Abstract
A linkage between hemostatic system components and tumor cell metastatic potential has been well established, but the underlying mechanism(s) by which various circulating and cell-associated coagulation factors and platelets promote tumor cell dissemination remains to be fully defined. One potential mechanism by which tumor cell-associated microthrombi might enhance metastatic potential is by interfering with the cytolytic elimination of tumor cell emboli by natural killer (NK) cells. In order to explore this hypothesis, we studied tumor dissemination in mice lacking either fibrinogen or Gαq, a G protein critical for platelet activation. Comparative studies of experimental lung metastasis in control and Gαq−/− mice showed that loss of platelet activation resulted in a two-orders-of-magnitude decrease in pulmonary metastatic foci formed by either Lewis lung carcinoma or B16 melanoma. The difference in metastatic success was not the result of differences in tumor growth rate, as tumors transplanted into the dorsal subcutis of Gαq−/− and wildtype animals grew at similar rates. Rather, tumor cell fate analyses using radiolabeled tumor cells showed that the survival of tumor cells within the lung was significantly improved in mice that retained platelet activation function relative to Gαq−/− mice with a profound platelet activation defect. In order to examine the potential interplay between platelet activation and natural killer cell function, we compared pulmonary tumor cell survival in cohorts of control and Gαq−/− mice immuno-depleted of NK cells with an anti-asialo GM1 antibody. Remarkably, platelet function was no longer a determinant of metastatic potential in mice lacking NK cells. Given that fibrin(ogen) is also an established determinant of metastatic success we explored whether the influence of this key hemostatic factor on tumor cell dissemination was also mechanistically-coupled to natural killer cell function. We interbred fibrinogen-deficient mice with Gz-Ly49A transgenic mice known to have a constitutive deficit in NK cells. In those cohorts of mice with normal NK cells, we affirmed the earlier finding that fibrinogen deficiency resulted in a significant diminution in metastatic potential. However, consistent with our findings in mice with defective platelet activation, fibrinogen was found to no longer be a determinant of metastatic potential in mice lacking NK cells. These data establish another important link between innate immune surveillance and the hemostatic system. Further, it appears that at least one mechanism by which tumor cell-associated microthrombi increase metastatic potential is by restricting NK cell-mediated tumor cell elimination. Given that NK cell cytotoxicity requires direct contact with any target cell, one attractive model presently being explored is that tumor cell-associated platelets physically block NK cell access to tumor cell emboli.
Natural killer cells in children with malignant solid tumors. Effect of recombinant interferon-α and interleukin-2 on natural killer cell function against tumor cell lines