The methylation of cytosine plays a fundamental role in mediating transcriptional repression. However, 5-methylcytosine can undergo spontaneous hydrolytic delamination to form thymine. It the resulting T:G mismatch is replicated prior to being repaired, a C:G to T:A transition mutation will be present in one of the two daughter DNA molecules. Methyl-CpG-binding domain protein 4 (MBD4) is a DNA glycosylase that can excise thymine from mismatches with guanine, and that acts preferentially on T:G mismatches within CpG dinucleotides <i>in vitro</i>. In order to test the hypothesis that MBD4 repairs the product of 5-methylcytosine delamination <i>in vivo</i>, MBD4-deficient mice were crossed onto the Big Blue genetic background. This background enables the frequency and spectrum of <i>in vivo</i> mutations in a bacteriophage lambda <i>cII</i> transgene to be determined. As predicted, <i>Mbd4^-/-</i> mice have a significantly increased frequency of C:G to T:A mutation sat CpG dinucleotides. T:G mismatch-specific thymine DNA glycosine (TDG) can also attempt to excise thymine from T:G mismatches within CpG dinucleotides <i>in vitro</i>. In an attempt to determine the relative contributions of MBD4 and TDG to the repair of 5-methylcytosine delamination-induced T:G mismatches <i>in vivo</i>, the mutation frequencies and spectra in cell lines lacking MBD4 and/or TDG were measured. An additional line of research focused on potential mechanisms by which the DNA repair activity of MBD4 may be regulated. A novel protein has previously been shown to interact with MBD4 in a yeast two-hybrid screen that used MBD4 as the bait protein. This interaction was further characterised by mapping of the interaction domains using the yeast two-hybrid assay, and by immunocytochemistry. Finally, it was also shown that MBD4 may be post-translationally modified by sumoylation.

Mi-2β, encoded by the <i>Chd4 </i>gene, is an ATP-dependent chromatin remodelling protein belonging to the CHD protein family. Structurally, Mi-2β contains two PHD finger domains, two chromodomains and an ATPase helicase domain. It is a key component of the NuRD (nucleosome remodelling and histone deacetylase) complex. Due to its deacetylase activities, the NuRD complex is considered to be a transcriptional repressor complex. In my thesis, I biochemically and functionally characterised Mi-2β and the associated NuRD complex. I used a gene-trap strategy to analyse the role of Mi-2β in development. <i>Chd4 </i>gene-trap homozygous embryos were not evident in post-implantation development, indicating that Mi-2β is essential in early embryonic development. <i>Ex vivo</i> analysis revealed a critical role of Mi-2β in the viability of early embryos, highlighting its importance in development. Additionally, I investigated the biochemical function of the PHD fingers and chromodomains. The chromodomains of Mi-2β do not appear to bind di-methylated lysine 4 of histone H3 or unmodified histone H3. Additionally, the PHD fingers of Mi-2β may be important in binding histone tails, irrespective of the modification state of these tails. I also employed a biotinylation tagging strategy to biochemically characterise the NuRD complex in embryonic stem (ES) cells. This method identified the subunits of the complex in ES cells and revealed that the complex was a heterogeneous complex, potentially explaining the differences in previously reported purifications. Additionally, the biotinylation purification allowed the identification of interacting transcriptional regulators in ES cells, which included the pluripotency-associated protein Sall4. In addition to the identification of previously known interactors, the purification implied roles for the NuRD complex not previously acknowledged, including a potential role in transcription elongation.

This thesis describes the expression, purification, and biochemical and biophysical characterization of a number of FK506 binding proteins. The main focus of the work is to help characterize the roles of played FKBP by thus providing a better understanding of their biological functions. <i>Neurospora crassa</i> <i>FKBP22</i> – Expression, purification, crystallization, and partial structure determination of the 22 kDa FKBP protein were carried out. The enzymatic PPIase activity of the N-terminal FKBP domain was characterized with the <i>k_ca/K_m</i> value of 4 x 10⁶ M^-1s^-1. Sequence analysis showed high homology of this domain to the 12 kDa FKBP with the highly conserved peptidly-prolyl active site. The C-terminal sequence appeared to be unique among other FKBP sequences; being highly charged with a stretch of predicted amphipathic helical conformation. Dynamic light scattering studies suggest that the protein may also exist as a dimer in solution. <i>Caenorhabiditis elegans FKBP48 </i>– The only TPR protein in <i>C. elegans</i> were cloned. Three natural breakdown products of FKBP48 were purified and characterized: an intact full length, a δ30 fragment missing final 30 C-terminal residues, and a ST fragment missing the C-terminal TPR domain. Results of small angle X-ray scattering revealed similarity of domain architecture between FKBP48 and FKBP51. <i>Caernorhabiditis elegans FKBP12</i>, <i>Caenorhabiditis elegans FKBP29, and Arabidopsis thaliana FKBP71 –</i> FKBP12/29 were well expressed. The enzymatic PPIase activity of FKBP29 was characterized with the <i>k_ca/K_m </i>value of 0.3 x 10⁶ M^-1s^-1. However the protein yield was insufficient for further studies. Purification of FKBP12 was difficult to manage in this work. Protein aggregatin was noticed during cleavage of fusion maltose binding protein. The recombinant 71 kDa FKBP protein was expressed at low level. This could be due to its mild toxicity. In addition quality of the protein was poor with inherent instability.

No description available.

MeCP2 is a transcriptional repressor that recruits the Sin3a chromatin remodelling complex to methylated loci. Sin3a-associated histone deacetylases contribute to the ability of MeCP2 to repress transcription and modulate chromatin structure. The biomedical importance of normal MeCP2 function is highlighted by the discovery that inactivating mutations in MeCP2 cause the severe neurological disease Rett syndrome. By deleting the <i>Mecp2 </i>gene, a mouse model of Rett syndrome has been generated and used to study the molecular and physiological outcome of MeCP2 deficiency. Inefficient regulation of neuronal gene expression may have a role in the etiology of Rett syndrome. By studying the biochemical properties of MeCP2 this thesis addresses in detail three basic questions; (1) what are the native biochemical properties of MeCP2? (2) what specific DNA sequences does MeCP2 bind? and (3) what are the affects of post-translational modification on MeCP2? To address the composition of any mammalian MeCP2 complexes, native MeCP2 was purified to near homogeneity from rat brain. Native MeCP2 is an elongated monomer that does not stably associated with other cofactors including Sin3a. Analysis of MeCP2 binding sites <i>in vivo </i>demonstrates that MeCP2 binds unique loci when compared to other MBP’s. Using an unbiased <i>in vitro </i>DNA binding site evolution assay, Methyl-SELEX, MeCP2 was shown to require methyl-CpG sequences containing a flanking run of A/T rich DNA for high affinity binding. Finally, biochemical fractionation of nuclear proteins revealed activities that phosphorylate MeCP2, and the potential affects of this modification were explored.

Previous work from our laboratory had suggested that the carboxyl-terminus of GroEL limited the ability of this protein to reverse the temperature-sensitive defects of <I>dnaA<SUP>ts</SUP></I> mutations in <I>E. coli</I>. This carboxyl-terminus consists of a 13 amino acid string of glycine and methionine residues, a motif that is highly conserved amongst many of the GroEL homologues sequenced so far. A clone was constructed which expressed a form of GroEL lacking this motif. Surprisingly this clone (which also expressed wild-type <I>groES</I>) was unable to suppress <I>dnaA</I><SUP>ts</SUP> mutations when overexpressed, but was able to complement <I>groEL</I><SUP>ts</SUP> mutations (even when present in single copy). A <I>groE</I> deletion mutant strain was constructed and this too was complemented by the truncated form of <I>groEL</I>. The resulting strain, carrying only truncated <I>groEL,</I> was extensively characterized and found to behave identically to an isogenic strain carrying wild type <I>groEL</I>, in that growth rates, temperature dependence, carbon source utilization, bacteriophage sensitivity, ethanol sensitivity, UV sensitivity and β-lactamase excretion were all found to be identical between the two strains. However, competition experiments involving co-culturing of the strains carrying <I>groEL </I>+ <I>groEL</I><SUB>tr</SUB><SUB> </SUB>demonstrated an advantage to the cells expressing the wild-type gene when grown at 42<SUP>o</SUP>C. The advantage was found to be due to the strain encoding the truncated <I>groEL </I>exiting stationary phase and entering log-phase growth more slowly than cells expressing <I>groEL</I>+. It was also found that strains expressing <I>groEL</I><SUB>tr </SUB>are more sensitive to the dye crystal violet, and low levels of some other biocide at 42<SUP>0</SUP>C. This suggests a role for the GroE proteins in membrane biogenesis and.or maintenance as well as an involvement of the glycine-methionine sequence.

The aims of this study were to characterise the molecular function of LSH and attempt to relate this to its role in DNA methylation <i>in vivo</i>. To address these aims the work described in this thesis uses a variety of approaches to address two key questions. (1) Is LSH an active SNF2 ATPase? (2) How does LSH interaction with, and modulate, the DNA methylation machinery? To determine if LSH is an active SNF2 ATPase, recombinant LSH was purified from insect cells and biochemically characterised. These experiments revealed that LSH can hydrolyse ATP and its activity is stimulated by DNA. However, the rate of ATP hydrolysis is low and LSH does not exhibit chromatin remodelling activity. Thus, either LSH is a relatively weak SNF2 ATPase that cannot remodel chromatin or recombinant LSH may not be fully active. To identify proteins that interact with LSH, biophysical analysis of the native protein was performed. This indicated that the majority of LSH was present as a free monomeric peptide <i>in vivo</i>. However, I was able to demonstrate that LSH is an HDAC transcriptional repressor <i>in vivo</i>. Also, a weak, transient or low abundance complex including LSH, DNMT3B, DNMT1, HDAC1 and HDAC2 was identified. Thus, how LSH interacts with the DNA methylation machinery was demonstrated. Finally, I attempted to investigate how LSH modulates the activity of DNMTs <i>in vitro</i>. These experiments did not identify a role of LSH in stimulating DNMTs <i>in vitro</i>. These studies shed new light on the role of LSH in DNA methylation.

No description available.

This thesis describes the study of the Hsp90 proteins from <i>Homo sapiens </i>(Hsp90α) and <i>Caenorhabditis elegans. </i>The attempt to crystallise the C-terminal proteins is also described. The overall goal of the project was to biochemically and biophysically characterise various Hsp90 constructs and use this information to elucidate the biological role of this important family of proteins. The C-terminal domain of human Hsp90α. The work involved over-expression, purification and characterisation of the C-terminus of human Hsp90 α protein. The over-expression and purification led to the production of two reproducibly pure C-terminal human Hsp90 α protein constructs. The characterisation of these proteins focused on the interaction of the C-terminus with the immunophilin Cyp40 and also with ligands such as ATP and novobiocin. <i>C. elegans</i> Hsp90. The work involved the cloning, over-expression and purification of the N and C-terminus of <i>C. elegans </i>Hsp90.  The cloning of genes for the N and C-termini was successful. Purification of the proteins only led to the production of one C-terminal protein but no purified N-terminal protein could be obtained. Similar characterisation studies were carried out to the C-terminus of <i>C. elegans </i>Hsp90. As for the human C-terminal protein, the <i>C. elegans</i> C-terminal protein was found to bind to Cyp40 with a dissociation constant in the micromolar concentration range. The protein also binds to ATP and novobiocin with an affinity very similar to that of the C-terminal proteins of human Hsp90α. The C-terminus of <i>C. elegans</i> also exhibits ATPase activity but the activity was ten-fold lower than that of the C-terminal of human Hsp90α. The work presented in this thesis provides conclusive evidence of the existence of an ATP binding site in the C-terminal domains of the Hsp90 class of proteins. The biological relevance of this finding is also discussed.

No description available.