[Regulation of matrix-degrading enzymes in gynecologic cancer tissues and cells]

Studies of tumor invasion and metastases have focused on the degradation of the basement membrane, which is predominantly made up of type IV collagen, laminin, and heparan sulfate proteoglycans. Matrix metalloproteinase-2 (MMP-2) and MMP-9, which can degrade type IV collagen, are implicated in cancer invasion and metastasis. Released and activated MMPs are controlled by specific tissue inhibitors of metalloproteinase (TIMP). In the present study, we have examined gelatinolytic and TIMP activity in the conditioned medium of human normal and cancer tissues by zymography and reverse zymography.

1) Tissues. Tissues were obtained at operation after informed consent was got from each patient. Sliced tissues were incubated in serum-free medium for 4 or 24 h at 37 degrees C. Human ovarian cancer cells (SAOV) were cultured for 24 h in serum-free medium containing conditioned medium of stromal tissues. After washing by PBS 3 times, SAOV cells were cultured for a further 24 h. 2) Zymography. Conditioned medium was subjected to SDS polyacrylamide gel containing 0.3 mg/ml of gelatin in zymography, and purified MMPs were added further in reverse zymography. After electrophoresis the gel was washed with Triton X-100, and incubated for 20 h at 37 degrees C in the reaction buffer. The gel was then stained with Coomassie brilliant blue. The gelatinase and TIMP activities were detected as unstained and stained bands, respectively. The photographs of the gels were scanned with a densitometer. 3) Other method. TIMP-1 levels of conditioned medium were assayed by ELISA kit. 4) Statistics. Statistical comparisons were made by Mann-Whiteny U test.

We have examined the gelatinolytic activity in gynecologic normal and cancer tissues by zymography and reverse zymography. Ovarian, cervical, and endometrial cancer tissues demonstrated higher gelatinolytic activity than normal tissues. The major gelatinases were those with molecular weight of 92 and 72kD, which corresponded to MMP-9 and MMP-2, respectively. The ratio of MMP 9 to MMP-2 was significantly higher in 3 types of cancer tissues than in normal tissues. Reverse zymography demonstrated that TIMP-1 and TIMP-2 were present in all tissues, and the ratio of TIMP-1 to TIMP-2 was significantly higher in 3 types of cancer tissues than in normal tissues. These findings suggested that MMP-9 and TIMP-1 were more associated with cancer phenotype than other types of MMP and TIMP. The influence of human stromal tissues (peritoneum, myometrium, ovary) on the secretion of MMPs and TIMPs was examined by addition of these stromal tissues culture medium to human ovarian cancer cells (SAOV). All conditioned medium of stromal tissues could increase in both MMP-2, MMP-9, TIMP-1, and TIMP-2 activity in SAOV cells. Fraction (> 100kD) of conditioned medium of peritoneum could increase remarkably in MMP-9, and this increase could be inhibited by anti alpha 5 antibody, which is the most popular receptor of fibronectin. Furthermore, the addition of fibronectin to SAOV cells induced increase in the secretion of MMP-9. These results demonstrated that one of the factors included in conditioned medium of peritoneum was fibronectin. We found that interferon beta could suppressed the secretion of MMP-2 and invasion in choriocarcinoma cells. However, no effect of interferon beta was observed in SAOV cells. Several flavonoids were screened to have ability to suppress the secretion of MMPs. All trans retinoic acid (RA) could suppress the secretion of MMPs in SAOV cells in time and concentration dependent manners. Further, RA could inhibited the invasion of SAOV cells by invasion assay using boyden chamber coated with matrigel.