Protease activity of urokinase and tumor progression in a syngeneic mammary cancer model

Background. We and others have previously shown that plasminogen activators generate endogenous angiogenesis inhibitors and induce antiangiogenic activity. Here we assessed the effects of plasminogen activator overexpression on tumor progression in a syngeneic mammary cancer model. Methods: Genes encoding murine tissue plasminogen activator (tPA), urokinase (uPA), and vector controls were stably transfected into 4T1 murine mammary cancer cells, and cell proliferation in vitro was analyzed. Cells were also implanted into female BALB/c mice (n = 12 per group), and tumor growth, lung metastases, and survival were compared. Tumor cell proliferation and microvessel formation were analyzed by immunohistochemistry using antibodies to proliferating cell nuclear antigen and CD31, respectively. 4TI cells transfected with proteolytically inactive uPA mutants (A and B) were assayed for proliferation in vitro and tumor growth in vivo by using the same syngeneic model (eight to 10 mice per group). All statistical tests were two-sided. Results: In vitro growth of uPA- and tPA-overexpressing and control 4TI cells was similar. In vivo, however, inhibition of tumor growth and lung metastasis were inhibited in the mice carrying tPA- and uPA-overexpressing tumors, compared with controls (tumor weight at day 34: control, mean = 1760 mg, 95% confidence interval [CI] = 1434 to 2087 mg; tPA, mean = 921, 95% CI = 624 to 1217 mg; P <.001; uPA, mean = 395 mg, 95% CI = 161 to 629 mg; P <.001; number of lung metastases at day 34: control, mean = 117, 95% CI = 74 to 159; tPA, mean = 33, 95% CI = 13 to 52; uPA, mean = 15, 95% CI= 4 to 25; P <.001). Median survival was 42 (95% CI = 36 to 44), 55 (95% CI = 48 to 61), and 73 (95% CI = 51 to 86) days in the control, tPA, and uPA groups, respectively (P <.001). uPA and tPA-expressing tumors had reduced angiogenesis and cell proliferation compared with controls. Tumors overexpressing uPA mutants grew faster than tumors expressing wild-type uPA (tumor volume at day 30: wild-type uPA, mean = 203, 95% CI = 121 to 285 mm(3); control, mean = 534, 95% CI = 460 to 608 mm(3); P <.001; mutant A, mean = 600, 95% CI = 520 to 679 mm(3); P <.001; and mutant B, mean = 435, 95% CI = 358.9 to 511 mm(3); P = .005). Conclusions: In this mouse model, uPA expression delayed tumor progression and had antiangiogenic and antiproliferative effects that may be mediated by uPA's protease activity. These results challenge the current dogma of proteases being exclusively tumor promoting and provide further rationale for exploring plasminogen activators as antitumor agents.