Cellular transformation is accomplished in vitro through the concerted action of growth factors and oncogenes. This association has demonstrated that malignant growth results from aberrations in pathways that normally operate to control proliferation. Activation of genes that code for growth factors, their receptors, and/or molecules essential in the transduction of signals from the cell surface to the nucleus are all potential mechanisms by which tumor cells could establish a selective growth advantage over normal cells. This dissertation addresses the question of what oncogenic mechanisms are important in the development and progression of human melanoma. These studies show that melanoma growth is regulated by endogenous substances produced by the melanoma cells themselves (autocrine stimulation), as well as by exogenous substances supplied by neighboring cells and platelets (paracrine stimulation). These factors work to drive the expression of the transformed phenotype for melanoma as evidenced by induction of serum-free soft agar growth. Human platelets were found to the the richest source of paracrine growth promoters. The factor from human platelets was characterized and partially purified. Melanoma cells respond to this 60,000 molecular weight, disulfide-bond-containing protein in colony formation assays. In addition, the protein has endothelial cell growth factor activity. Purified fractions which promoted optimal colony formation for human melanoma cells also maximally stimulated monolayer growth of bovine aortic endothelial cells, while melanocytes were nonresponsive. This implies that melanoma cells are expressing receptors for a protein which plays no known or apparent role in the normal growth of melanocytes. Melanoma cells are sensitive to growth regulatory molecules of autocrine and paracrine nature. This dissertation provides clues to the genetic lesions which have occurred in these melanoma cells to influence their proliferation. The aberrations appear to reside in those genes important in growth factor pathways at the level of endogenous production and misguided response to exogenous factors through receptor expression. We can not hope to fully inhibit the proliferation of tumor cells until we identify and understand those forces which drive their growth. These studies have increased our knowledge of those signals which stimulate melanoma cellular proliferation, and thus provide insight into important therapeutic targets.

University of Macau / Institute of Chinese Medical Sciences

University of Macau / Institute of Chinese Medical Sciences

published_or_final_version / Biochemistry / Doctoral / Doctor of Philosophy

published_or_final_version / Physiology / Master / Master of Philosophy

A two by six factorial design was used to investigate the effect of zinc and all-trans retinoic acid (RA) on the growth of two human breast cancer cell lines differing in their expression of metallothionein (MT) and of estrogen receptors; MCF7 cells express estrogen receptors, BT-20 cells do not. Cells were treated with zinc to induce MT then treated with six concentrations of RA. Cell proliferation, lipid peroxidation, MT protein, MT mRNA and glutathione concentrations were measured. / BT-20 cells expressed higher constitutive MT concentrations than MCF7 cells. MT was significantly induced by zinc treatment in BT-20 cells but not in MCF7 cells. Low RA concentrations stimulated growth proliferation but higher concentrations inhibited cell proliferation. High RA concentrations increased lipid peroxidation. There was a significant negative correlation between lipid peroxidation and cell proliferation. Growth inhibition and lipid peroxidation were reduced by zinc in BT-20 cells but not in MCF7 cells. Glutathione did not appear to be a significant factor. / Induction of metallothionein by zinc may modulate the growth inhibitory effects of all-trans retinoic acid in human breast cancer cells. One mechanism of growth inhibition may be through increased lipid peroxidation.

The signal transducer and activator of transcription (STAT) and protein inhibitor of STAT(PIAS) system represent an elegant regulatory mechanism of transcriptional control IN mammalian cytokine signalling. Abnormal activation of the system is associated with immune disorders and a large group of diverse tumours. PIAS3 is a multiple domain protein with distinct functions involved in regulation of cytokine-mediated gene activation pathways.Its over-expression significantly inhibits cell growth and renders cancer cells more sensitive to drugs. The objective of this study was to structurally and biochemically characterise the function of the PIAS3 protein using in silico, in vivo and in vitro analysis approaches.The conservation pattern of the PIAS protein family and critical conserved residues in the PINIT (Proline, Isoleucine, Asparagine, Isoleucine, Tyrosine) domain were identified. The PINIT domain model was generated based on the PINIT domain structure of yeast PIAS3 homologue Siz1 and structural determinants in the PIAS3-STAT3 interaction were evaluated.Guided by the in silico findings, in vivo analysis of the localisation of the PIAS3, mutantderivatives of PIAS3 (PIAS3-L97A, PIAS3-R99N, PIAS3-R99Q), PINIT and acidic domain was conducted. PIAS3 was completely localised in the nucleus while PIAS3 mutants appeared to exhibit diffuse cytoplasmic distribution. The PINIT domain was predominantly localised in the nucleus with some apparent perinuclear staining while the acidic domain exhibited a predominantly perinuclear staining pattern. Further analysis of the PINIT domain and the effect of the mutants on PIAS3-STAT3 interaction were assessed by in vitro analysis. Guided by in silico analysis, the PINIT domain and mutant derivatives of PINIT domain (PINIT-L97A, PINIT-R99N, and PINIT-R99Q) were heterologously expressed in Escherichia coli and subsequently purified using a combination of immobilized metal affinity and size exclusion based chromatography. The size and structural elements of the PINIT domain and its mutants were characterised. The 23 kDa PINIT domain was found to exist as a monomer in solution and its secondary structure was shown to consist of 66 % β-sheets by fourier transformed infrared spectroscopy consistent with the generated homology model.Using surface plasmonresonance spectroscopy (SPR) the PINIT domain was shown to bind to STAT3 in a specific concentration dependent manner. Recombinant PINIT-L97A,PINITR99N and PINIT-R99Q mutants, which exhibited similar structural integrity to the wildtype, were found to abrogate binding to STAT3. These findings suggest that these residues form part of a potential binding surface for stat3. In conclusion, this study has provided evidence that the PINIT domain is an important determinant of PIAS3 interaction with STAT3 and that the interaction is mediated by defined conserved residues directly involved in the PINITSTAT3 interaction.

Transcription factors (TFs) are key regulators of gene expression whose failure has been implicated in many diseases, including cancer. They bind at various sites at different specificity depending on the prevailing cellular conditions, disease, development stage or environmental conditions of the cell. TF binding specificity is how well a TF distinguishes functional sites from potential non-functional sites to form a useful regulatory network. Owing to its role in diseases, various techniques have been used to determine TF binding specificity in vitro and in vivo, including chromatin immuno-precipitation followed by massively parallel sequencing (ChIP-seq). ChIP-seq is an in vivo technique that considers how the chromatin landscape affects TF binding. Motif enrichment analysis (MEA) tools are used to identify motifs that are over-represented in ChIP-seq peak regions. One such tool, CentriMo, finds over-represented motifs at the center since peak calling software are biased to declaring binding regions centered at the TF binding site. In this study, we investigate the use of CentriMo and other MEA tools to determine the difference in motif enrichment attributed presence of Chronic Myeloid leukemia (CML)), treatment with Interferon (IFN) and Dexamethasone (DEX) compared to control based on Fisher’s exact test; using uniform peaks ChIP-seq data generated by the ENCODE consortium. CentriMo proved to be capable. We observed differential motif enrichment of TFs with tumor promoter activity: YY1, CEBPA, Egr1, Cmyc family, Gata1 and JunD in K562 while Stat1, Irf1, and Runx1 in Gm12878. Enrichment of CTCF in Gm12878 with YY1 as the immuno-precipitated (ChIP-ed) factor and the presence of significant spacing (SpaMo analysis) of CTCF and YY1 in Gm12878 but not in K562 could show that CTCF, as a repressor, helps in maintaining the required YY1 level in a normal cell line. IFN might reduce Cmyc and the Jun family of TFs binding via the repressive action of CTCF and E2f2. We also show that the concentration of DEX treatment affects motif enrichment with 50nm being an optimum concentration for Gr binding by maintaining open chromatin via AP1 TF. This study has demonstrated the usefulness of CentriMo for TF binding specificity analysis.

The termina1 enzyme in the linoleoyl-CoA desaturase enzyme complex, delta-6-desaturase was implied in the control of cell proliferation in cancer cells. One of the aims of this study was to isolate the terminal enzyme. It was decided that in order to isolate this enzyme it was first necessary to isolate the entire complex and then to enzymatically solubilise the first two components of the complex i e cytochrome b5 reductase and cytochrome b5 from the complex resulting in a pure delta-6-desaturase . The first two components were isolated and purified using simplified and easily reproducible methodologies which could be utilised in the final purification of delta-6- desaturase. The entire enzyme complex, linoleoyl-CoA desaturase was also isolated in a pure form and this pure complex was used to attempt to isolate delta-6-desaturase. The terminal enzyme was isolated with some cytochrome b5 still bound to it. The methods used had proven to be successful and with some modifications should yield a pure enzyme. Zinc and GLA were known to play a role in the inhibition of cancer cell proliferation and zinc was hypothesised to inhibit cell growth by stimulating the activity of the linoleoyl-CoA desaturase enzyme complex which is involved in the regulation of cell proliferation. GLA is the product of the reaction that this enzyme complex catalyses and GLA has been shown to inhibit cancer ce ll growth. The effect of GLA on cell growth and linoleoyl-CoA desaturase activity was thus investigated. Results showed that both zinc and GLA inhibited cell growth and that the combined addition of zinc and GLA generally resulted in the inhibition of cell growth and the activation of linoleoyl-CoA desaturase activity in the BL-6 cells while having a less pronounced effect on the LLCMK cells. The results of this study support the hypothesis that zinc may be a cofactor of linoleoyl-CoA desaturase.

No description available.