Conflicting results from clinical observations and murine models: what is the role of plasminogen activators in tumor growth?
Conflicting results from clinical observations and murine models: what is the role of plasminogen activators in tumor growth?
复制标题
临床观察和小鼠模型的相互矛盾的结果：纤溶酶原激活剂在肿瘤生长中的作用是什么？
DOI：
10.1093/jnci/djj227
复制
发表时间：
2006
期刊：
影响因子：
--
通讯作者：
Tosato,Giovanna
中科院分区：
文献类型：
--
作者：
Narazaki,Masashi;Tosato,Giovanna
Masashi Narazaki, Giovanna Tosato that uPA and tPA did not directly regulate cell proliferation. Unlike tumor cells that produce active uPA, tumor cells that overexpress enzymatically inactive forms of uPA were similar to control cells in their tumorigenicity, suggesting that protease activity is required to reduce tumor growth. Thus, the results of Merchan et al. appear at odds with clinical observations that link high tumor expression of uPA with poor prognosis, and they prompt several considerations. What are the differences between the experimental mouse tumor model presented by Merchan et al. and human cancer? To more closely mimic human breast cancer, the authors overexpressed the murine uPA gene in a murine mammary cancer cell line (BALB/c strain) and inoculated the cells orthotopically into immunocompetent BALB/c mice. Nonetheless, critical differences may exist between the model and the human disease. First, are enzyme levels achieved in the experimental system comparable to those in the human cancer? The authors have generated clones that overexpress the enzyme, in spite of the fact that the mammary carcinoma 4T1 cells express endogenous plasminogen activator (PA) mRNAs. Enzyme levels critically influence biological activities. Overproduction of uPA in endothelial cells was incompatible with normal capillary morphogenesis, which requires an appropriate balance between proteases and their inhibitors (18). At superphysiologic concentrations, PAI-1 inhibited tumor invasion and neovascularization, but tumor growth was reduced in PAI-1 null mice (19, 20). Second, the timing of enzyme expression in relation to tumor development may differ. The human correlative studies used tissue specimens from primary surgery (12, 13), and we do not know how enzyme levels fluctuated from tumor outset. In the experimental model of Merchan et al., high-level protease activity was present from the time the tumor cells were inoculated, which may have impaired initial cell attachment and growth (21, 22). Indeed, Merchan et al.(17) showed that, in vivo, PA-overexpressing tumor cells proliferated less than control cells from a very early time point. Third, the site of enzyme release may differ. Tissue specimens of human breast cancer from patients and from a xenograft model showed that expression of uPA is strongest in the stromal fibroblasts at the invasive front of the tumors (20, 23), whereas in the model described by Merchan et al., high PA expression in tumors was diffuse. Thus, enzyme levels, time of release, and location of release may constitute critical differences between the experimental tumor model of Merchan et al. and the human disease. Finally,
