Cardiac diseases are one of the leading causes of mortality in the western world. However, the molecular mechanisms responsible for these malformations or dysfunction remain poorly characterized. We have used the B type natriuretic peptide (BNP) as a marker to identify novel transcriptional pathways that regulate cardiac gene expression. Functional analysis of the BNP promoter had previously led to the identification of positive cis-regulatory elements and the isolation of the GATA-4 protein, which is an important trans-activator of several cardiac genes. The present study demonstrates that cell specific transcription from the BNP promoter is dependent on a dominant silencer activity in the distal upstream regulatory region, which serves to restrict its ectopic expression both, in vitro and in vivo. We have identified two distinct repressor domains R1 and R2 and show that while both repressor domains contribute additively to silencing of the native promoter, R2 alone is sufficient to repress the BNP promoter in heterologous cells. R2 is conserved across species and displays significant homology to a neuron restrictive silencer element (NRSE). We demonstrate sequence specific DNA binding activity on R2 that is present in a variety of non cardiac cells where the gene is repressed, but absent in cardiomyocytes, reiterating that repression is an important mechanism in the transcriptional regulation of BNP. / We have also investigated the potential role of cardiac enriched (GATA-4) and ubiquitous factors in regulation of BNP. We report that YY1, a known transcriptional repressor of muscle gene activity, is a positive regulator of BNP, and that it can synergise with the cardiac subgroup of GATA proteins (GATA-4 and GATA-6) to augment transcription. Collectively this work demonstrates that tissue restricted expression of BNP in vivo, is a result from the combinatorial interactions of both positive and negative factors, which include on the one hand, the GATA proteins that can activate the promoterin cardiac cells in conjunction with a variety of cell restricted and ubiquitous factors such as YY1 and on the other hand, repressor(s) that restrict its expression in non cardiac cells. / Our results are among the first to clearly demonstrate the in vivo importance of repressors in cardiac specific transcription, and emphasize how transcriptional diversity can be created by a limited set of transcription factors.

The basic-Helix-Loop-Helix PAS (bHLH-PAS) protein family is a growing family of transcription factors. Included in this family is the Drosophila single-minded (DSim) gene, which is critical for the development of the CNS midline cells. / We have isolated a murine homologue of the Drosophila Sim gene, mSim-2. The murine and Drosophila gene products share a high degree of homology in the bHLH and PAS regions but not within the carboxy terminus. Northern blot and RT-PCR analysis of adult murine tissues revealed that the mSim-2 is expressed at the highest levels in the kidneys and at lower levels is present in skeletal muscle, lung, testis, brain, and heart. In situ hybridisation experiments demonstrate that mSim-2 is also expressed in early fetal development in the central nervous system and in cartilage primordia. / We investigated the ability of mSIM-2 to associate with the Arnt gene product, a common dimerisation partner of a number of bHLH-PAS proteins. We found that the HLH and PAS motifs of both proteins are required for optimal association. We demonstrated the presence of two separable repression domains within the carboxy terminus of mSIM-2, in contrast to dSIM, which is a transactivator. We find that mSIM-2 is also capable of repressing activation by its binding partner, ARNT. We also demonstrate that mSIM-2 can functionally interfere with another bHLH-PAS transcription factor, HIF-1alpha, by competing for ARNT binding, providing a second mechanism by which mSIM-2 may inhibit transcription. / We also investigated the effects of mSIM-2 and its close paralogue mSIM-1, as heterodimers with ARNT, on reporter constructs containing native DNA binding sites. We find that mSIM-1 can effect transcriptional activation through it's association with ARNT. In contrast, mSIM-2/ARNT does not activate, as the mSIM-2 repression domains quench ARNT transactivation. We also find that the mSIM-2 can interfere with mSIM-1 mediated transactivation by competing for dimerisation with ARNT and for DNA binding site occupation. Our results suggest that mSIM-1 and mSIM-2 have similar dimerisation and DNA binding properties but different transcriptional effects and may therefore antagonise each other which may in turn be a mechanism of gene regulation by these two proteins.

Pax-3 is a transcription factors important in normal embryonic development, as highlighted by its mutation in independent alleles of the Splotch mouse mutant and human Waardenburg syndrome, both characterized by pigmentary disturbances in the heterozygous state and limb muscle defects in the homozygotes. Pax-3 contains two structurally independent DNA-binding domains, a paired domain and a homeodomain. Through the analysis of a number of naturally occurring mutations, we have shown that both DNA-binding domains are functionally interdependent, as independent mutations in either domain can affect the DNA-binding activity of the other. In support of this finding, modeling the deleterious mutations into the three dimensional structure of the paired domain suggests that a series of consecutive phosphate contacts are essential for DNA-binding by both the paired domain and homeodomain of Pax-3. In addition, the paired domain can regulate the DNA-binding specificity and dimerization potential of the homeodomain and we have exploited the possibility to transfer this effect to a heterologous homeodomain to identify the protein segments involved in this functional interaction. Characterization of a series of chimeric proteins containing stepwise deletions within the paired domain provided two key findings: (i) the C-terminal subdomain of the paired domain does not play a major role in the regulation of the homeodomain DNA-binding and (ii) the N-terminal subdomain and, in particular, the second alpha-helix are essential for the modulation of homeodomain DNA-binding. In addition to intermolecular interactions, the functional association with other transcription factors can also modulate Pax-3 target gene selection and regulation. We found that Pax-3 can recruit PEA3, a member of the ETS family, to form a ternary complex with DNA sequences containing either a paired domain or homeodomain recognition site and a sub-optimal Ets site to which PEA3 normally does not bind. We have

Cystic Fibrosis (CF), the most common fatal genetic disorder, is due to mutations of the cystic fibrosis transmembrane regulator protein (CFTR). Loss of functional CFTR leads to mucus thickening with the consequences of chronic pulmonary infection and inflammation, the most common fatal complication of this disorder. This thesis aimed to examine one aspect of the pathophysiology of CF, the potential role of glutathione (GSH) and oxidant stress. CFTR is permeable to GSH, at least in vitro, where evidence of the ability of CFTR to transport GSH is strong. Furthermore, extracellular GSH is diminished in CF airway lining fluid and it has been hypothesized that deficiency of CFTR leads to higher intracellular GSH coupled with lower extracellular GSH. This would be consistent with the observation of low extracellular GSH in CF patients and increased resistance to apoptosis of CF cell lines, where falling intracellular GSH concentrations ([GSH]) are required to initiate apoptosis. In addition to CFTR, other channels transport GSH, including Multi-drug Resistance Protein 1 (MRP1) which has structural and functional homology to CFTR. Higher expression of MRP1 has been associated with fewer symptoms in CF patients, suggesting that MRP1 may contribute to CF pathophysiology. We therefore set out to study GSH at the cellular level, with or without functional CFTR, in an airway epithelial cell model. Given the importance of oxidant stress in CF, we tested whether cells were more susceptible to stress in the absence of CFTR function. We hypothesized that lack of CFTR function would increase intracellular and decrease extracellular [GSH]. We also hypothesized that MRP1 function would influence intracellular and extracellular [GSH]. Blockade of CFTR function in Calu-3 cells failed to alter either intra- or extracellular [GSH], independent of oxidant stress conditions, suggesting the channel was not a determinant of [GSH]. However, inhibition of other anion channels did increase / La fibrose kystique (FK) est causée par des mutations dans la protéine CFTR (pour « Cystic fibrosis transmembrane conductance regulator »). La perte de fonction de CFTR mène à l’épaississement des muqueuses des voies respiratoires et l’avènement d’infections pulmonaires chroniques avec réponse inflammatoire, la cause primaire de mortalité chez les personnes affligées de FK. Cette thèse va examiner une partie de la pathophysiologie de la FK, celle du rôle du glutathion (GSH) et du stress oxydatif. Le CFTR est perméable au GSH et les données expérimentales in vitro suggèrent que le CFTR puisse transporter le GSH. De plus, le GSH extracellulaire est diminué dans le liquide de surface recouvrant les voies respiratoires des patients avec la FK et de cela découle l’hypothèse que la déficience en CFTR mène à une augmentation du GSH intracellulaire, avec une diminution concomitante du GSH extracellulaire. Cette hypothèse s’alignerait bien avec le bas taux de GSH extracellulaire observé chez les patients affectés par la FK et la résistance accrue contre l’apoptose vue dans les lignées cellulaires de la FK. Cette résistance découle du fait que la cellule a besoin d’une diminution de concentration de GSH ([GSH]) intracellulaire pour entamer le processus d’apoptose. À part du CFTR, il existe d’autres canaux membranaires qui transportent le GSH, tels que le Multi-drug Resistance Protein 1 (MRP1). Ce canal a une forte homologie structurelle et fonctionnelle avec le CFTR et son expression se trouve liée avec moins de symptômes cliniques chez les patients atteints de FK. Ces observations suggèrent que le MRP1 pourrait jouer un rôle important dans l’évolution de la physiopathologie de la maladie. Notre but est d’étudier l’état du GSH du point de vue de la cellule, avec ou sans stress oxydatif, dans une lignée de cellules épithéliales des voies respiratoires. Considérant l’importance du stress oxydatif dans la$

Carcinoembryonic antigen (CEA) is a tumor marker used clinically to assess post-operative recurrences of breast, lung and colon cancers. The CEA gene family, which is part of the immunoglobulin superfamily, is composed of several proteins crossreactive with anti-CEA antibodies. Carcinoembryonic antigen appears to function in the development of the gastro-intestinal tract as well in tumor formation; it is capable of mediating cell-cell adhesion in vitro which is consistent with its putative role of maintaining tissue architecture in vivo. CEA gene family members have also been identified in various tissues of the mouse. Two of the mouse proteins, mmCGMla and mmCGMlb, have been characterized. By sequence homology, mmCGMla and mmCGMlb are the mouse homologues of human biliary glycoprotein and of rat hepatocyte ecto-ATPase. Both of the mouse CEA related proteins function as adhesion molecules when expressed on the cell surface of transfectant cells; however, mmCGMla, unlike mmCGMlb, mediates cellular aggregation irrespective of calcium concentration or temperature. Sequence comparison of mmCGMla, mmCGMlb, and other cDNAs isolated by polymerase chain reaction techniques, demonstrates that there are at least eight possible transcripts encoding CEA-related proteins and that these transcripts are all produced by alternative splicing of one precursor messenger RNA.

A 1797 bp human EcoRI satellite II DNA sequence was cloned in vectors containing the thymidine kinase gene (HSV-1 tk) and the neomycin resistance gene, and introduced in a cell line deficient for these genes. We have observed that the electroporation efficiency of these plasmids depends on the location and/or the orientation of the satellite sequences within the transfected plasmid. Only one plasmid, pCFD1 containing one satellite fragment close to the neo gene, inhibited the formation of TK$ sp+$/NEO$ sp+$ transfectants. We have also shown that the instability of the TK$ sp+$ phenotype which was observed did not correlate with the presence of adjacent satellite DNA. In contrast, satellite DNA sequences within the transfected plasmid somehow interfered with the generation of stable TK$ sp+$ transfectants. / Moreover, we have detected (both in nuclear and partially purified HeLa whole cell extracts) the presence of proteins that specifically bind the human 1797 bp satellite II DNA sequence. Four proteins with molecular weights of 100, 93, 77 and 34 kDa were identified and named Satellite DNA-binding protein, Sbp-1, -2, -3 and -4, respectively. The function of these proteins is, as yet, unknown.

A previously identified rat cDNA (S1) that shares 78% nucleotide homology and a predicted 92% amino acid sequence homology with human EF-1$ alpha$ was used to express S1 in E. coli and to generate a polyclonal antibody to pS1. A recombinant plasmid pGEX-2T-S1 was constructed, containing the glutathione S-transferase gene. The expressed fusion protein was purified and digested with thrombin to produce a recombinant S1 protein (rpS1) containing slightly modified N-terminus. Purified rpS1 was used to raise a rabbit antiserum which recognized rpS1 on immunoblots. A polyclonal antiserum to EF-1$ alpha$ failed to react with rpS1. Similarly the anti-rpS1 does not recognize EF-1$ alpha$ on immunoblots. Anti-rpS1 therefore is able to distinguish pS1 from EF-1$ alpha$ despite their extensive amino acid sequence homology. Anti-rpS1 and anti-Ef-1$ alpha$ will be used to study the similarities and differences between pS1 and EF-1$ alpha$ in vivo and in vitro.

Growth hormone (GH) and IGF-I are essential for somatic growth, while exerting distinct effects on energy homeostasis. GH controls IGF-I production, but whether IGF-I exclusively mediates the growth effect is still debated. To explore the in vivo interactions of GH and IGF-I, we crossed IGF-I-overexpressing mice (MT-IGF) onto the GHR-/- background. We found that bitransgenic mice (GHR-/- and MT-IGF positive) exhibited fully restored body weight, lumbar, and body lengths, and normalized bone area, mineral content, and density. IGF-I overexpression also corrected glucose intolerance in GHR-/- mice. Significant changes in fasting glucose levels, glucose tolerance, lean/fat mass, and adipose histology were also observed. Intriguingly, on a GHR+/- background, overexpression of IGF-I significantly increased body weight. Our results establish that the growth defect caused by GHR gene deficiency can be restored by increasing IGF-I in vivo, and strongly supports the notion that IGF-I mediates most of the growth promoting effects of GH. / L'hormone de croissance et IGF-I jouent un rôle essentiel dans la croissance somatique et sont également impliques dans l'homéostasie énergétique. L'hormone de croissance (GH) contrôle la production d'IGF-I, mais l'action exclusive d'IGF-I sur la croissance reste discutée. Pour étudier in vivo les interactions entre GH et IGF-I, nous avons croisé un modèle de souris transgénique surexprimant IGF (MT-IGF) avec des souris knock-out GHR-/-. Nous avons trouvé une récupération complète du poids corporels, de la longueur des lombaires et de la taille des souris, ainsi qu'une normalisation de la composition minérale et de la densité osseuses, chez les souris bitransgeniques (GHR-/- et MT-IGF). Il y avait également une correction de l'intolérance au glucose chez les souris bitransgenique. Des changements significatifs de la glycémie à jeun, de la tolérance au glucose, des masses grasse et maigre ainsi que l'histologie du tissue adipeux ont également été observés. Étrangement, la surexpression d'IGF-I chez des souris transgéniques GHR+/-, augmente significativement leur poids corporel. Nos résultats établissent qu'une perturbation de la croissance induite par déficience en GHR peut être rétablie par l'augmentation in vivo d'IGF-I, et induisent fortement qu'IGF-I est un médiateur des effets promoteurs sur la croissance de GH.

Eukaryotic gene regulation by members of the nuclear receptor superfamily is crucial for a vast array of cellular and physiological processes including development, differentiation, proliferation and homeostasis. Regulation of the transcriptional activity of nuclear receptors occurs through the exchange of corepressor and coactivator complexes in a ligand-dependent manner. However, this may not be the case for orphan nuclear receptors, including RORalpha. Instead, RORalpha orphan nuclear receptor activity is modulated by Ca 2+/calmodulin-dependent protein kinase IV (CAMKIV), through a yet undetermined mechanism, and RORalpha itself regulates the expression of genes involved in the calcium signaling pathway. In this study, we report that RORalpha transcriptional activity is modulated by CaMKIV through direct phosphorylation of CIR, a novel nuclear receptor corepressor, thereby influencing its interactions with RORalpha. CIR exhibits ligand-oblivious and nuclear receptor AF-2 specific functional properties, similar to those of previously characterized RORalpha corepressor Hr. In contrast, the corepressors CIR and Hr display differential mechanisms of repression, comprising of HDAC-dependent and independent pathways, respectively. Furthermore, the calcium-dependent regulation of RORalpha transcriptional activity occuring through modulation of CIR does not extend to Hr, illustrating corepressor specificity. Thus, the molecular mechanisms governing active repression of RORalpha transcriptional activity involve distinct repression and signal transduction pathways, representing a dual mode of nuclear receptor regulation.

Numerous groups, including ours, have found that retinoids potently inhibit the growth of breast cancer cells, but the mechanisms by which growth regulation is achieved remains unclear. Although several of the effects of retinoids in breast cancer have been linked to the insulin-like growth factor (IGF) system, their effects on key signaling molecules in the IGF type-I receptor (IGF-IR) pathway have not been well characterized. This thesis project examined the hypothesis that retinoids mediate their growth inhibitory effects by targeting specific members of the IGF-IR signal transduction pathway. Although we did not observe regulation of IGF-IR itself, we found that all-trans retinoic acid (RA)-mediated growth inhibition is associated with a selective reduction in insulin receptor substrate 1 (IRS-1) protein and activity levels. We also present evidence that decreasing IRS-1 levels results in the selective down-regulation of the PI 3-kinase/AKT pathway in RA-treated MCF-7 cells. The relevance of IRS-1 regulation to the growth inhibitory action of RA is supported by the results showing that forced expression of IRS-1 abrogates the ability of RA to significantly inhibit MCF-7 cell growth. / Several studies have highlighted the importance of IRS-1 in breast cancer pathogenesis. High levels of IRS-1 in human breast tumors correlate with increased disease recurrence and constitutive IRS-1 signaling exists in breast tumors. This suggests that we may develop molecular strategies targeting IRS-1 by understanding the mechanisms controlling its expression and turnover. Since RA decreased IRS-1 protein levels without altering mRNA levels, we examined the hypothesis that RA-mediated regulation of IRS-1 levels was at the posttranslational level. Two proteasome inhibitors rescue the RA-mediated degradation of IRS-1, and RA increases the ubiquitination of IRS-1. We also found that RA increases the serine phosphorylation of IRS-1 and show that this occurs in a protein kinase C (PKC)-dependant manner, since PKC inhibitors block the RA-induced degradation and serine phosphorylation of IRS-1. We further demonstrate that RA activates PKC-delta in the sensitive, but not in the resistant cells, with a time course that is consistent with the RA-induced decrease of IRS-1. The involvement of PKC in the RA-mediated regulation of IRS-1 is supported by additional data showing that: (1) RA-activated PKC-delta phosphorylates IRS-1 in vitro, (2) PKC-delta and IRS-1 interact in RA-treated cells, and (3) mutation of three PKC-delta serine sites in IRS-1 to alanines results in no RA-induced in vitro phosphorylation of IRS-1. / Having identified IRS-1 as a novel target of RA and showing that RA regulates this protein via a mechanism involving the ubiquitin-proteasome pathway has contributed to an enhanced understanding of the effect of retinoids in human breast cancer cells.