Two hundred fmol ul'1 of horse heart myoglobin, HHM, showing an envelope of multiple charge states. -75- OONH 55 OOHH oooH oom cow ooh- mSH 55: L: «8 32 8:5: :3 an”? 8585 .mMmm .55 mum ms. 3. $5 $3 85 «mm :5 a? as a: Sea 82 $5 Sm SN. 53.. 5% L 35 4.2. 85 \ 82 .35 3m 25 NE .E 3.4 was 5:. 58‘ max 58 5.. HR mm... [as $5 mH< vaw 85 2.... 58 a... mvm .2: -77- I ; differ “not 1 press or a expa1 regio tlu's One characteristic of ESI instrumentation is the large pressure differential that is necessary for combining atmospheric pressure ionization with the low pressures required for mass analysis and ion detection. The pressure differential is usually achieved either with a sampling cone/skimmer or a capillary/skimmer arrangement. In either configuration as the gas expands through the sampling cone or capillary into the lower pressure region it forms a supersonic jet. At the pressures typically encountered in this region, a shock wave is formed due to collisions with background gas, and there is considerable scattering of the expanding beam. The position of the skimmer in the source relative to the cone or capillary is extremely important for ensuring minimal ion losses in this region.100 Potentials applied between the skimmer and the cone/capillary serve both to focus the ion beam through the skimmer orifice and to accelerate ions through this intermediate—pressure region. Raising the cone potential has been shown to improve signal-to-noise ratios in the ESI analysis of pesticides in water and also for the analysis of some large peptides. This has been attributed to improved ion focusing.101 Acceleration of ions through the intermediate- pressure region has also been found to cause collision-induced dissociation by increasing the energy of collisions with background gas.102:103 -73- Experimental: The in source fragmentation mass spectra were obtained on a Platform (Fisons, MA) single quadrupole mass spectrometer under control of a PC data system operating MassLynx. The fragmentations of the peptides occurred under conditions where the potential difference between the extraction cone and skimmer cone were increased to obtain maximum fragmentation. Typically, the potential was set to 55 V. All peptide samples were obtained from the Michigan State MacroMolecular Structure Facility. They were purified by HPLC utilizing a mobile phase of acetonitrile and water on an ODS column. Solutions of analytes were obtained by dissolving approximately 1 pg of sample in 1 ml of acetonitrile: water (1:1). These were introduced into the ESI source with a Sage syringe infusion pump at a flow rate of 20 pl min'l. The filter electrode was heated to 100° C, the nebulization gas flow rate was 18 l hrl, and the drying gas flow rate was 1501 hr'l. Collisional activations occurred with positively charged ions. The potential on the adjustable extraction cone to accomplish this was adjusted from thirty-seven to sixty-five volts in approximately five volt increments for selected peptides. The conditions were dependent upon whether molecular weight or sequence information was needed. Differences in charge state were taken into account only for interpretation of sequence and/or pseudo- molecular weight calculations. No attempts were made to influence the charge state of any species.1°4»105 Data acquisition occurred at 100 u 5:1 and -79- the instrument was set to observe the mass range from 100 to 1500 u. Resolution was adjusted such that for singly charged species the valley was 50% for unit resolution across the mass range. All other parameters were adjusted for optimum performance and sensitivity. Results and Discussion: For the twenty-one peptides utilized the average molecular mass was 1414 i: 174 u; the average degree of polymerization was 13 i 1. For these experiments, all pseudomolecular ions were first confirmed, by setting the potential on the cone to thirty-seven volts, optimizing all other parameters, and observing the, typically, doubly charged pseudomolecular ion (Figure 3.3). Species of higher charge were usually of considerably smaller abundance and, in most instances, the singly protonated species was also present, but at much lower intensity. Once the species of interest was confirmed, the extraction cone potential was raised to fifty-five volts (Figure 3.4). The choice of this potential was not arbitrary. The Optimum potential was obtained by observing the fragmentation pattern upon increasing the potential in five volt increments (Figure 3.5) for a couple of representative peptides. As was the case with the previously discussed saccharides, increasing the potential on the skimmer cone decreases the total ion current without a concomitant increase in signal-to-noise ratio obtained using a triple quadrupole. Although this alters the lower level of detection for the fragment ions, the sequence remains interpretable and comparable to -30- the data that can be obtained from ion trap and triple quadrupole instruments. Once optimum settings were determined, the potential was not changed for each species since it was desirable to find a standard set of values to apply to unknown peptides. It was determined, again, that 55 V was an optimum setting that allowed reasonable fragmentation with minimal amounts of adduct ions. The ion currents of individual oligomers were measured over a range of sampling cone potentials (Figure 3.5). The results are analogous to the results for the peptide in figure 3.4, perhaps suggesting that there may be a common effect dependent on both the mass and the charge of an ion rather than either the charge or the mass alone. There is, however, a difference, in regards to the exact dependency of ion current upon cone potential between the multiply charged ions of the single mass. From an inspection of the optimum extraction cone potentials for various ions in different spectra, Shiel et al.106 found that the optimum cone potential varied as approximately m/z (i.e., linear dependence) in the case of multiply charged ions of the single mass, but as approximately (m/z)"9 in the case of singly charged ions of multiple masses. The increase in peptide bond cleavage in multiply charged Species is more likely from increased charge repulsion. -31- .r. :55! .. Pal Figure 3.3: Example of a peptide (QKSQRSQAENPV) analyzed by the Fisons Platform mass spectrometer with the potential set to 37 V to obtain molecular weight information. -82- com? com. ooer OONF oooF N}: , nmw 22:55:..-52:...52.....:sm - 83 .. . _ L. I .. Nov. - . . . . A A '0— Nmmuwmhmw * m ngmgp 0,?va 89 59 58. an? ix. a «nap pump com mm” 000 oov CON wmm 1 oo— -1351- Figure 3.4: The same peptide as figure 3.3 fragmented using in- source fragmentation (cone potential set to 55 V). Nearly the entire sequence of the peptide is obtained, and the linkages not cleaved can be inferred from the data. -34- @Wmflwmflfl>m2w< L 5 Aw: cuIENO H on 3-50 x Ga Q2. 5 m 2 a: viii-ac; Ff EF-p-E- ELI LEFerL-rrrnlfstbr-r BR 000 .500 n ._ oo nmoo .500 m N.— h \ 58 vs 5 I v: s“ 5.5 as 55 as H. _ .:_Fe_Er-5EE_==_=___.LSL. Passer—var E. :_r_.-§E_E_-__._richi_c EFF-.1. PEI/rt r EN moo mmo woo wmo 550 Babe moo mmo moo mmo -85.. Figure 3.5: Ion current obtained for two peptides (QKSQRAQAENPV, solid line, and QKSQRSQDENPV, dashed line) obtained under conditions of increasing cone potential. -35- 6:00; on E 8 so mm 8 55 R LILIIIIlILII _ _ I_ _ 0958.0 . ..I.._. 1 Bio? 3 Ir 5+ mOON O. ._. . m> 3:00; -37- Table 3.1: Data from the in—source fragmentation of twenty one peptides. The type of fragment and the peptide bond that fragmented are indicated. -88- Same? v» Efiflcmcxoamm 83 w» £63 2 23:3ngamEfiS;2.3.8 Enact; cvfi mm m? fiogmfimqfiumo 82 mam.» w» ma when awoBmmmmqchmu 23 mi w» Saw; a» floammmfiflcmc 83 NE H3 3 mam fiwammmmqgoxu 83 «3 m5 ”56333305 33 adamafi a“? $33 9». SHE 303m§m<30mo 83 NENE Shah 2&3 guamfiflflomo 53 «E a» 3032x8330 53 m» m» >m2mm2mnammamma mHvH mm >m2moammcmxa mfiwfi a» >m2mnawxamxa mHvH m» mw. >m2mm2mmpem<. ume no mm am amxam=eem<. Hmmfi S 2 Sm amxofimoEEZ $2 mm mm oggqmormgg $3 ma Sm agmqmcwmka: mama ma a» amxamgmumzm O HO OH O OH ‘ o UDP-Gl UDP - , ° R-Sos . UDP-SQ UDP '0 SH 0 0H . H03 2 o oangc o + —-—b HO 0Ho (I) 0 HO OH UDP s lJDP 0:): O =53 O - Q R2 | SQDG I R1 DAG C.) o 0‘? o=c R2 1'21 Figure 4.4: Schematic for the sugar-nucleotide pathway of sulfoquinovose synthesis.131 Despite nearly forty years of efforts, our current knowledge about the reactions leading up to the formation of the sulfoquinovose (SQ) head group still falls short of the claim that the pathway of SQDG biosynthesis has been solved. The complexity of this classic problem of plant biochemistry becomes apparent if one considers that three major pathways contribute to the biosynthesis of SQDG: carbohydrate metabolism, sulfur assimilation, and fatty acid biosynthesis. Of these three metabolic processes, carbohydrate metabolism has been most intensely studied, and many ideas regarding the -107- possible pathway for SQDG biosynthesis have been derived by analogy to the common reactions of gluconeogenesis and glycolysis.132 There are many approaches that are used, singly or in combination, to work out the chemical details of a metabolic pathway. The first and most direct is to study the pathway in vitro, in a cell-free extract of a tissue capable of catalyzing the overall metabolic processes in question. Another important approach to elucidation of metabolic pathways is the study of genetic mutations in which a given enzyme fails to be synthesized in active form. Such a defect, if not lethal, may result in the accumulation and excretion of the substrate of the defective enzyme. Another powerful method for establishing the general outline of a metabolic pathway is to use an isotopic form of an element to label a given metabolite and then follow the label through its cycle. -108- Nod L NodM NodH NodPQ H3000 /— __ ‘0 0,803H HOHzC R3— 0 0 R4_ / HO ‘HO N N H OH N0dB///VO=C(CH \OszCH \O=\ 3 He” . NodC NodC CH3 lCH2)5 2'3 H u / (DH NOdFE (CH2)5 \ CH3 Figure 4.5: Schematic of sulfoquinovose with the nod genes indicated that are responsible for the formation of the various bonds. -109- Genetic Studies: In the study of the formation of Rhizobial head groups, a number of bacterial genes have been identified that are required for the formation of the nodulation factors. These nodulation genes (designated nod, no], and noe) of the symbiotic soil bacteria Rhizobia, may be divided into two classes. One class comprises genes which, when mutated, completely abolish nodulation on all legumes. Since genes in this group share significant sequence homology and can be complemented between different Rhizobial genera/species, they are often called “common.” Nod ABCIJ and nod D are the best known examples.131'134 Genes of the second class are necessary for the interaction with certain, but not all, legumes. Expression of these genes permits nodulation of additional hosts, and for this reason, they have been termed host specificity of nodulation (hsn) genes. By definition, they are unique to one or a few Rhizobia.108,109,135,136 In most cases, mutations cannot be fully complemented by the introduction of the corresponding genes fi‘om other Rhizobia. Mostly mutations result in alteration or extension of the host range. In 1990, Lerouge et al.137 showed that the products of the nod genes are N-acylated-D-glucosamine. Numerous other investigators have confirmed these findings.133:139 Since these substances are the products of the nod genes, they are called Nod factors. Principal differences among the Nod factors of the various Rhizobia concern the length of the core molecule as well -110- as the substitutions to both the reducing and nonreducing residues. Presumably, the hsn genes are responsible for these substitutions. Separation and Detection of SQ, SQDG, and Metabolites: Techniques for the separation and detection of phosphorilated and sulfated saccharides are similar to the techniques used for saccharides. Many of those that have been used include TLC on silica plates.140'144 These assays used mobile phases that were mixtures of organic and aqueous phases, typically a first dimension would contain chloroform: acetone: methanol: acetic acid: water, in various proportions. J oyard et al. used this as a first dimension and isopropyl ether: acetic acid as the second dimension for the detection of polar lipids. A deviation on this theme was used by Olsthoom et al. where butanol: ethanol: water was used, several reverse phase TLC assays, ODS coated TLC plates, using acetonitrile: water were also developed for lipooligosaccharides and some exclusively for SQDG.141v145’146 In all instances a radioactive nuclide, 13C, 35S or both, was detected by autoradiography. Attempts to determine the correct biosynthetic pathway for SQ and SQDG have also utilized ion exchange chromatography. Lee et a1.”7 using a PA 100 column with a mobile phase of 75mM KH2PO4 at a flow rate of 1m] min'1 with detection at 254 nm looked for methods to detect UDP- galactopyranose and UDP-galactofuranose. Also attempting to determine UDP-sugars were Hull and Montgomery.148 Their assay used a FAX-100 -111- column and two three step gradients all consisting of water, 200 mM NaOH, and 50 % isopropanol, conductometric detection was used. The goal was to separate UDP-sugars, nucleotides, and sugar phosphates. A related assay for sugar phosphates in biological extracts was developed by Smits et a1.”9 Yeast extracts were analyzed by solid-phase extraction and anion-exchange chromatography with a gradient of sodium acetate, from 100 mM to 500 mM and pulsed amperometric detection. In addition to HPAEC, reverse phase HPLC has been used for separating SQ, SQDG, its precursors, and in the search for metabolites. Sanjuan et al.150, to prove that a 2-O-methylfucose moiety is present in the lipo-oligosaccharide nodulation signal of B. japonicum, used a acetonitrile: water gradient at a flow rate of 1 ml min'1 to prove the purity of the nod factor prior to MS and NMR analysis. Synthetic standards have also been chromatographed by reverse phase HPLC. Rossak et al.151 used these standards to confirm the retention time of UDP-sulfoquinovose. By inactivating sqd D they caused an accumulation of UDP-sulfoquinovose and with the use of standards, confirmed that they co-chromatographed. Normal phase HPLC has also been used for the separation of sulfated glycans. Karamanos et al.152 separated 24 variously sulfated galactosaminoglycan- and hyaluronan-derived disaccharides on an Econosphere NH2 column. In order to obtain separation for all of these species, three separate mobile phases were needed. For non-sulfated A- -112- disaccharides, constituents on the column were eluted with 5 mM sodium dihydrogen orthophosphate buffered with orthophosphoric acid to pH 2.55, with a flow rate of 1.1 ml min-1. Monosulfated A-disaccharide’s required a mobile phase of 50 mM sodium dihydrogen orthophosphate buffered with orthophosphoric acid to pH 2.50 with the flow rate adjusted to 0.7 ml min'l. Finally, for over-sulfated species 50 mM sodium sulfate 10 mM sodium acetate buffered with acetic acid to pH 5.0 with a flow rate of 1.5 ml min-1 was required. HPLC is not capable of the high resolution of capillary electrophoreses. This technique has been used to separate neutral, charged, and highly charged oligosaccharides. Using open tubular capillary columns, Novotny, et al.153-155 have been able to obtain separations for all of these species. Isolation of SQ and Metabolites: Reverse phase HPLC is not strictly for analysis and purity conformation. Shultze, et al.156 used preparative reverse phase HPLC to purify sulfated lipooligosaccharides that exhibited different degrees of plant host specificity from R. meliloti. Their preparative scale (47 x 300 mm) separation used a mobile phase of 20 % acetonitrile: water with a flow rate of 120 ml min-1. Once purified, a family of sulfated lipooligosaccharides was obtained that exhibited different degrees of host plant specificity. A modification of this procedure was used by Poupot, et al.157v158 Their separation used a semi-preparative reverse phase HPLC -113- column (250 x 7.5 mm) and a linear gradient of 20 % acetonitrile: water to 100 % acetonitrile, at a flow rate of 2 m1 min-1. With these purified nod factors, they were able to determined that R. tropici nod factors are sulfated or nonsulfated chitopentasaccharides containing an N-methyl-N-acyl- glucosaminyl terminus, and that the nod factor sulfation is limited by the amount of sulfate. A similar separation was used by Mergaert, et al.159 to determine that modifications in a D-arabinosyl residue, an N-methyl and a carbamoyl group are present in the nod factors of a strain of A. caulinodans. In addition to reverse phase HPLC, column chromatography using silica gel has also been used to purify SQ, SQDG, and other nod factors. Sanjuan, et al.123v149 and Carlson et al.160 used silica gel 60 columns with a 60 % acetonitrile: water gradient for the purification of nod factors from B. japonicum. Both used this procedure to determine the structures and/or biological activities of lipooligosaccharide nodulation signals. -114- Chapter 5 SYNTHESIS OF SYNTHETIC STANDARDS OF SUSPECTED SULPHOQUINOVOSE METABOLITES Introduction: The study of metabolism has as its goal the complete, detailed description of the reactions by which the constituents of the cell are built up and broken down (intermediary metabolism) and of the mechanisms by which the chemical energy derived from the degradative reactions is conserved to be utilized in synthetic processes (energy metabolism or bioenergetics). In metabolism each consecutive step in the pathway brings about a small, specific chemical change, usually the removal, transfer, or addition of a specific atom, molecule, or functional group.161 As previously mentioned, there are several techniques for determining a biochemical pathway, other methods are available, but must be used in conjunction with the primary procedures. In order to discover the biological pathway for formation of critical biomolecules for Rhizobia symbiosis, the use of standard molecules can be invaluable. In the instance of SQ, there are currently two proposed pathways, one involves UDP-glucose (Figure 5.1131), and the other utilizes a modified glycolysis pathway. -115- Figure 5.1: One proposed pathway for a nucleotide sugar to form sulfolipids starting with either UDP-glucose or UDP-galactose. -116- CHzOH CH3 OU DP OH OU DP OH Pathway for synthesis of 6-deoxy sugars CHZSO3' 2- o so3 O OH OU DP OH 1 +2H 011,305 0 OOH OU DP OH UDP-sulphoquinovose +Diacylglyceroi Sulpholipid -117- Figure 5.2: An alternative route to the formation of sulfolipids using the gluconeogenesis pathway. -118- wrung—353% TEES—Asvowfifi 81!».- 80a segues-muggy“ §A§8€E§ hoax-2:53 2:3 38:0 . ram-1w \ . 28.6 row... “a «:2an 996% In _ 10w: 395???? OWE—u I~00 :00 :6me 12w: 99% I. I. \\\\. :60 . g olw olw anew \ g A.“ an? AI: ado if 5 .F< . 388351.. 2983 ovhan—Muuflofiax “jg .op-wmmmmm.u ._oaownmmoam I 0 IO I Om IO I 0 ID ID $.39 _ . . _ __ 6N6 IOn-VI All IOn—uI‘ll .nOlcwI .nOlow . ouw ofo :60 :«oo :60 son-ofo -119- Materials and Methods: The 3-Bromo-acetaldehyde-dimethyl acetal was obtained from Aldrich. The sodium sulfite used in all reactions was purchased from Sigma. The methyl a—D-glucopyranoside was obtained from Sigma. Carbon tetrabromide, triphenylphosphine (TPP), and sodium metaperiodate were all acquired from Aldrich. The derivatizing reagents, on- aminopyridine, p-nitrobenzylamine, and pyrenemethylamine were all obtained from Aldrich. All solvents were analytical grade from Baker. Synthesis: The formation of the sulfonic acid of acetaldehyde . dimethylacetal occurred by dissolving 2.11 g of bromoacetaldehyde dimethylacetal in 20 m1 of a water: methanol solution (4:1). To this solution 8 g of sodium sulfite (NaSO4) was added, and the reaction, shown in Figure 5.3A, occurred in one hour at 20 °C. The mother liquor from this reaction was evaporated and then dissolved in 18 ml of water and 2 ml of 12 M (concentrated) HCl. The mixture was left at room temperature overnight, for twelve hours, then concentrated to a syrupy solution to form the aldehyde, as shown in Figure 5.3B. This product was dissolved in 50 ml of water and calcium chloride (3.33 g) was added, to this solution. Sodium cyanide (1.50 g), dissolved in 20 ml of water, was then introduced. The mixture was stirred at room temperature for 36 hours and calcium hydroxide (2.22 g) was then added. Stirring was continued for a further 6 hours and then the mixture -120- was filtered and concentrated to form the sulfonic acid of lactic acid, as shown in Figure 5.3C. A so;2 in) BrCH2CH(0Me)2 —’ HO—fi—CHz-CH(OMe)2 o B o M HCl u HO—fi—CHz-CMOMe); ———> HO—fi—CHz-CHO o o C E.) NaCN Ca(OH)2 (I? I? ,o HO—fi-CHZ—CHO —» ——> HO—S—CHz-C—C: u —| o 0 OH OH Figure 5.3: A—B. Formation of the sulfonic acid of acetaldehyde. C. Formation of the sulfonic acid of lactic acid. The methyl glycoside of sulfoquinovose was synthesized starting with methyl a—D-glucopyranoside. The glycoside was suspended in dry pyridine (200 ml) and triphenylphosphine (TPP) (31.5 g, 0.12 mol) was added. The mixture was cooled to 10° C after which carbon tetrabromide (39.7 g) was introduced. The mixture was then heated to 100° C for 45 minutes, and the pyridine was removed on a rotary evaporator, the scheme is shown in Figure 5.4A. Water (500 ml) was added and the milky suspension was extracted 3 times (400 ml per extraction) with dichloromethane. The aqueous solution -121- was concentrated to a syrup which was taken up in water (150 ml) to which sodium sulfite (19 g) was added. The solution was heated on a steam bath for 18 hours and then concentrated to a syrup to which methanol (350 ml) was added. A white precipitate formed and the mixture was left at 10° C for 12 hours. It was then filtered and the white solid retained. An aliquot of the material (2.5 g) was dissolved in water (20 ml). Sodium metaperiodate (5.0 g) was then added, and the mixture was kept at 4° C for 12 hours in darkness. The reaction is shown in Figure 5.4B and is believed to take place via cyclic periodate intermediates.162 This formed the sodium salt of propanal-Z-ol-B- sulfonic acid. -122- .9959on 3&5th ma omo>oawsdom3m mo onmmoobwvmfioa 93 89a Bow oEom3m-m-~o.N-~m§Qoa mo Bow ammonium on» .«o 25398.25 .m .ogmonmbaoogwénfioa 89m omo>onm5vob=m .«o mnmmoobwafioa ofi mo ”85985.“ 9E- .< "Wm oar-m m..moo mo, w om om \olwamoJm-Io-wz Ml o x m o o o 9 z mzmo-mamo om m £00 £00 £00 on mam-H o - All v N Om Om «Z O #50 O aszmumo ammo om moamo om -123- Separations: The HPLC separation was performed on saccharides derivatized by reductive amination with glucose as the initial test molecule. A separation was developed involving three derivatizing reagents: (X- aminopyridine, p-nitrobenzylamine, and pyrenemethylamine. A baseline separation was obtained on a reverse phase, ODS, column with a gradient initially consisting of 10 % methanol modified with 0.1 % trifluoroacetic acid and adjusted linearly to 100 % methanol over 75 minutes. The initial flow rate was of 1 ml min'1 and decreased to 0.5 ml min'1 linearly over 20 minutes then remained at 0.5 ml min'1 for the remainder of the separation. Each derivatizing reagent/derivatized glucose pair was separated from the subsequent reagent and its derivatized glucose by ten to twenty minutes. All species were detected by a single wavelength UV detector operating at 240 nm. Confirmation of the elution order was obtained by collecting fractions, and acquiring a full spectrum on a scanning UV/V is spectrophotometer. The individual derivatizing reagents were also injected and their retention times determined. In addition to the reverse phase HPLC separation, since all molecules of interest are negatively charged due to the sulfonic acid group, an anion exchange assay was developed. For this separation a diethylaminoethyl (DEAE) column operating at a flow rate of 1.0 ml min'1 was used. The mobile phase consisted of 10 mM Na Borate modified with 2 % acetonitrile The pH was lowered to 4.4 using 1 M HCl. Complete separation of sulfoquinovose -124- from the sulfonic acid of acetaldehyde occurred in fifteen minutes. Refractive index detection was used to detect both species. Since all of these species are negatively charged, the final set of separations again involved electrophoresis. A complete separation of all synthesized species was obtained by a combination of paper electrophoresis and TLC. The paper electrophoresis was carried out on Watman # 3mm chr paper that was 40 cm long. The mobile phase consisted of 0.5 M ammonium acetate that was saturated with Ca(OH)2, and the electrophoresis occurred at 325 V for 1.5 hours. Bands were detected using silver nitrate to visualize the standards. The standards, and mixtures of the standards of interest, were removed from the paper by cutting and eluted into water. These eluted bands were lyophilized, and the resulting material dissolved in water: methanol (1:1) and spotted onto silica TLC plates. This dimension of separation occurred using a mobile phase of 1-propanol: methanol: ammonium hydroxide: water (1:1:1:1). To detect the resulting bands, the plates were developed with orcinol spray. -125- r‘i Figure 5.5: NMR first reaction product, the sulfonic acid of acetaldehyde dimethylacetal from the starting material of bromo- acetaldehyde-dimethylacetal. -126- an «.6 von D.n Don o.’ N.‘ Gov wov p.. L. . .. .L FL-PLItLlPirFP>u—ILLLLIr[I[I—Ir[LlhlrirLlri—IFELIEL.LL_-pLIrLl—iLLLLLiLLlp-rt LI.L-_... p-.. L r .-.-r b-— .LL.-.F -127- 4.3 H H203S—CH2- &— COZH 3.2 (5H ll'flllllil’llfilllll 4 3 2 1 Figure 5.6: The final reaction product of the first series, the sulfonic acid of lactic acid, one of the suspected metabolites of sulfoquinovose. Formation of the sulfonic acid of lactic acid after the reaction of the sulfonic acid of acetaldehyde with -CN and Ca(COs)2. -12s- Results and Discussion: Organic Reactions: The sulfonic acid of acetaldehyde, a.k.a.: 2-oxo- ethanesulfonic acid, has previously been prepared. Roy and Mewling,163 synthesized the sulfonic acid of acetaldehyde in several ways. In one instance, the method started with cysteine, which was reacted with barium chloride in aqueous sodium hydroxide. In addition Wang and Sayre.164 used as starting reagent 2-morpholin-4—yl-ethanesulfonic acid and obtained the same product in 24 hours at room temperature at pH 8. Since salts of sulfonic acids can be prepared by treatment of primary alkyl halides with sulfite ion165, our synthetic method starts from 2-bromo- acetaldehyde-dimethyl acetal this by a Strecker type reaction.166'167 It is clear from the resonances in figure 5.5B that the desired product was obtained. In particular the doublet at 3.13 ppm indicates the presence of a carbon with two protons next to a sulfonic acid group and a carbon with only one proton. In addition, the resonance at 4.75 ppm is a triplet indicating a proton on a carbon adjacent to a carbon with two protons. From the coupling of the resonances, these two carbons must be adjacent to one another. Its chemical shift is reasonable for a carbon next to two acetals. The large singlet gives evidence for the presence of the two acetals at 3.29 ppm, this did not deviate significantly fiom the starting material as shown in Figure 5.5A. Typically, the formation of a dehalosulfonation requires refluxing conditions.165 In this instance, likely due to resonance stabilization from the oxygen atoms beta to -129- the halogen, this reaction occurred at 4° C and was exothermic. In order to remove the dimethyl acetal, the resulting sulfonic acid was subjected to acid hydrolysis. The alkoxy group, -OMe, is not a leaving group, and so must be converted to the conjugate acids before they are hydrolyzed by dilute HCl. The reaction proceeds by an SNl mechanism since the carbocation formed is stabilized by resonance through the unshared electron pair on the oxygen adjacent to the cationic center.164 The proton NMR of the product is similar to the sulfonic acid acetaldehyde dimethylacetal, except for the loss of the resonances attributed to the acetals and the presence of an alcohol resonance at 3.27. This indicates that the aldehyde has been formed (figure 5.6). -130- ...mlm mi ...tmnummmm mmmlmillnl » 100 9 0 3 o 7 O 6 0 m Figure 5.7: The 13C spectrum fi'om the sulfonic acid of acetaldehyde dimethylacetal. In order to produce the sulfonic acid of lactic acid, the acetaldehyde sulfonic acid first must be converted to the cyanohydrin by addition of HCN. This reaction occurs very slowly when pure HCN is used, but rapidly when cyanide ion, CN', is added. This is because HCN is a weak acid (pKa = 9.1) and therefore is not significantly dissociated. The cyanide anion, however, is strongly nucleophilic, and addition occurs by a nucleophilic addition pathway. Protonation of the anionic tetrahedral intermediate yields the stable tetrahedral cyanohydrin product plus regenerated cyanide ion153. Cyanohydrin formation is useful because of the further chemistry that can be carried out. These functional groups can be hydrolyzed to yield carboxylic acids, thus providing a method for transforming an aldehyde into a different functional group while lengthening the carbon chain by one unit. -131- ...,..i Figure 5.8: NMR of the synthesized SQ-Me from the starting material of the glycopyranose methylglycoside. -132- 809 o.n N.n v.n u.n c.n O.v N.v v.v m.v o.v c.m . —. ,. . . _ . . . .._ ..... .. _ . ... . r . ._. _ .. r..r_L .LLLLL-.-.....T. ?r_-_LLIr.—.. r» .._ pLL._l_L .LL-.-»..LL._rrF—.r_!.LLLsPPFr—LLLF—Lu....-.._-r..r. _Li.L.p|_ . . r. —. .._ .. .1. 5,7 2,: , ...,. .1 1......- .1 :3-.. a. J ...I. .1. _4“ -133- Figure 5.9: NMR of the reaction product of SQ-Me after periodate oxidation, the formation of the sulfonic acid of propanal-2-01-3-sulfonic acid. -134- an n.n oov m.v o.n n.n 0.0 via 6.x. n4. o..¢ p—hn-n—thb—ph-I—DPPLI—bbLbr—thbhbbbbippnbb-—bbbb—-Irbh—hhhb~ _ I quu PIfii I] 3 -135- The reaction, to form sulfoquinovose, proceeds in two steps, intermediate formation and intermediate decomposition and involves three components. Intermediate formation involves the formation of the TPP-Br complex from TPP and CBr4. An ionic mechanism is responsible for the formation of the TPP-Br complex. It is initiated by the polarizing action of the permanent dipole of phosphane on the symmetrical but readily polarizable CBr4, and it leads to heterolytic bond cleavage by a direct interaction between the phosphorus atom and the carbon bromine bond.1‘59’170 If the —OH groups available for the reaction are chiral intermediate decomposition occurs with extensive inversion of configuration,. The final component of this system is a-methylglucopyranose. Since this is the methyl- glycoside this system can no longer enter into the straight chain configuration (Figure 1.1), and the only alcohol group present that can undergo halogenation is on carbon six. All other hydroxyls are blocked from the reaction by the closed ring. The intermediate decomposition with the TPP-Br complex is thus a method of converting a-methylglucopyranose into 6-bromo-oc-methylglucopyranose that can be achieved under very mild conditions. The sulfonato-de-halogenation reaction, the Strecker reaction, to form sulfoquinovose from 6-bromo-a—methylglucopyranose has previously been discussed.165 -136- Figure 5.10: Reverse phase HPLC separation with absorbance detection at 240 nm of oc-aminopyridine, p-nitrobenzylamine, and pyrenemethylamine after reductive amination using glucose as the test molecule. -137- =9 — Absorbance .82 ha .88 Time -138- (3.00 I 30.00 Figure 5.11: Anion exchange separation of SQ and the sulfonic acid of acetaldehyde. -139- Separations: As mentioned earlier, many HPLC methods have been developed for the separation of carbohydrates. These include separations that are specific for individual monomers, as well as oligomers, polymers, and charged species. Detectors for these separations have included refractive index, amperometry, and mass spectrometry. UV/Visible absorption detectors require derivatization to increase the sensitivity of the assay. Multiple derivatizations, to increase sensitivity as well as selectivity, have not been undertaken in the HPLC separation of carbohydrates and their metabolites. The test molecule in this study was glucose, an aldose that undergoes reductive amination. This technique was later extended to sulfoquinovose, one of the target molecules. The derivatizing reagents that were chosen are all primary amines with conjugated ring systems. They were (x-aminopyridine, p-nitrobenzylamine, and pyrenemethylamine. It has been previously mentioned that saccharides and modified saccharides require more than one separation dimension in order to assure accurate and precise identification. Reductive amination has been discussed previously and will not be reintroduced here (p. 36 and Figure 2.1). The order of elution for this separation was confirmed by several techniques. First, derivatizing reagents were injected separately to compare retention times. Then to confirm that there were no strong interactions between the derivatizing reagents they were co-injected. Finally, fi'actions were collected and a full UV/V is spectrum was obtained using a scanning spectrophotometer. Fractions were then -140- collected to determine the order of elution, whether the derivatized saccharide eluted first or the derivatizing reagent. These fractions were subjected to phenol/sulfuric acid to test for the presence of sugars. In addition, the retention time of the individual derivatizing reagents vs. the derivatized glucose were compared. These analyses determined that the derivitized saccharide was the first to elute in each pair of peaks. A typical chromatogram is shown in Figure 5.11, the peaks labeled 3 and 4 correspond to those for a-aminopyridine, 5 and 6 correspond to the derivatized glucose and excess derivatizing reagent of p-nitrobenzylamine, and the peaks labeled 7 and 8 correspond to those for pyrenemethylamine. In addition to reverse phase HPLC, an ion exchange chromatographic assay was developed (Figure 5.12) since all of the species of interest are negatively charged. Anion-exchange chromatography of innately anionic carbohydrates have been performed previously.171 Typically alkaline solutions are necessary for elution of carbohydrates in HPAEC. Sodium hydroxide is the base of choice at a concentration of 15-20 mM for monosaccharides.146v 170 For the purpose of this separation sodium hydroxide could not be used since N aOH has been shown to cause rearrangements upon drying of the sample, this would be unacceptable since further analysis with N MR is required. The requirement for complete structural analysis with NMR also precluded the use of pulsed amperometric detection. Although sensitivity and selectivity is greater for amperometric detection compared to -141- refractive index detectors, the technique may be destructive due to redox reactions. With the above limitations in mind, the following assay was developed. For the ions of interest for this assay a diethylaminoethyl (DEAE) column was used to separate the two most different molecules of interest. The mobile phase chosen to separate the molecules of interest consisted of 10 mM Na Borate modified with 2% acetonitrile at a pH of 4.4. This not only allowed for separation based on charge density, but also allowed for separation based on the size of the molecules. It is also advantageous that the separation was obtained using a borate buffer system, in that removal of buffer could be obtained by subjecting samples to dry ice prior to passing samples through a cation exchange resin. Choosing sulfoquinovose and the sulfonic acid of acetaldehyde a complete separation was obtained in fifteen minutes. A refractive index detector was used to detect eluting peaks. It was determined that the size of the molecules was critical for the separation, experiments that altered the pH were undertaken. As the pKa of the sulfonic acid group (pKa z 1.92) was approached, the retention time of the two species should have increased, due to increased interaction with the stationary phase. The hydrogen ion concentration at pH 2. 5 is critical and the separation is altered with changes in pH as small as 0.2 units. Below pH 4.5 it is no longer a factor and there is little change in the separation to pH 2. In order to confirm the order of elution, the standards were injected individually. Also, fractions -142- were collected, and desalted, for NMR analysis. It was confirmed that the first eluting band was sulfoquinovose, or the methylglycoside, and the second eluting band was the sulfonic acid of acetaldehyde. The final separations for the detection of sulfoquinovose and the suspected metabolites were paper electrophoresis and TLC. Paper electrophoresis was chosen, as was ion exchange chromatography, because all of the species of interest are negatively charged and any neutral or positively charged species would not interfere with the assay. Ammonium acetate was chosen as the buffer since it is volatile and, as repeatedly mentioned, a second dimension of separation would be required. In addition, calcium hydroxide was added. Although, Ca(OH)2 is not volatile, it did not interfere with the second separation. Also, Ca2+ chelates well with the hydroxyls on sugars. The separation occurred in 1.5 hours at 325 V on a water cooled stage. Here two dimensions of separation are required. Since the separation is based on charge density, any phosphorylated or sulfated sugar would produce the same migration in electrophoresis. Also, considering the low molecular weight of the expected metabolites a complete separation in the first dimension is not expected. To develop the separation, electrophoresis was done on the standards individually and in a mixture. The migration distances for the sulfonic acid of acetaldehyde and the sulfonic acid of lactic acid were the same, 4 cm. For SQ and the sulfonic acid of propanal-2-ol-3- sulfonic acid, the migration distance was 9 cm. Once the separation was -143- ifin complete the individual standard bands were visualized using silver nitrate solution. The bands were re-aligned with the lane containing a mixture of standards, and the individual bands were cut out, eluted into water, frozen, and finally lyophilized. The resulting solid was taken up in a water: methanol solution (1:1) and introduced onto TLC plates. Complete separation occurred with a mobile phase of l-propanol: methanol: ammonium hydroxide: water (1:1:1:1). Individual bands were visualized using orcinol spray. The final Rf values of the individual standards were 0.31, 0.43, and 0.76 for the sulfonic acids of acetaldehyde, lactic acid, propanal-2-ol-3-sulfonic acid and for SQ, respectively. -144- mfibwc‘flomc‘» '. I. 75.00 100.00 0.00 50300 Time Figure 5.12: Separation of SQ derivatized with oc-aminopyridine, p- nitrobenzylamine, and pyrenemethylamine. -145- Conclusion: Although the target molecules (standards) were all successfully synthesized, and each separation reached its individual goal, the final goal of obtaining a “gold standard” by which the pathway to determining sulfoquinovose biosynthesis was not achieved. The reverse phase HPLC assay is both sensitive and selective using a diode array detector, and was utilized in the development of the assay. The separation, although developed for glucose also works well for sulfoquinovose (Figure 5.13). The shortcoming is that not all of the suspected metabolites, from either proposed pathway, will undergo reductive amination. In addition, all other reducing sugars will undergo reductive amination, which complicates the interpretation of the chromatogram when attempting to identify sulfoquinovose and the detectable metabolites. Ion exchange chromatography with refi‘active index detection is not as sensitive an assay as the reverse phase chromatography described above, however all positively charged and neutral species are removed from the sample in the void volume, and the negatively charged species are separated. The molecules of interest can be collected and structural analysis can be performed on these fractions after removing the salts. The shortcoming here is that the detector is not selective. Overlapping bands may go unnoticed. The option again is to include a second dimension of detection, and here NMR was chosen. Due to sample loading limitations of the column, multiple -146- injections to collect fractions are required to obtain sufficient material for NMR analysis. Many two dimensional separations have been developed for TLC. Using paper electrophoresis as the first dimension allows for many neutral and positively charged species, such as other saccharides that are not of interest, to be removed. This allows for a second, or even third, dimension that is less complicated to interpret. By using preparative scale TLC, a sample that is possibly more pure may be obtained. Here, again, the difficulty is with sample loading limitations in the first dimension. Due to diffusive broadening of the bands, the initial spot must be small and concentration effects cause the migration of the sample to be altered. The standards would have to be spotted onto the paper at approximately the same concentrations as an unknown to assure that migration distances would correlate. Acknowledgements: I would like to thank Dr. Hollingsworth for the synthesis of sulfoquinovose and the sulfonic acid of lactic acid. I would also like to thank him for his assistance in removal of the methyl acetals from the sulfonic acid of acetaldehyde dimethylacetal. -147- Chapter 6 EXTRACTION AND STRUCTURE DETERMINATION OF SULFOQUINOVOSE Introduction: One important class of biomolecules is the oligosaccharides. These can be present in nature as free carbohydrates or as glycoconjugates (e.g., glycoproteins and glycolipids). The challenge in the separation and analysis of these materials is that they do not contain strong chromophores for absorption in the UV or visible region. They are also highly heterogeneous in terms of their degree of polymerization, base-unit composition, and even molecular shape (i.e., functional group modifications). The bioanalysis of oligosaccharides has been demonstrated in ion-pairing chromatography by conductivity detection, in ion-exchange chromatography by pulsed amperometric detection, in supercritical fluid chromatography by mass spectrometry, and by direct measurement using matrix assisted laser desorption mass spectrometry and fast-atom bombardment mass spectrometry. Mass spectrometry alone, particularly soft ionization techniques, cannot be used for complete structure determination of molecules. For structure determination, carbon and proton NMR are invaluable, although the amounts of sample required and relative purity of the sample necessary for interpretation are greater than for mass spectral analysis. -148- Materials and Methods: R. trifolii 1021 was grown at 29° C in 41 shaken flasks containing 21 of modified Bergensen’s, B III, media each.172 The cultures were supplemented with either the methyl glycoside of sulfoquinovose, as the carbon and sulfur source or H235803 as the sulfur source. To acclimate the bacteria to sulfoquinovose, as the carbon and sulfur source, starter cultures were used. Test tubes containing 5 ml of sterile media were serially diluted to 10 ml. The bacteria in each tube were incubated until late stationary phase before being transferred to the next tube, thereby increasing the concentration of SQ the cells were exposed to. Once acclimated to SQ-Me, these cells were then grown in two litre batches until they were in late stationary phase and were then harvested by centrifugation at 7.5 krpm for 25 minutes on a Sorvall RC2-B centrifuge equipped with a GSA rotor. The supernatant was removed and the resulting pellet was extracted by stirring with a mixture of ethanol: water (1:1) for three hours at 4° C. Cells were then centrifuged again to obtain a pellet and the supernatant was passed through a cation exchange column. The column was then washed with a mobile phase of ethanol: water (1:1, 500 ml) and the eluent collected. All fractions were evaporated on a rotary evaporator and redissolved in a solution of propanol: methanol: ammonium hydroxide (3:1:1). This was passed through a silica column (6x15 inches), and 5 m1 fractions were collected. These fi'actions were spotted onto silica TLC plates and -149- developed using a mobile phase of 1-propanol: methanol: ammonium hydroxide: water (1:1:1:1). Detection of the spots of interest was obtained by spraying with orcinol and heating, in the instance of the bacteria grown on SQ-Me, and by use of Autorad detection of radioactivity, for 35S. Fractions with similar purity were then combined and NMR analysis performed. -150- ‘1' o.)ul'-W" all Figure 6.1: First eluting fraction after sample was purified from R. 1021. -151- ...,. u f fit—PEPE gig-{4.3 .v o.m Egg-..» » 3...}: o c.r n.n v.n u.n o.n o.v N.v v.v a .. LEE-3:33.} -152- Figure 6.2: NMR of SQDG from the published work of Cedergren et a1. Results and Discussion: In order to harvest sufficient material for NMR analysis, none of the separations developed in the previous chapters was amenable to scale up. Once the bacteria had been lysed and centrifuged, to obtain a crude extract, which was passed through a cation exchange column. Most cations are removed, and the column does not retain neutral species or anions. An aliquot of the eluent was analyzed by NMR to determine whether molecules relevant to sulfoquinovose metabolism were -153- present in abundance. This was to be determined by comparing the crude extract to the synthetic standards. Further purification was then undertaken. The procedure used consisted of a silica flash column and an aqueous mobile phase of 1-propanol: methanol: ammonium hydroxide (3:1:1). This first dimension of separation resulted in several overlapping bands, determined by following the movement of 35S labeled peaks that were obtained from a second set of incubations of R. 1021. By running standards alongside the unknown sample on the same TLC plates, many of the Rf values corresponded to those fi‘om the bacterial samples. In Figure 6.1 as compared to Figure 6.2, it was determined that sulfoquinovose is present as a major negatively charged component obtained fi'om the biological samples and that the resonances from the NMR analysis of the column chromatography correspond well to those obtained previously fi'om preparative TLC. In particular, all of the resonances corresponding to the carbohydrate protons of SQ are present, and at similar frequencies. From the data presented in Figures 6.1 and 6.2, the published spectrum fi'om Cedergren et al.”3 of SQDG, as compared to Figure 5.9 from the synthesis of the methylpyranoside of sulfoquinovose, it is apparent that the synthesized material has been converted into sulfoquinovose, and, from the TLC retention factors, that the 35S has been taken up by the bacteria and included in its metabolism. -154- Conclusion: After more than forty years, the question still remains as to the biosynthesis of sulfoquinovose. The answer lies in obtaining sufficient material to obtain 2-D NMR and definitive proof for the structure of all metabolites. In addition, the enzymes responsible must be isolated and the reactions proven in vitro, so that no questions remain. Glycolysis and gluconeogenesis are two of the most well known examples of biosynthetic pathways. If sulfoquinovose is synthesized using this pathway, perhaps many of the enzymes involved are also involved in the synthesis of SQ. General forms of all of these enzymes are available commercially. The specific nutritional requirements for different microorganisms vary greatly. For example, heterotrophic microorganisms require organic matter, but autotrophic microorganisms do not. Some microorganisms are fastidious, having complex specific growth factor requirements, and therefore, these microorganisms are able to reproduce only under greatly restricted conditions. Other microorganisms are capable of reproducing under less restricted conditions. Some microorganisms have simple nutritional requirements and are able to reproduce by using a single carbon source as their growth substrate. Such organisms have the metabolic capability of interconnecting the various biochemical pathways so that a particular carbon compound can be transformed into the carbon backbones of all the required macromolecular constituents of those organisms. The fact that SQ and UDP- glucose can be used as substrates for growing Rhizobia is not surprising. -155- These organisms have been utilizing the carbon sources available for growth. The ability of Rhizobia to survive requires its adaptability and this extends to its sources for food. The fact that overlapping bands were obtained in the separation of sulfoquinovose and its metabolites reinforces the notion that multiple separations are required to obtain pure samples for analysis. One of these bands was sulfoquinovose (Figure 6.1) another has been tentatively identified as a sugar, although from the TLC it has the same Rf as UDP—glucose (Atima), it is not likely this compound. Since many of the assays for SQ and metabolites utilize TLC solely for identification based on Rf and the comparison with standards, this data indicates that this practice is prone to errors. Final identification of SQ and its metabolites requires obtaining sufficient material for detailed structural analysis. This will require that larger batch sizes, likely more than ten times the volume used in this study. Following the metabolism using 35S should be performed on a smaller scale, so that the elution of metabolites may be monitored. But for structural analysis of metabolites, the availability of larger quantities of material is crucial. This is especially true when the material, in the absence of mutants where intermediates build up, is present in only minute quantities. In addition cell free, in vitro, experiments need to be conducted. With the wealth of knowledge available on gluconeogenesis and glycolysis that have been accumulating for the past hundred 150 years, from Pastuer’s initial -156- experiments in determining the mechanisms of fermentation to the present, if the suspected metabolites follow this pathway the in vitro experiments are a logical next step. Acknowledgment: I would like to thank Dr. Hollingsworth for his help in acclimating the bacteria to sulfoquinovose. Atima Sharma’s help in starting and maintaining the bacterial cultures was invaluable. -157- APPENDICES APPENDIX A THE D-ALDOSES Aldom'ase F‘Q ‘1 ..‘dflfllh' H H 0 H H H O O I _ . _ _ ma .. H \ .c-.-._...._...._... a _ _ _ a. m m -c-._...._.-._.-.. H H H H m- o o H H o o m m n / I _ _ _ _ n: ._. ._. \. xii-74-...-. o; 4- ..u o. .- a a m m a \C.ICICIm m n W u." .9 ca ._a. m- 4 m m. / H H \CICInWIn—ulnfllm H O O H O or \ H w H o H ¢ _ _ _ w .mD H H \cIJ-Ji-fli mm. H H H H H H O H u o o o H o o a /. I _ _ a _ .. .m CICICICICIG W s m .- H H H m H w W I _ _ _ _ m: a a a \cl_-..-_i-_i-_i 0 H 0 O 0- \ H o o H H // _ _ _ H . ) H H 3...].-. m ( H H H H H nmnum/o’T-Xo. ClClClClClm H H H \ _ . _ _ o o o o. H H o H H /,c ._. ._. m u \l 2 P ‘ H "P "mu D 0 H W W Mun w / a _ _ _ _ m. H H H H \CIJ.I%I%I%I o o o o o. \ a o H H u a _ _ _ a We a Elfin-3.19am n a a a H H H H H o o o o o o / a _ _ _ _ a: CHICHCLIICIICHIC \ _ _ _ . H "H H H H D-‘Nocc (0d) D-ldooe -159- (Man) D-Mlnnose D—Ouioac D-Altrose n—Olucosc (Ole) D-Allose Kerommu APPENDIX B THE D-KETOSES Kerom'on (fl-1,0“ C I: O | (111,014 Dihydtoxyaccm cmon ' (I: = o H— ('3. — OH 04,011 D-Brythmlocc CH; OH Kelopemou: HO— H H — (I: "" OH 3mm D-Xylnlonc CH30H C1130}! mm ('3- O (I: - O Ho—c—H H—C—OH HO '- C — H HO '— (I: — H H—C—OH .H—C—OH (Izmoa 223,011 o-Tngm D—Sorboce -160- $11on c =- o I H—C—OH H—(IZ-OH CH,0H D-Ribulooe 01,011 (Lo “—22—“ H—(II—OH H—C-OH ‘CH,0H o-Puicoue AMINO ACIDS AND THEIR MASSES APPENDIX C Amino Acid Amino Acid 3 (1) Elemental Mono- Aver. Group Letter Composition isotopic Mass Code Mass Hydrophobic Alanine Ala (A) CaHsNO 71.03711 71.08 (nonpolar) R Valine Val (V) C5H9NO 99.06841 99.13 Groups Leucine Leu (L) CeHuNO 113.08406 113.16 Proline Pro (P) C5H7NO 97.05276 97.12 Phenylalanine Phe (F) C9H9NO 147.06841 147.18 Tryptophan Trp (W) CnH10N20 186.07931 186.21 Methionine Met (M) CngNOS 131.04049 131.20 Hydrophilic Glycine Gly (G) C2H3NO 57.02146 57 .05 (polar) R Serine Ser (S) C3H5N 02 87.03203 87.08 groups Threonine Thr (T) C4H7N02 101.04768 101.11 Cysteine Cys (C) C3H5NOS 103.00919 103.15 Asparagine Asn (N) CeH12N4O 156.10111 156.19 Glutamine Gln (Q) CsH8N202 128.05858 128.13 Tyrosine Tyr (Y) C9H9N 02 163.06333 163.18 -161- APPENDIX C (cont.) AMINO ACIDS AND THEIR MASSES Amino Acid Amino Acid 3 (1) Elemental Mono- Aver. Group Letter Composition isotopic Mass Code Mass Basic R Histidine His (H) C6H7N20 137.05891 137.14 groups Lysine Lys (K) C6H12N 20 128.09496 128.17 Argim'ne Arg (r) CsH12N4O 156.10111 156.19 Acidic R Aspartic Acid Asp (D) C4H5N03 1 15.02694 1 15.09 groups Glutamate Glu (E) C5H7N03 129.04259 129.12 -162- APPENDIX D GLYCOLYSIS AND GLUCONEOGENESIS HO OH OH Hexokinase < > Glucose 6-phosphatase '2o3po “0% HO OH OH 'zoapo l I CH20H o H OH OH 902' HCOPO3'2 . Fructose CH Phosphofructokmase—l 1,6-bisphosphatase ' 2. 002 403? 0130;.2 002' O 0 (I302. 2 A '-o HCOP03' .- H OH 8H2 V CH3 H2? OH H1?“ .— cHo Hoopoa'2 HCOH —> HCOH ' CH20P03 CH20P03 1 ll -2 H OH HICOPO3 2? HCOH —" H9011 ('3H20P03'2 CHzOPoa'2 -163- APPENDIX E SPECIES SPECIFIC MODIFICATIONS TO SQDG 30 m £0 m m 3.6 m.” 598 mSEEBS .< Smamododa 02-04“ 03.2 Sm-:.0m-o-v .mz-o-m so m Shag?“ m .36 m £202 32am .% .m 8.32-0-“ m m 3m m 320 mg $3.95 5.3 .m 0932-0-“ m m 8m m 36 Ha 8.4an 8828.... .m 8.56 so m m 95.326.” m .36 m 335: 88833 .m m m Damage-“ m .36 m 22me seesaw... .m m £0 mSm m :36 m $930 Ema: .m m v «<30 0950 m «<30 m m mSm m 35 3 :8 £58. .m OOan a v «<20 m m anew m 2:6 34 3M4 asses .m 3430 _ m m m 0060 = 2 36 Ha 083mm 922., 3 Emossas .m ad £ £ a a : qabm mfiabm 5:535 gamma a concave moutwnoonmo :8 a mo mougosbw -164- BIBLIOGRAPHY ~165- Bibliography 1 Gene Rodenbery, StarTrek series. 2 G. Hu and B. L. Vallee; Anal. Biochem.; 218; 185-91 (1994). 3 P. Jackson; Anal. Biochem.; 216; 243-52 ( 1994). 4 C. W. Sutton, J. A. O’Neill, and J. S. Cottrell; Anal. Biochem.; 218; 34-46 (1994). 5 D. Knapp; Analytical Derivatization Reactions; Wiley-Interscience Publication; New York; 146-224 (1979). 6 N imura, N.; Kinoshita, T.; Yoshida, T.; Uetake, A.; Nakai, C. Anal. Chem. ; 60, 2067-70 (1988). 7 Z. El Rassi, J. Postlewait, Y. Mechref, G. Ostrander; Anal. Biochem.; 244; 283-90 (1997). 3 J. Quirke, C. Adams, and G. Van Berke]; Anal. Chem.; 66, 1302-15 (1994). 9 D. Knapp; Analytical Derivatization Reactions; Wiley-Interscience Publication; New York; 25-64 (1979). 10 Bayliss, M. A. J .; Gomer, R. 3.; Shepherd, J. Chromatogr.; 445; 393-402 (1988). 11 D. Knapp; Analytical Derivatization Reactions; Wiley-Interscience Publication; New York; 338-46 (1979). ' 12 A. G. McInnes et al., J. Chromatogr. 1, 556 (1958). 13 C. C. Sweeley, R. Bentley, M. Makita, W. Wells; J. Am. Chem. Soc.; 85; 2497 (1963). 14 J. Lonngren and S. Svensson, Mass Spectrometry in Structural Analysis, 41-106 15 H. Kwon, J. Kim; Anal. Biochem.; 215; 243-52 (1993). 15 D. Fu, R. O’Neill; Anal. Biochem.; 227, 377-84 (1995). 17 S. Honda, E. Akao, S. Suzuki, M. Okuda, K. Kakehi, J. Nakamura; Anal. Biochem.; 180; 351-7 (1989). 13 K. Sato, K. Sato, A. Okubo, S. Yamazaki; Anal. Biochem.; 251; 119-21 (1997). 19 H. Lis, N. Sharon; Eur. J. Biochem.; 218; 1-27 (1993). 20 M. Spellman; Anal. Chem.; 62; 1714-22 (1990). 21 V. Reinhold; B. Reinhold, C. Costello; Anal. Chem.; 67; 1772-84 (1995). 22 P. Jackson; Anal. Biochem.; 216; 243-52 (1994). 23 Y. Lee; Anal. Biochem.; 189; 151-62 (1990). 24 M. Hardy, R. Townsend; Proc. Natl. Acad. Sci. USA; 85; 3289-93 (1988). 25 K. Koizumi, Y. Kobata, T. Tanimoto, Y. Okada; J. Chromatogr.; 464; 365-73 (1989). 26 W. Wang, K. Erlansson, F. Lindh, T. Lundgren, D. Zopf; Anal. Biochem.; 190; 182-7 (1990). 27 A. Manzi, S. Diaz, A. Varki; Anal. Biochem.; 188; 20-32 (1990). -166- 23 D. Whitfield, S. Stojkovski, H. Pang, J. Baptista, B. Sarkar; Anal. Biochem.; 194; 259-67 (1988). 29 R. Townsend, M. Hardy, O. Hindsgaul, Y. Lee; Anal. Biochem.; 174; 459-70 ( 1988). 30 T. Hayase, M. Sheykhanazari, V. Bhavanandan, A. Savage, Y. Lee; Anal. Biochem.; 211; 72-80 (1993). 31 A. Clarke; Anal. Biochem.; 212; 344-50 (1993). 32 Y. Lee; Anal. Biochem.; 189; 151-62 (1990). 33 G. Guile, S. Wong, R. Dwek; Anal. Biochem.; 222; 231-5 (1994). 34 M. Field, D. Papac, A. Jones; Anal. Biochem.; 239; 92-8 (1996). 35 J. Baenziger, M. Natowicz; Anal. Biochem.; 114; 276-80 (1981). 36 S. Mellis, J. Baenziger; Anal. Biochem.; 114; 276-80 (1983). 37 S. Mellis, J. Baenziger; Anal. Biochem; 134; 442-9 ( 1983). 38 S. Hase, S. Hara, Y. Matsushima; J. Biochem. 85; 217 -20 (1979). 39 A. Kondo, M. Kiso, A. Hasegawa, I. Kato; Anal. Biochem.; 219; 21-5 (1994). 40 N. Kuraya, S. Hase; Anal. Biochem.; 233; 205-11 (1996). 41 Y. Makino, N. Kuraya, K Omichi, S. Hase; Anal. Biochem.; 238; 54-9 ( 1996). 42 N. Tomiya, J. Awaya, M. Kurono, S. Endo, Y. Arata, N. Takahashi; Anal. Biochem.; 171; 73-90 (1988). 43 N. Takahashi, Y. Wada, J. Awaya, M. Kurono, N. Tomiya; Anal. Biochem.; 208, 96-109 (1993). 44 Y. Zhang, R. Cedergren, T. Nieuwenhuis, R. Hollingsworth; Anal. Biochem.; 208; 363-71 (1993). 45 K. Anumula; Anal. Biochem.; 230; 24-30 (1995). 45 J. Bigge, T. Patel, J. Bruce, P. Goulding, S. Charles, R. Parekh; Anal. Biochem.; 230, 229-38 (1995). 47 V. Ling, A. Guzzetta, E. Canova-Davis, J. Stults, W. Hancock; Anal. Chem.; 63; 2909-15 (1991). 48 P. Thibault, D. Watson, N. Young; In Proceedings of the 418t ASMS Conference on Mass Spectrometry and Allied Topics; San Francisco, CA, 613 (1993). 49 S. Carr, M. Huddleston, M. Bean; Protein Science; 2; 183 (1993). 50 K. Medzihradszky, C. Settineri, D. Maltby, A. Burlingame; In Techniques in Protein Chemistry IV; Angeletti, R. H.; Ed. Academic Press: San Diego, 117-25 (1993). 51 M. Huddleston, M. Bean, S. Carr; Anal. Chem. 65; 877-84 (1993). 52 S. Carr, J. Barr, G. Roberts, K. Anumula, P. Taylor; In Methods in Enzymology, Vol. 193; McCloskey, J. A.; Ed. Academic Press: San Diego, 501- 18 (1990). 53 K. Medzihradszky, D. Maltby, S. Hall, C. Settineri, A. Burlingame; J. Am. Soc. For Mass Spectrom.; 94; 350-8 (1994). -167- 54 T. Creighton; Proteins, Structures and Molecular Principles; W. H. Freeman and Co.; New York, 1984. 55 P. Edman, G. Begg; Eur. J. Biochem; 1; 80-91 (1967). 56 W. Pearson; In Methods in Enzymology, Vol. 183; Doolittle, R. F., Ed.; Academic: San Diego, 1990; pp. 63-98. 57 K. Biemann; Ann. Rev. Biochem.; 61; 977-1010 (1992). 53 T. Covey, R. Bonner, B. Shushan, J. Henion; Rapid Commun. Mass Spectrom.; 2; 249 (1988). 59 J. Loo, H. Udseth, R. Smith; Anal. Biochem.; 176; 404 (1989). 60 R. Smith, J. Loo, C. Edmonds, C. Barinaga, H. Udseth; Anal. Chem.; 62; 882-99 (1990). 61 A. Rockwood, M. Busman, R. Smith; Int. J. Mass Spectrom. Ion Processes; 111; 103-29 (1991). 62 M. Busman, A. Rockwood, R. Smith; J. Phys. Chem.; 96; 2397-400 (1992). 63 J. Loo, C. Edmonds, R. Smith; Anal. Chem.; 63; 2488-99 (1989). 64 B. Domon, C. Costello; Glycoconjugate J .; 5; 397-409 (1988). 65 A. Varki; Glycobiology; 3; 97 (1993). 66 V. Reinhold, B. Reinhold, C. Costello; Anal. Chem.; 67 ; 17 72-84 (1995). 67 M. R. Hardy, R. R. Townsend; Y. C. Lee; Methods Enzym. 179, 65-76 (1989). 63 L. J. Basa; M. W. Spellman; J. Chromatogr.; 499; 205-11 (1990). 69 K. Yamashida; T. Mizuochi; A. Kobata; Methods Enzymol.; 83; 105-26 (1982). 70 M. Spellman; Anal. Chem.; 62; 1714-22 (1990). 71 S. Hofl'stetteer-Kuhn; A. Paulus; E. Gassmann, H. M. Widmer; Anal. Chem.; 63; 1541-7 (1991). 72 R. A. Evangelista, M. Liu, F. Chen; Anal. Chem.; 67; 2239-45 (1995). 73 R. Townsend, M. Hardy, D. Cuming, J. carver, B. Bendiak; Anal. Biochem.; 182; 1-8 (1989). 74 L. Snyder; J. Kirkland; (1979) Introduction to Modern Liquid Chromatography, 2nd ed., pp. 281-9, Wiley, New York. 75 Y. Lee; Anal. Biochem.; 189; 151-62 (1990). 76 C. Yulp; V. Manam; R. Hepler; J. Hennessey; Anal. Biochem.; 201; 343-9 (1992). 77 Y. Lee in Methods in Enzymology (Ginsburg, V., Ed.) Vol. 28, pp. 63-7 3 (1972) Academic Press, New York. 73 G. Neuburger, D. Johnson; Anal. Chem.; 59; 150-4 (1987). 79 W. Lu, R. Cassidy; Anal. Chem.; 65; 2878-81 (1993). 80 L. Colon, R. Dadoo, R. Zare; Anal. Chem.; 65; 476-81 (1993). 31 T. O’Shea, S. Lunte; Anal. Chem.; 65; 948-51 (1993). 32 R. Rocklin, A. Clarke, M. Weitzhandler; Anal. Chem.; 70; 1496-1501 (1998). 33 H. Kwon; J. Kim; Anal. Biochem.; 215; 243-52 (1993). 34 P. Jackson; Anal. Biochem.; 216; 243-52 (1994). -168- 35 R. Pertel; Int. J. Mass Spectrom. Ion Processes; 16; 39 (1975). 36 R. Voykner, T. Pack; Rapid Commun. Mass Spectrom.; 63; 580 (1991). 37 V. Katta, S. Chowdhury, B. Chait; Anal. Chem.; 63; 174 (1991). 38 Fisons Platform Guide (1993). 39 C. Cresenzi, A. Di Corcia, S. Marchese, R. Samperi; Anal. Chem.; 67, 1968 (1995). 90 F. McLafi'erty; Registry of Mass Spectral Data. 91 V. Reinhold; B. Reinhold, C. Costello; Anal. Chem.; 1995; 67 ; 1772-84. 92 C. Barinaga, C. Edmonds, H. Udseth, R. Smith; Rapid Commun. Mass Spectrom.; 3; 160-4 (1989). 93 S. Carr, M. Hemling, M. Bean, G. Roberts; Anal. Chem.; 63; 2802-24 (1991). 94 R. Smith, J. Loo; R. Ogorzalek Loo; M. Busman, H. Udseth; Mass Spectrom. Rev.; 10; 249-51 (1991). 95 V. Datta, S. Chowdhury, B. Chait; Anal. Chem.; 63; 174-8 (1991). 96 C. Meng, C. McEwen, B. Larsen; Rapid Commun. Mass Spectrom.; 4; 151-5 ( 1990). 97 M. Barber, D. Bell, M. Morris, L. Tetler, M. Woods, J. Monaghan, W. Morden; Org. Mass Spectrom.; 24; 504-10 ( 1989). 93 R. Orlando, C. Fenselau; Org. Mass Spectrom.; 26; 664-6 ( 1991). 99 R. Orlando, R. Boyd; Org. Mass Spectrom.; 27; 151-5 (1992). 100 R. Pertel; Int. J. Mass Spectrom. Ion Processes; 16; 39 ( 1975). 101 C. Cresenzi, A. Di Corcia, S.. Marchese, R. Samperi; Anal. Chem.; 67 ; 1968 (1995). 102 R. Voykner, T. Pack; Rapid Commun. Mass Spectrom.; 63; 580 (1991). 103 V. Katta, S. Chowdhury, B. Chait; Anal. Chem.; 63; 174 (1991). 104 G. Neubauer, R. Anderegg; Anal. Chem.; 66; 1056-61 ( 1994). 105 P. Schnier, D. Gross, E. Williams; J. Am. Soc. Mass Spectrom.; 6; 1086-97 (1995). 106 S. Hunt, M. Sheil, M. Belov, P. Derrick; Anal. Chem.; 70; 1812-22 (1998). 107 D. Noble; Anal. Chem.; 66; 1239A-42A (1994). 103 R. Atlas; Microbiology; Macmillan Publishing Co.; NY; 757-71 (1984). 109 I. Vijn, L. das Neves, A. van Kammen, H. Franssen, T. Bisseling; Science; 260; 1764-5 (1993). 110 J. Denarie, J. Cullimore; Cell; 74; 951-4 (1993). 111 R. Dixon, C. Wheeler; Nitrogen Fixation in Plants; Chapman and Hall; NY (1986). 112 I. Tikhonovich, N. Provorov, V. Romanov, W. Newton; Nitrogen Fixation: Fundamentals and Applications; Dluwer Academic Publishers; Boston (1995). 113 Kloepper (1992). 114 Gunnera 115 Nap and Bisseling, 1990. -169- 115 A. van Brussel, K Recourt, E. Pees, H. Spaink, T. Tak, C. Wijfl'elman, J. Kijne, B. Lugtenberg; J. Bacteriol.; 172; 5394-401 (1990). 117 A. van Brussel, S. Zaat, H. Canter-Cremers, C. Wijfelman, E. Pees, B. Lugtenberg; J. Bacteriol.; 165; 517-22 (1986). 113 M. Bec-Ferte, H. Krishnan, D. Prome, A. Savagnac, S. Pueppke, J. Prome; Biochemistry; 33; 11782-88 (1994). 119 R. Carlson, S. Juan, U. Bhat, J. Glushka, H. Spaink, A. Wijfies, A. van Brussel, T. Stokkermans, N. Peters, G. Stacey; J. Biol. Chem.; 268; 1837 2-81 (1993). 120 P. Lerouge, P. Roche, C. Faucher, F. Maillet, G. Truchet, J. Prome, J. Denarie; Nature; 344; 781-4 (1990). 121 P. Mergaert, M. Van Montagu, J. Prome, M. Holsters; Proc. Matl. Acad. Sci.; 90; 1551-5 (1993). 122 R. Poupot, E. Martinez-Romero, J. Prome; Biochemistry; 32; 10430-5 (1993). 123 N. Price, B. Relic, F. Talmont, A. Lewin, D. Prome, S. Pueppke, F. Maillet, J. Denarie, J. Prome, W. Broughton; Mol. Microbiol.; 6; 3575-84 (1992). 124 J. Sanjuan, W. Carlson, H. Spaink, U. Bhat, W. Barbour, J. Glushka, G. Stacey; Proc. Natl. Acad. Sci.; 89; 8789-93 (1992). 125 M. Schultze, B. Quiclet-Sir, E. Kondorosi, H. Virelizier, J. Glushka, G. Endre, S. Gero, A. Kondorosi; Proc. Natl. Acad. Sci. 89; 192-6 (1992). 126 H. Spaink, D. Sheeley, A. van Brussel, J. Glushka, W. York, T. Tak, O. Geiger, E. Kennedy, V. Reinhold, B. Lugtenberg; Nature; 354; 125-30 (1991). 127 J. Denarie, J. Cullimore; Cell; 74; 951-4 (1993). 123 B. Bohool, E. Schmidt; Science; 185; 269-71 (1974). 129 J. Hamblin, S. Kent; Nature; 254; 28-30 (1973). 130 S. Puepppke, Plant Physiol.; 75; 924-8 (1984). 131 S. Vesper, W. Bauer; Appl. Environ. Microbiol.; 52; 134-41 (1986). 132 C. Benning; Annu. Rev. Plant Physiol. Plant Mol. Biol.; 49; 53-75 (1998). 133 P. van Rhijn, J. Vanderleyden; Microbiological Reviews; 59; 124-42 ( 1995). 134 M. Schultze, A. Kondorosi; Annu Rev Genet; 32; 33-57 (1998). 135 S Jabbouri, B. Relic, M. Hanin, P. Kamalaprija, U. Burger, D. Prome, J. Prome, W. Broughton; The J. of Biol. Chem.; 273; 12047 -55 (1998). 135 R. Poupot, E. Martinez-Romero, F. Maillet, J. Prome; FEBS Letters; 368; 536-40 (1995). 137 P. Lerouge, P. Roche, C. Faucher, F. Maillet, G. Truchet, J. Prome, J. Denarie; Nature; 344; 781-4 (1990). 138 J. Denarie, F. Debelle, J. Prome; Annu. Rev. Biochem.; 65; 503-36 ( 1996). 139 P. Mergaert, M. van Montague, M. Holsters; Mol. Microbiol.; 25; 811-7 (1997). 140 J. J oyard, E. Blee, R. Douce; Biochemica et Biophysica Acta; 879; 78-87 (1986). 141 C. Pugh, A. Roh, T. Hawkes, J. Harwood; Biochem. J .; 309; 513-9 (1995). -170- 142 M. Olsthoorn, I. Lopez-Lara, B. Peterson, K. Bock, J. Haverkamp, H. Spaink, J. Thomas-Oates; Biochem.; 37; 9024-32 ( 1998). 143 A. Roy, J. Harwood; Biochem. J .; 344; 185-7 (1999). 144 C. Benning, C. Somerville; J. Bacteriology; 174; 6479-87 (1992). 145 R. Carlson, J. Sanjuan, U. Bhatt, J. Glushka, H. Spaink, A. Wijfies, A. van Brussel, T. Stokkermans, N. Peters, G. Stacey; The J. of Biol. Chem., 268; 18372-81 (1993). 146 P. Mergaert, M. Van Montagu, J. Prome, M. Holsters; Proc. Natl. Aca. Sci. USA; 90; 1551-5 (1993). 147 R. Lee, D. Monsey, A. Weston, K. Duncan, C. Rithner, M. McNeil; Anal. Biochem.; 242: 1-7 (1996). 148 S. Hull, R. Montgomery; Anal. Biochem.; 222; 49-54 (1994). 149 H. Smits, A. Cohen, T. Buttler, J. Nielsen, L. Olsson; Anal. Biochem.; 261; 36-42 (1998). 150 J. Sanjuan, R. Carlson, H. Spaink, U. Bhat, W. Barbour, J. Glushka, G. Stacey; Proc. Natl. Acad. Sci. USA; 89; 8789-93 (1992). 151 M. Rossak, C. Tietji, E. Heinz, C. Benning; The J. of Biol. Chem.; 270; 25792-7 (1995). 152 N. Karamanos, A. Syrokou, P. Vanky, M. Nurminen, A. Hjerpe; Anal. Biochem.; 221; 189-99 (1994). 153 M. Stefansson, M. Novotny; Anal. Chem.; 66; 1134-40 (1994). 154 M. Stefansson, M. Novotny; Anal. Chem.; 66; 3466-71 (1994). 155 J. Sudor, M. Novotny; Anal. Chem.; 67 ; 4205-9 (1995). 156 M. Schultze, B. Quiclet-Sire, E. Kondorosi, H. Virelizier, J. Glushka, G. Endre, S. Gero, A. Kondorosi; Proc. Natl. Acad. Sci. USA; 89; 192-6 (1992). 157 R. Poupot, E. Martinez-Romero, J. Prome; Biochemistry; 32; 10430-5 (1993). 153 R. Poupot, E. Martinez-Romero, F. Maillet, J. Promet; FEBS letters; 386; 536-40 (1995). 159 P. Mergaert, M Van Montagu, J Prome, M. Holsters; Proc. Natl. Acad. Sci. USA; 90; 1551-5 (1993). 160 R. Carlson, J. Sanjuan, U. Bhat, J. Glushka, H. Spaink, A. Wijfies, A. van Brussel, T. Stokkermans, N. Peters, G. Stacey; The J. of Biol. Chem.; 268; 18372-81 (1993). 161 A. Lehninger; Principles of Biochemistry; Worth Publishers, Inc., NY, NY; p.348(1982) 162 J. McMurry; Organic Chemistry; Brooks/Cole Publishing Company; CA; 178 (1984). 163 A. Roy, M. Hewlins; Carbohydr. Res.; 302; 113-118 (1997). 164 F. Wang, L Sayre; J. Amer. Chem. Soc.; 114; 248-55 (1992). 165 J. March; Advanced Organic Chemistry; pp. 363-4; John Wiley & Sons; NY (1985). -171- 166 Gilbert; “Sulfonation and Related Reactions;” pp 136-48; Interscience, NY (1965). 167 R. Reed, H. Tartar; J. Am. Chem. Soc.; 57; 570 (1935). 168 J. McMurry; Organic Chemistry; p 300; Brooks/Cole; CA (1984). 159 J. Slagle, T. Huang, B. Franzus; J. Org. Chem.; 46; 3526-30 (1981). 170 R. Appel; Angw. Chem. Intemat. Edit.; 14; 801-11 (197 5). 171 Y. Lee; Anal. Biochem.; 189; 151-62 (1990). 172 173 R. Cedergren, R. Hollingsworth; ~172-