The human saposins are four homologous activator proteins that are essential for the lysosomal degradation of sphingolipids (SLs) with small headgroups. They function in part to increase the solvent accessibility of these SLs to specific acid hydrolases by two proposed modes of action. In the solubilizer model, saposins extract SLs into soluble saposin-lipid complexes, and in the liftase model, the saposins reorganize lipids within membranes. The four saposins, called saposins A, B, C and D, have dual characters and exist as soluble proteins and in membrane-associated states. In this thesis, I first present the crystal structure of saposin D in a lipid-free state. The structure exhibits a closed, monomeric fold as previously described for saposins A and C. Next, I examine the lipid interaction properties of saposin A and determine a crystal structure of SapA in a lipid-bound complex. The complex forms a discoidal lipoprotein particle composed of highly ordered bilayer-like hydrophobic core surrounded by a protein belt consisting of two copies of saposin A in an open conformation. The saposin A-lipid discs are most likely the effective substrate-presenting particles in galactosylceramide hydrolysis. Finally, I compare the lipid interaction properties of the four saposins and address membrane perturbation, liposome binding, lipid solubilization and lipoprotein particle formation for each protein. Each saposin displays a unique behavior in the presence of liposomes under conditions that mimic the lysosomal environment. In particular, saposin D reveals simultaneous formation of different sized protein-lipid complexes, which appear to be primarily dependent on the lipid to protein molar ratio. A comparison of the available structures of saposins A, B and C in the “closed” and “open” conformations reveals structural hinge regions that likely shape the different types of saposin self-association. These states are directly related to the protein-lipid solubilization and/or membrane association properties of the saposins. Collectively, these findings present a more complete understanding of the lipid interaction properties of the saposin proteins and provide new insights into their role as activator proteins.

The human saposins are four homologous activator proteins that are essential for the lysosomal degradation of sphingolipids (SLs) with small headgroups. They function in part to increase the solvent accessibility of these SLs to specific acid hydrolases by two proposed modes of action. In the solubilizer model, saposins extract SLs into soluble saposin-lipid complexes, and in the liftase model, the saposins reorganize lipids within membranes. The four saposins, called saposins A, B, C and D, have dual characters and exist as soluble proteins and in membrane-associated states. In this thesis, I first present the crystal structure of saposin D in a lipid-free state. The structure exhibits a closed, monomeric fold as previously described for saposins A and C. Next, I examine the lipid interaction properties of saposin A and determine a crystal structure of SapA in a lipid-bound complex. The complex forms a discoidal lipoprotein particle composed of highly ordered bilayer-like hydrophobic core surrounded by a protein belt consisting of two copies of saposin A in an open conformation. The saposin A-lipid discs are most likely the effective substrate-presenting particles in galactosylceramide hydrolysis. Finally, I compare the lipid interaction properties of the four saposins and address membrane perturbation, liposome binding, lipid solubilization and lipoprotein particle formation for each protein. Each saposin displays a unique behavior in the presence of liposomes under conditions that mimic the lysosomal environment. In particular, saposin D reveals simultaneous formation of different sized protein-lipid complexes, which appear to be primarily dependent on the lipid to protein molar ratio. A comparison of the available structures of saposins A, B and C in the “closed” and “open” conformations reveals structural hinge regions that likely shape the different types of saposin self-association. These states are directly related to the protein-lipid solubilization and/or membrane association properties of the saposins. Collectively, these findings present a more complete understanding of the lipid interaction properties of the saposin proteins and provide new insights into their role as activator proteins.

No description available.

No description available.

An investigation into the structures of the antigen-binding fragments (Fab) of mouse monoclonal immunoglobulins by X-ray crystallography, is presented. The family of immunoglobulins studied, Gloops 1-5, possess the ability to bind to both the peptide antigen and the parent protein: Hen Egg-white Lysozyme. The Gloop1 and Gloop4 Fabs were generated by proteolytic cleavage of the antibody, and crystallisation trials yielded crystals of uncomplexed Fabs of both species. Attempts to grow Fab:antigen complex crystals proved unsuccessful. Partial purification of the heterogeneous Gloop4 Fab was found to be essential for the success of the crystallisations. Data were collected on crystal forms of Gloop1, Gloop2 and Gloop4 on an area detector. The structures of three Gloop2 crystal forms were solved by the molecular replacement technique, using models consisting of Fab domain pairs. Two of the crystal forms were refined by molecular dynamics methods at maximum resolutions of 3.3Å and 2.8Å, and the third by rigid body methods alone at 3.5Å resolution. Analysis of the Gloop2 structures between the crystal forms and with other Fabs indicated no atypical features in the domains. The elbow bend of the Gloop2 Fabs differed by up to 7°. The relative association of variable and constant domain pairs was also seen to vary between the Fabs. The changes in domain pairing caused significant differences in the relative disposition of the complementarity determining regions (CDRs), although there was no evidence for conformational change within individual CDRs. The Gloop2 combining site is dominated by a groove of approximate dimensions 12Åx9Åx7Å, containing many aromatic side-chains and a pair of glutamic acid residues in analogous positions to a those found in a FabrHen egg-white Lysozyme complex.

Thesis (Ph. D.)--University of Washington. / Bibliography: l. 157-163.

Thesis (Ph.D.)--University of Melbourne, School of Physics, 2006. / Typescript. Includes bibliographical references (leaves 151-159).

No description available.

The molecular structures of 18 Group VI organometallic compounds were determined by single crystal X-ray diffraction. Most of these complexes either contain the 'Cp'M(NO)' fragment (where Cp'=Cp(ɳ⁵-C₅H₅ ) or Cp⁻(ɳ⁵-C₅Me₅ ) ) or were derived from such compounds. The purpose of this work was to determine the structures of representative types of Group VI nitrosyl complexes and related compounds and to compare their features. Also, an effort was made to study any structural changes that accompanied reactions involving ancillary ligands. Fourteen of the structures that were determined contain a metal-nitrosyl linkage that is nearly linear in all cases. One of these compounds is the first example of a Group VI nitrosyl complex containing two different alkyl groups. The structures of two of its reaction products show that one of the alkyl groups can form a second metal to carbon bond via an orthometallation process. Five other nitrosyl alkyl complexes that were studied contain benzyl groups. In these structures, the benzyl ligands bond to the central metal atoms in ɳ²-type linkages, through both the methylene carbon and the ipso carbon atoms. In the compounds studied, this unusual mode of attachment is not influenced by the nature of other ligands on the metal or by methyl substitution on the phenyl ring. Other neutral nitrosyl complexes studied were the products of reactions involving Group VI diene compounds with acetone. Two structures are the result of similar insertion of an acetone molecule into one of the metal-diene bonds with the formation of a new metal to oxygen bond. The orientation of the diene ligands, however, is different in these structures due to steric interactions between these groups and cyclopentadienyl ligands in the complex. A third compound, produced upon prolonged reaction of the parent diene nitrosyl complex with acetone, has a trimeric structure containing bridging acetone groups and no diene ligands. Three cationic nitrosyl structures showed some small differences from the other nitrosyl complexes studied. The metal to nitrosyl linkages are nearly linear but the bond lengths are slightly different from those found in other compounds. Metal to nitrogen distances are slightly longer and nitrogen to oxygen bond lengths are slightly shorter in these cationic complexes. This observation suggests that in these cations, the metal to nitrosyl π backbonding is weaker than in other complexes. Three of the structures presented were not derivatives of nitrosyl containing complexes. One of the compounds was isolated upon oxidation of a dioxoalkyltungsten complex. Its structure is a rare example of an organometallic oxo peroxo complex. Reaction of this product with tetracyanoethylene gave a charge-transfer complex; structural studies of this compound identified a 2:1 organometallic:TCNE complex with interactions involving the peroxo groups of the organometallic molecules and the central ethylene carbon atoms of the TCNE molecule. Treatment of a dioxoalkyltungsten complex with a substituted isocyanate did not give the anticipated diamido compound. Instead, the starting complex reacted with three isocyanate molecules; the structure of the product is essentially that of a solvated diamido complex, with one amido group and a substituted urea ligand. In the nitrosyl complexes studied, a range of ancillary ligands (alkyl groups such as trimethylsilylmethyl, triphenylethyl, benzyl; diene ligands; chloro or acetonitrile groups) does not alter the linear nature of the metal-nitrosyl fragment. In all cases, the nitrosyl group is attached to the metal centre through only the nitrogen atom and the M-N-O bond angle does not deviate significantly from 180°. / Science, Faculty of / Chemistry, Department of / Graduate

Dimethyltellurium diiodide is known in two forms; the crystal structure of the β-form has been determined by X-ray diffraction; Mo-Kα scintillation counter data were used for this analysis in which the heavy atoms were located from the Patterson function, the carbon atoms by a difference synthesis and refinement was by least-squaresmethods. β-dimethyltellurium diiodide is ionic, [Me₃Te]⁺ [MeTeI₄]⁻, and is built up from trigonal pyramidal Me₃Te⁺ cations with Te - C = 2.07 Å, C - Te - C = 95°, and square pyramidal MeTeI₄⁻ anions with Te - C = 2.15 Å, Te - I = 2.84 - 2.98 Å. The ions are bridged by four weak Te... I interactions (distances 3.84, 3.88, 3.97, 4.00 Å), which complete a distorted octahedral environment around each tellurium atom. The crystal and molecular structures of 2-biphenylylferrocene and 4-biphenylylferrocene have been investigated in order to compare the configuration of the rings of the biphenyl and ferrocene groups in these two molecules. The structure of 2-biphenylylferrocenje, has been determined with visual Gu-Kα data. The iron atom position was found by Patterson methods, the carbon positions from successive Fourier summations. The positional and thermal parameters were refined by least-squares. The cyclopentadienyl rings are eclipsed, the first six-membered ring of the biphenyl group is rotated 43° out of the cyclopentadienyl plane and the outer six-membered ring is rotated 58° out of the plane of the first six-membered ring. These rotations relieve the strain which would exist in a planar model for the C₅H₄ . C₆H₄ . C₆H₅ group. The mean bond distances are Fe - G = 2.05 Å, C - C (cyclopentadienyl) = 1.44 Å. Using Fe-Kα scintillation counter data, the structure of 4-biphenylylferrocene has been determined by Patterson and Fourier methods and refined.by least-squares. The two crystallographically independent molecules in the unit cell have slightly different conformations. In one the cyclopentadienyl rings are oriented about midway between the eclipsed and staggered conformations and the first six-membered ring is rotated 6° out of the plane of the cyclopentadienyl ring to which it is bonded, with the second six-membered ring rotated a further 9°. In the second molecule the cyclopentadienyl rings are only about 5° form the fully eclipsed position and the six-membered ring rotations are 0° and 10°. The mean bond distances are Fe-C = 2.07 Å, C-C (cyclopentadienyl) = 1.48 Å, C-C (biphenylyl) = 1.43 Å, C-C (between rings) = 1.48 Å. The intermolecular separations correspond to, van der Waals' interactions. / Science, Faculty of / Chemistry, Department of / Graduate