Immune modulation of mammographic density and breast cancer risk

Abstract

Mammographic density (MD) is a strong risk factor for breast cancer that can increase breast cancer risk by 4-6 fold independently of age and BMI. High MD is characterised by breast tissue containing high proportions of stroma, containing fibroblasts, collagen and immune cells, indicative of an inflammatory pro-tumour environment. Currently, the biological mechanisms that drive MD and the associated breast cancer risk are not yet understood. Immune signalling factors such as monocyte chemotactic protein 1 (CCL2), peroxidase enzymes, transforming growth factor beta (TGFB) and tumour necrosis factor alpha (TNFA) have been implicated in breast cancer risk, and are also known to influence functions of stromal fibroblasts. The work in this thesis aims to investigate how these immune signalling factors and the immune microenvironment may act as drivers of MD through their interactions with mammary fibroblasts in vitro, and this effect mammary tumorigenesis in a transgenic tumour mouse model. Firstly, primary mammary fibroblasts from women with high and low MD were treated in vitro with immune factors myeloperoxidase (MPO), eosinophil peroxidase (EPO), TGFB, TNFA and CCL2 for 72 hours. The abundance of mRNA encoding cancer associated fibroblast (CAF) markers and genes involved in extracellular matrix (ECM) regulation were investigated. Production of soluble collagen 1 and insoluble collagen fibres were also measured. No significant differences were observed in gene expression and collagen production between fibroblasts from women with low or high mammographic density. MPO and EPO significantly increased production of collagen 1 in mammary fibroblasts. TGFB and TNFA induced variable changes in CAF and ECM gene expression. TGFB and CCL2 increased deposition of insoluble collagen fibres. Further investigation of the crosstalk between mammary fibroblasts and macrophages utilised an indirect co-culture of primary mammary fibroblasts with THP-1 derived macrophages in the presence or absence of CCL2 for 72 hours. THP-1 derived macrophages underwent differentiation to an M2 phenotype, and had high expression of genes involved in tissue remodelling including MMP2, TIMP1 and VEGF. Co-culture with THP-1 macrophages induced high expression of ECM remodelling genes by mammary fibroblasts such as MMP1, MMP3, MMP9 and TIMP1, as well as inflammatory genes COX2, IL6 and IL8. Production of insoluble collagen fibres was increased in mammary fibroblasts in co-culture with THP-1 macrophages and CCL2. Thirdly, the effect of CCL2 on mammary tumourigenesis and global gene expression was investigated. Mmtv-CCL2 mice overexpressing CCL2 in the mammary gland were crossbred with Mmtv-PyMT tumour model mice. Tumour development was monitored until 9 and 12 weeks of age. CCL2 overexpressing mice were found to have increased macrophage infiltration to primary tumours and increased number of areas of early tumorigenesis in the mammary gland at 9 weeks age. No differences were observed in tumour latency, tumour burden, tumour grade, or pulmonary metastasis between CCL2 overexpressing mice and controls. Global gene expression was analysed by RNAseq in mammary glands from 12 week old Mmtv-CCL2 mice and FVB controls. Mmtv-CCL2 mice exhibited increased expression of genes involved in cancer, DNA damage and extracellular matrix deposition. Expression of genes involved in fatty acid metabolism and adaptive immunity were reduced in Mmtv-CCL2 mice. These results suggest that fibroblasts from women with high mammographic density are not inherently different to those from women with low mammographic density. The role of fibroblasts in MD and breast cancer risk may be a result of immune signals from surrounding cells in the breast microenvironment, which may be through interactions with macrophages and inflammation driven by CCL2.