Includes bibliographical references. / An RNA-dependent RNA polymerase from ribosomes of maturing chicken erythrocytes was investigated. Concurrent with an increase in globin and RNA synthesis during phenylhydrazine induced anaemia, the total and specific activities of ribosome bound RNA-dependent RNA polymerase increased 40 and 9.4 times respectively, indicating the possible involvement of this enzyme activity in the synthesis or control of synthesis of globin. A partially purified RNA-dependent RNA polymerase was prepared from ribosomes of immature chicken erythrocytes. This preparation was shown to be a predominantly primer dependent enzyme activity, incorporating UTP, CTP and to a lesser extent ATP and GTP into homopoly-ribonucleotide material. The ribosome-bound enzyme preparation demonstrated in addition, the synthesis of heteropolyribonucleotide material. A complementary DNA copy of chicken globin mRNA containing-9S RNP, was prepared. Unlabelled complementary DNA was used to demonstrate that no globin mRNA replicase activity was detectable in the ribosome fraction of immature chicken erythrocytes. The radioactively labelled product obtained after incubation of ribosomes with [³H]-UTP was shown by sucrose density gradient centrifugation to have a sedimentation coefficient of 4S; the ratio of non-terminal to terminal incorporation of UTP was 6.08. The product synthesized after incubation of intact cells in the presence of [³H]-uridine and actinomycin D was shown to be very similar with respect to S-value and the ratio of UMP/uridine incorporated. This similarity suggests the presence of a ribosome bound RNA-dependent RNA polymerase activity in intact cells.

Includes bibliographical references. / The antigenic determinants of the protein subunit of tobacco mosaic virus CTMV) have been studied by inhibition of complement fixation, inhibition of micro-precipitin and direct binding experiments. The viral subunit has been found to possess six antigenic determinants. Two of these, situated in tryptic peptide I (residues 1-41), were found to be a cryptotope (i.e. an antigenic determinant absent from the outer surface of the assembled viral capsid) and a viral antigenic determinant. Tryptic peptides 4 (residues 62-68) and 8(residues 93-112) also contained cryptotopes which were situated in the region of residues 63-65 and 108-112 respectively. Tryptic peptide 12 (residues 142-158) contained both a cryptotope and a neotope (i.e. an antigenic determinant which is only expressed on the outer surface of the viral capsid), which are situated in the C-terminal region of the polypeptide chain of the TMV protein, residue156 being associated with the cryptotope and residue 158 with the neotope. No antigenic activity could be demonstrated in tryptic peptide I I (residues 135-141 ). When the results are analyzed in terms of the three-dimensional structure of the viral subunit, it appears that all the antigenic reactive regions occupy highly accessible locations on the surface of the protein.

Includes bibliographical references. / This dissertation addressed the hypothesis that macrophages have an alternative killing mechanism that is independent of superoxide and nitric oxide but dependent on IFN-γ, TNF and C/EBPβ. Since the mechanism and the genes involved in this alternative pathway are mostly unknown, the aim of this dissertation was to identify these macrophage effector genes and to functionally characterize their role during infection utilizing gene deficient mouse models. Since mice deficient for C/EBPβ (C/EBPβ-/-) expressed normal levels of IFN-y and TNF during Listeria monocytogenes infection, the macrophage effector genes involved in confinement and killing of L. monocytogenes were postulated to be downstream of C/EBPβ. Furthermore, C/EBPβ-/- mice are highly susceptible L. monocytogenes due to impaired listericidal activity. Comparison of the gene expression profiles of WT and C/EBPβ-/- macrophages infected with L. monocytogenes was postulated to increase the probability of identifying these effector genes, which would be differentially expressed between the two groups. Comparative gene expression profiling by DNA microarrays between L. monocytogenes in infected WT and C/EBPβ-/- macrophages, successfully identified 1268 genes to be differentially expressed between the two groups. A focussed functional clustering strategy reduced the number of candidate genes to 220. PKCδ was selected for further study since it was involved in humoral defense, immune signalling, production of superoxide, regulation of transcription and may be putatively transcriptionally regulated by C/EBPβ. Furthermore, PKCδ was indirectly shown to promote L. monocytogenes escape from the phagosome and to negatively regulate transcription activity of C/EBPβ. In addition, since PKCδ was un-regulated, as shown by microarray and confirmed by RT-PCR, in L. monocytogenes infected C/EBPβ-/- macrophages, it was therefore thought to play a detrimental role during L. monocytogenes. However, since this premise has never been investigated directly, the role PKCδ during innate immunity against L monocytogenes was examined using the PKCδ deficient (PKCδ-/-) mouse model. Data in this dissertation provides new insight into the role of PKCδ during innate immunity to L. monocytogenes. PKCδ-/- mice were highly susceptible to L. monocytogenes due to enhanced listerial escape and impaired listericidal activity. Despite full macrophage activation and production of nitric oxide, PKCδ-/- mice displayed uncontrolled bacterial growth and dissemination of L. monocytogenes, which led to early death of the mice. In contrast, PKCδ-/- mice were able to control Mycobacterium infection as well as WT mice, suggesting that the activity of PKCδ may be negatively regulated by L. monocytogenes. A systems biology approach generated the hypothesis that PKCδ may promote Rab5a activation, which together with localized release of superoxide into the phagosome and activation of C/EBPβ by PKCδ, resulted in the confinement of the L. monocytogenes within the phagosome. Alternatively, PKCδ may act in a separate pathway that confines L monocytogenes within the phagosome, by activating and/or synergizing with unidentified proteins to neutralize that activity of listerial LLO and PI-PLC. Data in this dissertation clearly demonstrates that PKCδ is critical for confinement of L monocytogenes within phagosomes and may be part of a listericidal mechanism that is independent or nitric oxide, superoxide and pro-inflammatory cytokines.

Bibliography: leaves 128-146. / Cysteine proteinase inhibitors have shown potential as biocontrol agents for the protection of plants against insect and pathogen attack. With the advent of protein and genetic engineering such inhibitors can now be modified in order to improve their effectiveness. Because cystatins have already been isolated from plants. they provide a good starting point for developing modifications which may improve their function as biocontrol agents. The purpose of this project, therefore, was to design a potentially improved analogue of the rice cysteine proteinase inhibitor, oryzacystatin I, through molecular modelling studies. The gene sequence for this modified protein was then synthesised and expressed for kinetic analysis and insect trial assays. A prediction of the oryzacystatin I (OC I) tertiary structure was made using Biograf software on an Evans and Sutherland workstation. This structure was based on the known structures of stefin B and chicken cystatin ho se co-ordinates are published in the Brookhaven data files. Chicken cystatin is one of the most potent inhibitors of papain in the cystatin superfamily. This is believed to be due, in part, to an increased binding of the cystatin to papain through its amino-terminal region with the residues Leu7 to Gly9 playing a particularly important role.

The aim of this study is to establish a correlation between the physical maps of plasmid DNA (in the form of calorimetric profiles, thermal denaturation profiles and electron micrographs of partly melted DNA sequences) and genetic maps of these DNAs and thus deal with questions which were not answered by previous researchers. viz: Is there a correlation between base sequence function and a measurable physical property which can be assigned to biologically important units such as promoters or coding sequences? Is there a correlation between the denaturation of gene sequences and cooperative transitions observed in a given temperature interval? To answer these questions, a systematic study was initiated based on two families of plasmids with three different genes incorporated, namely the pGV 403 family which contains a Chloramphenicol resistance gene on one side and the pUC9 family which contains an Ampicillin and a Tetracycline resistance gene on the other side. Three different techniques were used to address the stated problems i.e. differential scanning calorimetry, high resolution thermal denaturation and electron microscopy. The reason for using three techniques instead of only one or two as in previous studies is that each technique gives specific results which can be supplemented by the other techniques and only in this way it will be possible to approach a deeper understanding of changes induced by perturbing the sequence based structural integrity by elevating temperature. In addition to measuring the experimentally observable parameters listed. a theoretical model was developed to predict the changes. This approach is termed local compositional complexity (LCC) analysis. The final goal of this investigation was to establish whether there is any correlation between the local compositional complexity and these selected genetic units. Based on the calorimetric experiments an improved table of thermodynamic data including the stacking energy for ten different combinations of basepairs is presented. The prediction of a melting curve based on primary structure information can be based on the enthalpy individual combination of basepairs[41.50.58]. The tables of the thermodynamic data published in the literature are given in Appendix D. In this thesis a slightly different approach to predict tm's was chosen (cf. p 67). The results obtained by this combined approach showed that there is indeed a correlation between the specific base sequence of a given plasmid DNA and its biologically important units (genes) and thus confirms that there is a semi-empirical correlation between genes and the observed cooperative melting units.

This Thesis will show that third strands can be incorporated into the four-way junctions combining the properties of the triple helices and those of branched structures within one system without compromising either of the two. It is now known that folding into secondary and tertiary structures by nucleic acids is crucial for their biological functions. However, what remain to be clarified are the mechanisms involved in the folding of nucleic acids into -noncanonical structures. It requires a thorough understanding of the chemical and physical properties of the structure in question. This in tum will improve the design of new secondary and tertiary structures that may add to the DNA nanotechnology. With this aim, thermodynamic and structural properties of two triple-helical DNA fourway junctions (JTIT3 and JT2T4) are reported and discussed. JTIT3 and JT2T4 differ only in the polarity of the third strands (and/or position of the loops). Both junctions contain the same Watson-Crick double-helical four-way junction, named Js, as a core structure. Js was constructed from four 20-mer oligomers, two of which consists of purines and the other two strands pyrimidines. Extending each of the pyrimidine strands of Js at the 3' end by four cytosines followed by twenty pyrimidine bases results in JTIT3. Similarly, JT2T4 is formed by extending the same pyrimidine strands at the 5' end. The junctions are named according to the position of the C4-loops. JTIT3 and JT2T4 are further simplified into JT1, JT2, JT3 and JT4. Lowering the pH from 12 to 2 allows the oligomers to fold sequentially from random coil into the double-helical four-way junction, Js, and finally into the triple-helical four-way junction. The analysis of the structures discussed is based on the biochemical methods such as native polyacrylamide gel electrophoresis and chemical footprinting using osmium tertroxide as a probe. The analysis is also based on the physical methods, UV spectroscopy and DSC. The native polyacrylamide gel electrophoresis has been used to verify the formation of the complete four-way junctions. Chemical footprinting has been used to detect the formation of the junctions as well as to indicate the conformations these junctions assume under different pH and/or salt conditions. The phase diagrams enthalpies and entropies of the structures are determined mainly by DSC. The results indicate that all the junctions are highly sensitive to salt concentrations and/or pH. The Tm vs. [Na+] results show that above 0.4M Na+, the structures adopt a conformation that suggests that the junctions are folded into stacked helix structures. The differences in thermal stabilities of the junctions JT1, JT2, JT3 and JT are due to the sequence composition of the arms and not the loops. JT1, JT2, JT3 and JT4 are thermally more stable than the underlying double-helical junction, Js. Similarly; the complete triple-helical four-way junctions JTIT3 and JT2T4 are thermally more stable than their substructures JT1, JT2, JT3 and JT4, The compiled transition enthalpies obtained for the individual arms of Js, JT1, JT2, JT3 and JT4 are less than the transition enthalpies associated with the melting of the complete junctions. The higher calorimetric enthalpies of the structures are due in part to the contribution from the single strands resulting from the partly unfolded arms. The overall results show that third strands have a stabilizing effect on the structure of the four-way junction.

Bibliography: pages 118-143. / The role that the sea urchin, Parechinus angulosus, embryo and sperm histone variants play in chromatin structure has been investigated. Chromatin structure has been determined at different levels of resolution in sperm and in developing embryos using micrococcal nuclease, pancreatic deoxyribonuclease (DNase I) and restriction endonucleases. Micrococcal nuclease and restriction endonuclease digestions of sea urchin gastrula chromatin have been analysed and it is shown that it is not possible to isolate large polynucleosomal chromatin complexes which are soluble in low ionic strength buffers. The nucleosomal DNA repeat lengths for sea urchin blastula, gastrula and sperm have been determined using micrococcal nuclease. The repeat length for sperm is significantly larger than blastula and gastrula repeat lengths whereas blastula and gastrula repeat lengths are not significantly different. Nucleosomal core particles have been isolated from early blastula, gastrula and sperm of sea urchins. After DNase I digestion of 51-labelled core particles the rate constants of cutting of the DNA at the susceptible sites on these core particles have been determined. The DNase I digestion kinetics of blastula and gastrula core particles are similar whereas sperm core particles are digested at a slower rate, mainly at the sites which are closest to the ends of the core particle DNA. Also, a site, which is 5 bases on the outside of the core particle and which is partially protected from nuclease attack, has been identified. The implications of these findings in relation to the histone variants in embryos and sperm of sea urchins are discussed.

Bibliography: pages 93-108. / The aim of this thesis was to isolate and characterize a homeobox-containing gene of the South African sea urchin Parechinus angulosus. This was achieved by constructing a genomic library of several individuals and screening this library using a probe containing the Antennapedia homeobox. Eight clones were isolated and shown to represent different alleles of the same gene. One clone was sequenced, revealing a homeobox which was termed PaHboxl. This homeobox was compared to published homeobox sequences and shown to be a member of the Antp (Hoxl.l) subclass (table 1.1). A splice donor site was identified 23 bp upstream of the homeobox and the observation confirmed by RNAase mapping. PaHboxl is situated in a genomic area showing a significantly higher degree of restriction fragment polymorphism than expected. This was shown by a statistical analysis which should be of general value in the interpretation of such polymorphisms. The expression of PaHboxl was examined by RNAase protection assays and Northern blotting. Two distinct phases of expression were observed - during embryogenesis PaHboxl is expressed transiently at low levels in 11,5 hr mesenchyme blastula stage embryos (44 ± 8 transcripts per embryo) with levels 3-5 fold lower 2,5 hr before and after this stage. Expression is observed again at up to 160 fold higher levels in the adult with maximal expression in testis (11 transcripts per 10 pg total RNA), and increasingly lower levels in intestines, ovary and Aristotle's lantern. Two transcripts of size 5,2 and 5,7 kbp were observed. Expression in Aristotle's lantern and embryonic stages could not be detected by Northern analysis.

. / The effect of different dietary fats on the serum cholesterol concentration has been discussed. In a series of experiments, conditions have been created under which rats can be induced to imitate humans in this respect. Using the rat as an experimental animal, the hypocholesterolaemic property of dietary sunflower seed oil has been investigated by the chemical and physical fractionation of this oil. By feeding the saponifiable and unsaponifiable matter of the sunflower seed oil, it was found that the latter fraction was partly responsible for the serum regulating property of this oil. This finding was confirmed by altering the composition of the unsaponifiable matter by the physical fractionation process of liquid propane segregation. Several fractions of oils with similar fatty acid compositions but differing in both quality and quantity of the unsaponifiable matter were found to have different effects on the serum cholesterol level.

Histones H2B have been purified from sperm of the sea urchin Parechinus angulosus and Psammechinus miliaris. Three H2B variants have been completely sequenced from P. angulosus and two H2B variants from P. miliaris have been partially sequenced. The sequences of the sperm histones H2B are compared to the known sequences Of other H2B histones. The sea urchin sperm histones H2B are all considerably more arginine rich than other histones H2B and contain an extended amino-terminal region containing reiterated pentapeptides.