Zinc cluster proteins represent a subclass of zinc finger proteins and function as transcriptional regulators. An in vivo genetic screening system was developed in yeast to identify DNA binding sites and specificities for these proteins. / An oligonucleotide library of 200 000 clones was constructed. Control screening trials with Hap1p and Gal4p demonstrated effectiveness in recovering binding sites. Sequencing of isolated clones showed correlation with published target sequences and binding was confirmed by electrophoretic mobility shift assay (EMSA). / Screening over 100 000 clones of the library with the YLR228c gene product allowed the isolation of 10 clones. Mutational EMSA studies were performed to identify nucleotides important for binding to derive a consensus sequence. A CGG triplet was found to be significant for binding. It can be hypothesized that Ylr228p may bind as a monomer to its targets.

A growing repertoire of proteins is involved in the molecular events surrounding clathrin-coated vesicle (CCV) formation, which is seminal to the regeneration of synaptic vesicles in nerve terminals, as well as to more universal membrane trafficking mechanisms. Amphiphysin-I and endophilin-I are two nerve terminal-enriched proteins thought to function in CCV formation. In the first study of this thesis, we identified a novel amphiphysin-I isoform (termed amphiphysin-II), through a homology search of the expressed sequence tag (EST) database. We demonstrated that like amphiphysin-I, amphiphysin-II is highly enriched in nerve terminals, where it interacts with dynamin-I and synaptojanin-I through its SH3 domain. Further, both amphiphysins were found to bind directly to the vesicle coat protein, clathrin, in an SH3 domain-independent manner. / In the next two studies, we cloned a number of amphiphysin-II splice variants that were found to display differential clathrin binding. The use of these splice variants, together with amphiphysin-II truncation and deletion mutants, allowed us to identify two distinct clathrin-binding domains that are also conserved in amphiphysin-I. Using these amphiphysin constructs we also determined that the amphiphysins are targeted to the plasma membrane through their N-termini, which also appears to be important in facilitating amphiphysin oligomerization. / In the final study of this thesis, we turned our attention to a related endocytic SH3 domain-containing protein, endophilin-I. In an attempt to further understand its complete role within the cell, we screened a brain expression library with its SH3 domain, and identified the rat germinal center kinase-like kinase (rGLK), a member of the germinal center kinase (GCK) family of c-jun N-terminal kinase (JNK) activating enzymes, as a novel endophilin-I-binding partner. We characterized the interaction both in vitro and in vivo and determined the sequence on rGLK that mediates the interaction. Importantly, overexpression of full-length endophilin-I was found to enhance rGLK-mediated JNK activation, whereas fragments of endophilin-I blocked JNK activation, suggesting a novel role for endophilin-I in JNK intracellular signaling. / Taken together, the data in this thesis not only adds to the growing list of endocytic accessory proteins, but also contributes to our understanding of their individual molecular interactions and functions within the cell.

Unlike in somatic cells, cancer cells adopt an aberrant pattern of methylation as well as histone acetylation, and therefore distort the chromatin structure. Chapters 2--4 of this thesis look at mechanisms carried out by the recently cloned DNA demethylase, how its demethylation activity is closely linked with the semblance of acetylation of chromatin, and how this relationship can be skewed in cancer. The three intriguing mechanisms described provide attractive models by which to explain general genome wide demethylation, site specific demethylation of genes upon their activation, and the relationship between aberrant methylation and histone acetylation in cancer. The thesis begins by characterizing the mechanism of demethylation carried out by the bona fida DNA demethylase---an enzyme identified and cloned in our laboratory found to demethylate both hemi and double-stranded DNA in vitro. This enzyme manifests the removal of methyl groups from DNA without damaging the DNA and is therefore a candidate protein responsible for hypomethylation seen during development as well as in transformed cells. One essential property of an enzyme that removes methylation from wide regions of the genome could be processivity. Southern blot analysis and sodium bisulfite mapping experiments determine that purified demethylase demethylates DNA in a processive manner in vitro. Experiments in Chapter 3 demonstrate how an active demethylase enzyme is involved in shaping patterns of methylation relative to the state of histone acetylation. We present evidence suggesting demethylase activity is directed by the state of histone acetylation, therefore contrasting the accepted dogma, and suggesting that the local histone acetylation state determines the resulting methylation pattern. Aberrant DNA methylation and histone deacetylation are frequently associated with silencing of tumor suppressor genes in cancer and yet cannot simply be explained by the level of methyltransferase(s) enzyme(s)

The CDP/Cux transcription factor contains four DNA binding domains: three Cut repeats (CR1, CR2 and CR3) and the Cut homeodomain (HD). The CCAAT-displacement protein (CDP) was first identified as it represses transcription of certain promoters by competing for the occupancy of the CCAAT sequence. CDP was then found to be the ortholog of the Drosophila Cut protein (Cux: Cu t homeobox&barbelow;). The goal of my project was to define the modes of interactions of CDP/Cux with DNA. I demonstrated that high affinity DNA binding requires the cooperation between at least two of the CDP/Cux DNA binding domains. Among all combinations of domains, Cut repeats 1 and 2 (CR1CR2) and Cut repeat 3 and the Cut homeodomain (CR3HD) exhibited the highest DNA binding affinities, but with different kinetics and specificities. Whereas CR1CR2 bound with fast kinetics to dimers of CART or CGAT, CR3HD bound to ATCGAT with slow kinetics. CR1CR2 was shown to be responsible for the CCAAT displacement activity of CDP/Cux. Surprisingly, the full-length CDP/Cux protein, p200, exhibited DNA binding properties similar to that of CR1CR2, indicating that CR3HD is inactive in this context. However, an amino terminally truncated 110 KDa CDP/Cux isoform capable of stable interaction with DNA was identified in cellular extracts. The p110 isoform was shown to be generated by proteolytic cleavage of p200 at the G1/S transition. The p200 and p110 isoforms displayed different transcriptional activities in reporter assays, as only p110 was able to activate transcription from the DNA polymerase a gene promoter. Interestingly, expression of p110 was found to be increased in human uterine leiomyomas as compared to the adjacent normal myometrium, raising the possibility that proteolytic processing of CDP/Cux is activated in cancer cells. Finally, cyclin D/CDK4 was shown to interact with CDP/Cux, phosphorylate it on several serine residues and inhibit proteolytic processing as well as DNA binding by CR3HD. Repla

Telomerase is a ribonucleoprotein (RNP) enzyme responsible for the replenishment of repetitive DNA sequences present at the ends of most eukaryotic chromosomes. Telomerase is minimally composed of a protein catalytic subunit, the telomerase reverse transcriptase (TERT), and an RNA subunit. Using a small single-stranded segment (7--11 nucleotides) of the telomerase RNA as a template, the active site of the TERT catalytic subunit adds complementary nucleotides onto telomeric DNA. The primary goal of the work presented in this thesis was to biochemically and functionally characterize the human telomerase reverse transcriptase (hTERT) and the human telomerase RNA (hTR), as well as to identify novel telomerase-associated proteins. First, we demonstrated that the budding yeast Saccharomyces cerevisiae possesses the cellular machinery to fully assemble a catalytically active human telomerase RNP in vivo. We also analyzed telomerase activity and binding of hTR to hTERT in rabbit reticulocyte lysates by expressing different hTERT and hTR variants. Our results identified two distinct regions of hTR that can independently bind hTERT in vitro. Furthermore, sequences or structures that include the conserved CR4--CR5 domain of hTR were found to be important for hTERT-hTR interactions and telomerase activity reconstitution. Human TERT carboxy- and amino-terminal amino acid deletions indicated that the polymerase and RNA binding functions of hTERT are separable. We also found that the product of the survival of motor neuron ( SMN) gene, a protein involved in the biogenesis of certain RNPs, is a telomerase-associated protein. Our results demonstrate that the human TR and the human TERT are not associated with Sm proteins, in contrast to Saccharomyces cerevisiae telomerase. Taken together, the work presented in this thesis indicate that the reconstitution of human telomerase activity in vitro requires regions of hTERT that (i) are distinct from the conserved reverse transcriptase

We have shown that nine cDNA clones, which were derived from an oligo (dT) primed cDNA library of human embryonic lung fibroblast (IMR90), are capable of supporting autonomous replication of bacterial plasmid in mammalian (HeLa) cells. Two of these cDNA clones contain a region homologous to "O"-family middle repetitive sequences; the other seven clone hybridize with neither "O"-family nor Alu family sequences and are called NOA clones. Oligo (dT) primed cDNA of IMR90 has been demonstrated to be an enriched source for autonomously replicating sequences. Two out of three "O"-family homologous cDNA clones and seven out of ten NOA clones were capable of autonomous replication. In contrast, none of the five clones, which contained Alu repetitive sequences, demonstrated replicative potential. The two autonomously replicating, "O"-family homologous cDNA, 343 and 363, were characterized in the greatest detail. In particular, sequence and structural features of a 448 bp fragment of cDNA 343, which was capable of supporting autonomous replication activity, were: extensive asymmetrical A/T-rich regions; 10/11 match to yeast ARS consensus; scaffold attachment region consensus of Drosophila; bent DNA; a DNA unwinding element and an in vitro matrix binding activity. We have used a PCR-based mapping strategy to identify an in vivo initiation zone located in a region of approximately 1.6 kb which is inclusive of cDNA 343. The 343 sequence has been localized to the long arm (6q22-6qter) of human chromosome 6 by both Southern analysis of hamster-human cell hybrids and in situ hybridization. Transcripts homologous to 343 have been detected in both poly(A$ sp+$) and non-poly(A$ sp+$) RNA fractions to be 1.3 kb and 5 kb, respectively; the 1.3 kb transcript is greatly enriched in poly(A$ sp+$) RNA fraction, and the 5 kb transcript is present predominantly in non-poly(A$ sp+$) RNA fraction. The detailed characterization of 343 genomic locus demonstrated that the 343 sequence is conserv

Chromogranin A (CgA) is the major member of a family of acidic secretory glycoproteins (collectively known as granins) that are expressed in all endocrine and neuroendocrine cells. Granins have been proposed to play multiple roles, both intracellular and extracellular, in the secretory process. The regulation of synthesis and secretion of CgA in primary cultures of bovine parathyroid cells was studied and compared to that of parathyroid hormone (PTH). The effects of both extracellular calcium and the active metabolite of vitamin D$ sb3$, 1,25-dihydroxyvitamin D$ sb3$ (1,25(OH)$ rm sb2D sb3$) were shown to modulate CgA synthesis and secretion in the parathyroid. Whereas a reduced medium calcium level stimulated both CgA and PTH secretion, it had no effect on CgA mRNA levels. In contrast, 1,25(OH)$ rm sb2D sb3$ stimulated CgA mRNA levels by increasing the CgA gene transcription rate. Thus, 1,25(OH)$ rm sb2D sb3$ had opposite actions on the CgA and PTH genes in the parathyroid cell. The relationship between 1,25(OH)$ rm sb2D sb3$-stimulated CgA mRNA levels and total CgA protein levels (intracellular and extracellular) was not simple. Although 1,25(OH)$ rm sb2D sb3$ increased CgA mRNA levels several-fold, there was only a modest increase in CgA protein. This was the result of a specific effect of 1,25(OH)$ rm sb2D sb3$ on decreasing the translational efficiency of CgA mRNA as demonstrated by an increased ribosome transit time for CgA mRNA. To better understand the functional domains of CgA, and its evolution, and to study the basis of its neuroendocrine cell-specific gene expression, the human CgA gene was cloned and characterized. It spans 15 kb and contains eight exons. There is high conservation between the human, mouse, and bovine CgA genes with respect to exon number and placement of exon/intron borders. Transient transfection studies with constructs containing portions of the 5$ sp prime$ gene flanking region upstream of a chloramphenicol acetyl transferase rep

Human mitochondrial NAD-dependent methylenetetrahydrofolate dehydrogenase-methenyltetrahydrofolate cyclohydrolase was expressed in a soluble and active form in Escherichia coli, purified to homogeneity for the first time and characterized. Phosphate is absolutely required for the dehydrogenase activity. NADP can be utilized by the dehydrogenase with a low turnover rate. The properties of the enzyme suggest that it has evolved from NADP to NAD specificity by the unique mechanism of substituting inorganic phosphate for the 2$ sp prime$-phosphate of NADP. The role of Mg$ sp{++}$ in this process is not understood, but it is required for activity with both dinucleotides. This enzyme is proposed to be a mammalian homolog of the product of the yeast MIS1 gene, functioning in providing formyltetrahydrofolate for the synthesis of formylmethionyl tRNA required for the initiation of protein synthesis in mitochondria. The two activities of this bifunctional enzyme are resolved significantly by diethylpyrocarbonate modification of at least two histidines. Most striking is the ability to activate the cyclohydrolase up to 50% independently of the dehydrogenase by modification with diethylpyrocarbonate in the presence of ligands: Mg$ sp{++}/$phosphate/NAD/folate.

Our search for protein tyrosine phosphatases (PTPases) that are involved in murine development led to the isolation of a novel PTPase called MPTP-PEST. To elucidate the function of MPTP-PEST, several biological aspects of the protein were investigated. It was determined that the MPTP-PEST enzyme is a stable cytosolic protein tyrosine phosphatase of 112 kDa that is ubiquitously expressed both in the adult and in the embryo. MPTP-PEST is composed of a single amino-terminus catalytic domain which is active against phosphorylated substrates in vitro. The gene structure and chromosomal localization of MPTP-PEST were determined using a series of $ lambda$ phage clones isolated from a mouse genomic library. Analysis of the MPTP-PEST locus indicated that the gene spans over 90 kb of the mouse genome and is composed of 18 exons, 10 of which constitute the catalytic phosphatase domain. Fluorescence in situ hybridization with MPTP-PEST genomic DNA defines the map position of MPTP-PEST to mouse chromosome 5 region A3-B. To gain mechanistic insights into the function of MPTP-PEST, the association of proteins with MPTP-PEST was investigated using several different in vitro and in vivo binding assays. It was shown that the proto-oncoprotein SHC and the adaptor protein Grb2 associate with MPTP-PEST through specific carboxy-terminus sequences. In addition, the SHC/MPTP-PEST association was shown to be mediated by a novel type of protein-protein interaction. Both SHC and Grb2 function downstream of receptor type and cyloplasmic protein tyrosine kinases and have been shown to mediate several signal transduction events. The association of MPTP-PEST with these signaling proteins may therefore represent a function for MPTP-PEST in signaling events.

Translation is regulated on multiple levels in the cell. The complexity of translation in eukaryotes has led to the evolution of numerous factors with major or minor effects on the final level of mRNA expression. These regulatory factors can be in cis or trans to the mRNA with mRNA-specific or non-specific effect. As the 3' tail of the eukaryotic mRNA, poly(A) is one of the cis regulatory factors of mRNA, significantly affecting its expression level. Poly(A) has been shown to have synergistic effect with another cis element of the eukaryotic mRNA, 5' cap structure. Among the protein factors that affect mRNA translation, PABP has the most fundamental roles. As the binding partner of poly(A), PABP is involved in the formation of the pre-initiation complexes. The two binding partners of PABP, Paip1 and 2 regulate the rate of translation. Here, using structural biology techniques, such as X-ray crystallography, I have studied the structure of the poly(A) RNA and the assembly of PABP complexes. I further quantified these interactions using thermodynamic analyses and dissected the mechanisms of the complex formation. I could show the self assembly of poly(A) and cooperative assembly of a ternary complex of poly(A)/PABP/eIF4G. The latter is regulated by the interdomain allostery of PABP, which is further controlled by its interactions with Paip1 and 2. These studies suggest the importance of inter- and intra-molecular interactions in regulating cellular processes. / La traduction est régulée à de multiples niveaux dans les cellules. La complexité de la traduction chez les eucaryotes a conduit à l'évolution de nombreux facteurs qui contrôlent l'expression des ARNs messagers (ARNm). Ces facteurs de régulation peuvent fonctionner en cis ou en trans avec des effets spécifiques ou non-spécifiques pour les ARNm. À l'extrémité 3' de l'ARNm eucaryote, la queue poly(A) est l'un des facteurs de régulation cis de l'ARNm qui affecte de manière significative son niveau d'expression. La queue poly(A) a un effet synergique avec un autre élément de l'ARNm eucaryote: le complexe de pré-initiation situé à l'extremité 5'.Parmi les facteurs protéiques qui affectent la traduction des ARNm, la PABP joue les rôles les plus fondamentaux. Comme partenaire de liaison avec le poly(A), la PABP est impliquée dans la formation des complexes de pré-initiation. Les protéines Paip1 et Paip2 se lient à la PABP pour contrôler le niveau de traduction. Ici, par des techniques de biologie structurale, comme la cristallographie par rayons X, j'ai étudié la structure de l'ARN poly(A) et l'assemblage des complexes de la PABP. J'ai également quantifié ces interactions au moyen d'analyses thermodynamiques ce qui a permis de dévoiler les mécanismes de la formation du complexe. J'ai pu démontrer l'auto-assemblage du poly(A) et l'assemblage coopératif d'un complexe ternaire de poly(A)/PABP/eIF4G. Ce dernier est assujetti à l'allostérie inter- domaines de PABP, qui est en outre contrôlée par ses interactions avec Paip1 et Paip2. Ces études suggèrent l'importance des interactions inter- et intra-moléculaires dans la régulation des processus cellulaires.