Expression of membrane type 1 matrix metalloproteinase (MMP-14) in epithelial ovarian cancer: high level expression in clear cell carcinoma

Abstract

Objective: Clear cell carcinomas of the ovary constitute approximately 5% of all ovarian neoplasms and have a distinct gene expression profile relative to other ovarian carcinoma histotypes. Tumors often present as an early stage large pelvic mass with a high degree of recurrence and frequent early metastasis. Matrix metalloproteinases (MMPs) play a role in intraperitoneal metastasis through breakdown of cell-cell and cell-matrix barriers, enabling anchoring of secondary lesions and promoting proliferation in a geometrically constrained matrix environment. The objective of this study was to evaluate MMP expression in ovarian clear cell carcinoma.

Methods: Immunohistochemistry was used to evaluate expression of membrane type 1 MMP (MMP-14), MMP-2 and MMP-9 in a panel of ovarian tumors. Western blotting and gelatin zymography were used to examine MMP-14 expression and activity in the clear cell carcinoma cell line ES2. The ability of ES2 cells to invade and proliferate within three-dimensional collagen gels was evaluated.

Results: High level expression of MMP-14 and MMP-2 were observed in ovarian clear cell carcinoma relative to other histotypes (94-95% strong positive). MMP-14 was expressed and active in cultured ES2 cells. ES2 cells also exhibited MMP-dependent invasion of and proliferation within three-dimensional collagen gels.

Conclusions: The high level expression of MMP-14 together with in vitro functional analyses suggest that MMP-14 may contribute to both the proliferative capacity and the enhanced parenchymal metastasis of ovarian clear cell carcinoma.

Figure 1. Immunohistochemical analysis of serial human clear cell ovarian tumor sections for MMP-14 and MMP-2

Serial sections of primary ovarian tumor samples were stained with (A) H&E or antibodies to (B) MMP-2 or (C-F) MMP-14 (as indicated) and scored as described in “Materials and Methods”. Strong positive staining for both MMP-14 and MMP-2 was observed in 94% (17/18) of tumors analyzed and examination of serial tumor sections revealed co-localization of staining. Panels (D-F) show examples of tumors with staining intensity scored as weak (1+), moderate (2+) and strong (3+), respectively. 400X magnification.

Figure 2. MMP-14 protein expression in ES2 cells

Cells were grown to confluence, serum-starved for 3 hours and incubated in the presence or absence of concanavalin A (20 ug/ml) for 24 hr in serum-free media. (A) Whole cell lysates were probed for MMP-14 protein using anti-MT1-MMP hinge Ab (1:1000 dilution) and HRP-conjugated secondary antibody (1:5000 dilution) followed by ECL detection. (B) Cell surface expression of MMP-14 was determined by labeling cell surface proteins with a non-cell permeable biotin analog followed by streptavidin precipitation and probing the labeled protein pool for MMP-14 as described above. In control experiments, DOV13 cells were used for comparison.

Figure 3. MMP-14-catalyzed proMMP-2 activation

(A) Cells were grown to confluence, serum-starved for 3 hours and incubated in the presence or absence of concanavalin A (20 μg/ml) or LPA (80μM) for 24hr in serum-free media. Gelatinase activity of conditioned media was detected using zymography. The migration positions of pro- and active MMP-2 are indicated. (B) Effect of three dimensional collagen culture on MMP-14 activity. Cells were grown to 70-80% confluence in 6-well culture plate. Serum-free media (1 ml) or type I collagen (0.8 mg/ml) diluted in serum-free media, containing concanavalin A (20μg/ml) or vehicle, was added to each well. After 48 hours incubation, conditioned media or collagen gel was collected and centrifuged. Presence of active MMP-2 in conditioned media was determined using gelatin zymography. Control samples contained the broad spectrum MMP inhibitor GM6001 (25 uM, designated GM).

Figure 3. MMP-14-catalyzed proMMP-2 activation

(A) Cells were grown to confluence, serum-starved for 3 hours and incubated in the presence or absence of concanavalin A (20 μg/ml) or LPA (80μM) for 24hr in serum-free media. Gelatinase activity of conditioned media was detected using zymography. The migration positions of pro- and active MMP-2 are indicated. (B) Effect of three dimensional collagen culture on MMP-14 activity. Cells were grown to 70-80% confluence in 6-well culture plate. Serum-free media (1 ml) or type I collagen (0.8 mg/ml) diluted in serum-free media, containing concanavalin A (20μg/ml) or vehicle, was added to each well. After 48 hours incubation, conditioned media or collagen gel was collected and centrifuged. Presence of active MMP-2 in conditioned media was determined using gelatin zymography. Control samples contained the broad spectrum MMP inhibitor GM6001 (25 uM, designated GM).

Figure 4. Invasion of type I collagen gels by ES2 cells

Cells were grown to confluence and serum starved for 3 hours before trypsinization. Equal numbers of cells in serum-free media were added to the inner chamber of collagen-coated inserts (100μg) and incubated for 20 hours. Invading cells were fixed, stained, and counted. Results are the average of two experiments of triplicate wells. Error bars indicate standard deviation.

Figure 5. ES2 cells demonstrate a growth advantage in 3D collagen

Cells (1×10³) were plated in neutralized collagen (500μl) at a concentration of 2mg/ml. Collagen mixture was incubated at 37°C for 30 minutes to allow gel to form. Growth media was added to each well. Cells within collagen gel were photographed using phase-contrast microscopy 1 and 5 days after plating.