Mouse Swiss 3T3 fibroblasts maintained in serum-free medium lose their actin stress fibres and focal adhesions. Addition of serum or lysophosphatidic acid (LPA) to these cells induced the rapid reformation of these structures, and the increased tyrosine phosphorylation of ppl25Fak and paxillin. The formation of focal adhesions and actin stress fibres was shown to require tyrosine phosphorylation; it was blocked by the tyrosine kinase inhibitor genistein, and stimulated by the tyrosine phosphatase inhibitors sodium orthovandate and phenyl arsine oxide. Moreover, the activity of a tyrosine phosphatase directed against ppl25Fak and paxillin was elevated in serum-starved cells. Despite the absence of focal adhesions and actin stress fibres, the adhesion of serum- starved cells was shown to be mediated by integrins, as peptides corresponding to the cell-binding motif of fibronectin caused cell detachment from the extracellular matrix. In addition, the integrin-mediated formation of actin stress fibres and protein tyrosine phosphorylation was blocked by introducing C3 transferase into cells, a result which establishes a role for the small GTP-binding protein rho in signalling via integrins. PKC was also required to maintain adhesion in serum-starved cells as the PKC inhibitor calphostin C caused cell retraction, a process that was accompanied by the formation of actin stress fibres. In contrast, activation of PKC with phorbol esters suppressed the LPA-induced assembly of actin stress fibres and focal adhesions, and caused loss of these structures in cells grown in serum. The results suggest that PKC negatively regulates the formation of actin stress fibres. The proposed role of vinculin phosphorylation in the formation of focal adhesions was found to be inconsistent with the finding that the phosphorylated form of the protein is not associated with the Triton X-100 insoluble fraction in Swiss 3T3 cells. Vinculin was a poor substrate for PKC in vitro, although a cryptic C-terminal phosphorylation site was exposed by PIP2.

Chlorella vulgaris, strain Brannon no.1 and Chlamydomonas reinhardi can convert acetate or ethanol into cell materials by the reactions of the glyoxylate cycle. In Chlorella this was found to operate during growth on these substrates but not during growth on glucose in the dark or on carbon dioxide in the light. This is reflected in the difference in the levels of malate synthase under these two growth conditions. However, the level of isocitrate lyase does not vary significantly according to the nature of the substrate, which is surprising, since in Chlamydomonas and most other organisms, glucose represses isocitrate lyase formation. The constitutive production of this enzyme in Chlorella cannot be explained by postulating a need to compensate for some enzymic deficiency in the tricarboxylic acid cycle or in routes leading to its replenishment, since these pathways were shown to operate quite normally in this strain. Partially purified isocitrate lyases from acetate- and glucose-grown cells were found to be so similar in charge, molecular weight, affinity for isocitrate and sensitivity to inhibition by phosphoenolpyruvate that they were considered to be identical molecules and not isoenzymes. However, the enzyme was found to be entirely soluble in glucose-grown cells, whereas in acetate-grown cells a large proportion of it was particle-bound. The excretion of glycollate by Brannon no.1 during growth on glucose in the dark provided evidence that the soluble enzyme is active in vivo even though it no longer has an anaplerotic function. The discovery of an organism unable to control the amount of isocitrate lyase formed but able to altar its spatial arrangement according to growth conditions suggests that there are two separate coarse control mechanisms associated with this enzyme, one controlling its level and the second determining its distribution.

Arthrobacter globiformis is able to utilize glycine as sole carbon and energy source for growth. It has been shown that the glycine is first converted to pyruvate by way of the serine pathway. Two of the enzymes in this pathway, namely the tetrahydropteroylglutamate-dependent glycine cleavage system and L-serine dehydratase, increase in activity as the bacterium adapts to growth on glycine. A. globiformis possesses a pyruvate carboxylase and a functioning tricarboxylic acid cycle. Therefore the pyruvate formed from glycine should serve both as a source of energy and of biosynthetic precursors. The pyruvate carboxylase shows an absolute requirement for catalytic amounts of acetyl CoA. It is cold labile and protected from cold inactivation by acetyl CoA. A. globiformis possesses a phosphoenolpyruvate carboxykinase which is considerably more active with inosine or guanosine nucleotides than with adenosine nucleotides. During growth on glycine and other substrates not degraded to phosphoenolpyruvate, the level of this enzyme is relatively high, suggesting that its function is probably to provide the phosphoenolpyruvate required under these growth conditions for gluconeogenesis and other biosynthetic purposes. The L-serine dehydratase extracted from A. globiformis is slowly activated by its substrate, L-serine, for which it exhibits a sigmoid saturation curve. The rate of activation of the enzyme is increased by the presence of Mg2+, and, most probably, by that of any one of a variety of other cations. The enzyme has been shown to indergo a reversible dimerization in the presence of L-serine. Kinetic studies suggest that dimerization of the enzyme is the rate limiting reaction in its activation by L-serine. In contrast to the extracted enzyme, the enzyme in toluene- treated cells shows no evidence of substrate activation. Most of the unusual properties of the extracted enzyme have been explained in terms of an extended form of the Monod-Wyman-Changeux model for allosteric proteins.

Axenically-grown myxamoebae of Dictyostelium discoideum strain. Ax-2, containing various amounts of glycogen, develop in a manner chronologically and morphologically similar to myxamoebae of Dictyostelium discoideum strain NC-4 grown on Aerobacter aerogenes or Escherichia coli. Isotopic and other studies indicated the presence of two incompletely separate pools of cellular glycogen during this development: (1) Myxamoebal glycogen; present in the cell cytoplasm in granular form unbounded by membranes and degraded immediately upon the onset of development, probably by combined amylase and maltase action, to products which are mainly oxidised to carbon dioxide. (2) Developmental glycogen; synthesised in an unknown cellular location during aggregation, regardless of myxamoebal glycogen content, using hexose derived at least in part via gluconeogenesis. Net degradation of this glycogen occurred only during sorocarp construction, possibly by the action of glycogen phosphorylase which reaches peak specific activity at this time. This data is inconsistent with the model of Wright et al (1968) and was discussed in this context. Whilst end-product saccharides (trehalose, a mucopolysaccharide, cell wall polysaccharide) may in part be synthesised from developmental glycogen, analyses revealed that myxamoebal glycogen can also act as precursor. Investigations of this system showed that the pathways of end-product saccharide synthesis are not coordinately controlled. Moreover, trehalose accumulation is regulated not by the cellular content of trehalose-6-phosphate synthase nor trehalase but rather by the cellular concentrations of UDP-glucose and glucose-6-phosphate, the substrates for trehalose synthesis. Although some myxamoebal glycogen can be used for end-product saccharide synthesis, most was oxidised to carbon dioxide probably with the production of metabolically useful energy, but this failed to inhibit developmental protein or RNA degradation and oxidation. It is therefore suggested that protein and RNA degradation is essential for successful development for reasons other than energy production. The remarkable versatility of metabolism during the developmental phase of Dictyostelium discoideum is discussed.

An E.coli Chloramphenicol Acetyltransferase Type III variant has been purified and characterized. The variant has been shown to be related to, but not identical, to the E.coli Type I and Type II variants. Inhibition studies with the reagents iodoacetamide and iodoacetate revealed that iodoacetamide was 15 times as effective as an inhibitor as was iodoacetate. Experiments with [14C] iodoacetamide gave an incorporation of approximately 1 mole of [14C] label per mole of enzyme monomer. Two unique radioactive peptides were isolated and sequence analysis indicated that a histidine and a cysteine residue were modified. The substrates, chloramphenicol and acetyl-S-CoA were both able to protect against the loss of activity. To test the hypothesis that a negatively charged residue was near the active site of the variant the 5,5 dithiobis (2-nitrobenzoic acid) analogues, 5,5 dithiobis (2-nitromethylbenzoate) and 5,5 dithio-bis (2-nitrotoluene) were characterized. The latter was found not to be suitable as an inhibitor but the former rapidly inhibited the enzyme. Chloramphenicol was able to protect against the loss of activity, supporting the hypothesis that a reactive sulphydryl group existed near or in the chloramphenicol binding site. The histidine residue was further investigated using the reagents diethylpyrocarbonate (DEP) and methyl nitrobenzene sulphonate (MNBS). DEP modified two histidine residues, one within the chloramphenicol binding site and one outside the substrate binding sites. Radioactively labelled MNBS was synthesised and analysis of the modified enzyme revealed the formation of 3-methylhistidine. The acetyl-S-CoA analogues, acetonyl-S-CoA and methyl-S-CoA, were synthesised and characterized. Methyl-S-CoA was found to accelerate the rate of inhibition seen with MNBS and increased the dissociation constant of chloramphenicol with respect to the enzyme. Preliminary studies were also carried out with the reagents rose bengal, butanedione and an epoxide antibiotic. The data suggests that a reactive histidine and cysteine residue exists within or near the active site of the variant. A proposed reaction mechanism with the histidine residue acting as a general base is postulated.

The discovery that the L-serine dehydratase (E.C. 4.2.1.13) from Escherichia coli B could be stabilised by iron and dithiothreitol has allowed this enzyme to be purified to a level approaching homogeneity. The purification scheme involved chemical treatments with streptomycin sulphate and ammonium sulphate, and elution from DEAE cellulose, pentyl agarose. Mono Q, Reactive Green and Phenyl superose chromatography columns. The purified dehydratase had a specific activity of 1653 ?moles of pyruvate min-l mg-1 enzyme and was, as judged by SDS PAGE, about 90 % pure. Overall recoveries were in the region of 4 % of the starting activity. A series of spectroscopic investigations and inhibitor studies showed that L- serine dehydratase did not utilise pyridoxal phosphate as cofactor. However, the purified enzyme did show an absolute requirement for iron and dithiothreitol for activity. The activation produced by these reagents was characterised and found to be slow, markedly influenced by both temperature and pH, and could be prevented, or reversed, by metal chelators, such as EDTA and o-phenanthroline. The activation process was also oxygen-dependent, and appeared to involve the production of an oxygen radical, since it was subject to inhibition by catalase and stimulation by hydrogen peroxide. Activation of L-serine dehydratase by iron and DTT also appeared to involve iron binding, at a ratio of 2 - 3 ?moles of Fe per ?mole enzyme. However, UV/visible and EPR investigations were unable to identify the structural form in which this bound iron existed. L-Serine dehydratase was found to be specific for L-serine; D-serine, L-threonine and L-cysteine were not deaminated. The timecourse of pyruvate formation was found to be non-linear, and the substrate saturation curve for L-serine sigmoidal, with an S[0.5] value of 2.6 mM, and a Hill coefficient of 2.13. The dehydratase could be activated by its substrate, L-serine, or substrate analogue, D-serine, which resulted in the production of a linear timecourse and hyperbolic substrate saturation profile (S[0.5] 2.8 mM, Hill coefficient 1.13). The molecular basis of this substrate activation process was investigated, and appeared to have its origins in a slow, serine-dependent rearrangement of the tertiary structure of the enzyme rather, than had previously been suggested from studies of the dehdyratase in crude extracts, a dimerisation reaction. In common with other microbial L-serine dehydratases, the purified E. coli B enzyme showed a broad pH optimum for pyruvate production, with maximal activity occurring between pH 7.8 and 8.2. It was inhibited competitively by L-cysteine and D-serine, with Ki values of 1.6 and 4.2 mM, respectively, and irreversibly by sulphydryl-active agents such as DTNB, N-ethylmaleimide and HgC12. In addition, the N-terminal amino acid sequence of the E. coli B L-serine dehydratase was analysed, and was found to show a high level of similarity with the predicted N-terminal sequence of the L-serine dehydratase from E. coli K12.

The Escherichia coli trp repressor is a member of the "helix-turn-helix" family of DNA-binding proteins, which is allosterically activated by the binding of L-tryptophan. In this thesis, I have used NMR spectroscopy to investigate the mechanism of allosteric activation of the trp repressor. I report here the results of the binding to trp repressor of tryptophan analogues such as 5-methyltryptophan, L- and D-tryptophan, which are corepressors, and inducers such as indole-3-propionic acid and 5-methylindole. A comparison of the mode of binding of indole-3-propionic acid with other analogues of tryptophan (inducers and corepressors) has also been carried out. The chemical shifts of the indole NH and 4-protons differs between corepressors and the inducer, indole-3-propionic acid. The pattern of intermolecular NOEs between protons of the repressor and those of the ligand also differ between the two classes of ligand. These results suggests that the orientation of the indole ring in the binding site differs by approximately 180 between the two kinds of ligand. The chemical shifts and intermolecular NOE patterns for the inducer, 5-methylindole, suggests that this ligand can bind in both L-tryptophan and indole-3-propionic orientations. In order to facilitate the determination of ligand induced shifts, protein-ligand NOEs and spectral analysis, I have also used isotopic labelling strategies such as random fractional and selective protonation to minimise spectral overlap. These labelling strategies will eventually be useful for NMR studies of the repressor-oligonucleotide complex, which has a Mr of 38 kDa and should also improve spectral resolution. However, for the purpose of efficient isotopic labelling, strong expression systems in E. coli are required. Therefore, I have constructed an improved T7 expression plasmid for producing isotopically labelled trp repressor.

Pacemaker cells of the mammalian sino-atrial node (SAN) contain a hyperpolarization-activated, non-specific cationic current, If which is an important component involved in the initiation and neurotransmitter-mediated control of cardiac rhythm. cAMP can directly modulate If by a mechanism independent of phosphorylation, demonstrating that cyclic nucleotide-sensitive ion channel genes are expressed within cardiac pacemaker cells. Through a combination of library screening methods based on cross-hybridising cyclic nucleotide-gated channel probes and a PCR 'fingerprint' employing primers designed to sequences encoding an ion channel cyclic nucleotide-binding domain, partial cDNA clones were isolated from a prepared sino-atrial node regional-specific cDNA library, which were either homologues of previously identified ion channels shown to be expressed in sensory tissues or putative new channel clones. Isolate rscNGC 1 following retrieval of a full coding region by anchor-PCR, demonstrated 90.4% sequence identity to the a-subunit of the rod photoreceptor cGMP-gated channel. The PCR 'fingerprint' identified a SAN homologue of the olfactory neuron cAMP-gated channel within library aliquots. This was the first demonstration that two distinct cyclic nucleotide-gated ion channel genes were expressed in SAN region of heart. Heterologous expression of rscNGC 1 following micro-injection of capped cRNA in Xenopus oocytes, gave rise to cGMP-stimulated channel activity exhibiting electrophysiological properties similar to the characterised a-subunit of the rod photoreceptor cGMP-gated channel. A reconstituted second messenger-pathway mediating endogenous receptor coupling to heterologously expressed cAMP-gated ion channels - shown to be present within native nodal tissue - was attempted within MEL cells. However, the absence of endogenous receptors positively-coupled to adenylyl cyclase within MEL cells, and the inability to functional characterise cAMP-stimulated cationic conductances via electrophysiological methods, prevented such studies. Thus demonstrating the inappropriateness of the MEL cell, as a heterologous system for studying receptor-mediated second messenger coupling to cNG channels. Although cyclic nucleotide-gated ion channels are obligatorily coupled to intracellular signalling agonists commonly found in heart, they have yet to be described in functional terms within SAN or any other cardiac subregion. It is postulated that they may have a role in vasculature - underlying mechanisms of smooth muscle relaxation.

Insulin-like growth factor binding proteins (IGFBPs) are important modulators of IGF action. It is becoming clear that they play an important role in processes such as differentiation, growth and development. Although the six IGFBPs show a high degree of homology, little is known of their structure, interactions or functions. IGFBP1 is produced in large amounts by uterine decidua in human pregnancy. The aim of the project was to identify its ligand binding domain and to pinpoint residues which are responsible for specific binding of IGFs. This was attempted using phage display, a relatively new technique, which couples mutagenesis to functional screening. In principle, this allows desirable mutants to be selected from large pools. The method is especially suitable for studying IGFBP1, since existing knowledge of its structure is inadequate for any strategy involving directed mutations. Mammalian and bacterial expression systems were evaluated using wild-type IGFBP1, in preparation for production of mutants selected by phage display, with a view to testing structure-function relationships of IGFBP1 in vitro and in vivo to elucidate its role in pregnancy. Wild-type IGFBP1 was displayed on fd-phage and retained its IGF-binding properties. Several schemes were devised to select for IGFBP1 molecules with altered affinities for IGFI and/or IGFII, but these proved unsatisfactory for selection of IGFBP1 as IGF appears to be sequestered within the binding protein. Various random mutagenesis methods were unsuccessful, probably due to the high GC content of IGFBP1 cDNA. Hence, although the combination of random mutagenesis and phage display is a powerful technique for the screening and selection of large numbers of mutants, the technical difficulties could not be resolved in the time available.

Alcaligenes eutrophus NCIMB 40529 was grown in continuous culture (D=0.1h-1) under glucose, ammonium, potassium and oxygen limitation. The rates of PHB production (qPHB) were highest during ammonium-limited growth and lowest during glucose-limited growth. The PHB content of cells grown under ammonium limitation exhibited an inverse relationship to the dilution rate (D), whilst the rate of substrate (glucose) utilisation (qS) increased with D. It was concluded that glucose is utilised increasingly for the production of non-polymer biomass (NPB) as D increases, with the resulting diminution of PHB synthesis. PHB production from alternative carbon substrates was also studied in ammonium-limited chemostat culture (D=0.1h-1). PHB content, the rate of PHB production (qPHB) and qS varied substantially with the nature of the carbon source in the order lactate pyruvate glucose/gluconate fructose. qPHB was not related to umax in batch culture or to the relative oxidation-reduction states of the substrates, but was directly related to the qSsurplus i.e. the rate of substrate utilisation above that required solely for the production of NPB. The substantially higher qPHB from lactate than from glucose in ammonium limited chemostat culture reflected the inability of the cell to regulate lactate utilisation at D umax, whereas glucose utilisation is partially repressed under these conditions. Washed cells prepared from, glucose- and lactate-limited chemostat cultures (D=0.1h-1) metabolised glucose and lactate respectively at significantly faster rates than the cultures from which they were derived. Both types of washed cells synthesized PHB, indicating that the biosynthetic enzymes are constitutive; this was confirmed by assay of a representative biosynthetic enzyme in broken cells prepared from carbon- and ammonium-limited chemostat cultures. Washed cells also excreted pyruvate. It is suggested that carbon-limited cells are derepressed for carbon substrate uptake and catabolism, and that pyruvate excretion reflects the limited capacity of non-growing cells for pyruvate utilisation. Citrate synthase, a potentially important regulatory enzyme of the tricarboxylic acid cycle in A. eutrophus, was purified. The enzyme consisted of a single type of subunit with a Mr of 47,000, and exhibited a native Mr of 227,000. Citrate synthase activity displayed a hyperbolic dependence on both oxaloacetate (Km=1.2mM) and acetyl Co A (Km= 94uM) concentration, and a sigmoidal dependence of inhibition by NADH (I50=0.62mM). It is concluded that, apart from its insensitivity to AMP, the enzyme is similar to citrate synthases purified from other Gram-negative aerobes.