• endometrial adenocarcinoma;
  • matrix metalloproteinases;
  • tissue inhibitors of metalloproteinases;
  • active gelatinases;
  • tumor invasion


  1. Top of page
  2. Abstract
  6. Acknowledgements


The actions of the extracellular-matrix degrading enzymes, matrix metalloproteinases (MMPs), are implicated in tumorigenesis. The cellular localization of MMP-2, MMP-9, membrane type 1 (MT1)-MMP, tissue inhibitors of metalloproteinases (TIMPs) 1-3, and the presence of active gelatinases were investigated in endometrial carcinoma.


Endometrial carcinomas were grouped according to histologic grade (Grades 1-3), depth of myometrial invasion (0, < 50%, > 50%) and the presence of vascular/lymphatic invasion. Twenty-nine endometrial carcinoma biopsies were investigated immunohistochemically to determine the tissue localization of MMP-2 (gelatinase A), MMP-9 (gelatinase B), MT1-MMP, and TIMPs 1-3. In situ hybridization was performed to localize MMP-2 and MMP-9 mRNA. The presence of active gelatinases was assessed using in situ zymography.


Epithelial tumor cells were the main site of MMP-2, MMP-9, and MT1-MMP protein. Variable stromal cell localization was also observed, particularly in areas adjacent to tumor nests. Semiquantitative analysis revealed increases in MMP-9 and MMP-2 but not MT1-MMP staining scores in tumor epithelial cells in the transition from histologic Grade 1 to Grades 2 and 3. Matrix metalloproteinase-9 and MT1-MMP staining scores in tumor cells were significantly associated with the presence of myometrial invasion and vascular/lymphatic invasion, while MMP-2 did not correlate with these factors. In addition, MT1-MMP was co-localized with MMP-2, supporting its role in the activation of proMMP-2. Tumor cells from all histologic grades stained intensely for TIMP-2 and TIMP-3 proteins, while variable stromal staining was observed. In Grade 1 carcinomas TIMP-1 was predominantly immunolocalized to the stromal compartment with variable tumor cell localization being observed in Grades 2 and 3 carcinomas. Matrix metalloproteinase-9 and MMP-2 mRNAs were predominantly observed in tumor epithelial cells as well as in the stroma to varying degrees. In situ zymography revealed active forms of gelatinases at the cellular surface and in association with tumor epithelial cells within endometrial carcinoma tissues.


These data suggest that increasing expression of MMPs and endometrial carcinoma progression are closely related. Active gelatinases are present in endometrial carcinoma, resulting in alterations to the microenvironment that promote tumor invasion and metastasis. Cancer 2002;94:1466–75. © 2002 American Cancer Society.

DOI 10.1002/cncr.10355

Endometrial carcinoma is the most common malignant tumor of the female genital tract. It typically affects postmenopausal women and is increasingly frequent in many advanced countries, composing 6% of all malignancies in women in the U.S. and 4% in Australia.1, 2 The invasion of endometrial cancer cells through the myometrium and to nearby lymph nodes is a key factor related to poor prognosis and involves the escape of tumor cells from their site of origin, their penetration of blood or lymph vessel walls, and their subsequent growth at a new site. Metastatic spread requires the interaction of tumor cells with components of the extracellular matrix (ECM) and with other cells. Such interactions depend on factors from the cellular surface, such as surface bound proteolytic enzymes, growth factors, and growth factor receptors and cell adhesion molecules. The ECM is the main physical barrier to the movement of cells in each step of tumor progression. This comprises both the interstitial ECM, whose fibrillar framework is composed mainly of Types I, II and III collagens, and the basement membrane, comprising predominantly Type IV collagen along with laminin and nidogen.

The matrix metalloproteinases (MMPs) are a family of zinc-dependent endopeptidases which participate in the degradation of collagens and other extracellular matrix macromolecules3 and are therefore implicated in tissue destruction in various pathologic conditions, including inflammation, tumor invasion, and metastasis. In addition, MMPs play a role in releasing and activating certain bioactive molecules, which in turn alter cell function.4, 5 Matrix metalloproteinases can be subgrouped based on substrate specificities and sequence characteristics into the stromelysins, collagenases, gelatinases, and the membrane type MMPs. Most MMPs are secreted as inactive zymogens (proMMPs) and are activated extracellularly; membrane-type MMPs are an exception, as they are anchored within the cellular membrane. The activation process is a key step for the action of the proteinases in vivo and generally occurs extracellularly. The tissue inhibitors of metalloproteinases (TIMPs) partly regulate MMP activity,6–8 which form high affinity, non-covalent 1:1 complexes with active forms of all MMPs.9 In addition, complexes are formed between specific TIMPs and the C-termini of progelatinases: TIMP-1 with proMMP-9 and TIMP-2 with proMMP-2.6, 10 It is now clear that, along with membrane type 1 (MT1)-MMP, TIMP-2 plays an important role in the activation of pro-MMP-2 in cell membranes.11 As it is associated with the insoluble ECM, TIMP-3 differs from the other family members.12

The enhanced expression of many MMPs has been correlated with the malignant phenotype of human cancer cells both in vivo and in vitro.13 In particular, the gelatinases are implicated with tumor invasion, as they degrade constituents of both the interstitial stroma and the subendothelial basement membrane matrix. Clinicopathologic studies have shown a correlation between the expression of MMP-2 and MMP-9 mRNA and the invasion and metastasis of human carcinomas, such as gastric,14 thyroid,15 breast,16 and ovarian carcinomas.17 Studies have shown MT1-MMP to play a significant role in the activation of proMMP-2 in breast18 and thyroid carcinomas.19 The immunoreactive enzyme proteins likewise have been shown in a wide variety of human carcinomas and to stroma or to tumor cells within these carcinomas,14, 16, 18, 20 the pattern and intensity of immunostaining varying with each carcinoma type. In many instances, immunostaining may reflect MMP binding to various ECM-bound molecules, such as integrins, fibronectin, and laminin.

Several studies have reported mRNA synthesis or the immunohistochemic localization of the gelatinases in the malignant endometrium. Such studies have localized MMP-2 and MMP-9 mRNA to stromal cells, including endothelial cells, macrophages, and fibroblasts, particularly in areas within or surrounding tumor aggregates,21, 22 and also, to varying degrees, to epithelial cells of tumor origin.21 Furthermore, the frequency of MMP-2 and MMP-9 expression increases with advancing histologic grade and with increasing depth of myometrial invasion.22 Studies on MMP immunolocalization in endometrial carcinoma are limited: matrilysin (MMP-7) was localized predominantly to carcinoma cells,19 MMP-2, MMP-1, and TIMP-1 proteins were correlated with histologic stage, depth of infiltration, histologic type and age,23 while MMP-9 proteins were mainly localized to tumor cells.24 Studies have localized TIMP-1 mRNA to the stromal compartment of the malignant endometrium, while TIMP-2, TIMP-3, and MT1-MMP mRNA have been localized to both the stromal and epithelial compartments.20 As in the case of the gelatinases, increased mRNA expression was observed in peripheral regions of the tumors and in relation to the lower differentiated tumors.20 However, to our knowledge, no studies have examined both the cellular site of synthesis (mRNA) and the location of the proteins for these enzymes in the same tissues. Neither is it known whether the gelatinases are active in endometrial carcinoma tissues.

In the current study we examined the cellular localization of MMP-2, MMP-9, MT1-MMP, TIMP-1, TIMP-2, and TIMP-3 proteins in human endometrial carcinomas of varying histologic grades by immunohistochemistry, along with MMP-2 and MMP-9 mRNA by in situ hybridization, and have shown correlations with increasing tumor grade and invasion. To our knowledge, we have shown for the first time the presence of active forms of gelatinases in endometrial carcinoma tissues.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Patients and Tissues

The current study utilized tissue from 29 patients undergoing total abdominal hysterectomy for endometrial carcinoma at the Monash Medical Centre and the Royal Women's Hospital, Melbourne, Australia. At each hospital, the Human Ethics Committee approved the research project and informed consent was obtained from each patient. Endometrial carcinoma tissue, taken at hysterectomy, was fixed in formalin overnight and processed to paraffin wax. Some tissue was also frozen in OCT compound (Tissue-Tek, Sakura Finetek Inc., Torrance, CA) for in situ zymography.

All tissues were examined by a specialist gynecologic pathologist. The presence of endometrial carcinoma was confirmed in all patients and tumors were graded histologically according to the International Federation of Gynecology and Obstetrics. In this system Grade 1 designates a well differentiated tumor with < 5% solid growth pattern, Grade 2 a moderately differentiated tumor between 5 and 50% solid growth, and Grade 3 a poorly differentiated tumor with > 50% solid tumor of nonmorular pattern. The presence of vascular/lymphatic invasion was noted and the depth of myometrial invasion was classified as either: no invasion, < 50% myometrial invasion, or > 50% myometrial invasion.

Immunohistochemistry and In Situ Hybridization

Formalin-fixed tissues were embedded in paraffin wax and cut at 5 μM thickness onto 3-aminopropyltriethoxy silane (AAS, Sigma Chemical Co., St Louis, MO) coated slides. Sections were deparaffinized in histosol and rehydrated in a graded series of ethanol. Endogenous peroxidase activity was blocked via incubation in 3% H2O2 in Tris-buffered saline (TBS) for 10 minutes at room temperature. Following a blocking step for non-specific staining, and primary and secondary antibody incubations, sections were washed with TBS (four times at 5 minutes each) and signals amplified using the Strept-ABC system (Dako, Glostrup, Denmark) for 30 minutes at room temperature. Detection was via the liquid DAB-plus substrate kit (Zymed, San Francisco, CA) for 5-7 minutes. In the case of MMP-9, the alternative New Fuchsin development kit (Dako) was utilized for 30 minutes at room temperature. Sections were then counterstained in Harris hematoxylin diluted 1:10 for 3 minutes, dehydrated in a series of ethanol, and mounted from histosol with DPX. For every tissue, a second section on the same slide was used as a negative control. A single block of normal human endometrium was used as a positive control, and one section from this block was included in each immunostaining run to provide quality control.

Detection of MMP-2, MMP-9, and MT1-MMP protein

Non-specific staining was blocked using normal goat serum diluted 1:10 with TBS (MMP-9) or normal rabbit serum diluted 1:20 (MMP-2). The following mouse monoclonal antibodies (Oncogene Research Products, Cambridge, MA) were used: anti-human MMP-2 (clone 42-5D11), anti-human MMP-9 (clone 1B-7013), and anti-human MT1-MMP (clone 114-6G6). Primary antibody was applied at final concentrations of 2 μg/mL (MMP-9), 0.5 μg/mL (MMP-2), and 2 μg/mL (MT1-MMP) diluted in 10% normal human serum/TBS. Mouse immunoglobulin (IgG, Oncogene Research Products) was used as a negative control, diluted to the same protein concentrations as the primary antibodies. Sections were incubated overnight in a humid chamber at 4 °C. Anti-mouse IgG (Oncogene Research Products) was used as a secondary antibody, applied at 1:50 in TBS and incubated at room temperature for 30 minutes.

Detection of MMP-9 and MMP-2 mRNA

Sense and antisense digoxygenin (DIG) labeled RNA probes against a 340bp fragment of human MMP-2 cDNA (courtesy of D. Edwards, Ph.D., Norwich, England) and a 255bp fragment of human MMP-9 cDNA (courtesy H. Nagase, Ph.D., London, England) were generated using the DIG RNA Labeling Kit (Roche, Castle Hill, Australia). Plasmid DNA was removed using the DNA-free kit (Ambion Inc, Austin, TX). Labeled RNA was quantified and stored at −80 °C until further use.

Formalin-fixed tissues were cut at 5 μM onto AAS coated slides. Sections were deparaffinized in histosol and rehydrated. Cells were made permeable with proteinase K (12 μg/mL for MMP-9 and 24 μg/mL for MMP-2) at 37 °C for 25 minutes and then fixed in 4% paraformaldehyde for 30 minutes at 4 °C. Following dehydration in a graded series of ethanol, sections were hybridized overnight in a humid chamber at 42 °C with 10 ng of DIG labeled RNA diluted in hybridization buffer consisting of × 2 sodium chloride sodium citrate (SSC), 10% dextran sulphate, 50% deionized formamide, 2% sheared salmon sperm DNA, and 0.02% sodium dodecyl sulfate (SDS). After hybridization, tissue sections were washed in increasingly stringent solutions of SSC (× 2 to × 0.1 SSC) containing 20% deionized formamide at 45 °C in a shaking waterbath. Non-specific binding was abolished by treatment with 20 μg/mL RNAse for 30 minutes at 37 °C. Bound probe was subsequently visualized using an anti-DIG antibody (Roche) conjugated to alkaline phosphatase diluted in blocking serum (10% fetal calf serum, 10% normal sheep serum, 0.1% Triton-X-100), with Nitro Blue tetrazolium/S-bromo-Δ-chloro-3-indolyl phosphatase (Dako) as a chromogen to produce a blue-purple stain. Sections were mounted with glycerol gelatin (Sigma Chemical Co.). Sense probes were applied at the same concentration as the antisense probes and treated identically to act as negative controls. Premenstrual normal human endometrium was used as a positive control and was included in every staining run to provide quality control.

Detection of TIMP-1, -2 and -3 proteins

Non-specific binding was blocked using 10% normal human serum and 10% normal rabbit serum or normal sheep serum in TBS/0.1% Tween 20 (pH 7.6) for 30 minutes at room temperature. The sheep anti-human TIMP-1 antibody (courtesy H. Nagase) was used at a dilution of 1:1000. Rabbit polyclonal anti-TIMP-2 and TIMP-3 (Triple Point Biologics, Forest Grove, OR) were used at 0.5 μg/mL and 0.25 μg/mL respectively. Anti-TIMP-1 was diluted in 10% normal sheep serum/TBS, while anti-TIMP-2 and anti-TIMP-3 were diluted in 10% normal rabbit serum/TBS. Either normal sheep IgG (Serotec, Oxford, UK) or normal rabbit IgG (Dako) was used as a negative control, diluted to the same concentration as the primary antibody. Incubation was for one hour at room temperature. Biotinylated secondary antibodies (Vector Laboratories, Burlingame, CA) were rabbit anti-sheep IgG and goat anti-rabbit IgG; incubation was for 30 minutes, again at room temperature.

Semi-quantitative analysis of immunohistochemistry

Positive immunostaining was scored semi-quantitatively by two independent observers. An intensity score for each cellular compartment (stroma and epithelial tumor) was assigned from 0 (no staining) to 3 (maximal staining intensity). The percentage of positive cells in each cellular compartment was also specified. The intensity score was multiplied by the percentage score in each case to give a total staining score of between 0 (minimum) to 300 (maximum). Results are presented as mean ± standard error of the mean. Statistical analysis was performed using the one-way ANOVA and Tukey's post hoc test, with P < 0.05 considered significant.

In situ zymography

This technique utilizes a specific substrate which fluoresces only when cleaved by MMP-2 or MMP-9 and thus identifies the location of active forms of these enzymes in a tissue. Substrate (DQ-collagen, Molecular Probes, OR) was dissolved to a final concentration of 25 μg/mL in a mixture of 2% gelatin and 2% sucrose in phosphate-buffered saline with 0.02% sodium azide. Frozen sections (7 μM) were cut directly onto slides overlaid with 100 μL of substrate solution and incubated at 37 °C overnight in a darkened, humid chamber. Slides were viewed under fluorescence using a fluoroscein isothyocyanate (FITC) filter and microscope. Control slides were coated with substrate, to which the MMP inhibitor 1,10-phenanthroline had been added to a final concentration of 10 nM.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Clinicopathologic Features of Patients

There were 29 cancer patients, with an age range of 45-88 years, with the mean age being 64 years (standard deviation = 10.8). Ten biopsies were collected from each of histologic Grades 1 and 2 carcinomas, and nine biopsies were collected from Grade 3 carcinomas. All patients were diagnosed with endometrioid adenocarcinoma tumors, except for two patients, one of whom also had squamous differentiation, and another who was diagnosed with clear cell carcinoma of the endometrium. Myometrial invasion was present in 78% of patients; of these, 52% had invasion to less than 50% of the myometrium, and 26% had invasion to 50% or more of the myometrium. The presence of vascular/lymphatic invasion, as assessed by tumor histopathology, was apparent in 33% of patients.

MMP-2, MMP-9, and MT1-MMP

Matrix metalloproteinase-2 and MMP-9 immunohistochemic staining was present in all cancer tissues examined. In all cases, the gelatinases were predominantly localized to epithelial cells of tumor origin (Fig. 1). Protein was also evident in the stromal compartment to varying degrees, although staining intensity was overall stronger in stromal cells close to epithelial tumor cells. Overall, the intensity of MMP-9 and MT1-MMP immunoreactivity in stromal cells was less than that for MMP-2. Membrane type 1-MMP protein was detected in 20 out of 29 tissues examined, and, in most cases, its localization paralleled that of MMP-2 (Fig. 1). The gelatinases were also expressed by endothelial cells of peritumor and intratumor microvessels. Tissue leukocytes stained positive for the gelatinases as well as MT1-MMP (Fig. 1). Matrix metaloproteinase-9 staining was particularly intense in clusters of macrophages.

thumbnail image

Figure 1. Immunohistochemic and in situ hybridization detection of matrix metalloproteinase (MMP)-2, MMP-9, and membrane type 1 (MT1)-MMP in endometrial carcinoma tissue. A) Matrix metalloproteinase-9 protein immunolocalized to abnormal uterine glands (arrowed) in Grade 1 carcinoma. Stromal staining (S) is also evident. B) Matrix metalloproteinase-9 predominantly immunolocalized to tumor tissue (T) in Grade 2 carcinomas. Variable stromal staining is evident particularly in areas adjacent to tumor tissue. C) Matrix metalloproteinase-9 immunolocalized to tumor tissue in Grade 3 carcinoma. D) Matrix metalloproteinase-2 immunolocalized to tumor tissue and to the stromal compartment to varying degrees in Grade 2 carcinoma. E) Membrane type 1-MMP immunolocalization predominantly paralleled that of MMP-2. F) Negative control. G, H, I) Tissue leukocytes in endometrial carcinoma tissue are positive for MMP-2, MT1-MMP, and MMP-9 protein, respectively. (× 40). J) Matrix metalloproteinase-2 mRNA localized to abnormal endometrial glands (arrowed) in Grade 1 carcinoma using in situ hybridization. K) Matrix metalloproteinase-9 mRNA localized to tumor tissue in Grade 2 carcinoma. Variable stromal localization is also evident. L) Negative control. All figures taken at × 10 magnification unless otherwise stated.

Download figure to PowerPoint

Semiquantitative analysis revealed that the staining scores for all three MMP proteins were significantly higher in epithelial tumor cells in comparison to stromal cells (Fig. 2). Increases in MMP-9 and MMP-2 but not MT1-MMP staining scores in tumor epithelial cells were evident in the transition from histologic Grade 1 to histologic Grades 2 and 3 (Fig. 2 and Table 1). Matrix metalloproteinase-9 and MT1-MMP staining scores in tumor cells were significantly associated with the presence of myometrial invasion and vascular/lymphatic invasion, while MMP-2 did not correlate with these parameters (Table 1). There were no variations in stromal staining in relation to histologic grades, depth of myometrial invasion, or presence of vascular/lymphatic invasion.

thumbnail image

Figure 2. Semiquantitative analysis of matrix metalloproteinase (MMP)-9, MMP-2, and membrane type 1 (MT1)-MMP in both tumor epithelial cells and stromal cells in endometrial carcinoma tissues. A) Matrix metalloproteinase-9, MMP-2, and MT1-MMP proteins are present in significantly higher amounts in tumor epithelial cells in comparison to stromal cells. B) Matrix metalloproteinase-9 and MMP-2, but not MT1-MMP, are correlated with higher tumor histologic grade (a = P < 0.05).

Download figure to PowerPoint

Table 1. Correlation of Immunohistochemic Staining Scores in Tumor Epithelial Cells with Clinicopathologic Features
Clinicopathologic featureComparison of immunohistochemic staining score (P value)a
  • MMP: matrix metalloproteinase; FIGO: International Federation of Gynecology and Obstetrics; NS: not significant.

  • a

    Obtained by ANOVAR and Tukey's post hoc test.

  • Matrix metalloproteinase-9 and MMP-2 are associated with increasing tumor histologic grade. MMP-9 and MT1-MMP, but not MMP-2 are correlated with the presence of myometrial invasion and vascular/lymphatic invasion.

Histologic grade (FIGO)
 Grade 1 vs. Grade 2P < 0.0001P < 0.05NS
 Grade 1 vs. Grade 3P < 0.0001P < 0.05NS
 Grade 2 vs. Grade 3NSNSNS
Depth of myometrial invasion (mi)
 no invasion vs. < 50% miP < 0.0001NSP < 0.05
 no invasion vs. > 50% miP < 0.0001NSP < 0.05
 < 50% mi vs. > 50% miNSNSNS
Vascular/lymphatic invasion
 no invasion vs. vascular invasionP < 0.0001NSP < 0.05

Matrix metalloproteinase-9 and MMP-2 mRNA were detected in all 29 tissues examined. Matrix metalloproteinase-2 mRNA was present in epithelial tumor cells and stromal cells (Fig. 1). Matrix metalloproteinase-2 mRNA localization did not always correlate with protein localization, suggesting cellular uptake or ECM-binding of secreted proteins. Matrix metalloproteinase-9 mRNA was predominantly present in tumor cells (Fig. 1) however, some stromal localization was also observed, particularly surrounding tumor nests.

TIMP-1, TIMP-2, and TIMP-3

The TIMPs were immunohistochemically detected in three tissues from each grade of carcinoma. Tissue inhibitor of metalloproteinases-1 was predominantly immunolocalized to the stromal compartment in Grade 1 carcinomas (not shown). However, variable tumor cell localization for TIMP-1 was observed in Grades 2 and 3 carcinomas (Fig. 3). The reverse was observed for TIMPs –2 and –3. Tumor cells from all histologic grades stained intensely for TIMP-2 (Fig. 3) and TIMP-3 (Fig. 3) proteins, while variable stromal staining was observed.

thumbnail image

Figure 3. Immunohistochemic detection of tissue inhibitors of metalloproteinases 1-3 and in situ zymography on endometrial carcinoma tissue. A) Tissue inhibitor of metalloproteinases-1 immunolocalized to stroma and tumor compartments in Grade 3 carcinoma (T: tumor tissue; S: stromal staining). B) Tissue inhibitor of metalloproteinases-2 immunolocalized to epithelial tumor cells in Grade 2 carcinoma. C) Tissue inhibitor of metalloproteinases-3 immunolocalized to epithelial tumor cells and stromal cells in Grade 3 carcinoma. D) Negative control. E, F) Fluorescence detected from in situ zymography is indicative of gelatinase activity in Grades 1 and 2 carcinomas, respectively. G) Gelatinase activity in single cells of Grade 2 carcinoma (× 40). Addition of an MMP inhibitor, 1, 10 phenanthroline, abolished fluorescence (not shown). All figures taken at ×10 magnification unless otherwise stated.

Download figure to PowerPoint

Active Enzymes in Endometrial Carcinoma Tissue

Fluorescence, representing cleavage of the gelatin substrate and the presence of active forms of gelatinases, was observed in tissue sections from Grades 1 and 2 carcinomas (Fig. 3), suggesting that a proportion of the gelatinases detected by immunohistochemistry are in active forms. Due to the lack of sufficient tissue, we were unable to perform this technique on Grade 3 endometrial carcinoma tissue. Fluorescence was observed close to the cellular surface and in association with tumor epithelial cells.


  1. Top of page
  2. Abstract
  6. Acknowledgements

The current study has shown, firstly, that increasing MMP-9 and MMP-2 protein production in endometrial tumor cells is associated with advanced histologic grade and, secondly, that MMP-9 and MT1-MMP, but not MMP-2, are associated with the presence of myometrial invasion and with vascular/lymphatic invasion. A key determinant of invasive behavior is MMP activation. In situ zymography has shown, to our knowledge for the first time, that a proportion of MMPs detected in endometrial carcinoma tissues are in their active forms and that this activity is associated with tumor cells.

These data are overall in accord with other studies investigating MMPs in endometrial carcinoma. Simultaneous angiogenesis and over-expression of MMP-2 and MMP-9 mRNA have been shown during endometrial carcinoma progression. In particular, MMP-2 and MMP-9 detected by in situ hybridization were significantly over-expressed from histologic Grade 1 to Grade 3 carcinomas and with increasing depth of myometrial invasion.22 Inoue et al.23 showed that MMP-9 protein was associated with histologic Grade 3, vessel invasion, and lymph node metastasis. No previous studies have examined MT1-MMP protein in endometrial carcinomas.

Using immunohistochemistry, we have shown that the gelatinases are predominantly localized to tumor epithelial cells. However, some stromal staining was also observed at both the mRNA and protein levels, and this was identified in stromal fibroblasts, endothelial cells, and leukocytes. Matrix metalloproteinase-9 protein was particularly associated with tissue macrophages, in agreement with the study by Iurlaro et al.,22 who localized MMP-2 and MMP-9 mRNA to these stromal compartments. Therefore the regulation of tumor progression in endometrial carcinoma may involve several cell types, including both malignant and non-malignant cells in the tumor stroma.

In the current study, MT1-MMP protein was mainly localized to neoplastic epithelial cells, as also reported in lung and gastric carincomas.24, 25 The current data support previous studies showing that endometrial tumor cells are the main producers of MT1-MMP mRNA in endometrial carcinoma26 and that high MT1-MMP gene expression in primary endometrial carcinoma tissue correlates with MT1-MMP expression in metastatic lesions.27 We observed a correlation of MT1-MMP protein expression with both increasing depth of myometrial invasion and the presence of vascular/lymphatic invasion. Collectively, these studies suggest important roles for MT1-MMP in endometrial carcinoma progression. One likely role is the activation of proMMP-2, supported by the observation of MT1-MMP, MMP-2, and TIMP-2 co-localization. The presence of gelatinase activity at the cellular surface, as shown by in situ zymography, further supports this model. However, MT1-MMP also functions in degradation of ECM components.28, 29

The pattern of MMP expression is usually similar in different cases of the same type of cancer, while it varies between different types of cancer. For example, in neuroblastomas, MMP-9 is expressed by stromal cells,30 in tumors of the brain, it is expressed by malignant epithelium.31 In prostate carcinoma, MMP-2 is expressed by malignant epithelial cells only,32 while in breast carcinoma, MMP-2 is generally expressed ubiquitously by both epithelial and stromal cells.33 Tissue-specific differences in MMP expression may reflect local microenvironments, possibly in response to matrix components or from signals derived from an inflammatory reaction which often accompanies the establishment of a malignant tumor. In the case of endometrial carcinoma, the predominance of MMP-2, MMP-9, and MT1-MMP expression by epithelial cells may reflect the fact that these proteinases are expressed by this cell type at various stages of the normal menstrual cycle.34, 35

In general, MMPs are secreted in pro-enzyme form, which are then activated following limited proteolysis. However, over-production of pro-enzyme and generation of proteolytic activity are not equivalent. Therefore, the detection of active MMPs is critical. Active gelatinases have recently been detected in carcinoma cell nests of oral squamous cell carcinoma36 and in ovarian neoplasms37 using in situ zymography. Using this technique, we report the presence of active gelatinases in endometrial carcinoma. The current results show significant gelatinase activity even at the early stages of tumor progression (i.e., histologic Grades 1 and 2). It is likely that such early gelatinase activity results in local ECM breakdown in the endometrium, which is necessary for local tumor invasion. Significant gelatinase activity was also observed in malignant endometrium from tumors which had penetrated the myometrium and invaded the lymph nodes.

The regulation of MMP expression can be linked to a number of inducing agents, including growth factors, cytokines, oncogenes, and ECM-derived signals.38 As well as ECM degradation, MMPs process and release a variety of molecules, which are regulators of vascular growth or function, including fibroblast growth factor receptor type I,39 tumor necrosis factor-α,40 and heparin-binding-epidermal growth factor.41 Thus the metalloproteinases are not limited to simply degrading structural proteins that surround the cell, but have a more generalized role in the interactions of cells with their matrix environment, affecting basic cellular processes such as differentiation, proliferation and apoptosis.

Matrix metalloproteinases are not only regulated at the level of gene transcription and activation but are also regulated locally by the TIMPs. An imbalance between MMPs and TIMPs is linked to the degradation of the extracellular matrix in tumor progression. Using immunohistochemistry we localized TIMP-1 to the stromal compartment in all grades of endometrial carcinoma, with variable tumor epithelial cell localization in Grades 2 and 3 carcinomas only. Tumor cells from all histologic grades stained intensely for TIMP-2 and –3, with varying degrees of stromal staining also observed. These data partially support studies of Määtta et al.,26 who found TIMPs-2 and –3 mRNA in tumor epithelial cells, with the latter also being expressed by stromal and endothelial cells. They also found TIMP-1 mRNA localized to stromal and endothelial cells with no clear epithelial cell expression.25 The TIMPs are expressed in most tissues, including the non-malignant endometrium, where high levels of TIMPs 1-3 presumably maintain tissue integrity and ECM homeostasis.42 However, TIMPs have additional functions in cellular proliferation43, 44 and cell survival pathways,45 with TIMPs-1 and –2 having specific roles in controlling the activations of proMMP-9 and proMMP-2 respectively.46–48

In conclusion, MMPs, along with TIMPs, are likely to play a complex role in the progression of endometrial carcinoma. Semiquantitative analysis has shown correlations between MMP-2, MMP-9, and MT1-MMP immunoreactive protein and clinicopathologic features of endometrial carcinoma, such as the presence of myometrial invasion, vascular/lymphatic invasion, and increasing histologic grade. Most importantly, active gelatinases are present, which would facilitate tumor invasion and subsequent metastasis. In order to fully understand the complex roles that MMPs play in endometrial carcinoma, further questions need to be addressed, including the precise mechanisms of MMP regulation/activation and the interactions of MMPs/TIMPs with other molecules.


  1. Top of page
  2. Abstract
  6. Acknowledgements

The authors acknowledge the technical assistance of Franca Pansino and thank Helen Sell and the gynecologic oncology staff at Monash Medical Centre and the Royal Women's Hospital for assistance in obtaining tissue samples.


  1. Top of page
  2. Abstract
  6. Acknowledgements
  • 1
    North American Association of Central Cancer Registries., 1999.
  • 2
    Australian Institute of Health & Welfare., 1999.
  • 3
    Woessner JF. Matrix metalloproteinases and their inhibitors in connective tissue remodeling. FASEB J 1991; 5: 21452154.
  • 4
    Edwards DR, Murphy G. Cancer. Proteases—invasion and more. Nature 1998;394: 527528.
  • 5
    Shapiro SD. Matrix metalloproteinase degradation of extracellular matrix: biological consequences. Curr Opin Cell Biol 1998; 10: 602608.
  • 6
    Birkedal-Hansen H, Moore WGI, Bodden MK, et al. Matrix metalloproteinases: A Review. Crit Rev Oral Biol Med 1993; 4: 197250.
  • 7
    Nagase H, Okada Y. Proteinases and Matrix Degradation. In: KelleyWN, HarrisED, RuddyS, SledgeWB, editors. Textbook of rheumatology. Philadelphia: W.B Saunders, 1997: 32341.
  • 8
    Greene J, Wang M, Liu YE, Raymond LA, Rosen C, Shi YE. Molecular cloning and characterization of human tissue inhibitor of metalloproteinase 4. J Biol Chem 1996; 271: 3037530380.
  • 9
    Apte SS, Olsen BR, Murphy G. The gene structure of tissue inhibitor of metalloproteinases (timp)-3 and its inhibitory activities define the distinct timp gene family. J Biol Chem 1995; 270: 1431314318.
  • 10
    Ray JM, Stetler-Stevenson WG. The role of matrix metalloproteases and their inhibitors in tumour invasion, metastasis and angiogenesis. Eur Respir J 1994; 7: 20622072.
  • 11
    Butler GS, Butler MJ, Atkinson SJ, et al. The TIMP2 membrane type 1 metalloproteinase “receptor” regulates the concentration and efficient activation of progelatinase A. J Biol Chem 1998; 273: 871880.
  • 12
    Leco KJ, Khokha R, Pavloff N, Hawkes SP, Edwards DR. Tissue inhibitor of metalloproteinases-3 (TIMP-3) is an extracellular matrix-associated protein with a distinctive pattern of expression in mouse cells and tissues. J Biol Chem 1994; 269: 93529360.
  • 13
    Ray JM, Stetler-Stevenson WG. The role of matrix metalloproteases and their inhibitors in tumour invasion, metastasis and angiogenesis. Eur Respir J 1994; 7: 20622072.
  • 14
    Nomura H, Fujimoto N, Seiki M, Mai M, Okada, Y. Enhanced production of matrix metalloproteinases and activation of matrix metalloproteinase 2 (gelatinase A) in human gastric carcinomas. Int J Cancer 1996;69: 916.
  • 15
    Maeta H, Ohgi S, Terada T. Protein expression of matrix metalloproteinases 2 and 9 and tissue inhibitors of metalloproteinase 1 and 2 in papillary thyroid carcinomas. Virchows Arch 2001; 438: 121128.
  • 16
    Jones JL, Glynn P, Walker RA. Expression of MMP-2 and MMP-9, their inhibitors and the activator MT1-MMP in primary breast carcinomas. J Pathol 1999; 189: 161168.
  • 17
    Sakata K, Shigemasa K, Nagai N, Ohama K. Expression of matrix metalloproteinases (MMP-2, MMP-9, MT1-MMP) and their inhibitors (TIMP-1, TIMP-2) in common epithelial tumors of the ovary. Int J Oncol 2000; 17: 673681.
  • 18
    Ueno H, Nakamura H, Inoue M, et al. Expression and tissue localization of membrane-types 1, 2, and 3 matrix metalloproteinases in human invasive breast carcinomas. Cancer Res 1997; 57: 20552060.
  • 19
    Nakamura H, Ueno H, Yamashita K, et al. Enhanced production and activation of progelatinase A mediated by membrane-type I matrix metalloproteinase in human papillary thyroid carcinomas. Cancer Res 1999; 59: 467– 473.
  • 20
    Määtta M, Soini Y, Liakka A, Autio-Harmainen H. Differential expression of matrix metalloproteinase (MMP)-2, MMP-9, and membrane type 1-MMP in hepatocellular and pancreatic adenocarcinoma: implications for tumor progression and clinical prognosis. Clin Cancer Res 2000; 6: 27262734.
  • 21
    Soini Y, Alarakkola E, Autio-Harmainen H. Expression of messenger RNAs for metalloproteinases 2 and 9, type IV collagen, and laminin in nonneoplastic and neoplastic endometrium. Hum Pathol 1997; 28: 220226.
  • 22
    Iurlaro M, Loverro G, Vacca A, et al. Angiogenesis extent and expression of matrix metalloproteinase-2 and -9 correlate with upgrading and myometrial invasion in endometrial carcinoma. Eur J Clin Invest 1999; 29: 793801.
    Direct Link:
  • 23
    Inoue Y, Abe K, Obata K, et al. Immunohistochemical studies on matrix metalloproteinase-9 (MMP-9) and type IV collagen in endometrial carcinoma. J Obstet Gynaecol Res 1997; 23: 139145.
  • 24
    Sato H, Takino T, Okada Y, et al. A matrix metalloprotease expressed on the surface of invasive tumour cells. Nature 1994; 370: 6165.
  • 25
    Nomura H, Sato H, Seiki M, Mai M, Okada Y. Expression of membrane-type matrix metalloproteinase in human gastric carcinomas. Cancer Res 1995; 55: 32633266.
  • 26
    Määtta, M, Soini Y, Liakka A, Autio-Harmainen H. Localization of MT1-MMP, TIMP-1, TIMP-2, and TIMP-3 messenger RNA in normal, hyperplastic, and neoplastic endometrium. Am J Clin Pathol 2000;114: 402411.
  • 27
    Sakamoto H, Kinoshita K, Nakayama Y, et al. Gene expression in primary, metastatic and recurrent lesions of endometrial cancer. Gynecol Obstet Invest 1999; 48: 215220.
  • 28
    Ohuchi E, Imai K, Fujii Y, Sato H, Seiki M, Okada A. Membrane type 1 matrix metalloproteinase digests interstitial collagens and other extracellular matrix macromolecules. J Biol Chem 1997; 272: 24462451.
  • 29
    d'Ortho M, Stanton H, Butler M, Atkinson SJ, Murphy G, Hembry RM. MT1-MMP on the cell surface causes focal degradation of gelatin films. FEBS Letters 1998; 421: 159164.
  • 30
    Sugiura Y, Shimada H, Seeger RC, Laug WE, DeClerck YA. Matrix metalloproteinases-2 and -9 are expressed in human neuroblastoma: contribution of stromal cells to their production and correlation with metastasis. Cancer Res 1998; 58: 22092216.
  • 31
    Rao JS, Yamamoto M, Mohaman S, et al. Expression and localization of 92 kDa type IV collagenase/gelatinase B (MMP-9) in human gliomas. Clin Exp Metastasis 1996; 14: 1218.
  • 32
    Still K, Robson CN, Autzen P, Robinson MC, Hamdy FC. Localization and quantification of mRNA for matrix metalloproteinase-2 (MMP-2) and tissue inhibitor of matrix metalloproteinase-2 (TIMP-2) in human benign and malignant prostatic tissue. Prostate 2000; 42: 1825.
  • 33
    Hoyhtya M, Fridman R, Komarek D, et al. Immunohistochemical localization of matrix metalloproteinase 2 and its specific inhibitor TIMP-2 in neoplastic tissues with monoclonal antibodies. Int J Cancer 1994; 56: 500505.
  • 34
    Jeziorska M, Nagase H, Salamonsen LA, Woolley DE. Immunolocalisation of the matrix metalloproteinases gelatinase B and stromelysin 1 in human endometrium throughout the menstrual cycle. J Reprod Fertil 1996; 107: 4351.
  • 35
    Zhang J, Hampton AL, Nie G, Salamonsen LA. Progesterone inhibits activation of latent matrix metalloproteinase (MMP)-2 by membrane-type 1 MMP: enzymes coordinately expressed in human endometrium. Biol Reprod 2000; 62: 8594.
  • 36
    Shimada T, Nakamura H, Yamashita K, et al. Enhanced production and activation of progelatinase A mediated by membrane-type I matrix metalloproteinase in human oral squamous cell carcinomas: implications for lymph node metastasis. Clin Exp Metastasis 2000; 18: 179188.
  • 37
    Furuya M, Ishikura H, Nemori R, Shibata M, Fujimoto S, Yoshiki T. Clarification of the active gelatinolytic sites in human ovarian neoplasms using in situ zymography. Hum Pathol 2001; 32: 163168.
  • 38
    Stetler-Stevenson WG, Aznavoorian S, Liotta LA. Tumor cell interactions with the extracellular matrix during invasion and metastasis. Annu Rev Cell Biol 1993; 9: 541573.
  • 39
    Levi E, Fridman R, Miao HQ, Ma YS, Yayon A, Vlodavsky I. Matrix metalloproteinase 2 releases active soluble ectodomain of fibroblast growth factor receptor 1. Proc Natl Acad Sci USA 1996; 93: 70697074.
  • 40
    Gearing AJ, Beckett P, Christodoulou M, et al. Processing of tumour necrosis factor-α precursor by metalloproteinases. Nature 1994; 370: 555557.
  • 41
    Suzuki M, Raab G, Moses MA, Fernandez CA, Klagsbrun M. Matrix metalloproteinase-3 releases active heparin-binding EGF-like growth factor by cleavage at a specific juxtamembrane site. J Biol Chem 1997; 272: 3173031737.
  • 42
    Zhang J, Salamonsen LA. Tissue inhibitor of metalloproteinases (TIMP)-1, -2 and -3 in human endometrium during the menstrual cycle. Mol Hum Reprod 1997; 3: 735741.
  • 43
    Hayakawa T, Yamashita K, Tanzawa K, Uchijima E, Iwata K. Growth-promoting activity of tissue inhibitor of metalloproteinase-1 (TIMP-1) for a wide range of cells. A possible new growth factor in serum. FEBS Lett 1992;298: 2932.
  • 44
    Hayakawa T, Yamashita K, Ohuchi E, Shinagawa A. Cell growth-promoting activity of tissue inhibitor of metalloproteinase-2 (TIMP-2). J Cell Sci 1994; 107: 23732379.
  • 45
    Li G, Fridman R, Kim HR. Tissue inhibitor of metalloproteinase-1 inhibits apoptosis of human breast epithelial cells. Cancer Res 1999; 59: 62676275.
  • 46
    Goldberg GI, Strongin A, Collier IE, Genrich LT, Marmer BL. Interaction of 92kDa type IV collagenase with the tissue inhibitor of metalloproteinases prevents dimerization, complex formation with interstitial collagenases and activation of the proenzyme with stromelysin. J Biol Chem 1992; 267: 45834591.
  • 47
    Murphy G, Willenbrock F, Ward RV, Cockett MI, Eaton D, Docherty AJ. The C-terminal domain of 72kDa gelatinase A is not required for catalysis, but is essential for membrane activation and modulates interactions with tissue inhibitors of metalloproteinases. Biochem J 1992; 283: 637641.
  • 48
    Kleiner DE Jr., Tuuttila A, Tryggvason K, Stetler-Stevenson WG. Stability analysis of latent and active 72-kDa type IV collagenase: the role of tissue inhibitor of metalloproteinases-2 (TIMP-2). Biochemistry 1993;32: 15831592.