The orthogonally protected trisaccharide benzyl 4-<i>O-</i>[4'-<i>O-</i>(2",3",4",6"-tetra-<i>O</i>-acetyl-b-D-glucopyranosyl)-2',3'-di-<i>O</i>-allyl-6'-<i>O</i>-benzyl-a-D-glucopyranosyl]-2,3,6-tri-<i>O</i>-allyl-b-D-glucopyranoside X has been synthesised from maltose and 1,2,3,4,6-penta<i>O</i>-acetyl-b-D-glucopyranoside in 11 steps with an average yield of 76% and a scale of one gram. Allyl ether protection could not be removed successfully so another target was designed. The orthogonally protected trisaccharide pent-4-enyl 2,3,6-tri-<i>O</i>-benzyl-4-<i>O</i>-{2',3'-di-<i>O</i>-benzyl-4'-<i>O</i>-(2",3", 4", 6"-tetra-<i>O</i>-acetyl-b-D-glucopyranosyl)-6'-<i>O</i>-pivaloyl-a-D-glucopyranosyl}-b-D-glucopyranoside X was synthesised on a scale of 2.6g from maltose and 1,2,3,4,6-penta<i>O</i>-acetyl-b-D-glucopyranoside in only 12 steps with an average yield of 86%. (Fig. 16083) This trisaccharide was sequentially deprotected in order to generate six partially protected analogues. Each of these analogues was oxidised selectivity using the catalytic oxoammonium radical 2,2,6,6-tetramethylpiperidinoxy (TEMPO) and the oxidant(s) sodium chlorite and/or sodium hypochlorite. The TEMPO oxidation was developed in both aqueous and biphasic conditions in order to enable the oxidation of substrates with hydrophilic, hydrophobic and intermediate polarities. Several oxidation protocols were developed and the reactions proceeded on mmolar and mmolar scales and could be monitored by HPLC. Six oxidised trisaccharides were synthesised including mono, di and tri-uronate species.

No description available.

No description available.

No description available.

Nitrile oxide-isoxazoline methodology has been employed in a novel convergent approach towards 2-ulosonic acid analogues, in particular those of 3-deoxy-D-<I>arabino</I>-2-heptulosonic acid (DAH) and 3-deoxy-D-<I>manno</I>-octulosonic acid (KDO). It involves regio- and diastereoselective [3+2] cycloaddition of a nitrile oxide to carbohydrate alk-1-enes containing four, five or six carbons, yielding 2-isoxazolines. Subsequent deprotection followed by reductive hydrolytic cleavage affords the 2-ulosonic acid analogues. The four carbon alkene 1,2-dideoxy-3,4-<I>0</I>-cyclohexylidene-D-<I>glycero</I>-1-enitol was chosen as a model alkene on which to establish the methodology which could then be applied directly to the two target classes of compounds. Cycloadditions were performed using four nitrile oxides: ethoxycarbonylformonitrile oxide, benzonitrile oxide, acetonitrile oxide and (diethoxyphosphoryl)acetonitrile oxide. These proceeded in 40-88% yield with moderate π-facial selectivity (54-65% dc) in favour of <I>Erythro</I> adducts, with the major product in each case possessing S-configuration at the new asymmetric centre, C-5. This selectivity can be rationalised in terms of the 'inside alkoxy effect' proposed by Houk <I>et al</I>, and the 'homoallylic' modification offered by De Micheli <I>et al.</I> In an attempt to control selectivity in the nitrile oxide cycloaddition reactions, a dispiroketal protecting group was utilised. However, the increase in steric bulk and presence of a six-membered ring as the protecting group incorporating the diol, rather than the five-membered ring, does not appear to influence π-facial selectivity. Cycloadditions were also performed using the five carbon alkenes 1,2-dideoxy-3,5-<I>0</I>-ethylidene-D-<I>erythro</I>-pent-1-enitol and 1,2-dideoxy-3,5-<I>0</I>-benzylidene-D-<I>threo</I>-pent-1-enitol. With the former alkene these proceeded in 40-80% yield but with minimal π-facial selectivity (4-6% DE). However, changing the homoallylic substituent from axial to equatorial in the later alkene significantly increased the selectivity (to 62-64% de). In both instances the major product was shown by X-ray crystallography to possess <I>S</I>-configuration at the new chiral centre (C-5) corresponding to a <I>threo</I> relationship about the C-5/C-6 bond. The corresponding acetate derivatives also yielded <I>threo</I> adducts preferentially.

A convergent route, based on nitrile oxide/isoxazoline chemistry to carbon-linked disaccharides (<I>C</I>-disaccharides) and <I>C</I>-glycosides with functionalised linkages has been investigated. The sequence involves three steps. Firstly, the cycloaddition of a sugar-derived nitrile oxide to a terminal alkene provides a 3,5-disubstituted 2-isoxazoline. Next, the substituents attached to the isoxazoline may be modified, and finally the heterocycle subjected to either hydrogenolysis or reductive cleavage to afford a <I>C</I>-glycoside with a carbonyl or an amino functionalised linkage. Four terminal alkenes were selected; two ω-unsaturated monosaccharide derivatives, methyl 5, 6-dideoxy-2,3-<I>0</I>-isopropylidene-α-D-<I>lyxo</I>-hex-5-enofuranoside (93) and 3-<I>O</I>-benzyl-5, 6-dideoxy-2, 3-<I>O</I>-isopropylidene-α-D-<I>xylo</I>-hex-5-enofuranose (84), and two non-carbohydrate model alkenes, methylenecyclohexane and styrene. The <I>in situ</I> Mukaiyama dehydration of the peracetylated 2,6-anhydro-1-deoxy-1-nitroalditol derivatives from nitromethane and D-xylose, D-galactose and D-mannose provided the source of nitrile oxides (79), (89) and (90) respectively. Cycloadditions of the xylopyranosyl and galactopyranosyl nitrile oxides (79) and (89) to alkenes (93) and (84) proceeded regiospecifically and with a high degree of π-facial selectivity (40-64% d.e.). In each case the major adduct was found to possess <I>R</I>-configuration at the newly formed chiral centre C-5, corresponding to an <I>erythro </I>relationship between it and the adjacent C-4 position of the attached furanose unit. The structure of the isoxazoline formed from xylopyranosyl nitrile oxide (79) and D-mannose derived alkene (93) was established by X-ray crystallography. The observed selectivies may be rationalised in terms of the "inside alkoxy effect" proposed by Houk and the "homoallylic" modification of De Micheli <I>et al.</I>

No description available.

We have developed a novel chemoenzymatic strategy to GDP-azidodeoxymannoses which differs from the natural substrates of mannosyltransferases of the Leloir type due to the presence of an azide group instead of hydroxyl functionality on the mannopyranose ring. We first chemically synthesized 2-, 3-, 4- and 6-azidodeoxy-α-D-mannose-1-phosphates, and then tested them as substrates for a recombinant GDP-Mannose Pyrophosphorylase from <i>Salmonela enterica </i>produced as polyhistidine-tagged fusion protein in <i>E. coli. </i>This enzymatic step yielded all the corresponding GDP-azidodeoxymannoses in 41%, 52%, 55% and 63% yields respectively, which were characterized by high resolution MS, ¹H- ¹³C and ³¹P NMR. Furthermore, we have optimized a high throughput colourimetric screen (via malachite green dye) which allows for indirect monitoring of this enzymatic reaction, and could be applied in future studies of enzymatic directed evolution to improve the efficiency of unnatural products formation. The GDP-azidodeoxymannoses so formed were tested as substrates for the soluble fragment of an α-1,2-Mannosyltransferase (Kre2p/Mnt1p) from <i>S. cerevisiae </i>which we produced in <i>E. coli. </i>Our preliminary data demonstrated that GDP-4-azidodeoxymannose was the most efficient donor substrate when transferred to a mannose α-methyl glycoside acceptor, as monitored by electrospray ionization mass spectrometry, and we foresee that the other isomers may find similar applications on further enzymes. Consequently, GDP-azidodeoxymannoses may serve as valuable tools for glycobiological studies using bioorthogonal azide chemistry.

No description available.

This thesis is concerned with the structural studies of Brucine and some adducts. The use of brucine as a resolving agent is well known, but the use of the brucine a co-crystallising agent for molecules which crystallise with difficult or not all is unknown. Carbohydrates are notorious for their difficult crystallisations if they can be crystallised at all. One aim of this thesis, explained in Chapter 1, is to investigate this use of brucine by crystallographic means. The experimental methods by which this was achieved are described in Chapter 2. Chapter 3 describes the simplest of the structures; the brucine hydrates. This comprises three brucine structures; the anhydrous, di-hydrate and tetra-hydrate. The description of these structures by means of crystallography, thermal analysis and powder diffraction gives a solid basis for further discussion. In particular, anhydrous brucine taken as a 'groundstate' is important in this aspect. Chapters 4 and 5 discuss the co-crystallisation of brucine with uronates and nucleic acids. These classes of molecules are treated as precursors of carbohydrates and show the tendency of brucine to form ribbons in crystalls. They crystallise readily with brucine. The conformation of each acid is compared to conformations found in other crystal structures of these acids and found to be related. Chapter 6 describes the dicarboxylic acid series (oxalic acid to glutaric acid) and derivatives pyruvic, lactic and oxamic acid. These acids were chosen to co-crystallise with brucine to investigate the effect of slight changes between them and on structural methods. Brucine is shown again to favour crystallisation in ribbons, although with a few notable exceptions. In Chapter 7 the same dicarboxylic acids are crystallised with strychnine to make a comparison between brucine and strychnine co-crystallisations. Here the predominant strychnine packing is a bilayer, but crystallisations with uronates and nucleic acids are not favourable.