Human prostate tumor angiogenesis in nude mice: Metalloprotease and plasminogen activator activities during tumor growth and neovascularization of subcutaneously injected Matrigel impregnated with human prostate tumor cells

Background: A critical aspect for growth of solid tumors is the development of a blood supply. Our objective was to establish a model for the study of angiogenesis of human prostate tumors by examining the growth of microvessels into Matrigel containing human prostate tumor cells implanted subcutaneously in nude mice. Methods: Human prostate tumor cell lines PC-3 and LNCaP were injected in Matrigel under the abdominal skin of nude mice and were harvested at 4, 8, and 14 days post-injection. The growth of tumor cells and blood vessels was examined histologically and by immunohistochemical localization of von Willibrand Factor VIII (vWF). Since plasminogen activators and matrix metalloproteases are associated with angiogenesis, the activities and molecular forms of these proteases were determined in Matrigel control and Matrigel-tumor cell subcutaneous implants. Results: Blood vessel formation in the Matrigel implants containing LNCaP and PC-3 cells was demonstrable at 8 days post-injection. However, the pattern of blood vessel formation by the two tumor cell lines was different; PC-3 tumors showed a more invasive phenotype and smaller diameter blood vessels, whereas LNCaP tumors grew as large cellular spheroids surrounded by large, dilated blood vessels. Many blood vessels of PC-3 tumors expressed vWF by day 14 of growth, whereas most blood vessels in LNCaP tumors were immunohistochemically negative for this antigen. Mouse skin blood vessels in the same PC-3 and LNCaP tumor histological sections were positive for vWF. Matrigel contained both plasminogen activator and metalloprotease activities. The plasminogen activator activity in Matrigel control implants was totally inhibited by 4 days post-injection, indicating the presence of an inhibitor provided by the host mouse. LNCaP tumor cells injected did not have appreciable plasminogen activator activity, nor did LNCaP tumors develop plasminogen activator activity with tumor growth post-injection. PC-3 cellS did have plasminogen activator activities, which were partially negated after subcutaneous injection (4 days), but then increased again by 8 days post-injection. This increase in plasminogen activator activity was due to urokinase (about 54 kDa) produced by the tumor and not by the mouse host (mouse urine urokinase about 44 kDa). Matrigel alone demonstrated gelatinase B (about 95 kDa) activity in zymograms, and gained considerable gelatinase A (about 70 and 74 kDa) activity after subcutaneous implantation. No metalloprotease activity from the tumor cells could be distinguished over that contributed by the mouse host cells in the Matrigel. Matrigel also contains caseinolytic activities of approximately 56, 80, 85, and 89 kDa. After subcutaneous injection of Matrigel, the 89 kDa form increases considerably in activity and the others are diminished. This pattern is also observed in LNCaP and PC-3 tumors post-injection, except the PC-3 tumors demonstrate increased 56 kDa activity. Conclusion: The subcutaneous growth of LNCaP and PC-3 prostate tumor cells in Matrigel in nude mice can be used to study tumor-induced angiogenesis. However, the organization of LNCaP and PC-3 tumor growth and the pattern of microvessels associated with each tumor are different in this system, implying that each tumor has unique influences on the pattern of microvessel development. The mode of action by which this is brought about is not known, but may be due to specific factors produced/released by the tumor cells.