Methylenetetrahydrofolate reductase (MTHFR) is an important enzyme in the metabolism of folate and methionine. MTHFR catalyzes the conversion of 5,10 methylenetetrahydrofolate (5,10-methyleneTHF) to 5-methyltetrahydrofolate (5-methylTHF). 5-MethylTHF is a co-substrate for homocysteine remethylation to methionine catalyzed by the vitamin B12-dependent enzyme methionine synthase. A common MTHFR variant, 677C → T substitution resulting in the conversion of an alanine to a valine residue, has been shown to be a thermolabile form of MTHFR and to have reduced activity (Frosst et al., 1995). Since many diverse cancer cells have been documented to be methionine-dependent in culture, the effects of MTHFR downregulation on cell survival of transformed cells was examined. In addition, the influence of the MTHFR 677C → T polymorphism on the sensitivity of transformed lines to antifolate drug treatment was studied. / To test the effect of decreased MTHFR expression on cell viability, we transfected antisense oligonucleotides (ASOs) complementary to the MTHFR mRNA into the colon carcinoma cell line SW620. (Abstract shortened by UMI.)

DNA tumor virus oncoproteins have evolved to regulate the p53 tumor suppressor. They must overcome p53-dependent cell cycle arrest and apoptosis, which would interfere with viral replication. Upon cellular stress, signalling to p53 takes the form of a large increase in the stability of the protein, by inhibition of the MDM2 protein which normally targets p53 for rapid degradation. Expression of the adenovirus type 5 early region 1 A (E1A) polypeptides can stimulate a quiescent cell to reenter the cell cycle and induces the accumulation of p53 protein and p53-dependent cell death under certain conditions. We mapped the regions of E1A necessary to induce p53 stabilization. Binding of either the p300 or pRb family of proteins can signal to p53, and these are the same activities required for E1A to push the cell to enter S phase. To replicate, the virus must combat this accumulation of p53, and we discovered that two other early viral proteins, E1B55K and E4orf6, can target p53 for ubiquitin-mediated degradation by the 26S proteasome. We found that when cells are infected with adenovirus, no significant accumulation of p53 occurs, and this is due to a large reduction in the half-life of the protein. E4orf6 and E1B55K each bind p53 and also bind each other, and we generated a series of E4orf6 mutations to study the regions necessary for this interaction. We identified a region at the amino terminus of E4orf6 that is the minimal domain permitting interaction with E1B55K. We also determined that binding E1B55K along with functions requiring most of the E4orf6 carboxy terminus have to be intact for E4orf6 to mediate p53 degradation. We tested the activity of the viral proteins in MDM2-deleted cells, and found that E4orf6/E1B55K don't require MDM2 to induce p53 turnover. A series of cellular proteins that interact with E4orf6 were discovered in our lab, and we identified this complex as Cullin 5, Elongin B, Elongin C, Rbx1, E4orf6 and E1B55K. This complex is very similar

Human carcinoembryonic antigen (CEA) is the prototypic member of a large family of highly related cell surface glycoproteins that includes CEACAM6 (formerly NCA) and CEACAM1 (formerly BGP). The extracellular domains of CEA/CEACAM6 are bound to the external surface of the plasma membrane through a glycosylphosphatidyl inositol (GPI) anchor and are over-expressed in more than 50% of all human cancers. In contrast, CEACAM1 contains extracellular, transmembrane and cytoplasmic domains, and its level of expression is down-regulated in human tumors of the colon and prostate. When over-expressed on the surface of various cell types in model systems, CEA/CEACAM6, but not CEACAM1, function as pan-inhibitors of cell differentiation and cell polarization and cause a distortion of tissue architecture. Anoikis is a quality control mechanism that must be inhibited in cancer cells for such a distortion to persist. This thesis presents data demonstrating that CEA/CEACAM6 over-expression on the surface of a variety of cell lines inhibited anoikis. The molecular basis for the inhibitory effects of CEA/CEACAM6 on both anoikis and differentiation is shown to be correlated with perturbation of the function of certain integrins. In contrast to CEA/CEACAM6, the expression of the CEACAM1 glycoprotein neither perturbed integrin function nor prevented anoikis and, consistent with this, inhibited tumor growth. As a conclusion, we propose that CEA/CEACAM6, but not CEACAM1, over-expression on the surface of cancer cells inhibits cell differentiation and anoikis through perturbation of integrin functions. These inhibitory effects could instrumentally contribute to tumor formation and progression.

Urinary plasminogen activator (uPA), a member of the serine protease family, is implicated in the progression of various cancers including breast cancer due to its ability to provoke malignant cell invasion. uPA production is reported to be much higher in late stage, estrogen receptor (ER) negative breast cancer patients than those with benign hyperplasia. Since all existing evidence points to a role for uPA in breast cancer progression, exploring the mechanisms regulating its gene expression will be of immense value. Two human breast cancer cell lines were selected for this study. MDA-MB-231 represents late stage breast cancer. This cell line has high uPA expression and is highly invasive. In contrast, the MCF-7 cell line represents early stage breast cancer and fails to express detectable levels of uPA mRNA. I have demonstrated by methylation specific PCR (MSP) that the differential expression of the uPA gene in MDA MB-231 and MCF-7 cells closely correlates with the amount of methylated cytosines present within the uPA promoter in these cells. The observation that the DNA methylation status of the uPA promoter directly affects the expression of the uPA gene was then confirmed using an in vitro luciferase reporter assay. Results suggest that the accessibility of the transcription factor Ets-1 is limited by DNA methylation. I further reported increased demethylase (DMase) activity with decreased maintenance activity of methyltransferases (DNMTs), which together favor the generation of a hypomethylated uPA promoter in these highly invasive MDA-MB-231 breast cancer cells. Given the pivotal role of uPA in breast cancer progression, I then disrupted the function of uPA and studied its effects on cancer progression. The effects of an 8-mer synthetic peptide (A6) derived from the non-receptor-binding region of uPA were investigated. This peptide inhibits cancer cell invasion of both human (MDA-MB-231) and rat breast cancer cells (Mat B-III) using an in vitro cell invas

OV90, a spontaneously immortalized epithelial ovarian cancer (EOC) cell line, has been shown to contain a homogeneously staining region (HSR) originating from chromosome 22. To identify the amplified genes, differential gene expression was assessed using high-density oligonucleotide array (chip) technology. Genomic differential PCR and dot blot analysis showed that overexpression is consistent with gene amplification, as all sequences tested within the 1 Mb region of 22q11.21 are present in increased copy number. However, overexpression was not common, as only two of 69 EOC samples evaluated using the Hu6800 chip showed overexpression of a gene amplified in OV90, and three other EOC cell lines did not have increased expression in this region. This study tested the approach of chip technology to identify genes involved in gene amplification, in a rapid and reliable manner.

Over the past two decades, novel targets for anticancer agents have been identified. One such target, the epidermal growth factor receptor (EGFR) that is overexpressed in a large number of carcinomas including breast, ovarian, and prostate, is a marker for tumour invasiveness and poor prognosis. Agents of the quinazoline class have been developed that block EGFR-mediated signaling and induce antitumour activity in the clinic. However, the major deficiency of these compounds is that they are cytostatic agents that induce reversible antiproliferative activities. To circumvent these problems, we designed a novel tumour targeting approach termed "the combi-targeting" concept. This theory is based on the fundamental premise that compounds capable of interfering with multiple targets in the cancer cells are more efficient antitumour agents than the single-targeted ones. Thus, the "combi-targeting" concept proposes the design of molecules termed "combi-molecules" or TZ-I to not only bind to the tumour target but also to be allowed to degrade to another inhibitor I of the same target + a DNA damaging species TZ, leading to small molecules capable of repetitively blocking one target while damaging another. Using EGFR as a tumour target, the first TZ-I prototype SMA41, an anilinoquinazoline containing a triazene tail at the 6-position, was synthesized. We demonstrate herein that the compound enters the cell by passive diffusion, degrades under physiological conditions to yield an intact TK inhibitor ("SMA52") (I) + a short-lived DNA-damaging methyldiazonium species (TZ). In the EGFR-expressing human tumour cells, SMA41 (TZ-I) behaved as a binary targeted drugs with ability to: (a) inhibit EGF-induced receptor autophosphorylation, cell cycle progression and growth, (b) induce dose-dependent DNA damage, and (c) inhibit cell proliferation with greater potency than the released reversible inhibitor (I) both in vitro and in vivo. This work represents the

The ErbB-2/neu receptor tyrosine kinase is involved in normal tissue development. However, this receptor has been implicated in the genesis of human breast and renal carcinomas, where ErbB2 is amplified in 20--30% of human breast cancers and correlates with poor prognosis. Using the non-transformed MDCK epithelial cell model, I have established that a deregulated activated ErbB2/Neu receptor (NeuNT) but not overexpression of the wild type (WT) receptor induces cell dispersal and motility, accompanied by the breakdown of cell-cell junctions and E-cadherin internalization, in addition to reorganization of the actin cytoskeleton. This phenotype can be reversed following treatment of the cells with a pharmacological inhibitor of MEK, indicating that MEK dependent pathways are involved in the NeuNT-induced remodeling of cell-cell junctions. In three-dimensional cultures of MDCK cells, NeuNT but not WT ErbB2 triggers a morphogenic response that correlates with recruitment and increased phosphorylation levels of the Shc adapter protein. This demonstrates that the deregulated ErbB2/NT receptor induces a distinct biological response when compared to the wild type receptor and induces the loss of epithelial architecture observed in carcinomas. / Invasive morphogenesis downstream from the Met/HGF receptor is modulated through a sustained phosphorylation of the Gab1 docking protein and of downstream kinase (Erk). In contrast, a transient phosphorylation of Gab1 and Erk induced by EGF is not sufficient to promote a morphogenic response. In Chapter III, I demonstrate that NeuNT but not the WT ErbB2 receptor display elevated and sustained levels of Gab1 and Erk phosphorylation which correlates with their ability to promote invasive morphogenesis. In addition, co-immunoprecipitation analyses provide evidence for the recruitment of Gab1 to ErbB2/Neu in a Grb2-dependent and Grb2-independent manner. / To identify physiologically relevant factors that synergize with ErbB2, I established that HGF, the Met receptor ligand, promotes the disruption and invasion of NeuNT-induced epithelial structures in three dimensional matrix cultures. Moreover HGF synergizes with NeuNT, enhancing the invasive potential of NeuNT expressing cells ten fold through Matrigel. HGF treatment of NeuNT expressing cells promotes a decrease in E-cadherin protein, and can be blocked or reversed by treatment with the MEK inhibitor, UO126, establishing the involvement of MEK-dependent pathways in this process. These results demonstrate that physiological signals downstream from HGF/Met cooperate with deregulated ErbB2/Neu to enhance the malignant phenotype promoting a more stable epithelial-mesenchymal transition and enhanced cell invasion.

Pancreatic cancer is considered the fifth leading cause of cancer deaths in Canada and one of the most fatal diseases in the world. Its definite underlying cause is still unidentified, and its actual cell of origin remains unclear. Unfortunately, most of the current research on the pancreas is focused on one disease only, namely diabetes with much less consideration for other pancreatic diseases. Diabetes has been extensively studied from a developmental aspect, and continues to attract the interest of numerous researchers. On the contrary, few accomplishments have been done to decode the developmental errors occurring in pancreatic cancer. It is therefore necessary to allocate more research resources to address this disease from a developmental aspect. / This study provides a literature review of the pancreas concerning its anatomy and function, transcription factors and signaling pathways controlling its development, and the role of these signaling pathways in pancreatic cancer. The review provides distinct emphasis on three important aspects. First, a review of pancreas development is provided, with a focus on different transcription factors and signaling pathways involved in this process. Second, it addresses how the signaling pathways which play a role in pancreas development are the same signaling pathways that play a role in pancreatic cancer, additional emphasis is placed on describing the genetic alterations occurring in pancreatic cancer. Third, a methodology of approaching pancreatic cancer research from a developmental aspect is presented. Using an example of one gene, Anterior gradient 2 (Agr2), is highly expressed in pancreatic cancer in ductal cells only, and might play a role in ductal cell development of the pancreas. Thus, the main objective of this review is to provide a developmental framework for the analysis of pancreatic cancer.

Prolactin is a key mammary gland differentiation factor. However, the contribution of prolactin (PRL) to breast carcinogenesis is less clear. Accumulating evidences indicate that in established breast carcinomas autocrine/paracrine PRL can enhance growth/viability of breast cancer cells. Still, it is not known whether the ascribed pro-oncogenic activity of PRL describes fully the role of PRL in regulating breast carcinogenesis. On the other hand a critical role for Ras-Erk1/2 and TGF-beta (Transforming Growth Factor beta) pathway in breast cancer progression has already been established. Our results indicate that blocking PRL signal leads to activation of Ras-Erk1/2 pathway and TGF-beta pathway, two key pathways contributing to breast cancer metastasis. I showed that modulation of PRL signaling in breast cancer cells alters their morphogenic program. My results highlight a critical role for PRL in regulating epithelial plasticity and implicate PRL as invasive suppressor hormone in breast cancer cells.

Prostate carcinoma (CaP) is the most frequently diagnosed malignancy in men in Western countries. Increased activity of the androgen receptor (AR) and/or AR gene (AR) amplification in the majority of both androgen-dependent and androgen-independent prostate cancers suggest that the AR plays an important role in CaP. / The polymorphic AR CAGn, repeat correlates inversely with AR transactivation. To better understand the contribution of this repeat to CaP, we propose to replace the stable mouse AR CAG/CAA/CAC tract with various lengths of CAG, using gene knock-in methodology. The neo-NTR-HSV-tk cassette from the pPGKneoNTRtkpA vector was introduced into the ploxPneo-1 vector such that loxP sites flanked it. Various CAG repeats were then cloned into mouse AR genomic 5' end clone. Finally, the 5' and 3' mouse AR genomic DNA fragments were cloned into the targeting vector, upstream and downstream of the cassette with flanking loxP sites. Knock-in mice with various CAG repeat lengths will be generated using these constructs.