The aim of this work has been to use linkage analysis to determine the chromosomal location of genes involved in chloroquine resistance in the rodent malaria parasite <I>Plasmodium chabaudi</I>. The <I>P. chabaudi</I> genome was found to contain 14 chromosomes. A genetic map of each chromosome was made using DNA markers. Most of the markers were known genes from other species of <I>Plasmodium</I>. Other markers were developed by the RAPD-PCR (random amplified polymorphic DNA-polymerase chain reaction) technique, the first time this method has been developed for use with <I>Plasmodium</I> parasites. In total, more than 100 markers were mapped to individual <I>P. chabaudi</I> chromosomes. Two important markers were cloned from <I>P. chabaudi</I> DNA using PCR. (i) the <I>P. chabaudi</I> homologue (<I>pcmdr1</I>) of the multiple drug resistance gene of <I>P.falciparum (pfmdr 1</I>); this has been implicated in the mechanism of chloroquine-resistance in <I>P. falciparum</I>. (ii) a possible homologue of the <I>P. falciparum</I> marker pS590.7, which has been claimed to be linked to a chloroquine-resistance locus in <I>P. falciparum</I>. 13 genetically distinct clones from the cross were phenotyped for their susceptibility to chloroquine. 8 were found to be resistant and 5 sensitive. These clones were analysed for their inheritance of 46 polymorphic markers. This revealed that neither <I>pcmdr 1</I> nor the putative ps590.7 homologue were linked to chloroquine resistance in this cross. 12 of the 13 progeny showed linkage disequilibrium with a locus on chromosome 11. Statistical analysis showed the linkage to be significant, with a probability of 0.05>P>0.03. The work submitted here represents the first in-depth genetic analysis of a <I>P. chabaudi</I> cross and identifies a locus which may be involved in the genetic mechanism of chloroquine resistance.

The immediate aims of the project, as set out in the introduction, were 1) to separate the <i>lacZ</i> and <i>lacY</i> genes of the lactose operon such that they could be controlled/induced independently 2) to maintain the expression construct in the <i>E.coli</i> chromosome. The <i>lacY</i> gene was subcloned into plasmid PBN372 downstream of the <i>S.marsescens trp</i> promoter. The flanking <i>E.coli trp</i> genes were exploited to integrate the construct into the <i>E.coli</i> chromosome at the <i>trpB</i> locus via homologous recombination. Homologous recombinants should be <i>trp<SUP>{-}</SUP></i>. Two approaches were employed to achieve integration: 1) transformation of a <i>recD</i> strain (which was also a <i>lacY</i> deletion) 2) transduction with phage lambda. The first method was unsuccessful, only spontaneous <i>trp</i><SUP>-</SUP> mutations were isolated. The second method yielded several integrants, one of which was used in subsequent growth experiments. Since the constructed strain was rendered <i>trp</i><SUP>-</SUP>, the internal tryptophan concentration could be influenced by the concentration of tryptophan in the medium and the level of induction could be set by the addition of differing amounts of the antirepressor, IAA. The growth rate of the constructed strain in minimal lactose media was comparable with that of wild type <i>E.coli</i>. The permease activity of the constructed strain was seen to vary when assayed in the presence of varying amounts of IAA. The expression of permease was also demonstrated to be independent of galactosidase activity. The constructed strain therefore met all the initial requirements for the experimental system set out in this thesis. Difficulties were encountered during the analysis of permease induction in batch culture. This was due to the competition between antirepressor (IAA) and corepressor (tryptophan). These difficulties would have been minimal had the analysis been carried out in chemostat experiments. It would then have been possible to maintain a constant, low level of tryptophan throughout the experiment. In batch culture, the tryptophan level is constantly changing, initially being relatively high and becoming depleted as the experiment progresses.

To gain further insight into the pre-rRNA processing and maturation pathway, I identified three putative ribosome synthesis factors, <i> NOP53, NOP9</i> and <i>SDO1 </i>through a bioinformatics approach. In cells depleted of Nop53p, synthesis of mature rRNA components of the 60S ribosomal subunit is severely inhibited. This aberration causes defective pre-60S subunits to accumulate in the nucleus. These particles are likely to contain the 7S and 25S’ pre-rRNAs, both of which strongly accumulate in a Nop53p depleted strain. Together this suggests that Nop53p is required for pre-60S particles to gain export competence. Nop53p has been identified as a protein that co-precipitated with Trf4p and Mtr4p, components of the TRAMP complex, which function together with the exosome in the surveillance and degradation of defective pre-ribosomes. It appears that transcripts which accumulate in Nop53p depleted cells are not subject to polyadenylation, which we hypothesise prevents efficient targeting of late pre-60S ribosomes for degradation by the TRAMP/exosome system. Nop9p is a nucleolar protein that co-purifies with the 20S pre-rRNA. On depletion of Nop9p the synthesis of mature 18S rRNA is severely inhibited due to a delay in early cleavage steps, whereby 20S fails to by synthesised.  Movement of 90S and/or pre-40S particles form the nucleolus to the nucleoplasm and cytoplasm is inhibited on depletion of Nop9p. Nop9p contains a pumilio-like RNA binding motif and <i>in vitro</i> binding assays demonstrate that Nop9p is capable of binding RNA. Sdo1p was identified as being required for the processing of the 35S pre-rRNA transcript. The human homologue of Sdo1p, <i>SBDS</i>, encodes a protein mutated in the marrow failure condition, Shwachman-Bodian Diamond Syndrome (SBDS). Depletion of Sdo1p results in a short delay in the synthesis of 5.8S rRNA.

<I>Eco</I>KI cuts or modifies DNA according to the methylation states of specific adenine residues in its target recognition site. If these residues are unmethylated the DNA is cut, often thousands of base pairs away from the recognition site. Communication between <I>Eco</I>KI bound to the recognition site and the cleavage site occurs by DNA translocation, which is driven by ATP hydrolysis. The amino acid sequence of the subunit required for restriction contains seven motifs that are conserved in the DEAD box family of proteins. This family is a sub-group of the superfamily of DNA and RNA helicases. Previous studies on DEAD box proteins have found these motifs are involved in the ATPase and helicase activities of these proteins. To assess the importance of the DEAD box motifs in the restriction of DNA by <I>Eco</I>KI, amino acid residues in each of the seven motifs were changed and the effects of these substitutions on restriction were investigated. Eight proteins, each containing a different amino acid substitution, were purified and the DNA binding abilities, nuclease activities and ATPase activities of the proteins were studied. All changes had an effect on restriction except those changes in motif IV. Motif IV was defined prior to the discovery of a frame-shift in the <I>hsdR </I>DNA sequence and these results suggest it has been incorrectly identified. An A619G substitution in motif III slightly impaired restriction, but other substitutions at this position (A619D and A619V) abolished restriction. All other changes prevented any DNA restriction. The nuclease assays with the purified proteins confirmed the <I>in vivo </I>results. None of the available evidence indicates that the amino acid substitutions prevent interaction with ATP. The protein with a D577H substitution in motif II showed a reduced DNA binding ability, none of the other changes investigated affected DNA binding.

Lines of mice carrying an ovine β-lactoglobulin (BLG) transgene secrete large quantities of BLG protein into their milk. To explore the stability of transgene expression, a systematic study of expression levels in three BLG transgenic mouse lines was performed. Unexpectedly, two of these lines exhibited variable levels of transgene expression. Copy number within lines appeared to be stable and there was no evidence of transgene rearrangement. Studies on the most variable line showed that BLG production levels were stable within individual mice in two successive lactations. Backcross experiments demonstrated that the genetic background did not contribute significantly to the variation of expression levels. Tissue <I>in situ</I> hybridisation experiments revealed mosaic patterns of transgene expression within individual mammary glands from the two variable lines; in low expressing animals, discrete patches of cells expressing the transgene were observed. The concentration of transgene protein in milk reflected the proportion of mammary epithelial cells expressing BLG mRNA. Furthermore, <I>in situ</I> hybridisation to metaphase chromosomes indicated that the transgene arrays in both these lines are situated close to the centromere. These experiments suggest that variable mosaicism of transgene expression among individuals within a transgenic line is a consequence of the chromosomal location and/or the nature of the primary transgenic integration event.

The main aims of this work to set up two new mouse models, by conventional and conditional gene targeting, to investigate the functions of ERCC1, which has not, thus far, been associated with any known human disorders. To facilitate gene targeting studies, a series of overlapping ERCC1 genomic clones were retrieved from a bacteriophage library. Analysis of the mouse ERCC1 5'-flanking region revealed a novel transcription initiation site. Furthermore, several potential liver- and brain-specific transcription factor binding sites were present. A novel ERCC1-deficient mouse line, designated ET#209, was generated by the targeted deletion of the ERCC1 exon 3-5 region. In addition to the aberrant nuclei in the KT#209 ERCC1-/- livers, as reported in previous ERCC1 knockouts, we also found microvesicular steatosis and mitochondrial abnormalities in livers of both our ERCC1-deficient lines. These findings suggested that oxidative stress may damage nuclear and/or mitochondrial DNA, leading to the abnormal lipid deposits. We also found that other repair proteins (MSH2, PMS2 and PCNA) and an immediate early gene product, c-fos, were elevated in both ERCC1-deficient liver extracts. A delay in cerebellar development was found in both our ERCC1-deficient lines and the possibility of a cerebellar disorder was investigated. Also, we have demonstrated that some thymocyte markers were changes in flow cytometric profiles, suggesting a possible role for ECC1 in V(D)J recombination. To overcome the limitations of premature death in the ERCC1-deficient mice, we used a conditional gene targeting strategy. We reconstructed a functional ERCC1 allele with loxP sites flanking exons 3 and 5. In vitro analysis demonstrated that the ERCC1 exon 3-5 region between the loxP sites was deleted after transient expression of Cre recombinase. We have generated mice with this floxed ERCC1 allele, which will be used for Cre/loxP-mediated recombination in a tissue- and/or developmentally-specific manner.

The <i>ftsK </i>gene of <i>Escherichia coli </i>encodes a 147kDa peptide, which consists of three distinct domains; namely a conserved N-terminal domain, a proline and glutamine rich region, and a C-terminal domain with consensus nucleotide-binding pocket (Begg <i>et al., </i>1995). The N-terminus is essential for cell division (Begg <i>et al.,</i> 1995; Draper <i>et al</i> 1998; Wang and Lutkenhaus 1998) and targets the whole protein to the septum (Yu <i>et al., </i>1998 and Wang and Lutkenhaus, 1998). It was found in this thesis that depletion of the C-terminal domain resulted in the appearance of cells with abnormally located chromosomes and also in cell division-dependent SOS induction. Further study in this thesis showed that it was required only for chromosomal dimer resolution. The phenomenon of cell division-dependent SOS induction in FtsK-depleted GLC600 cells suggested that chromosome partition mutants should have continuous SOS induction. A study of the mutants obtained in this way showed that the RuvABC proteins, which are required for the resolution of recombination intermediates, are involved in chromosome partition. Over-expression of RuvAB could cause inhibition of cell division even in <i>recA^-</i> cells, slow down the elongation of chromosome replication and cause the repression of transcription of cell division genes in the 2min region of the <i>E. coli</i> chromosome. To confirm that cell division and chromosome replication are coupled, nalidixic acid treatment and thymine starvation were used to block chromosome replication and it was found that cell division was coupled with chromosome replication at the level of transcription. It was also found that over-expression of DsbB, a protein required for protein disulfide bond formation in <i>E. coli </i>(Bardwell <i>et al., </i>1993), resulted in the loss of the rod shape of <i>E. coli </i>cells. Over-expression of another gene named <i>friL </i>caused inhibition of cell division at a very early stage. It is proposed that FriL might be a regulator of cell division.

Elastase is irreversibly inhibited by compounds incorporating a β-lactam ring and this thesis describes structural studies on three classes of β-lactam inhibitor. The mechanism of inhibition begins with nucleophilic attack by the catalytic serine on the carbonyl which is at position 2 of all the β-lactams. In all cases this results in the opening of the ring between position one and two and the departure of the position four substituents. X-ray crystal structures and mass spectra obtained of elastase in complex with each of the ligands described the bound products of each reaction. β-lactams having geminal diethyl groups at position 3 of the ring were proven to react by a different mechanism to those with a hydroxyl ethyl at position 3. Hydroxy ethyl ligands are proposed to result in amore stable bound product. A ligand which had an ethyl group attached by a double bond at position 3 was found to be hydrolysed by the enzyme to give a breakdown product which remains non-covalently bound in the active site. A crystal structure of the non-covalently bound product was determined. The determination of 3-D structures of elastase with gold triethyl-phosphine and rhenate is also reported. Factor XIII is the last enzyme in the blood coagulation cascade, stabilising blood clots by cross-linking fibrin. It is a transglutaminase, linking a lysine from one peptide substrate to a glutamine from the other via e(γ-glutamyl)lysyl isopeptide bonds. In plasma, factor XIII exists in the inactive form as a heterotetramer, having two A subunits and two B subunits. Upon activation the two B subunits fall away and the A-A dimer performs the catalysis. The activation process also involves an unknown conformational change and thrombin cleavage which allows the departure of an activation peptide. Attempts made to purify the active form of the enzyme for crystallisation are described. A crystal structure of the inactive form of factor XIII has been used as the basis for computer modelling of the interaction of peptide ligands and the active form of the enzyme.

PRP8 protein is a component of the nuclear pre-mRNA splicing machinery. The <i>PRP8</i> gene had previously been cloned from yeast and antibodies raised against the protein. Using these it had been shown that PRP8 protein is a component of the U5 snRNP, U4/U6.U5 tri-snRNPs and of the spliceosome, in which it contacts the substrate RNA. In this thesis I present further work on the characterisation of the roles of the PRP8 protein in splicing complex assembly and splicing using these antibodies and two other approaches: genetic depletion of the protein <i>in vivo</i> and heat-inactivation of temperature-sensitive forms of the protein. The <i>in vivo</i> depletion of an intrinsic snRNP protein has not previously been reported and this approach allowed a more definitive investigation of the function of PRP8 protein.(i) Antibodies which recognise the native PRP8 protein inhibited splicing <i>in vitro</i> and this was shown to be due to a block in spliceosome assembly. A pre-spliceosome complex containing the U1 and U2 snRNPs accumulated in splicing reactions inhibited by these antibodies. Anti-PRP8 antibodies also detected a change in the interactions of PRP8 protein in the spliceosome as the active complex formed.(ii) A yeast strain conditionally producing PRP8 protein was generated, and several temperature-sensitive <i>prp8</i> mutations were outcrossed from mutagenised backgrounds and the mutations mapped within the genes. Extracts in which PRP8 protein was either depleted or heat-inactivated were made from these strains and shown to be inactive for splicing. Splicing reactions carried out with these extracts accumulated a pre-spliceosome complex similar to that seen when PRP8 function was blocked by antibodies.(iii) Depletion and heat-inactivation of PRP8 protein were shown to result in loss of U4/U6.U5 tri-snRNPs, consistent with a requirement for PRP8 activity for the stable formation of tri-snRNPs, without which spliceosomes fail to form. As a consequence of PRP8 depletion the levels of U4, U5 and U6 snRNAs declined dramatically. From this result, and known genetic interactions between <i>PRP8</i> and several putative RNA helicase genes, it was proposed that without PRP8 activity aberrant tri-snRNPs form, on or in which helicase activities act to unwind RNA structures, thereby exposing the snRNAs to the action of nucleases.(iv) <i>In vivo</i> depletion of U5 snRNA was found to have little effect on either PRP8 protein or the U4 and U6 snRNAs; in fact, the levels of U4 and U6 snRNAs rose as U5 became depleted. This indicated that the loss of U4 and U6 snRNAs seen on depletion of PRP8 protein was not a consequence of loss of U5 snRNAs and favoured the model proposed in (iii). These results together with previously published data suggest that PRP8 protein activity is required for : 1) the formation of complete and functional U5 snRNPs; 2) the formation of stable tri-snRNPs; 3) interaction of tri-snRNPs with the pre-spliceosome; 4) the regulation of helicase activities associated with the tri-snRNP and the spliceosome.

The murine casein locus has been characterised by long range restriction mapping and the analysis of long fragment genomic clones. Cloned sequences from five mouse casein genes (α, β, γ, ?, κ) were used to screen a murine (strain 129) genomic library in a bacterial artificial chromosome vector (BAC). Of the nine clones isolated, two contained 3 casein genes α, β, γ and γ, ?, κ respectively. The following combinations were found in other clones - α + β; β + γ; γ + ?. Thus, I deduced that the gene order in the locus is α- β- γ- ?-κ. I have confirmed this order by restriction analysis of the clones. Expression studies of casein genes at various time points during pregnancy and lactation revealed a co-ordinate expression pattern for the three (α, β and γ) calcium sensitive genes from mid pregnancy to parturition. The genetic variation in the casein loci of <I>Mus musculus </I>(eight different strains) and <I>Mus spretus </I>was also examined as restriction fragment length variations (RFLV) using five restriction enzymes, <I>Bam</I>HI, <I>Sfi</I>I, <I>Hind</I>III, <I>Eco</I>RI and <I>Xba</I>I. The information obtained from the present mapping study and the clones isolated (129 strain) can be used to manipulate the casein locus in embryonic stem (ES) cells as most gene targeting experiments are carried out in ES cells isolated from 129 mice. Gene targeting is more effective if the targeting DNA is prepared from isogenic DNA. It is also advantageous to use isogenic DNA derived maps in targeting experiments. Thus, the results of this study could not only contribute to basic studies on genome structure and function but also in the longer term underpin applications in biotechnology and agriculture/industry.