No description available.

No description available.

Understanding metabolic pathways is one of the most important fields in bioscience in the post-genomic era, but curating metabolic pathways requires considerable man-power. As such there is a lack of reliable experimentally verified metabolic pathways in databases and databases are forced to predict all but the most immediately useful pathways by inheriting annotations from other organisms where the pathway has been curated. Due to the lack of curated data there has been no large scale study to assess the accuracy of current methods for inheriting metabolic pathway annotations. In this thesis I describe the development of the Literature Metabolic Pathway Extraction Tool (LiMPET), a text-mining tool designed for the automated extraction of metabolic pathways from article abstracts and full-text open-access articles. I propose the use of LiMPET by metabolic pathway curators to increase the rate of curation and by individual researchers interested in a particular pathway. The mining of metabolic pathways from the literature has been largely neglected by the textmining community. The work described in this thesis shows the tractability of the problem, however, and it is my hope that it attracts more research into the area.

Binding sites for GnRH in the human placenta differ from the well-characterised pituitary GnRH receptor. The pituitary superagonists Buserelin and Tryptorelin were found by the placenta, as were the salmon and chicken-II isoforms of GnRH, however placental binding of the GnRH agonist analogues was not enhanced compared to the two isoforms, which were not bound by the pituitary receptor. Specific activity of binding of GnRH, GnRH isoforms and GnRH agonist analogues to the human placental binding site varied with stage of gestation. The high levels observed in early trophoblast (6-8 weeks) fell to almost non-detectable levels at 12-18 weeks, before increasing to maximal levels by term. Rebinding data suggested that differential tracer degradation could not account for gestational age-related variations in binding, and decreased binding was not related to contamination of the membrane fraction by non-villous tissue. Although the specific activity of binding to placental binding sites varied with gestation, specificity was unchanged: Tryptorelin, Buserelin, chicken GnRH-II, salmon GnRH > mammalian GnRH π lamprey GnRH, chicken GnRH-I. Buserelin, Tryptorelin and chicken GnRH-II were displaced from the membrane binding site by cytosolic extracts of placenta of all stages of gestation. Cytosol fractions from 12-18 weeks displaced iodinated GnRH tracers from the binding sites of membrane preparations from the same stage of gestation to a greater extent than an excess of non-iodinated Buserelin, but did not displace tracer from term membrane binding sites to the same degree. Membrane binding and the displacement effect of cytosol were reduced in the presence of protease inhibitors, however this was due to the presence of ethanol (as a solvent) in the inhibitor cocktail. The exact nature of this ethanol effect is unknown.

One approach to understand the underlying mechanisms for transcriptional repression by DNA methylation is through the study of mammalian methyl-CpG binding proteins. MeCP1 and MeCP2 are two methyl-CpG binding activities previously characterised in this laboratory. The molecular nature of MeCP1 is unknown, but the protein responsible for MeCP2 has been identified. Both <I>in vitro</I> and <I>in vivo</I> data strongly suggest that MeCP1 is a methyl-CpG dependent transcriptional repressor. MeCp2 is also a transcriptional repressor and contains a transcriptional repression domain. Recent efforts to identify other novel methyl-CpG binding proteins by searching the EST databases for proteins with methyl-CpG DNA binding domain of MeCP2 were successful. MBD1, MBD2 and MBD4 have been shown to bind specifically to methylated DNA <I>in vitro</I>. MBD2 is a DNA repair protein and therefore unlikely to be involved in transcriptional repression. The thesis describes collective studies of these methyl-CpG binding proteins (MeCP2, MBD1, and MBD2). MeCP2 is associated with the mSin3 corepressor complex which contains histone deacetylase subunits. Repression by the transcriptional repression domain of MeCp2 is sensitive to the histone deacetylase inhibitor, Trichostatin A (TSA), indicating that deacetylation is a critical component of the repression mechanism of MeCP2. Interestingly, MBD2 was found to be the DNA binding component of the long sought MeCP1 complex by several experimental criteria. MBD2 is also associated with histone deacetylases, and can repress transcription when tethered to the DNA. Repression of certain promoters by MBD2 can be relieved by treatment with TSA. MBD1 can also repress transcription and contains a potent repression domain. The repression by this novel domain was found to be sensitive to TSA, suggesting that deacetylation may again be involved. Altogether, these studies provide a molecular link to account for the long known relationship between DNA methylation, transcriptional repression and chromatin modification.

Magnesium transport across membranes was investigated in ferret red cells. Both Mg efflux and influx were measured. Mg transport was found to have the following properties. 1. Mg efflux from unloaded, untreated ferret red cells is 44 ± 2 μmol (1 cell)<SUP>-1</SUP>h<SUP>-1</SUP> (mean ± S.E.M. n = 38). 2. Mg efflux is partially inhibited by amiloride, quinidine, quinine, imipramine and divalent cations. Efflux is stimulated by SITS (4-acetamdio-4'-isothiocyanato-stilbene-2,2'-disulfonic acid) and unaffected by vanadate. 3. Mg efflux is stimulated by the depletion of cell ATP. Reducing the cell ATP content reduces the Mg buffering capacity of the cells and thus increases the internal free Mg concentration ([Mg<SUP>2+</SUP>]<SUB>i</SUB>). This increase in [Mg<SUP>2+</SUP>]<SUB>i</SUB> may account for some of the stimulation of Mg efflux. 4. Reducing external Na ([Na<SUB>o</SUB>]) from 145 to 10 mM (Na replaced with choline choride) stimulates Mg efflux. However Mg efflux is inhibited when [Na<SUB>o</SUB>] is below 10 mM. Net uptake of Mg is measured when [Na<SUB>o</SUB>] is maintained below 1 mM even if the extracellular medium contains only contaminant Mg (Mg is transported against its electrochemical gradient). Changes in the membrane potential are not responsible for this net uptake. These results show that the system is capable of reversing direction and mediating active transport. 5. Net Mg uptake in media with low [Na<SUB>o</SUB>] is stimulated when [Mg<SUB>o</SUB>] is greater than 1 mM. The stimulated uptake is partially inhibited by amiloride, quinidine, imipramine, NEM (N-ethylmaleimide) and cobalt. Net uptake is stimulated by vanadate and inhibited by reducing cell ATP.

Numerous hormones and neurotransmitters have been reported to modulate steroid from the adrenal cortex. In particular, acetylcholine (ACh) and angiotensin II (AII) stimulate both aldosterone and cortisol secretion from the adrenocortical zona glomerulosa (zg) and zona fasciculata/reticularis (zfr) respectively. This study has investigated the cellular mechanisms by which ACh and AII stimulated cortisol secretion from bovine adrenal zfr cells maintained in primary culture. The ability of ACh and AII to stimulate cortisol secretion from zfr cells was assessed over experimental days 1 (initial isolation of cells) to 5. Freshly isolated cells secreted cortisol in response to ACh and AII. This steroidogenic response was significantly reduced or absent on day 2, and recovered over days 3 and 4, at which time cells were maximally responsive to these agents. The reduced secretory response on day 2 was accompanied by a paradoxical increase in phospholipase C( PLC) activation, indicating that stimulation of PLC by ACh and AII becomes uncoupled from stimulation of steroidogenesis at this time. In contrast, steroidogenic and second messenger responsiveness to adrenocorticotropin (ACTH) and adrenaline, which activate adenylate cyclase, increased over experimental days 1 to 4. This loss of steroidogenic responsiveness to ACh and AII was shown to occur through a defect in the protein kinase C-mediated stimulation of the steroidogenic pathway. The receptor subtype(s) mediating the steroidogenic responses to ACh and AII were characterised using selective antagonists. Hexahydro-sila-difenidol and p-fluoro-hexahydro-sila-difenidol were potent competitive antagonists of ACh-stimulated cortisol secretion with pA<SUB>2</SUB> values of 8.68 and 7.96 respectively.

The Sec18 protein (Sec18p) of the yeast <I>Saccharomyces cerevisiae</I> has been identified as a component involved in the vesicular transport of proteins through the secretory and endocytotic pathways. Sec18p is a homologue of the mammalian protein NSF which has been shown, using a number of <I>in vitro</I> transport assay systems and affinity purification procedures, to interact with other proteins in a multisubunit protein complex. This work represents two approaches taken with the aim of identifying proteins that interact with Sec18p in the yeast <I>Saccharomyces cerevisiae</I>. Isolation of protein complexes was first attempted by affinity purification of a tagged version of Sec18p. The protein was C-terminally tagged with a protein A moiety from <I>Staphylococcus aureus</I> containing IgG binding domains. It was hoped that the affinity of protein A for IgG Sepharose could be used to isolate protein complexes that formed <I>in vivo</I> with the Sec18p. Although the fusion construct was shown to be active <I>in vivo</I>, specific complexing proteins could not be isolated due to the large amount of non-specific binding of yeast proteins to the protein A moiety. A second genetical approach was used where the <I>SEC18</I> gene was randomly mutagenised and yeast cells harbouring these mutagenized genes were screened for a dominant negative phenotype. Dominant negative mutant forms of the Sec18p interfere with the normal function of the wild-type protein <I>in vivo</I>. Five such mutants were isolated and classified into two main groups using a number of biochemical and morphological criteria. Class I mutants show a classical secretory mutant phenotype whereas the Class II mutant has a novel phenotype. A number of mutants in which the dominant negative phenotype was suppressed were isolated but the genes responsible for this phenotype could not be identified. It is hoped that alternative strategies can be employed in the future to identify extragenic suppressors of these mutants.

It was the aim of this thesis to characterise α2-adrenoceptor subtypes using radiolabelled agonist and antagonist ligands in a variety of tissue preparations. RS-15385-197 is a high affinity and selective α2-adrenoceptor antagonist, and the compound was shown, by the rank order of affinity of a number of competing ligands, to label α2A and α2B-adrenoceptor subtypes in human platelet and rat neonatal lung membranes, and a subtype in rat cortex which shows greatest similarity with the α2D-adrenoceptor subtype. Thus the receptor in rat brain was shown to form a distinct subtype. Differentiation of the α2-adrenoceptor into 2A and 2B subtypes could not be demonstrated however with the agonist ligand, [<SUP>3</SUP>H]-adrenaline, under normal assay conditions. The functional consquences of α2A-adrenoceptor activation were addressed in a model of α2-adrenoceptor mediated inhibition of cAMP accumulation. As a result of this work, [<SUP>3</SUP>H]-idazoxan an α2-adrenoceptor antagonist with an imidazoline structure, in addition to labelling α2-adrenoceptors, was also shown to label a population of imidazoline binding sites in rat kidney which were not adrenoceptors based on the low affinity of noradrenaline and RS-15385-197. Characterisation of these sites with [<SUP>3</SUP>H]-idazoxan and another imidazoline ligand [<SUP>3</SUP>H]-p-aminoclonidine suggested that the imidazoline sites labelled by these ligands were heterogeneous and were located over discrete areas of rat brain. As a direct consequence of this work a novel compound was identified which had greater than 10,000 fold selectivity for imidazoline sites over α2-adrenoceptors. In the hamster adipocyte, a tissue which I have shown previously to contain both α2-adrenoceptors and imidazoline binding sites, the inhibition of glycerol release by UK14304 was reversed by compounds showing selectivity for α2-adrenoceptors and not by imidazoline selective agents, suggesting that imidazoline sites are not involved in the UK14304 mediated inhibition of lipolysis in this tissue.

No description available.