Abstract B65: Milky spots are required for ovarian cancer metastatic colonization of peritoneal adipose depots

Abstract

Abstract Introduction: In ovarian cancer, attachment of cells to the omentum represents a rate-limiting step in metastasis formation. Thus, there is significant interest in identifying the tissue microenvironments involved in cancer cell colonization of this preferred metastatic site. The omentum is composed of translucent membranes and adipose, which contains immune structures known as milky spots. Two general models of omental colonization can be discerned from the published literature. In the “milky spot-driven” model, cancer cell lodging is due to factors produced by milky spots which serve as a colonization niche1. In contrast, the “adipose-driven” model proposes that adipocyte-derived factors are solely responsible for cancer cell lodging2. By taking advantage of peritoneal adipose deposits which either contain or lack milky spots, we set forth to test the hypothesis that these unusual immune structures are required for cancer cell colonization of adipose-rich tissues such as the omentum. Methods: Ovarian cancer cell migration, lodging and growth was assessed in a panel of peritoneal fat depots which either contain milky spots (omentum, splenoportal fat) or lack them (uterine fat, mesentery, and gonadal fat). The presence or absence of milky spots was confirmed by histology and immunohistochemistry (IHC) for CD45. Ovarian cancer cell lodging was evaluated via experimental metastasis assays using SKOV3ip.1 (human) or ID8 (mouse) ovarian cancer cells injected into Athymic Nude or C57BL/6 mice, respectively. At 7 days post injection (dpi) peritoneal fat depots were harvested and evaluated for the presence of cancer cells using histology and pan-cytokeratin IHC. The relative chemoattractant ability of tissue-conditioned media was evaluated by transwell migration assays using SKOV3ip.1 and ID8 cells. Results: The presence of milky spots in omental and splenoportal fat and their absence in uterine, gonadal, and mesenteric fat was confirmed by histology and IHC. In metastasis assays, cancer foci were observed in omental and splenoportal milky spots at 7 dpi. In contrast, cancer cells were never observed in adipose that lacked milky spots; even at 63 dpi. Consistent with this, conditioned media from milky-spot containing tissues had a significantly increased ability to promote ovarian cancer cells migration . Conclusions: In sum, our findings show that milky spots are an absolute requirement for cancer cell lodging on peritoneal adipose. Consistent with this, migration assays show that the presence of milky spots causes a significant enhancement in the ability of peritoneal adipose to promote directed migration. Taken together, these finding suggest that while adipose may secrete a general chemoattrative signal, additional contributions from the milky spots are required for localization and invasion. Current experiments include investigating the chemoattractant(s) used by the milky spots to attract cancer cells and the cell type(s) responsible for their secretion. Competing interests: The authors declare no competing interests.