Endometriosis is a benign inflammatory disorder, defined by the presence of endometrial tissue outside the uterus with lesions typically found on the pelvic peritoneum in close association with the peritoneal mesothelium. The prevalence of endometriosis is estimated at 6-10% of women of reproductive age and it is associated with chronic pelvic pain, dysmenorrhoea, dyspareunia and infertility. Surgical excision can provide symptom relief, but symptoms recur in up to 75% of surgical cases and available medical treatments have undesirable side effects. New treatments are limited due to our poor understanding of the aetiology of endometriosis. To date, the majority of research has focused on changes within the ectopic endometrial tissue to explain the development of endometriosis lesions, however, there is increasing evidence that the peritoneal mesothelium plays an important role. According to Sampson’s theory of retrograde menstruation, ectopic endometrial cells must first have to attach to the surface of the peritoneum before undergoing invasion, proliferation, and neoangiogenesis. TGF-β1 is an inflammatory growth factor that regulates a variety of cellular functions including; cell adhesion, cell invasion and angiogenesis. Levels of TGF-β1 are increased in the peritoneal fluid of women with endometriosis compared to controls and research using a mouse model of endometriosis has demonstrated TGF-β1-null mice to develop smaller and fewer endometriosis lesions than their wild-type controls. Together these studies suggest that TGF-β1 plays a major role in the development of peritoneal endometriosis lesions and that targeting this pathway may be of therapeutic potential. However the functional role that TGF-β1 plays in peritoneal endometriosis is still unclear. The overall aim of this thesis was to determine if the peritoneal expression of TGF-β and its target genes are disrupted in women with endometriosis and whether this could contribute to the development of endometriosis lesions. Our first aim was to determine if the peritoneum was a source of TGF-β expression and if reception and/or signalling were altered in women with endometriosis. We found that the peritoneal fluid of women with endometriosis contained increased concentrations of TGF-β1 and that peritoneal mesothelial cells adjacent to endometriosis lesions expressed significantly higher levels of TGF-β1 mRNA. Analysis of TGF-β signalling targets within the peritoneum showed that women with endometriosis express significantly higher levels of TGF-β targeted genes associated with tumourigenesis processes including; EMT, invasion and angiogenesis. We next asked if there are changes in the metabolic phenotype of endometriosis lesions and peritoneum in women with endometriosis, similar to the metabolic changes seen in tumour cells. Endometriosis lesions expressed markers of aerobic glycolysis, including HIF-1α, suggesting that lesions may metabolise in a similar fashion to tumours. Furthermore, peritoneum adjacent to endometriosis lesions expressed significantly higher levels of markers of aerobic glycolysis, including HIF-1α, suggesting that the peritoneum may feed forward high-energy lactate, a by-product of glycolysis, to the endometriosis lesions. These observations were supported by significantly increased lactate concentrations within the peritoneal fluid of women with endometriosis that positively correlated with levels of TGF- β1. TGF-β1 was shown to increase expression of glycolysis markers and lactate expression in peritoneal mesothelial cells in-vitro, suggesting TGF-β may regulate this change. We then determined if TGF-β1 was responsible for the change in peritoneal mesothelial cell metabolism by signalling through the ID-HIF-1α pathway. ID proteins are transcription factors, whose expression is regulated by the TGF-β superfamily. We found expression of ID1 mRNA to be increased in the peritoneum of women with endometriosis and that TGF-β1 significantly increased ID1 but decreased ID2 expression in the peritoneal mesothelial cells in-vitro. ID1 siRNA knockdown decreased glycolysis initiator HIF-1α mRNA and ID2 siRNA knockdown increased HIF-1α mRNA and lactate expression, suggesting TGF-β1 regulates mesothelial cell metabolism, at least in part, through the ID pathway. ID transcription factors are also known to regulate VEGF-A expression, therefore we next determined if TGF-β1 induced ID1 and/or reduced ID2 expression in the peritoneum promoted VEGF-A mRNA and protein expression. VEGF-A, a cytokine essential for angiogenesis, was significantly increased in the peritoneal fluid of women with endometriosis and levels positively correlated with TGF-β1. TGF-β1 increased VEGF-A expression in-vitro and siRNA knockdown of ID1 decreased and siRNA knockdown of ID2 increased VEGF-A mRNA and protein expression in the peritoneal mesothelial cell. Lastly we aimed to determine if TGF-β1 induces EMT in peritoneal mesothelial cells. The peritoneum of women with endometriosis expressed higher levels of EMT markers. Exposure of peritoneal mesothelial cells to TGF-β1 in-vitro induced EMT-like changes, including; changes to cell morphology, gene expression and protein localisation. Peritoneal mesothelial cells were more migratory and invasive suggesting that TGF-β1 induced EMT may disrupt the mesothelial cell monolayer allowing ectopic endometrial cells to invade the peritoneal tissue or for peritoneal mesothelial cells to migrate into endometriosis lesions. In summary, the novel data presented in this thesis provides evidence that the pelvic peritoneum and in particular the peritoneal mesothelial cell may play a critical role in the aetiology of peritoneal endometriosis. Expression of TGF-β1 and its transcriptional target genes are dysregulated in the peritoneum of women with endometriosis. TGF-β regulated ID expression may induce changes in cell metabolism and promote neoangiogenesis, prompting peritoneal endometriosis lesion development. Furthermore other TGF-β1 transcriptional targets, such as those involved in EMT, are also altered in the peritoneum of women with endometriosis and may contribute the development and maintenance of lesion formation. These results point to a central role for TGF-β1 expression and signalling in the aetiology of peritoneal endometriosis. Furthermore it is likely that changes within the expression profile and morphology of the peritoneal mesothelial cells contribute to peritoneal lesion formation. Drugs that target these pathways may provide new therapies for women with endometriosis.

Metastatic dissemination of ovarian tumors involves the invasion of multi-cellular tumor cell clusters into the mesothelial cell lining of organs in the peritoneal cavity. We developed an in vitro assay that models this initial step of ovarian cancer metastasis to investigate the mechanisms of invasion. Pre-clustered ovarian cancer multicellular spheroids are incubated with GFP-expressing mesothelial monolayers and the extent of mesothelial invasion is monitored by time-lapse video microscopy.

Epithelial Ovarian Cancer (EOC) kills more women annually in the United Kingdom than any other gynaecological cancer. Survival rates for women diagnosed with EOC have not improved over the past 30 years, due to the often advanced stage at presentation, where widespread intra-peritoneal dissemination has occurred. The natural history of the disease remains uncertain but the ovarian surface epithelium (OSE) is a strong candidate for the tissue of origin. The OSE undergoes cyclical damage and repair in women of reproductive age following ovulation, which can be considered an acute inflammatory event. Factors that prevent ovulation (pregnancy, breastfeeding and contraceptive pill use) also protect against the development of EOC. Previously published data show that the OSE is able to upregulate the enzyme 11-beta hydroxysteroid dehydrogenase type 1 (11βHSD1) in response to inflammation, the enzyme responsible for converting inactive cortisone to anti-inflammatory cortisol. This thesis hypothesises that 11βHSD isozymes are deregulated in ovarian cancer; that the peritoneal surface epithelium (PSE) is indistinguishable from the OSE in its response to inflammation and should be considered a potential source of some “ovarian cancers”; and finally that the expression of the tumour suppressor gene OPCML (OPioid binding Cell adhesion Molecule-Like) is altered by inflammation. These hypotheses were examined at three levels. Firstly, primary cultures of EOC were established, and glucocorticoid metabolism and the response to inflammation was compared to normal OSE. Results from these investigations reveal that the11βHSD1 response to IL-1α stimulation is impaired in EOC compared to normal OSE at the mRNA level but there is no significant difference when 11βHSD1 enzyme activity is measured in these tissues. When basal levels of 11βHSD1, 11βHSD2 and COX2 are compared amongst untreated samples of EOC and OSE, there was a significant correlation between 11βHSD1 and COX2 mRNA expression (P<0.001). 11βHSD2 mRNA expression was significantly higher in the EOC specimens compared to OSE (P<0.05). Secondly the response to inflammation was compared in primary cultures of human peritoneal surface epithelial (PSE) cells and OSE. The data suggest that the mRNA response to inflammation was similar in OSE and PSE, but that the 11βHSD1 enzyme activity was reduced in PSE (P<0.05), which may result in differences in tissue healing. Finally, the effect of inflammation on the expression of the ovarian cancer associated tumour suppressor gene (TSG), OPCML (OPioid binding Cell adhesion Molecule-Like) and the other members of the IgLON family, was examined in OSE. These results suggest that OPCML mRNA expression can be induced by IL-1α, an effect that is inhibited by cortisol.