| dc.description.abstract | The primary cause of mortality associated with cancer is metastasis, i.e., the spread of transformed cells from the site of oncogenesis to secondary loci within the body. Epithelial ovarian cancer (EOC) is a debilitating gynaecological disease in which transformed ovarian surface epithelia metastasize by shedding into the peritoneum and disseminating through the ascites (abnormally accumulated peritoneal fluid) to adhere to, and colonize, specific peritoneal locales (such as the mesentery and omentum) and constitute secondary micrometastatic foci. Metastasis can be understood as the result of a dynamic exchange of cues between disseminated cancer cells (seeds) and the peritoneal microenvironment (soil). Within this metaphorical conceptualization, the dynamics of the seed (cancer cell) has been under active investigation. In contrast, the importance of the soil (prospective metastatic site microenvironment), is relatively less well understood and is only beginning to be appreciated. In this thesis, I have examined the role of distinct components of the peritoneal microenvironment in the process of EOC metastasis.
The body cavity (coelom) is lined by mesothelia except where lymphatic drainage occurs, wherein the underlying extracellular matrix (ECM) is exposed. Therefore, mesothelia represent the principal cellular barrier against incoming cancer cells and the clearance of the former by the latter has been demonstrated, although mechanisms underlying the same are not known. Using cocultures of untransformed immortalized human mesothelia and EOC cell lines, I show that confluence of mesothelia plays a contextual role in the process: dense monolayers resist spheroidal adhesion and their own clearance, even if their migration is pharmacologically inhibited. Even subconfluent mesothelial monolayers quickly grow confluent and confine the spreading cancer spheroids; however, inhibiting the actin remodeling Arp2/3 and Rho activity renders them susceptible to clearance by being pushed by an expanding cancer spheroid. I then use human cancer cell-murine mesenteric explant cocultures to show that the confluence of mesothelia may be regulated by secretions of cancer spheroids: exposure of murine mesenteries to these secretions results in the depletion of their mesothelial content, suggesting that a combination of global and local mechanisms may be employed by EOC cells to break down the cellular peritoneal barrier.
I have next focussed on the role of Fibronectin and Collagen I, the principal peritoneal ECM molecules in EOC colonization. I observe increased levels of fibronectin (using immunohistochemistry) and fibrillar Collagen (using two-photon excited fluorescence (TPEF) microscopy and second harmonic generation imaging) in EOC patient omenta with cancer deposits, compared to non-cancerous patient omental tissues. Using 3D ECM scaffold cultures, I show that increased concentrations of Fibronectin and Collagen I cause greater spreading of EOC spheroids. As the concentration of Collagen I is increased, stronger fibrillar organization is evidenced through second harmonic generation (SHG) microscopy and elastic modulus is increased as measured using atomic force microscopy. For both substrata, Arp2/3 and Rho activity were found to be crucial for spheroidal spreading. Furthermore, I have investigated the role in ovarian cancer colonization of Decorin, a dermatan sulfate proteoglycan (DSPG) that binds to Collagen I, alters its fibrillar geometry, and is secreted abundantly by peritoneal mesothelia, making it a constituent of the untransformed omental ECM. Immunohistochemistry reveals lower expression of Decorin relative to uncolonized human omenta and regions within colonized omenta which lacked EOC cell deposits. Scaffolds of Collagen I polymerized with Decorin show lower spreading of EOC spheroids. Concomitant with my earlier observations, presence of Decorin weakens the organization of Collagen I and decreases its stiffness, suggesting that altered architecture, biochemical, and biophysical properties of ECM substrata are correlated with the spread kinetics of ovarian cancer spheroids.
Interestingly, EOC cells themselves express Decorin to significantly lower extents than untransformed coelomic epithelia. I confirm this in vivo by observing lower levels of Decorin in the ovarian tumors relative to untransformed ovarian tissue. To probe whether the depletion of Decorin regulates EOC progression, I stably overexpress the protein in the SKOV-3 ovarian cancer line. In addition, I also develop a line overexpressing a mutant Decorin that lacks conjugation to dermatan sulfate glycosaminoglycan. SKOV-3 cells with Decorin overexpression form smaller spheroids and migrate inefficaciously on Collagen I scaffolds compared with control cells. Cells from Decorin overexpressing spheroids show slower migration velocities on Fibronectin and Collagen I. Interestingly, these traits are unaffected in cells with expression of the mutant Decorin, suggesting the glycosaminoglycan chain may play an important role in the suppression of colonization attributed to Decorin.
Taken together, my results suggest that ovarian cancer metastasis is accompanied by suppression of key microenvironmental soil constituents that impede colonization, while levels of components that aid the process are increased. The potential remodelling of the soil by the seed adds an intriguing layer of complexity to the metaphor; while its molecular mechanisms remain to be further probed, the principle may hold true across other disseminated cancers. | en_US |
