Coactivators for the superfamily of nuclear receptors are defined as factors that enhance their transcriptional activity. Most of these coactivators exert their action through the receptor ligand-dependent activation function 2 (AF-2). This interaction occurs between a coactivator LXXLL motif (NR-box) and a hydrophobic cleft located on the ligand-binding domain of the receptor. Here we describe the cloning and characterization of CIA, a novel C&barbelow;oactivator I&barbelow;ndependent of A&barbelow;F-2 function. CIA displays specific interaction with the RVR orphan nuclear receptor and both specific interaction and transcriptional coactivation potential with the estrogen receptors (ER) alpha and beta. The interaction with ERalpha and beta is strongly enhanced by its natural ligand, estradiol (E2) and surprisingly also by pure antiestrogens EM 800 and ICI 164,384. While the E2-dependent CIA-ERalpha interaction requires an intact CIA LXXLL motif, CIA also interacts with an AF-2 null mutant of ERalpha. Therefore, CIA constitutes the first example of a novel type of ligand-dependent but AF-2 independent nuclear receptor coactivator that may play a specific role in the ER physiology via selective ER modulators (SERMs).

Nuclear receptor coactivators are factors that enhance the transcriptional activity of the receptor. Most coactivators characterized to date work in a ligand-dependent manner by interacting with the activation function-2 (AF-2) of the receptor. A newly characterized coactivator termed steroid RNA activator (SRA) was cloned as an AF-1-dependent and ligand-independent coactivator. SRA was shown to enhance the transcriptional activity of the steroid receptors by interacting with their AF-1. Here, we describe the effect of SRA on the transcriptional activity of the estrogen receptor alpha (ERalpha) and beta (ERbeta). SRA coactivates ERalpha and ERbeta in a ligand-dependent manner and the transcriptional activity of ERalpha can be enhanced by SRA in a ligand-independent manner through the AF-1 domain of the receptor. The integrity of the serine residue at position 118 (S118) in ERalpha AF-1 is required for the complete coactivation of ERalpha by SRA. Activation of MAPK induces ligand-independent activation of ERalpha by SRA, a mechanism independent of the AF-2 activity of the receptor. Mutation of S118 to an alanine residue abolishes the MAPK-induced activation of ERalpha, suggesting that phosphorylation of S118 by MAPK may be important for the ligand-independent effect of SRA on ERalpha AF-1. Thus SRA may play an important role in the regulation of ER transcriptional activity.

Gap junctions consist of a hexamer of connexin proteins which are involved in intercellular communication. Connexin 43 (Cx43) is one of the most abundantly expressed connexins and its expression, either at the RNA or protein level, is often disregulated in most cancers. Pseudogenes are sequences in the genome which are homologous to other genes however, they are often considered to be inactive due to the accumulation of frameshift mutations and/or deletions. The connexin 43 pseudogene (ΨCx43) is 97% identical in the open reading frame to Cx43. It was originally thought that ΨCx43 had no function. In chapter 2, I report that ΨCx43 is transcribed in breast cancer cell lines, and can be translated both in vitro and in vivo. In addition to this, it was determined that overexpressing ΨCx43 results in growth inhibition, similar to Cx43, however it does not enhance gap junction intercellular communication. In chapter 3, I investigate whether endogenous ΨCx43 is translated by determining if the pseudogene is present in polysomal RNA. Using Hs578T cells (no endogenous ΨCx43) and MDA 435 (high level of ΨCx43) cells, ΨCx43 transcript was found in MDA 435 polysomal RNA and not in the Hs578T cells. Upon transfecting Hs578T cells with ΨCx43, a shift in the Cx43 polysomal transcripts towards the monosomes and early polysomes was observed indicating a possible regulation of Cx43 by its pseudogene. Using siRNA to knock down both Cx43 and ΨCx43 in MCF7 cells, which endogenously express both Cx43 and ΨCx43, an increase in Cx43 RNA and protein was demonstrated in response to ΨCx43 siRNA treatment. Furthermore, since previous reports indicated an increase in sensitivity of canc / Les jonctions communicantes consistent d'un hexamère de protéines nommées connexin qui sont impliquées dans la communication intercellulaire. Parmi les membres de la famille des connexins, Connexin 43 (Cx43) est celle dont l'expression cellulaire est la plus abondante. Tant au niveau de l'ARN qu'au niveau protéique, l'expression de Connexin 43 est déréglée dans la grande majorité des cancers. Les pseudo gènes sont des séquences dans le génome qui sont homologues à d'autres gènes. Ces pseudo gènes sont souvent considérés inactifs du à l'accumulation de mutations de changements de phases et/ou délétions. Le cadre ouvert de la lecture du pseudo gène de connexin 43 (ΨCx43) est 97% identique à celui de Cx43. Par le passé, la fonction de ΨCx43 était inconnue. Dans le chapitre 2, je démontre que ΨCx43 est transcrit dans les lignées cellulaires de cancer du sein, et peut être traduit autant in vitro que in vivo. De plus, il a été démontré que la surexpression de ΨCx43 cause l'inhibition de la croissance, un effet similaire à celui de Cx43, mais par contre, ΨCx43 ne peut augmenter la communication intercellulaire à travers les jonctions communicantes. Dans le chapitre 3, j'investigue si la protéine endogène de ΨCx43 est traduite en déterminant si le pseudo gène est présent dans l'ARN polysomal. En utilisant les cellules Hs578T (ne contiennent pas de protéine ΨCx43 endogène) et les cellules MDA 435 (contiennent un niveau élevé de ΨCx43), le produit de la transcription de ΨCx43 a été détecté dans l'ARN polysomal des cellules MDA 435 et pas dans celui des cellules Hs578T. Lors de la transfection de ΨCx43 dans les cellules Hs5

E1B 19kDa protein (19K), a member of Bcl-2 family of protein, is an adenovirus early gene product. It functions to suppress apoptosis of the host cell caused by viral infection. The exact role of E1B 19kDa in regulating apoptosis in response to many different death stimuli has not been investigated. / We found that hygromycin B, an aminoglycoside antibiotic that is used to establish stable mammalian cell lines, kill cells by apoptosis. Apoptosis induced by Hygromycin B is p53-independent and can be blocked both by E1B 19kDa and Bcl-2, showing that E1B 19kDa, like Bcl-2, may confer protection against apoptosis in a variety of circumstances. / The sequence similarities between E1B 19kDa and Bcl-2 are largely restricted to the BH1 and BH3 domains. Our sequence alignment analysis shows that E1B 19kDa also contains a BH4 region overlapping its BH3 domain. We find that the BH1 domain of E1B 19kDa contributes to its anti-apoptotic function and its ability to interact with Bax. A conserved residue Gly 87 in the BH1 domain of E1B 19kDa that is critical for Bcl-2 and Bax interaction is also important for the E1B 19kDa and Bax interaction. We do not detect an interaction between E1B 19kDa and Bad, suggesting that E1B 19kDa is a structural and functional homolog of Bcl-2 but also has its unique features. / Bap31, a 28kDa polytoptic integral protein of the endoplasmic reticulum, is part of an ER protein complex that also includes Bcl-2/Bcl-xL, procaspase-8, and a CED-4-like protein. E1B 19kDa does not interact with Bap31; nor does it interact with Bcl-2/Bcl-xL. We have made a set of chimeric E1B 19kDa proteins carrying different BH domains from Bcl-2 to test their protein-protein interaction properties. We show that the BH3 domain of Bcl-2, when substituted for the homologous region of E1B 19kDa, confers the properties of interaction with Bap31 and Bcl-xL onto 19K. The subcellular distribution and antiapoptotic function of the 19K/Bcl-2BH3 is very similar to that of wild type 19K. These results show that the BH3 domain of Bcl-2/Bcl-xL has an important role in mediating Bcl-2/Bcl-xL and Bap31 interaction.

Carcinoembryonic antigen (CEA) family members in humans can be divided into two subgroups according to the type of anchorage to the cell membrane: Glycophosphatidyl inositol (GPI)-linked and transmembrane (TM) linked. The mode of membrane anchorage can profoundly affect the biological functions of CEA family molecules, including the ability to inhibit differentiation. All information so far, suggests that the GPI-anchored CEA family members evolved recently presumably from a more primordial TM-anchored family member such as CEACAM1. The results of this study indicate that very few mutations in the TM domain are required to effect this change. The introduction of a stop codon in the CEACAM1 TM domain at the position corresponding to that of all GPI-linked CEA family members plus two more amino acid changes in this domain resulted in highly efficiently GPI-processed protein. The GPI-anchored mutant CEACAM1 protein is still capable of mediating cell-cell adhesion but, unlike TM-linked CEACAM1, could block myogenic differentiation. Screening of genomic DNA from various mammalian groups by PCR showed convergent evolution of GPI anchorage in the CEA family, in that it evolved twice independently by different packages of mutations during the primate radiation. The same group of GPI-generating mutations as in human CEA family members evolved in the common ancestor of primitive primate groups of tarsiers and New World monkeys, and a novel group of GPI producing mutations evolved later in Cebidae family of New World monkeys. These independent groups of mutations that gave rise to the same structural feature imply adaptive evolution of GPI-anchor acquisition in CEA family, This supports the notion that GPI-anchoring of CEA family members is functionally important in vivo and therefore, has been positively selected during evolution. A chimeric cDNA of the novel GPI anchor-generating TM domain linked to the extracellular domains of human CEACAMI-4L was efficiently processed t

This work is aimed at understanding the mechanisms that underlie the functions of human immunodeficiency virus type 1 (HIV-1) nucleocapsid protein (NCp7) during reverse transcription initiation as well as during specific viral genomic RNA encapsidation. To pursue the first goal, we developed an in vitro reverse transcription reaction system, in which synthetic NCp7 was employed to place natural tRNALys3 onto an HIV-1 RNA template containing the primer binding site (PBS) and its flanking sequences. After placement, NCp7 was removed by Proteinase K digestion and phenol: chloroform extraction, so that the function of NCp7 in annealing could be separated from that in subsequent reverse transcription. Low concentrations of dNTPs were used in these reactions in order to detect very early pausing events during the initiation stage of reverse transcription. The results showed that initiation of HIV-1 (-) strand DNA synthesis was characterized by early pausing events at the +1 and +3 nt positions in the cell-free reverse transcription system. The +1 pausing was associated with the use of an RNA primer, while the +3 pausing event was attributable to the existence of a stem-loop structure upstream of the PBS. A mutant NCp7 molecule, devoid of Zn fingers, was able to place tRNALys3 onto vRNA as efficiently as wt NCp7; however, the tRNALys.3:vRNA complex formed with the assistance of wt NCp7 was far more efficient than that formed by the mutant NCp7 in both initiation and the subsequent transition from initiation to elongation during the synthesis of (-) strand DNA. / To study the role of NCp7 in viral RNA incorporation, we deleted the cis-acting packaging signals SL1 and SL3, which are located up- and down-stream of the 5' major splice donor (SD) site, respectively. These deletions led to defected viral replication kinetics. Long-term culture of these mutated viruses gave rise to second-site mutations in the p2 and NC domains of the Gag protein, that were able to rescue the aforementioned deletions. We further deliberately replaced the corresponding residues within p2 or NC by each of 19 other amino acids in the context of the SL1 deletion, and found that only hydrophobic amino acids with long side chains (including V, L, I, and M) were able to rescue this deletion. In addition, it was observed that compensation of the SL1 and SL3 deletions involved changes at different residue positions in the p2 and NC proteins, indicating that RNA packaging signals up- or down-stream of the 5' SD site bind to Gag protein in different ways during viral RNA packaging.

Polyomavirus large T antigen binds to four adjacent sites on viral DNA, denoted 1/2, A, B and C. Binding to site 1/2 leads to the initiation of viral DNA replication. RNA polymerases transcribing the late strand of the circular, 5.3 kb viral DNA are blocked just upstream of the replication origin and stalled RNA polymerases accumulate in that region (Skarnes et al., J. Mol. Biol. 203: 153-171, 1988). This accumulation of stalled RNA polymerases depends on binding of large T antigen to site A but not B or C (Bertin et al., J. Virol. 67:5766-5775, 1993). To investigate whether this accumulation depends on sequences other than site A, we cloned an insert containing sites A and 1/2, and the adjacent enhancer region, in the small T antigen intron or between two polyadenylation signals of a polyomavirus mutant. / Most of the insertion mutants located in the small T antigen intron were nonviable. However, those located between the two polyadenylation signals were viable. Although stalling did not occur within the inserts, a major transcript within the replication origin was found while examining runon RNAs with viral transcription complexes isolated from cells infected with insertion mutants. Labelled RNA was hybridized with a single-stranded DNA, complementary to late-strand RNA, that extends through the replication origin; a major 80-nt RNA species was protected against nuclease treatment. Hybridization with two other DNAs that extend 18 and 41 nt farther in the direction of late transcription protected 98- and 121-nt RNAs, respectively. These results place the 5$ sp prime$ end of this transcript near the early side of site 1/2. Control experiments were carried out in which labelled RNAs synthesized in vitro were hybridized with the same DNAs; no RNAs with these 5$ sp prime$ ends were detected. / S1 mapping of nuclear RNA from virus-infected cells detected no such transcript. The RNA polymerases that synthesize this RNA could terminate in vivo shortly downstream of the initiation site, or the RNA could be unstable. This "ori-RNA" is sensitive to $ alpha$-amanitin, resistant to aphidicolin, and independent of functional large T antigen. Ori-RNA synthesis could help stabilize the unwinding of DNA at the replication origin and enhance initiation of replication.

The c-myc gene was previously shown to be regulated by a conditional block to transcription elongation and sequences from its promoter were implicated in this regulation. The objective of this project was to define the promoter elements involved in the control of transcription elongation. Using heterologous promoter constructs, the ME1a1 protein binding site located in the c-myc promoter was shown to be required for the block to transcription elongation. From mutation analysis, a correlation was established between the binding of nuclear factors to ME1a1 sites in vitro and the ability of these sites to confer block to transcription elongation in vivo, strongly implicating trans-acting factors in this process. Fractionation studies demonstrated that three nuclear factors interact with the ME1a1 site, thus generating three protein-DNA complexes termed "a", "b", and "c". These factors were characterized and a cDNA encoding the protein responsible for complex "c" was isolated. This protein was shown to represent the human homologue of the Drosophila Cut homeodomain protein (hu-Cut) and to repress expression from the c-myc promoter in transient transfection assays.

Viral enzymes that catalyze replication of the viral genome remain the main subject of currently available antiviral therapies. This work focuses on three anti-viral agents that target viral DNA polymerases: foscarnet, acyclovir and entecavir. Though the broad spectrum of antiviral activities of foscarnet has been recognized for decades, only recently has it been shown that the anti-herpesvirus drug acyclovir and the anti-hepatitis B virus drug entecavir are also active against the human immunodeficiency virus (HIV) 1. The clinical benefits of the multiple antiviral activities of foscarnet have been demonstrated in treating HIV/herpesvirus co-infections. Foscarnet is currently approved for treating human cytomegalovirus infections. But the precise molecular mechanism of foscarnet action and resistance remains poorly understood. The study of foscarnet's mode of action against the HCMV DNA polymerase (UL54) is complicated in part by the difficulty in expressing and purifying the viral enzyme. To address the first issue, I conducted biochemical studies of foscarnet/UL54 interactions using an unpurified viral enzyme. I then proposed a model based on these studies for the presumptive foscarnet binding site within UL54. To address the second issue I generated the RB69 bacteriophage/HCMV polymerase chimera, which is easily purifiable and displays the UL54 phenotype with respect to foscarnet and acyclovir. The recent discovery of the anti-HIV activity of acyclovir has been followed by a report showing the selection of the V75I mutation within HIV-1 reverse transcriptase (RT) under the selective pressure of acyclovir. My own biochemical studies have revealed that the major effect of V75I substitution involves reducing the catalytic rate of acyclovir incorporation. The recent report of entecavir anti-HIV-1 activity has demonstrated the selection of M184V-containing HIV-1. Subsequent biochemical studies have reveal / Les enzymes virales qui sont responsables pour la réplication du génome viral constituent l'objet principal des thérapies antivirales actuelles. Ce travail se concentre sur trois agents anti-viraux qui inhibitent les polymérases d' ADN virales. Malgré le fait que plusieurs activités antivirales du foscarnet sont connues depuis des décennies, il a éte démontré seulement récemment que l'acyclovir, un agent contre le virus de l'herpes, et entecavir, un agent contre le virus de l'hepatite B, ont une activité anti-VIH-1. Les avantages cliniques des multiples activités antivirales du foscarnet dans le traitement des co-infections VIH/herpes ont été démontrés. Le foscarnet est présentement approuvé pour traiter les infections du cytomégalovirus humain. Le mécanisme précis de l'action antivirale du foscarnet, ainsi que le mécanisme de résistance au foscarnet demeurent mal compris. L'étude du mécanisme d'action de foscarnet contre l'ADN polymérase du VCMH (UL54) est compliquée en partie par deux problématiques : l'expression de l'enzyme virale et la purification de cette protéine. Pour adresser le premier problème, j'ai poursuivis des études biochimiques sur les interactions foscarnet-UL54 en utilisant une enzyme virale non-purifiée. En se basant sur ces études, j'ai ensuite proposé un model du site de liaison du foscarnet. Pour addresser le deuxième problème, j'ai généré une protéine chimère bactériophage/HCMV polymérase. Cette protéine est facilement purifiable et possède le profile de la UL54 polymérase envers foscarnet et acyclovir. La découverte récente de l'activité anti-VIH de l'acyclovir a été suivie d'une étude démontrant la sélection de la mutation V75I dans la trans criptase inverse (TI) du VIH-1 sous la pression sélective de l'acyclovir. Mes propres études biochimiques ont révélées que l'effet majeur de la substitution V75I implique la réduc

RNA interference (RNAi) is a gene silencing mechanism occurring in a wide variety of organisms. Different RNAi pathways exist in C. elegans for the regulation of various endogenous processes. The ERI-mediated RNAi pathway involves the ERI proteins and C. elegans Dicer, DCR-1, suggested to be involved in initiating endogenous RNAi. In this study, it was observed that many of the ERI proteins co-fractionate with DCR-1 in a complex of ~885 kDa in size. It was shown that the RdRP RRF-3 and the Tudor domain protein ERI-5 require each other in order to associate with DCR-1, but independently of ERI proteins ERI-1 and ERI-3. Furthermore, it was also observed that the helicase DRH-3 requires RRF-3 and ERI-5 in order to associate to DCR-1. The data suggested that at least two distinct binding sites for ERI proteins are found on DCR-1. Preliminary GST pulldown data confirmed distinct binding regions on DCR-1 for ERI-5 and ERI-3. / L'ARN interference (RNAi) est un mécanisme de "silencing" de l'expression génique se produisant chez de nombreux organismes. Chez C. elegans, il existe différentes voies RNAi régulant de nombreuses fonctions biologiques. La voie dépendant d'ERI implique les protéines de la famille ERI et DCR-1, l'orthologue de Dicer chez C. elegans, qui pourraient être engagées dans l'initiation du processus RNAi endogène. Durant mon Master, j'ai pu observer que beaucoup de protéines ERI co-fractionnent avec DCR-1 au sein de complexes de 885kDa. La protéine RdRP RRF-3 et ERI-5, une protéine possédant deux domaines Tudor, sont interdépendantes pour l'association avec DCR-1 mais de manière indépendante d'ERI-1 et ERI-3. En outre, j'ai pu également démontrer que l'hélicase DRH-3 requiert RRF-3 et ERI-5 pour se fixer à DCR-1. L'ensemble de ces données suggère qu'il existe au moins deux sites de fixation aux protéines ERI sur la protéine DCR-1. Enfin, mes résultats préliminaires de GST-pulldown semblent confirmer la fixation d'ERI-3 et ERI-5 au niveau de régions distinctes de DCR-1.