In order to identify cellular factors that affect trinucleotide repeat stability, changes in the length of a (CTG)₄₃ repeat were studied over 140 generations in wild-type <i>Escherichia coli</i> and in strains that are deficient in post-replicative mismatch repair, secondary structure repair and homologous recombination. It is shown that (CTG)₄₃ inserted into pUC18 expands and contracts in wild-type <i>E. coli</i> in an orientation-dependent manner that is unaffected by transcription. In cells deficient in post-replicative mismatch repair (CTG)₄₃ repeat instability is greater than in wild-type cells but orientation-independent. The observation of single trinucleotide insertions and deletions in these mutator mutants indicates that replication slippages of 3 bp occur <i>in vivo</i> leading to repeat expansion and contraction if left unrepaired. Compared to wild-type cells large deletions are reduced in these mutator mutants, but only if the CTG sequence serves as the lagging strand. Based on the opposing effects of mismatch repair a model is proposed in which orientation-dependent CTG repeat instability in mismatch repair proficient cells is caused by the repair of 3-bp slippages. This leads to the creation of larger deletions during repair synthesis due to the formation of unusual secondary structures by the CTG sequence on the lagging strand. Mutations in the <i>recA</i> and <i>sbcCD </i>genes do not affect the stability of plasmid-borne CTG repeats. Similarly the viability of <i>recA</i>-deficient strains carrying chromosomal insertions of (CTG)₂₅ and (CTG)₄₃ suggests that, unlike long palindromes, these trinucleotide repeats are not substrates for the structure-directed nuclease complex SbcCD or, alternatively, they do not form secondary structures frequently enough to cause lethality in <i>recA-</i>deficient hosts. In contrast, a mutation in the <i>recG</i> gene, also involved in homologous recombination, severely destabilises the (CTG)₄₃ repeat in a strongly orientation-dependent manner that exceeds all other tested mutants. Possible explanations for this observation are discussed.

A novel <I>ERCC1</I> mRNA has been identified in mouse skin. Subsequent characterisation of the transcript demonstrated that the difference between the normal and skin-specific <I>ERCC1 </I>mRNA was at the 5' end and is due to differential initiation of transcription. As with the normal <I>ERCC1 </I>transcript the novel skin-specific transcript did not appear to be induced by UV irradiation. A functional <I>ERCC1 </I>minigene was constructed to facilitate subsequent analysis of the upstream promoter region and identification of the sequences involved in the regulation of the observed skin-specific pattern of expression. The minigene corrected the UV sensitivity of <I>ERCC1</I>-deficient cultured cells but did not exhibit the characteristic skin-specific expression pattern at the level of mRNA analysis. A pilot study for a potential gene therapy approach to increasing the lifespan of the <I>ERCC1 </I>knockout mice was completed. Recombinant ecotropic retroviruses containing <I>ERCC1 </I>coding sequences were produced using a murine leukaemia virus derived viral vector and viral packaging cell line. The resultant <I>ERCC1 </I>retrovirus was shown to partially correct the UV sensitivity associated with pools of <I>ERCC1</I>-deficient mouse embryonic fibroblasts. Clones, subsequently isolated from the transduced pools, were shown to be phenotypically corrected to wild type levels. The observed phenotypic correction correlated with expression of a retroviral <I>ERCC1 </I>mRNA. A transthyretin regulated <I>ERCC1 </I>transgene was used to bring about the targeted expression of <I>ERCC1 </I>in the liver. Pronuclear injection of the- transgene was used to produce a transgenic mouse line containing <I>~5 </I>copies of the transgene integrated at a single site. Expression of this transgene on an <I>ERCC1 </I>-deficient background resulted in the correction of the lethal liver phenotype. Transgene positive nulls were not as severely runted and survived for between 9 and 12 weeks compared to the three-week survival of the transgene negative <I>ERCC1</I>-deficient mice. The consequences of <I>ERCC1 </I>deficiency in other tissues were studied. Abnormalities were identified in the skin and kidneys of adult transgene positive nulls. An investigation into the consequences of DNA repair deficiency on UV-B induced immunosuppression revealed that antigen presenting epidermal Langerhans cells, in the transgene positive nulls, did not show the normal pattern of accumulation in the lymph nodes following UV irradiation.

An Indian muntjac cell line, SVM, is deficient in an array of DNA repair processes. It exhibits hypersensitivity to UV irradiation and alkylating agents. SVM was transformed with the DNA virus SV40, and some evidence in the literature suggests that this transformation process may of itself have an effect on the repair phenotype of a cell. In order to investigate this possibility for SVM, a spontaneous UV<SUP>r</SUP> partial revertant was isolated and aspects of the SV40 virus were characterized in the mutant cell line, SVM(M), and the revertant cell line, SVM(R). Investigation of the integration site of SV40 has revealed a change between SVM(M) and SVM(R). This change is due to an amplification of the control region of SV40 (containing an enhancer and an origin of replication) at the viral-cellular DNA junction. This amplification has no effect either qualitatively or quantitatively on the biologically important gene product of SV40, the large T-antigen. This study reveals novel T-antigens in the SVM cell lines apparent at 100kD and 76kD as compared to the reported wild-type produce at 94kD. The nature of the 100kD 'super T-antigen' was established as due to an internal duplication of SV40 sequences. The cellular sequence flanking the amplification was cloned and characterized. Northern blot analysis, utilizing probes generated from this region revealed that this cellular DNA encoded a 11kb transcript which was aberrantly overexpressed in SVM(M) and SVM(R) compared to CDM, a cell line considered normal with respect to its DNA repair capacity. Furthermore, one probe from this region revealed a 2.8kb transcript, which was also abundant in SVM cell lines compared to CDM. This transcript may also differ quantitatively between SVM(M) and SVM(R). In order to investigate whether aberrant overexpression of the 11kb transcript was responsible for the DNA repair phenotype of SVM(M), an experiment to decrease levels of this transcript utilising sense/antisense technology was performed. A model for the cause of the spontaneous reversion event in SVM(R) is discussed, and the means by which overexpression of a transcript can lead to a DNA repair deficient phenotype are considered.

The work undertaken has involved an investigation of the relationship between different aspects of chromosome behaviour and their relation to chromosome banding patterns. Firstly, unmethylated CpG islands, associated with just over 50% of human genes, were used as gene-specific markers. By exploiting their methylation-sensitive restriction enzyme cleavage sites, digestion with three such enzymes effectively yielded kilobase to megabase-sized inter-gene restriction fragments. To investigate the relationship between <I>Alu</I> repeat density and gene density, each size fraction was purified and subjected to <I>Alu </I>PCR (inter-<I>Alu</I> amplification). Although it was hoped that a greater density of <I>Alus</I> would lead to a greater number of amplification products, this approach proved fruitless due to unavoidable contamination between the original inter-island size fractions. However, it was found that the chromosome bands containing the highest number of mapped genes were also the most <I>Alu</I>-rich bands, according to published data. It was also established from such data that R bands contain 80% of the 2% of human genes mapped to date. To verify this link for a greater proportion of human genes, inter-island fragments were "painted" to metaphase chromosomes using fluorescence <I>in situ</I> suppression hybridisation (FISH) to display their bands of origin. Working with a single human chromosome 11, it was shown that each chromosome band has a characteristic range of gene density, with each type of band (e.g. R, G) having a similar gene density. It is apparent from this work that R bands and G bands represent separate compartments of the human euchromatic genome, with R bands containing the vast majority of genes. These findings also pinpoint the regions of the genome which will yield the highest density of coding sequence information. This work also paves the way to the further physical and biochemical characterisation of G and R band chromatin compartments.

The thesis considers two topics in the automation of cytogenetics and cytology: the automated allocation of human chromosomes to the twenty-four classes which humans possess; and the detection of abnormal cervical smear specimens. For chromosome allocation, the following work is presented and evaluated on a number of data sets derived from chromosome preparations of different quality:1. Three new procedures for modelling between-cell variation.2. Six ways of combining class information on variability in multivariate Normal discrimination.3. Covariance selection models for individual chromosome classes and an assumed common covariance structure for a number of classes.4. Some two-stage procedures for the calculation of discriminant scores in multivariate Normal discrimination.5. The application of some non-parametric and semi-parametric methods.6. The modelling of band-transition sequence probabilities. For the detection of abnormal cervical smear specimens, the use of a consensus probability of a specimen being abnormal, derived from a number of cytologists' assessments, is considered. The sequential use of multiple regression equations to try to predict the logit transformations of these consensus probabilities is described. Finally, the sequential use of features in multivariate discrimination is considered mainly for the case of two known multivariate Normal populations with equal covariance matrices.

As a result of checkpoint activation, a signalling cascade is initiated and a number of complexes between the checkpoint components are formed. This leads to the inhibition of the Anaphase Promoting Complex (APC), which is the ubiquitin ligase responsible for targeting mitotic proteins: securing and cyclin B for degradation by the 26S proteasome. The complexes formed include the MCC, or Mitotic Checkpoint Complex, which in fission yeast (Schizosaccharomyces pombe) consists of Mad2, Mad3 checkpoint proteins together with the APC activator, Slp1 (the Cdc20 homologue). The MCC has been shown to bind and inhibit the APC in HeLa cells. In my PhD I focused on the interactions between the MCC and the APC, in particular on Mad3 protein. Mad3 is a conserved checkpoint component, homologous to human BubR1. It carries 2 putative KEN boxes, motifs, which typically target proteins for degradation (like D-boxes). We mutated both KEN boxes in S. pombe Mad3 and show that they are essential for Mad3 checkpoint function. One of the two motifs is also required for MCC formation, MCC/APC binding and Mad3 turnover in G1 stage of cell cycle. We argue that KEN boxes mediate the inhibitory interactions between checkpoint proteins and the APC. The formation of the MCC requires Mad2 and Mad3. These proteins are also dependent on one another for binding to the APC. We study the formation of the MCC, as well as its binding to the APC in different checkpoint mutants, trying to understand dependencies between these proteins and the requirement for kinetochore localisation. Finally, we make an attempt to study the cellular localisation of S1p1 and the APC and find that while S1p1 colocalises with a kinetochore marker, the APC subunits Cut9 and Lid1 remain distributed around the fission yeast nucleus.

This project was designed to isolate and characterise interstitial telomere repeat containing loci from the human and mouse genomes and to investigate the nature of the mouse telomere. Cloning of the internal telomere repeat loci proved to be extremely difficult and so alternative methods such as restriction enzyme analysis, hybridisation analysis, inheritance studies, and mapping within recombinant inbred and backcross mouse strains were employed to characterise these regions within the mouse genome. Similar methods were used to characterise mouse telomeres. From these experiments it was shown that, in the mouse, <i>Trypanosoma</i>-like (TTAGGG)<SUB>n</SUB> telomere repeats are present at the telomeres and at interstitial sites. Within both of these regions, the (TTAGGG)<SUB>n</SUB> repeats are present within distinct genetic loci that are stably inherited through subsequent generations. New variant generation is observed at both types of loci, takes place at a significantly higher frequency at the telomeric compared to interstitial loci and occurs during gametogenesis. It is possible that the higher rate of new variant generation at mouse telomers compared to internal sites may relate to their position within the mouse genome. Restriction enzyme and hybridisation sequence analysis demonstrated that both classes of loci are composed of telomere-related repeats and that an undefined simple repeat may also be present. Direct sequencing is required before the nature and organisation of simple repetitive DNA within these loci can be determined. Mapping of the interstitial, (TTAGGG)<SUB>n</SUB> telomere repeat containing loci within the BxD Rl and <i>Mus spretus</i>/C57Bl/6 backcross mice demonstrated their presence within the protermini of chromosomes 9, 13 and X. It remains to be determined whether this distribution is functionally significant.

The gene encoding topoisomerase I (PfTopoI) from the malarial parasite, <I>Plasmodium falciparum </I>was isolated using two methods. (a) PCR of a cDNA library using oligonucleotides modelled on highly conserved regions of sequences from other species, isolated a 900bp fragment of the gene. (b) The entire PfTopoI open reading frame was obtained from a <I>Hin</I>DIII/<I>Eco</I>RI genomic library after screening with a random-labelled heterologous TopoI probe from <I>Saccharomyces cerevisiae. </I>DNA sequence analysis indicated an open reading frame of 2520 nucleotides, encoding a deduced protein of 839 amino acids with no detectable introns. Two large amino acid inserts were present in the PfTopoI gene sequence which were present in both the genomic and cDNA indicating that they were not introns. The predicted amino acid sequence was compared with homologous sequences of other type I topoisomerases. On average, PfTopoI showed around 40% identity with all the sequences examined, with the TopoI sequence from man showing the highest level of homology. The gene is present as a single copy on chromosome 5 and Northern analysis identified a transcript of 3.8kb. The PfTopoI open reading frame was cloned into prokaryotic expression vectors in order to express, purify and assay the recombinant protein for TopoI relaxation activity. Both the full length PfTopoI protein and fragments were expressed, although no activity could be detected. However TopoI and II activity was observed in both crude and nuclear parasite extracts. Inhibition of the TopoI relaxation activity was observed when the type I topoisomerase specific inhibitor, camptothecin, was added to the assay. Stage specific expression of PfTopoI during the intraerythrocytic stages of the <I>Plasmodium falciparum </I>life cycle was examined using synchronised parasites. Nuclear run on assays indicate that the promoter is inactive during the early ring stage and is active only in the later mixed trophozoite and schizont stages.

Expanded trinucleotide repeats (TNRs) (e.g. CAG, CTG, CCG) cause 40 different human diseases, however the molecular mechanism underlying the expansion of TNRs is poorly understood. This work describes the integration, in the chromosome of the bacterium <i>Escherichia coli</i> of differently sized CAG and CTG TNRs into the start of the <i>lacZ</i> gene and of a zeocin resistance recombination reporter substrate into the nearby gene<i>, cynX</i>.  We show that TNRs stimulate recombination at <i>cynX</i> in a length dependent manner. Furthermore, stimulation of recombination is dependent on TNR orientation with respect to the origin of replication. Experiments indicate that zeocin recombination is reduced in <i>E. coli</i> mutants for double strand break repair (<i>recA</i> and <i>recB</i>); but not for gap repair (<i>recR</i>). TNR induced stimulation of recombination is shown to be independent of the DNA hairpin nuclease SbcCD arguing against any role of secondary structure formed by TNRs. The formation of DSBs by TNRs is investigated using pulse field gel electrophoresis (PFGE). Since reversed replication forks can initiate homologous recombination (HR), this thesis investigates the possibility that TNRs lead to replication form reversal (RFR). We test this hypothesis by assaying for HR using zeocin recombination reporter substrates positioned proximal and distal with respect to the origin of replication. We show that TNR induced HR is lower at the distal site. Furthermore, the protein UvrD is known to be essential for RFR in certain replication mutants. TNR dependent stimulation of HR is lost in <i>uvrD</i> mutants. Both these data support the hypothesis that TNRs can cause RFR.

This study reports the isolation of the <I>nupC</I> gene of <I>E.coli</I> from the Kohara ordered phage lambda library. Lambda clones carrying <I>nupC</I> were identified by their ability to restore nucleoside transport to a nucleoside transport deficient mutant by reciprocal exclusion. Subsequent subcloning, complementation and sequencing studies identified an ORF of 1,200bp which codes for a hydrophobic polypeptide of 43,560Da. A NupC-β-lactamase fusion protein was constructed and identification of the subcellular location of this hybrid confirmed the NupC protein to be membrane associated. Analysis of the <I>nupC</I> promoter region revealed the presence of at least two putative CRP binding sites, centred at -40bp and -89bp, which probably flank a CytR binding site. In addition, downstream from the <I>nupC</I> ORF an adjacent IS<I>186</I> element was identified and found to reside within an underlying (potentially bidirectional) terminator structure. This arrangement is shown, by Southern blot analysis, to reflect the existing order in the chromosome of the <I>E.coli</I> strain W3110. The topological organisation of NupC within the cytoplasmic membrane has been analysed experimentally by construction of fusions to mature β-lactamase. From the combination of this data, with information from a computational hydropathy analysis, a two-dimensional model of NupC is presented which predicts that twelve transmembrane segments are likely to be present in this protein.