Introduction: Chromosome 12p13 has been reported to show gain/amplification in some breast cancers. This region contains putative stem cell-associated genes: DPPA3, EDR1, GDF3, and NANOG, but these genes have not been investigated in breast previously. Hence, the aim of this thesis was to study their role in breast carcinomas. Materials and methods: The mRNA expression was evaluated in normal and malignant breast tissues using TaqMan® gene expression assays. Western blotting and immunohistochemistry were used for determination of protein expression. Copy number variations (CNVs) were assessed by TaqMan® copy number assays (CNAs) and Affymetrix® genome-wide human single nucleotide polymorphism (SNP) array 6.0. Results: Expression of DPPA3, EDR1, and NANOG was undetectable in normal breast tissue, but there was variable expression in breast carcinomas (BC) where expression of these genes tended to be higher in surrounding normal breast tissue. GDF3 was not expressed in BC. At the 95% confidence interval, higher expression of DPPA3 in BC was related to axillary lymph node metastasis; lower expression of DPPA3, EDR1, and NANOG correlated with high grade; and lower expression of NANOG was found in tumours of size > 2.0 cm. Both TaqMan® CNAs targeting each gene individually and SNP 6.0 genome wide array revealed complicated patterns of CNVs for these genes. The majority of BC had gain of DPPA3 but loss (deletion) of EDR1 and NANOG. However, there was no significant correlation between CNVs and either mRNA expression or protein expression. Conclusion: Variable aberrations in copy number and expression of DPPA3, EDR1, and NANOG genes in the chromosome 12p13 region are associated with aggressive characteristics of breast carcinomas.

Amongst women diagnosed with early breast cancer, the probability of recurrence and subsequent death can vary substantially depending upon a number of known prognostic factors. With such heterogeneous levels of risk, the absolute effectiveness and consequently the cost-effectiveness of adjuvant chemotherapy, a treatment known to reduce the likelihood of recurrence, can also be expected to vary between patients.

HMGBl is an abundant and ubiquitously expressed non-histone nuclear protein. HMGB1 is a nuclear architectural factor, but can also be found in the extracellular environment. Necrotic cells, but not apoptotic, passively release HMGBl in the extracellular environment, whereas specific cell types, such as macrophages, actively secrete HMGB1. In the extracellular milieu HMGB1 works as a cytokine and a signal of tissue damage, stimulating cell migration by an NF-KB-dependent mechanism, cell proliferation and differentiation. HMGBl is over-expressed in several types of tumors and plays a role in cancer progression and metastasis, when present in the tumor microenvironment. Moreover HMGBl over-expression protects cells from different apoptotic sdmuli and is associated to expression of genes linked to metastasis. Analyzing a panel of human breast cancer biopsies, we found almost universal HMGB1 overexpression in tumor cells, whereas HMGBl cytoplasmic relocation appears to be linked to lymph node invasion.

The role of ERα in hormone responsive breast cancer is well established but that of ERβ remains to be determined. In this thesis, the effects of overexpression of ERβlong have been studied in MCF7 human breast cancer cells which possess high endogenous levels of ERα . Stable overexpression of ERβlong protein (2-fold) resulted in reduced fold induction (from 2.6 to 1.8) of an ERE-CAT reporter gene with 10-8M 17β-oestradiol but enhanced fold induction (from 1.18 to 2.12) by 10-5M genistein and an increased induction (from 4.4 to 11.4) of growth with 10-4M methylparaben.

Despite much research investigating the role of diet and lifestyle in breast cancer, the evidence for many dietary factors remains uncertain. The main focus of this thesis concerns the role of dietary patterns and fat in breast cancer risk and also recurrence and survival in breast cancer patients. Meta-analyses conducted in this thesis have suggested that a prudent dietary pattern may reduce the risk of breast cancer, whilst a drinker dietary pattern may increase risk. High saturated fat intake was also found to be associated with breast cancer mortality. Data obtained from participants of the DietCompLyf study, a prospective cohort of breast cancer patients in the UK, was used in this thesis to investigate dietary patterns and fat intake in relation to breast cancer recurrence and quality of life. The physical activity questionnaire used in the DietCompLyf study was also retrospectively validated in this thesis to determine how best to use the physical activity information as a confounder in the" diet and breast cancer recurrence and quality of life analyses in this thesis. In participants of the DietCompLyf study, adherence to a Traditional dietary pattern was associated with an increased risk of breast cancer recurrence. A prudent dietary pattern was associated with a better quality of life, whilst Western dietary patterns associated with a poorer quality of life in unadjusted analyses. The complicated nature of the diet and quality of life relationship was highlighted when adjustment for potential confounders modified and attenuated the observed associations. Results observed in this thesis suggest possible roles for diet in both risk of breast cancer and breast cancer mortality. Further analyses of data obtained from the DietCompLyf cohort should be conducted in the future to clarify theses observations.

Background: MicroRNAs (miRNAs) modulate gene expression by targeting mRNAs for cleavage or translational suppression. miRNAs have deregulated expression in human breast cancer (BC), and there is evidence suggesting that they function primarily as tumour suppressors. Dicer and Drosha are two proteins which play key roles in miRNA biogenesis to produce mature miRNAs from their precursor molecules, and are known to be deranged in human Be. The contribution of miRNAs and their maturation regulators in the initiation and progression of human Be can be exploited to achieve novel modifications in the existing diagnostic and therapeutic approaches in Be management. Methods: Protein expression levels of miRNA maturation regu lators, Dicer and Drosha, were assessed immunohistochemically in 2 sets of Be: 1) full-face sections of selected Be cases (n = 24) with normal breast parenchymal tissue (NBPT) and lesions representing different stages of Be progression (Ductal carcinoma in situ, DCIS; invasive breast cancer, IBC; and metastatic breast cancer, MBC) to assess their differential expression, 2) tissue microa rrays (TMAs) comprising a large and well-characterised series of primary IBC (n = 1902) to evaluate their biological, clinicopathological, prognostic and predictive significance. Additionally, miRNA expression profiling was explored in the 4 tissue components (NBPT, DCIS, IBC and MBC) of 7 BC cases using an Agilent microa rray-based miRNA profiling study, to investigate the differential expression of miRNAs in NBPT and different stages of BC progression. Results: Dicer and Drosha protein expression was observed to decrease with BC progression, which implies that loss of these two miRNA maturation regulators might have a role in the process of Be progression. In the IBC series, loss of Dicer expression was associated with features of aggressive behaviour including higher histological grade and loss of ER, PR and BRCA1 protein expression . Moreover, Dicer expression was revealed to be an independent predictor of a shorter disease free interval at 5 years, and predictive of better response to chemotherapy and to endocrine therapy. Loss of Drosha cytoplasmic expression was associated with higher tumour stage and loss of expression of BRCA1 and basal phenotype; and loss of Drosha nuclear expression was associated with larger tumour size, higher tumour grade, and loss of ER, PR, BRCA1, E-cadherin and CK14. Loss of Drosha cytoplasmic expression was associated with shorter Be specific survival, BC recurrence, and distant metastases, and th is correlation was maintained in ER-negative and HER2- negative cases. Loss of Drosha cytoplasmic expression was predictive of better response to systemic therapy in ER-negative patients. Furthermore, we present data revealing that some miRNAs are differentially expressed across the 4 tissue components (NBPT, DCIS, IBC and MBC). The comprehensive survey of miRNA expression profiling, confirmed differential expression of one miRNA previously reported to be deranged in BC, and identified 6 miRNAs downregulated du ring BC progression, 5 downregulated and 1 upregulated, which have not previously been linked to Be. Among the novel downregulated miRNA candidates, 2 (hsa-miR-1181 and hsa-miR-4322) were revealed to gradually reduce in amount across the 4 tissue components. These observatinos support the hypothesis that miRNAs play a role in tumourigenesis and in advancing BC tissues towards acquiring metastatic potential. Conclusions: The intimate involvement of miRNAs and their maturation regulators in malignant transformation, cellular invasion and metastases in BC makes them potentially relevant as future diagnostic, predictive and therapeutic targets. Although the current study is preliminary, we believe the results add to the present knowledge of Significance of miRNAs and their regulators in Be. Therefore, the results would serve as a robust initial set of potential biomarkers for validation in a larger cohort of patients.

It has become clear that the phenotypic alterations present in carcinoma-associated fibroblasts play an important role in the development and progression of breast carcinomas. To investigate these alterations, a panel of carcinoma-associated fibroblast and their counterpart fibroblast primary cell cultures were assessed for their expression of p53, p21 and survivin. The induction of p53 and p21 in response to gamma-irradiation was defective in most of the carcinoma-associated fibroblasts but not in their counterpart fibroblasts. In addition, the constitutive levels of p53 and p21 proteins were significantly lower in most of the carcinoma-associated fibroblasts and modulated in the majority of tumour counterpart fibroblasts compared to the normal breast fibroblasts. Survivin expression was higher in both carcinoma-associated fibroblasts and tumour counterpart fibroblasts as compared to in normal fibroblasts. This part of the work demonstrates that carcinoma-associated fibroblasts have undergone premalignant changes that make them resistant to irradiation. The cytokine, CXCL12 is secreted by stromal cells and can affect the growth and progression of breast tumours. In malignant breast epithelial cells, CXCL12 mRNA has been shown to be expressed also and to be increased by oestrogen. To evaluate the importance of CXCL12 in the oestrogen response of breast cancer cells, the expression of CXCL12 and both its receptors, CXCR4 and CXCR7, was assessed by quantitative real time PCR and western transfer analysis in a panel of eleven human breast cancer cell lines. Notably CXCL12 was expressed mostly in the oestrogen-responsive breast cancer cell lines, whereas CXCR4 and CXCR7 were expressed at high levels in MDA-MB-231, an oestrogen non-responsive breast cancer cell line. Furthermore, oestrogen increased CXCL12 mRNA expression in breast cancer cell lines, while the expression of CXCR4 and CXCR7 mRNAs was decreased in response to oestrogen. The effects of CXCL12 on the proliferation and migration of breast cancer cells were tested. The results of these experiments showed that CXCL12 has a motogenic effect in both oestrogen-responsive and oestrogen non-responsive breast cancer cells. Interestingly, the combined effect of CXCL12 and IGF-1 was more than additive on migration of breast cancer cells. Furthermore, phosphorylation of CXCR4, and its downstream effectors, Akt and MAPK, was studied in oestrogen-responsive breast cancer cells, ZR-75 cells, and oestrogen non-responsive breast cancer cells, MDA-MB-231 cells. Our results suggest that the phosphorylation of Akt and MAPK may be involved in mediating the effects of CXCL12 in breast cancer cells. The work described in the second part the thesis emphasis the importance of the interplay between oestrogen, CXCL12 and IGF-1 in the responsiveness of malignant breast epithelial cells.

Trefoil proteins, TFF1, TFF2 and TFF3 are small molecular mass proteins that are secreted by mucus secreting epithelia. They share a three-looped structure, trefoil domain which contains six conserved cysteine residues that form three intramolecular disulphide bonds. TFF1 and TFF3 contain a seventh conserved uncoupled cysteine residue. In gastrointestinal mucosa, trefoil proteins are involved in stabilisation of the mucus layer, restitution of damaged mucosa and stimulation of epithelial cell motility. Increased expression of trefoil proteins has been reported in several human epithelial tumours and it has been proposed that trefoil proteins predispose to tumour cell migration, invasion and metastasis. Expression of trefoil genes is regulated by oestrogen in human breast cancer cell lines and strong association between expression of oestrogen receptor and expression of TFF1 and TFF3 mRNAs has been demonstrated in breast tumours. The role of trefoil proteins in breast cancer has not been fully understood. The purpose of this study is to investigate the molecular forms of TFF1 and TFF3 in breast cancer cells and to evaluate the expression of TFF1 and TFF3 in primary and metastatic breast tumours. The molecular forms of TFF1 and TFF3 proteins were examined in protein lysates and conditioned medium prepared from MCF-7, EFM-19 and EFF-3 breast cancer cell lines by western transfer analysis under non-reducing conditions and the different molecular forms were immunoprecipitated with anti-TFF1 and anti-TFF3 antibodies. Two distinct molecular forms of TFF1 and TFF3 were detected in breast cancer cell lines. They have approximate molecular masses of 66 kDa and 14.5 kDa for TFF1 and 66 kDa and 20 kDa for TFF3. The 66 kDa forms represent heterdimers that constitute one molecule of TFF1 or one molecule v of TFF3 bound to another protein of approximately 60 kDa through an intermolecular disulphide bond. The expression of TFF1 and TFF3 proteins was measured in 296 primary breast tumours and 76 metastatic deposits of breast tumours from patients who presented in the Newcastle Upon Tyne hospitals between 2002 and 2003 by immunohistochemistry. The association of TFF1 and TFF3 expression with various clinico-pathological features and with expression of oestrogen receptor, progesterone receptor, CD31, CD34, E-cadherin, Bcl-2 and Bax proteins was tested statistically with SPSS software (p<0.05). Clear TFF1 and TFF3 expression was demonstrated in normal, benign and malignant breast epithelial cells. There is no expression in stromal, endothelial or immune cells. There is a strong association between expression of TFF1 and expression of TFF3 in normal, benign and malignant breast epithelial cells. Expression of TFF1 and TFF3 varied enormously between tumour samples. Both had highest expression in mucinous and tubular breast carcinomas and both proteins are expressed at higher levels in lobular than in ductal breast cancers. Expression of TFF1 and TFF3 proteins is associated strongly with expression of both oestrogen and progesterone receptors. There is a negative association between TFF1 and TFF3 expression and tumour grade. There is however a positive association between TFF3 expression and presence of vascular invasion and presence of axillary lymph node metastasis and TFF3 expression is higher in metastatic breast tumour cells than in primary breast tumour cells. High expression of TFF3 is associated with increased number of small blood vessels, with increased expression of Bcl-2 and with increased expression of E-cadherin in primary breast cancer tissue. This is the first report about the expression of TFF3 and its association with expression of TFF1 in primary breast cancer tissues. The strong association between expression of TFF1 and TFF3 and expression of oestrogen receptor suggests that trefoil protein expression is dependent on oestrogen in a significant proportion of breast tumours. The high expression vi of TFF3 in breast tumours with vascular invasion and in metastatic breast tumour cells supports the hypothesis that TFF3 predisposes towards breast cancer cell invasion. The high molecular mass protein partner of TFF1 and TFF3 could be involved in mediation of trefoil proteins effects in breast cancer.

Breast cancer is a common disease with 49,961 new cases being diagnosed and causing 11,633 deaths in the UK in 2010. One of the cellular processes associated with carcinogenesis is dysregulation of translation. Translation is a tightly controlled process regulated by a number of different mechanisms, a key factor being the availability of eukaryotic initiation factor 4E (eIF4E). eIF4E has been shown to be overexpressed in a variety of human malignancies and higher levels of expression are associated with a poorer prognosis in some cancers. The availability of eIF4E is determined in a large part by the phosphorylation status of the 4E-binding proteins (4E-BPs). I have analysed the expression, localisation and prognostic importance of eIF4E and the 4E-BPs in large cohorts of both male breast cancer (MBC) and female breast cancer (FBC). I have also attempted to determine whether translation of the 4E-BPs is responsive to changes in eIF4E expression.

ER~ is one of two estrogen receptors (ERs) which mediates estrogen action. ER~ is often down regulated in cancer compared with normal cells, but the molecular mechanisms involved remain unclear. The aims of this thesis were to examine whether ER~ and its splice variants are transcriptionally silenced by DNA methylation and post- transcriptionally regulated by miRs. The methylation status of the ER~ promoters (OK, ON and E1) and untranslated exons (OK and ON) were analyzed using bisulphite sequencing PCR (BSP) and methylation-specific PCR (MSP), in a panel of breast cancer cell lines of different ER~ status and clinical samples. The expression of ER~1, -~2 and -~5 mRNA was determined using (RQ) RT-PCR and correlated with the methylation data. ER~1 and -~2 expressions were inversely correlated with the methylation status of exon ON. However, no correlation was observed between promoters OK, ON and E1, and exon OK, ER~1 and -~2 expressions. ER~5 was abundantly expressed in these samples and did not correlate with the methylation of promoters and untranslated exons, suggesting that ER~5 was not regulated by DNA methylation. Pharmacological restoration of ER~ 1 and - ~2, but not -~5 mRNA, after in vitro DNA demethylation and histone deacetylation, using 5-aza-dc and TSA, provided experimental evidence that ER~1 and -~2 expressions were epigenetically regulated. microRNAs (miRs) are a class of non-protein coding small RNAs that regulate the expression of genes at post-transcriptional levels. In silico analysis using the miRGen Targets, MicroCosm Targets and RNA hybrid predicted a putative miR-92 target sequence within the 3'-untranslated region (UTR) of II ER~1. Expression analysis in breast cell lines and in a cohort of primary breast tumours confirmed a significant negative correlation between miR-92 and both ER~1 mRNA and protein, but not -~2, -~5 and ERa. Inhibition of miR-92 in MCF-7 cells increased the ER~1 expression in a dose-dependent manner, whereas miR-92 over expression led to ER~1 down-regulation. Reporter constructs containing candidate miR-92 binding sites in the 3'UTR of ER~1 confirmed that miR-92 down-regulated ER~1 via direct targeting of its 3'UTR. miR-92 expression was also modulated by the ER ligands 17~-estradiol and tamoxifen in MCF-7 cells. In summary, this study has provided new mechanistic insight on possible mechanisms that are. responsible for the down-regulation of ER~ in breast cancer. Besides the involvement of DNA methylation in regulation of the ER~ expression, this study has provided, for the first time, novel evidence that demonstrates the involvement of miR-92 in ER~1 regulation. These findings could provide the basis for potential therapeutic strategies for breast cancer, aimed at reactivating the expression of ER~1 through the manipulation of miR- 92 expression.