| dc.description |
Epithelial ovarian cancer (EOC) is the most lethal of all gynaecologic cancers, principally because the early stages are relatively asymptomatic, resulting in many patients having advanced, metastatic disease at diagnosis. A major site of metastasis is the omentum, a large abdominal adipose tissue bed mainly comprised of adipocytes encased in an outer mesothelial cell layer. For EOC cells metastasising to the omentum, via the transcoelomic route, the mesothelium is the first point of contact. After successful invasion through the mesothelium into the omental tissue, the EOC cells are able to activate angiogenesis i.e. new blood vessel formation, in the existing host microvasculature to supply the growing tumour with nutrients and oxygen. The initiation of angiogenesis requires activation of the endothelial cells (ECs) lining the omental microvasculature by pro-angiogenic factors, and it is possible that the production of these factors by both the metastasising tumour cells themselves and the resident cells within the omentum (e.g. adipocytes) jointly contribute to form a rich microenvironment for metastasis growth. However, the signalling interactions between the ovarian cancer cells and the cells of the omentum are poorly understood. Therefore, the aim of this thesis is to investigate the crosstalk between EOCs and omental adipocytes, mesothelial cells and microvascular ECs which may support the induction of metastasis and angiogenesis during secondary spread of EOC to the omentum.
Initially, improved and reliable methodologies for the isolation of primary human omental microvascular ECs (HOMECs) and omental mesothelial cells from donated human omentum were developed, thus providing disease relevant cell types for further study. The EOC cancer cell lines SKOV3 and A2780 were used as model cancer cells. HOMEC proliferation (to model EC activation during angiogenesis) was assessed by WST8 and BrdU assays. Commercially available antibody-based arrays and ELISA were used to investigate the secretome of omental adipose tissue and HOMEC signalling pathways.
Conditioned medium (CM) from omental adipose tissue significantly enhanced HOMEC proliferation via a signalling mechanism which involved STAT3 activation. Thus, secreted adipokines were screened to investigate the pro-proliferative factors responsible, highlighting adiponectin, cathepsin L, MIF, leptin and lipocalin-2. However, no increase in proliferation was observed when HOMECs were treated with any of these selected adipokines, either individually or in combination.
In contrast, treatment of HOMECs with CM collected from ovarian cancer cells which had been pre-incubated with the adipokines leptin or lipocalin-2 did significantly increase HOMEC proliferation compared with CM from untreated cancer cells. Activation of VEGFR2 by its ligand VEGF was implicated, as confirmed by ELISA and receptor inhibitor proliferation studies. Alongside VEGF/VEGFR2, the signalling mechanism may be potentiated by PDGFRβ activation leading to downstream activation of ERK1/2 and subsequent cell proliferation. Importantly, the ERK1/2 and STAT3 signalling pathways may be additive, suggesting that in the omentum, interaction between the adipocytes and the EOC cells can jointly drive angiogenic changes in the omental microvasculature during metastasis.
Cancer cell adhesion onto the mesothelial layer under experimental conditions mimicking peritoneal fluid shear stress was also examined. While CM from adipose tissue or EOCs had no effect, CM from mesothelial cells themselves decreased cancer cell adhesion (implying a protective effect) at a shorter timepoint.
The data presented suggest that while secreted omental adipokines do not impact the initial adhesion of metastasising EOC to the omentum, following invasion of the cancer cells into the omental tissue the microenvironment created is able to support angiogenesis and secondary tumor growth. Specifically, the adipocytes secrete (as yet unidentified) factors that can induce proliferation directly in HOMECs, in addition to secreting leptin and lipocalin-2 which act indirectly on the cancer cells to induce increased production of VEGF, which then also activates pro-angiogenic changes in the HOMECs. Thus, the omental microenvironment provides a rich metastatic niche for the growing secondary tumour, which may partly explain the high rates of omental metastasis in EOC patients.
In conclusion, tumour angiogenesis within the omentum is enhanced by interactions between adipokines secreted from resident adipocytes and metastatic cancer cells. A better understanding of the mechanisms involved may highlight reasons why current therapies are relatively ineffective and better direct the development of future therapies. |
|