OBJECTIVE: To study the efficacy of local therapy using overexpression of the mouse angiostatin gene to treat human hepatic carcinoma. To investigate the mechanism of any therapeutic effect. METHODS: Full-length mouse angiostatin cDNA was directly cloned into the eukaryotic expression plasmid pcDNA3.1(+). The recombinant plasmid was used to transfect HCC7721 human hepatic carcinoma cells (experimental group). The vector control group cells were transfected with pcDNA3.1(+), and the mock control group was not exposed to the plasmid. Angiostatin expression at the mRNA and protein levels was detected by using RNA dot blots and fluorescence-activated cell sorter (FACS), respectively. Cell growth in vitro was assayed with thiazoyl blue tetrazolium bromide and cell cycle distributions were analyzed with FACS. Angiostatin-expressing cells and the vector control cells were transplanted subcutaneously into nude BALB/c mice and tumor volumes were measured after 30 days. Microvessel densities (MVD) and angiostatin expression in the tumor tissues were detected by using immunohistochemical staining with the primary anti-vWF and anti-HA antibodies, respectively. RESULTS: The pcDNA3.1(+)-angio recombinant eukaryotic expression plasmid was successfully constructed. No significant differences existed in cell morphology, cell growth curve and cell cycle distributions among the three study groups in vitro . However, markedly smaller tumor volumes, lower microvessel densities in tumor tissues, and stronger angiostatin-positive staining were observed in the experimental group relative to the control group, indicating reduced tumorigenesis associated with angiostatin overexpression in vivo . CONCLUSIONS: Mouse angiostatin did not directly inhibit human hepatic carcinoma cell growth and proliferation in vitro. However, local production of angiostatin effectively inhibited tumorigenesis of HCC7721 in vivo , probably because of inhibition of tumor angiogenesis via a... [ABSTRACT FROM AUTHOR]