Assessing the responsiveness and reactivation of memory T cells in the microenvironment of human ovarian tumors and ascites fluid
MetadataShow full item record
One of the major conundrums in cancer is why human tumors are not rejected and progress despite the presence of inflammatory leukocytes in the tumor microenvironment. While studies addressing the mechanisms responsible for the failure of immunocompetent cells to control tumor progression have shed considerable light upon this issue, not enough is known about the mechanisms contributing to the regulation of tumor-derived T cells in the microenvironment of human tumors. T cells derived from both solid primary and metastatic human ovarian carcinoma and ovarian ascites fluid were assessed both phenotypically and functionally. The majority of the CD3+T cell populations derived from both the ascites and solid tumors have a phenotype that is consistent with an effector memory T cell namely CD45RO+, CD45RA-, CD44+, CD11a+, CXCR3+ and CD62L low or negative. In spite of this same phenotype, the ascites-derived and tumor-derived T cells initially appeared to demonstrate differential responsiveness when stimulated with anti-CD3/CD28. The memory T cells from the ascites translocated NF-κB, mobilized CD107a and proliferated, to the same degree as normal donor peripheral blood lymphocytes. However, only a portion of the tumor-derived T cells were found to be responsive by all three endpoints (i.e. NF-κB, CD107a expression and proliferation) of activation. The ability of the solid tumor-derived T cells to translocate NF-κB in response to TNF-α and PMA/Ionomycin demonstrated that these cells remain responsive to a CD3/CD28 independent pathway (TNF-α induced) and have an intact signaling pathway distal to diacylglycerol (PMA is a diacylglycerol analog). Furthermore, the tumor-derived T cells are able to translocate the transcription factor NFAT to the same degree as the T cells derived from the ovarian ascites and normal donor peripheral blood. Collectively these results suggested that the site of the CD3/CD28 signaling arrest, in the tumor-derived T cells, was localized to diacylglycerol. Attempts to reverse this tumor-derived T cell hyporesponsiveness were unsuccessful. The initial finding that the T cells from the ovarian ascites responded normally to activation was surprising given the assumption made by several investigators that both the solid tumor and tumor ascites microenvironments are immunosuppressive. Since our initial testing of was done with the ascites-derived T cells separated from the ascites fluid, further studies evaluated the responsiveness of these T cells in the original ovarian ascites supernatant. The ascites-derived T cells stimulated (with anti-CD3/CD28) in the ascites supernatant had a significant reduction in NF-κB translocation, proliferation and IFN-γ production, when compared to the ascites-derived T cells stimulated in ascites free medium. The ascites-induced suppression in anti-CD3/CD28 activation was not specific to malignancy as the activation of T cells derived from both the ovarian ascites and normal donor peripheral blood were suppressed by ascites fluids derived from patients with hepatic cirrhosis (non-malignant microenvironment). These data suggested that the reduction in T cell activation was not dependent on malignancy but rather related to common elements within chronic inflammatory microenvironments that have the capacity to suppress many if not all mature T cells types. Studies which evaluated NFAT translocation, and the phosphorylation of the upstream CD3/CD28 signaling events and ERK phosphorylation (which are downstream of DAG) further validated that the signaling defect induced by the ascites fluid was initiated downstream of PLC-γ and likely localized to the turnover or functional inactivation of DAG. Therefore, the effect of the ascites supernatant on activation very closely mirrors the molecular signature seen in the solid-tumor microenvironment. A novel human tumor xenograft model has been designed and characterized that is expected to make it possible to study T cell function within the tumor microenvironment. The implantation of non-disrupted pieces of human tumor from non-small lung carcinoma injected into NOD-scid IL-2cγR null mice resulted in the formation of xenografts in which the tumor microenvironment is maintained for prolonged periods. Early studies with this model have revealed that the hyporesponsive T cells within the tumor microenvironment can be reactivated by exogenous IL-12 delivered by liposomes and biodegradable microspheres. Collectively these data demonstrate that the reduction in the activation of T cells is initiated and maintained in environments of solid tumors, tumor ascites and non malignant chronic inflammatory microenvironments. Based upon our experimental findings we postulate that this hyporesponsiveness is regulated through diacylglycerol in the CD3/CD28 pathway. Future studies are required to identify the factor(s) that are initiating this defect in activation.