Clinical & Experimental Metastasis 20: 621–631, 2003. © 2003 Kluwer Academic Publishers. Printed in the Netherlands.
621
Matrix metalloproteinases (MMP), EMMPRIN (extracellular matrix metalloproteinase inducer) and mitogen-activated protein kinases (MAPK): Co-expression in metastatic serous ovarian carcinoma Ben Davidson1, Vered Givant-Horwitz2, Philip Lazarovici2,5 , Björn Risberg1 , Jahn M. Nesland1 , Claes G. Trope3 , Erik Schaefer4 & Reuven Reich2,5 1 Department
of Pathology, The Norwegian Radium Hospital, University of Oslo, Montebello Oslo, Norway; 2 Department of Pharmacology and Experimental Therapeutics, School of Pharmacy, Faculty of Medicine, The Hebrew University of Jerusalem, Jerusalem , Israel; 3 Signal Transduction Research and Development, QCB – a division of BioSource, Hopkinton, Massachusetts, USA; 4 Department of Gynecologic Oncology, The Norwegian Radium Hospital, University of Oslo, Montebello, Oslo, Norway; 5 Affiliated with the David R. Bloom Center for Pharmacy at the Hebrew University
Received 24 January 2003; accepted in revised form 9 May 2003
Key words: effusions, EMMPRIN, matrix metalloproteinases, mitogen-activated protein kinases, ovarian carcinoma
Abstract Activation or suppression of intracellular signaling via the mitogen-activated protein kinase (MAPK) family has been linked to expression of matrix metalloproteinases (MMP) in experimental models, but this association has not been demonstrated in clinical material. The objective of this study was to investigate the possible association between expression and activity of MMP, expression of the MMP inducer EMMPRIN, and the expression (level) and phosphorylation status (activity) of the extracellular-regulated kinase (ERK), c-Jun amino-terminal kinase (JNK) and high osmolarity glycerol response kinase (p38) in effusions from patients diagnosed with serous ovarian carcinoma. MAPK level and activity were studied in 55 effusions using immunoblotting. MMP-1, MMP-2, MMP-9 and EMMPRIN expression was studied using immunocytochemistry (ICC) and mRNA in situ hybridization (ISH). The gelatinolytic activity of MMP-2 and MMP-9 was measured by zymography. ERK and phospho-ERK (p-ERK) were detected in 54/55 (98%) and 50/55 (91%) specimens, respectively. JNK and p-JNK were detected in 53/55 (96%) and 38/55 (69%) specimens, respectively. p38 was expressed in 54/55 (98%) specimens, and its phosphorylated form was found in 51/55 (92%). MMP-2 mRNA expression (P = 0.048), protein expression (P = 0.046) and gelatinolytic activity (P = 0.039) correlated with ERK phosphorylative activity. MMP-2 activity also correlated with p38 activity (P = 0.017). MMP-9 protein expression correlated with phosphorylation of p38 (P = 0.046), but enzyme activity showed inverse relationship with both p-ERK (P = 0.05) and p-p38 (P = 0.033) expression. EMMPRIN expression correlated with MMP-1 (P < 0.001), MMP-2 (P = 0.042) and MMP-9 (P = 0.029) expression, as well as with ERK activity (P = 0.001). Our results present the first evidence of a possible link between MAPK signaling and MMP expression and activity in vivo. These data may expand our understanding regarding the mechanisms by which MMP synthesis is regulated in effusions and possibly affect treatment strategies for this form of malignancy.
Introduction The evolution and progression of cancer is a multi-step process, during which malignant cells develop pathophysiological transformations affecting major aspects of cellular function, such as adhesion, proliferation, differentiation and apoptosis [1]. Many of the characteristics of malignant cells, including the synthesis of metastasis-associated molecules (e.g., proteolytic enzymes, angiogenic factors), refractoriness to apoptotic signals and immortality originate from a variety of extracellular signals, including stress, growth factors, transforming factors, cytokines and mitogens, and Correspondence to: Dr Ben Davidson, Department of Pathology, The Norwegian Radium Hospital, Montebello N-0310 Oslo, Norway. Tel: +47-22934871; Fax: +47-22508554; E-mail:
[email protected]
are mediated via membrane receptors, such as tyrosine kinase receptors [2, 3]. These in turn relay their messages to the nucleus using a complex network of intracellular signaling pathways. The mitogen-activated protein kinase (MAPK) intracellular signaling mode is a four-kinase component cassette, in which each kinase activates the following kinase substrate through a complex network, enabling the cell to maintain diversity and specificity while responding to various extracellular stimuli [2–5]. The final level consists of 12 MAPK, including extracellular-regulated kinase (ERK1-5), c-jun amino-terminal kinase (JNK1-3) and the high osmolarity glycerol response kinase in its different isoforms (p38 α,β,γ ,δ). Tyrosine and threonine phosphorylation of MAPK occurs in a specific manner by seven MAPK, MEK1,2 and
622 MKK3-7. Activation of MAPK is followed by phosphorylation of a variety of cytosolic substrates, as well as their translocation to the nucleus, where they activate a large number of transcription factors, such as AP-1, p53, Elk-1, Ets-1, c-Myc and STATs [3]. This results in a variety of biological effects, some of which are induced by several members belonging to the above three groups of MAPK. The ERK subfamily of kinases is largely activated by growth factor signals, such as those mediated by receptor tyrosine kinases [3]. The net result is growth, differentiation and proliferation [3]. The JNK and p38 subfamily of kinases is activated by a large spectrum of stress-related stimuli. These include osmotic shock, inhibition of protein synthesis and formation of oxygen radical species [6]. Signaling by p38, for example, affects gene expression, signaling via the adrenergic, arachidonate and nitric oxide pathways, apoptosis and proliferation and differentiation, and is involved in the pathology of ischemic injury, infection and wound healing [7]. In many cases, p38 and JNK are thought to largely mediate apoptotic signals, while ERK promotes the opposing effect [8]. However, overlaps are now known to exist in these functions. Pathological interference with MAPK expression and/or activity has been increasingly recognized as a critical factor of stress-induced cellular responses and diseases [9–11]. Matrix metalloproteinases (MMP), a family of zinc- and calcium-dependent enzymes, are central mediators of tumor metastasis, owing to their ability to degrade basement membrane and extracellular matrix (ECM) components [12]. We have previously found higher MMP expression in ovarian carcinoma cells in effusions as compared to primary tumors [13]. In an additional study [14], we analyzed mRNA expression of membrane-type MMP, enzymes that are involved in activation of several MMP, and found MT1-MMP and MT2-MMP expression in the majority of effusions, in correlation with MMP-2 expression [14]. The association between MAPK activity and MMP expression and activity has been the subject of several studies. Induction of MMP-1 in fibroblasts has been shown to be mediated by p38 [15–17], JNK [17] and ERK [15, 18]. MMP-1 activation through p38 originated from EMMPRIN in one study [16]. A similar association has been shown for p38 and ERK in transformed keratinocytes [19] and chondrosarcoma cells [20], and for JNK in the A549 carcinoma cell line [21]. Induction of MMP-2 and MT1-MMP by phorbol 12-myristate 13-acetate (PMA) in glioma cells has been shown to be mediated by ERK [22], while a similar effect was mediated by p38, rather than ERK, in melanoma cells [23]. MMP-9 synthesis and/or activity have been shown to be associated with one or more of the three MAPK families in head and neck [24, 25], breast [26, 27] and ovarian [28, 29] carcinomas, as well as gliomas [30]. Despite this extensive research in vitro, no data is available regarding the possible association between MMP, EMMPRIN, and MAPK in clinical specimens. To initiate this study, we took advantage of a series of phosphorylation state-specific antibodies ideally suited for studying complex patterns of phosphoregulation [31]. The present study ana-
B. Davidson et al. lyzed the association between MAPK expression and the expression of MMP-1, MMP-2, MMP-9 and EMMPRIN in 55 malignant serous effusions from ovarian carcinoma patients.
Materials and methods Effusion specimens The material consisted of 55 fresh non-fixed peritoneal and pleural effusions submitted to the Division of Cytology, Department of Pathology, The Norwegian Radium Hospital, during the period of January 1998–September 2000. Specimens were obtained pre-operatively, intra-operatively, or at disease recurrence, from 47 patients diagnosed with serous ovarian carcinoma and 2 patients diagnosed with primary peritoneal carcinoma (PPC). Effusion specimens consisted of 37 peritoneal and 18 pleural effusions. All effusion specimens, as well as relevant clinical data, were obtained from the Department of Gynecologic Oncology, Norwegian Radium Hospital. In order to preserve physiological activity, specimens submitted to our laboratory arrived within minutes after tapping and were processed immediately. Cells were suspended and frozen in RPMI + DMSO at −70 ◦ C. Smears and cellblock sections from all specimens underwent morphological evaluation by three experienced cytopathologists, and were further characterized using immunocytochemistry with broad antibody panels against epithelial and mesothelial epitopes, as previously detailed [32–34]. Specimens were in addition divided into homogenous (containing a cancer cell population of 80% or more of the entire cell population) or mixed (containing a larger than 20% population of mesothelial or inflammatory cells) based on cytology smears. Clinicopathologic data are presented in Table 1. Western blotting Frozen effusion specimens were thawed and washed twice in phosphate buffered saline (PBS). Samples were subsequently lysed in 1% NP-40, 20 mM Tris-HCl (pH 7.5), 137 mM NaCl, 0.5 mM EDTA, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, 1 µg/ml aprotinin, 2 µg/ml leupeptin, 1 mM sodium orthovanadate and 0.1% SDS. After centrifugation, the supernatant was collected and protein content was evaluated by the Bradford assay. Fifteen micrograms from each sample, under reducing conditions, were loaded into each lane and separated by electrophoresis through SDS-10% polyacrylamide gels. After electrophoresis, proteins were transferred to Immobilon transfer membranes (Millipore, Bedford, Massachusetts). Membranes were blocked in TBST (10 mM Tris-HCl (pH 8.0), 150 mM NaCl and 0.1% Tween 20) containing 5% BSA (Sigma, St. Louis, Missouri) for 2 h at room temperature. Membranes were subsequently incubated overnight at 4 ◦ C in 5% BSA in TBST containing either anti-phospho-ERK (p-ERK), anti-phospho-JNK (p-JNK) or anti-phospho-p38 (p-p38).
MAPK, MMP and EMMPRIN in ovarian carcinoma Table 1. Clinicopathologic data of the study cohort. Parameter
Number of specimens
Percent
Effusion site Peritoneal Pleural
37 18
67 33
Grade I II III
6 20 29
11 36 53
Chemotherapya Yes No
26 29
47 53
Purityb Yes No
29 26
53 47
FIGO stage III IV
28 27
51 49
Agec 31-40 41-50 51-60 61-70 71-80
2 5 17 14 11
4 10 35 29 22
a Prior to sampling. b Specimens with a cancer cell population exceeding 80%
of cells were considered as pure. c For a total of 49 patients.
Activation of MAPK requires that these enzymes will be dually phosphorylated by the respective MEK on both the Thr and Tyr residues in the pTEpY (ERK), pTPpY (JNK) and pTGpY (p38) consensus sequences within the catalytic domains [35]. The phospho-MAPK antibodies used in the present study were developed as polyclonal antibodies in rabbit against dually phosphorylated synthetic peptides encompassing the above consensus sequences (residues Thr202/185 and Tyr-204/187 of ERK1 and ERK2, residues Thr183 and Tyr-185 of JNK, Thr-180 and Tyr-182 of p38). The antibodies were purified using a repetitive adsorption step to remove antibodies that recognize the non-phosphorylated peptide, followed by positive selection-affinity purification with the dually phosphorylated peptide to select for antibodies preferentially recognizing ERK 1/2 (44/42 kDa), JNK1 (49 kDa) and JNK2 (55 kDa) or p38α, p38β and p38γ . These phospho-antibodies were rigorously tested in Western blotting and immonohistochemistry in rat and human tissues and detect as little as 250 pg of active ERK or 1ng of phosphorylated recombinant JNK or p38 enzymes at the dilution of 1:1000-2000. The antibodies were intensively investigated in a variety of cell lines resulting in very high signal-to-noise ratios of 10 to 80-fold stimulation of antibody compared to untreated controls [36–38]. After incubation, membranes were washed three times for 10 min in TBST, followed by 1 h incubation with Peroxidase-cojugated AffiniPure Goat anti-Rabbit IgG
623 (Jackson ImmunoResearch, West Grove, Philadelphia) in TBST containing 5% BSA. After 4 washes of 10 min each in TBST, membranes were developed by enhanced chemiluminescence (Pierce, Rockford, Illinois), according to manufacturer’s specifications. Membranes were then washed for 10 min in TBST and stripped twice for 30 min in 0.2 M Glycine, 0.1% SDS and 1% Tween 20 (pH 2.2), followed by three washes of 5 min each in TBST. After blocking in TBST containing 5% BSA for 2 h at room temperature, membranes were incubated overnight at 4 ◦ C in 5% BSA in TBST containing anti-ERK (BioSource, Camarillo, California), anti-JNK (Biosource), or anti-p38 (StressGen Victoria, BC, Canada). The above-described procedure was performed to visualize the signal. A375SM melanoma cells were used as control in all the gels. Quantification of blotting results Gels were photographed by the KODAK EDAS 290 system. Densitometer analysis of films was performed using computerized image analysis (NIH IMAGE 1.62, 1999 version) program. The parameters analyzed were as follows: 1. Total enzyme level (using the ‘pan’ antibody). 2. Phosphorylated enzyme activity (using the phosphodirected antibody). For statistical purposes, the expression of ERK, p38 and JNK (pan- and phospho-scores) was grouped as 0–25%, 26–100% and > 100% of A375SM cell line values, corresponding to a score of 1–3. 3. Determination of isoenzyme expression: the presence or absence of pan- and p-forms of the high, medium and low molecular weight isoforms of each kinase family was determined. Expression was determined as positive (= 1) or negative (= 0) for each isoform. Immunocytochemical analysis (ICC) ICC was performed on all 55 specimens as previously described [13], using monoclonal antibodies against MMP1 (clone 41-IE5), MMP-2 (clone 42-5D11) and MMP-9 (clone 56-2A4) (Calbiochem, La Jolla, California). Staining was evaluated as negative, weak or intense. Staining for EMMPRIN was performed using a goat polyclonal antibody directed against the N-terminal domain of the protein (Santa Cruz Biotechnology, Santa Cruz, California), as previously described [39]. Negative controls for all experiments consisted of sections that underwent a similar staining procedure, with the exclusion of primary antibody application, or that were stained with mouse myeloma protein of the same isotype as the primary antibody used. Two ovarian carcinoma biopsies in which immunoreactivity for the studied antigens was previously demonstrated were used as positive controls. Twenty-five specimens were additionally stained using an antibody against p-ERK1/2 (Cell Signaling, Beverly, Massachusetts) in order to verify the cellular localization of the activated protein. Immunocytochemistry was performed by technicians blinded to any morphological or clinical data regarding the patients and the material.
624 Oligonucleotide probes Specific antisense oligonucleotide DNA probes for mRNA transcripts of MMP-2, MMP-9 and EMMPRIN were obtained from Research Genetics (Huntsville, Alabama). Probe sequences (5 –3 ) were as follows [40, 41]: MMP-2: 5 TGGGCTACGGCGCGGCGGCGTGGC 3 MMP-9: 5 CCGGTCCACCTCGCTGGCGCTCCGGU 3 EMMPRIN: 5 CAG CGC GAA TCC CAG CAG CAC GAA C 3 A poly d(T)20 oligonucleotide (Research Genetics) was used to verify the integrity and lack of degradation of mRNA in each sample. DNA probes for MMP-2 and MMP-9 were hyperbiotinylated. Stock dilution was prepared with a resulting equal concentration for both probes. The stock dilution was diluted with probe diluent (Research Genetics) immediately before use. Specific sense oligonucleotides were used for the evaluation of non-specific activity for each probe.
B. Davidson et al. Table 2. The association between MAPK and the cellular composition of effusion specimens.a Total expression and activity MAPKb
Homogenous specimens 1 2 3
Mixed specimens 1 2 3
P -value
Pan-ERK p-ERK Pan-JNK p-p38
0 8 7 19 0 15 5 4
7 10 21 8 11 12 14 6
P P P P
18 0 11 17
12 0 6 9
= 0.016 = 0.001 = 0.002 = 0.030
a Specimens with cancer cell population of 80% or more of the total cell
count were considered homogenous. b Level/activity of MAPK was grouped as related to A375SM cell line
values, as detailed in the text. Table 3. MAPK isoenzyme expression in the 55 effusions studied. Protein
Number of positive specimens High mW ERK Medium mW ERK Low mW ERK
mRNA in situ hybridization (ISH) ISH was performed on 46 specimens. Cellblock sections (four micron-thick) of formalin-fixed, paraffin-embedded specimens were mounted on ProbeOn Plus slides (Fisher Scientific, Pittsburgh, Pennsylvania). Sectioning was performed in RNAase-free water. ISH was carried out using the microprobe manual staining system (Fisher Scientific) [42]. Hybridization of the probes was carried out as previously described [43]. Known positive controls were used in each hybridization reaction. These consisted of two ovarian carcinomas for which positive hybridization was reproducible in a previous study. Controls for endogenous alkaline phosphatase for all probes included treatment of the sample in the absence of the probe and use of chromogen alone. Evaluation of ISH and ICC results Staining of carcinoma cells was scored. For both mRNA and protein expression, staining extent was scored using a cutoff of 20%. Staining of 20% or less of cells was scored as focal (= 1), while staining of more than 20% of cells was interpreted as diffuse (= 2). Evaluation was done without knowledge of MAPK results. Gelatinolytic assay Evaluation of MMP-2 and MMP-9 activity was performed on cell lysates from all 55 specimens, as previously described [44]. Densitometer analysis of films was performed using the NIH IMAGE program. Activity was scored as negative/low or high. Statistical analysis Statistical analysis was performed applying the SPSS-PC package (version 10.1, SPSS, Chicago, 2000). Probability of < 0.05 was considered significant. Studies of the association between the expression of MAPK, MMP and EMMPRIN, as well as with MMP activity, were undertaken using the twosided chi-square test. Analyses of the association of these molecules, cellular composition of specimens and effusion
pan-ERK p-ERK
43 (78%) 41(74%)
54 (98%) 51 (93%)
15 (27%) 15 (27%)
High mW JNK Medium mW JNK Low mW JNK pan-JNK p-JNK
pan-p38a p-p38
41 (75%) 32 (58%)
24 (44%) 7 (13%)
15 (27%) 7 (13%)
High mW p38
Medium mW p38
Low mW p38
54 (98%) 17 (31%)
Not studied 36 (66%)
Not studied 21 (38%)
a The antibodies directed against pan-p38 available to date recognize
only p38-α.
site were similarly performed using the two-sided chi-square test.
Results MAPK level and activity ERK and p-ERK were detected in 54/55 (98%) and 50/55 (91%) specimens, respectively (Figure 1). Values for panERK ranged from 0% to 393% of the level detected in A375SM cells. p-ERK activity showed a range of 0–98%. JNK and p-JNK were detected in 53/55 (96%) and 38/55 (69%) specimens, respectively (Figure 1). Values for panJNK ranged from 0% to 319% of the level detected in A375SM cells. P -JNK activity showed a similar range of 0–253%. Pan-p38 was expressed in 54/55 (98%) specimens, and its phosphorylated form was found in 51/55 (92%) (Figure 1). As for JNK, pan-p38 levels ranged from 0% to 225% of A375SM values. However, a larger number of specimens showed high phosphorylation activity, exceeding 500% of control values in 2 effusions. MAPK levels were significantly higher in homogenous specimens as compared to mixed ones, significantly so for pan-JNK (P = 0.002), pan-ERK (P = 0.016), P -ERK (P = 0.001) and p-p38 (P = 0.03) (Table 2).
MAPK, MMP and EMMPRIN in ovarian carcinoma
625
Figure 1. MAPK expression and activity in 4 pleural (lanes 1–4) and 8 peritoneal (lanes 5–12) effusions. Upper and lower rows show the pan- and phospho- forms of ERK, JNK and p38. Phosphorylation of p38 and ERK predominates over that of JNK (see text). Table 4. The association between MAPK isoenzyme expression and activity and the cellular composition of effusion specimens.a Isoenzyme expression and activity MAPK isoenzymeb
Homogenous No Yes
Mixed No Yes
P -value
pan-high mW ERK p-high mW ERK p-medium mW ERK p-low mW ERK pan-high mW JNK p-high mW JNK
2 2 0 15 2 7
10 12 4 25 12 16
P P P P P P
24 24 26 11 24 19
19 17 25 4 17 13
= 0.016 = 0.004 = 0.049 = 0.018 = 0.004 = 0.034
a Specimens with cancer cell population of 80% or more of the total cell count were considered homogenous. b Level/activity scored as negative or positive.
Isoenzyme expression Results with respect to expression of different MAPK isoforms are shown in Table 3. The high and medium molecular weight ERK isoforms were the major ERK isoforms expressed, but expression was accompanied by phosphorylation of all three isoforms in the majority of positive specimens (Table 3). Of the stress-induced kinases, the high molecular form of JNK was the major one expressed and activated, while the medium molecular weight p38 isoform was the major phosphorylated kinase from this family. Pan-high molecular weight ERK (P = 0.016) and panhigh molecular weight JNK (P = 0.004) levels, as well as phospho-high molecular weight ERK (P = 0.004), phospho- medium molecular weight ERK (P = 0.049), phospho-low molecular weight ERK (P = 0.018) and
phospho-high molecular weight JNK (P = 0.034) were significantly higher in homogenous specimens as compared to mixed ones (Table 4). Immunocytochemical analysis (ICC) ICC using the anti-p-ERK1/2 antibody, performed in 25 specimens, showed predominant expression in carcinoma cells, with expression range of 0–5% in reactive mesothelial and inflammatory cells (Figures 2A, 2B). Protein expression of MMP-1, MMP-2, MMP-9 and EMMPRIN was found in cancer cells in 45 (82%), 44 (80%), 22 (40%) and 48 (87%) effusions, respectively (Figures 2C–2F, 2H–2I). Expression of MMP-1 and MMP-2 was seen mainly in cancer cells, while MMP-9 expression was also observed in macrophages and less frequently in mesothelial cells (Figure 2G). Protein expression was comparable in pleural and peritoneal effusions (P > 0.05, data not shown). mRNA in situ hybridization (ISH) A positive signal using a poly d(T) probe was detected in all cases (Figure 3A). Specimens hybridized with the sense probe or without probe were negative (Figure 3B). mRNA expression of MMP-2, MMP-9 and EMMPRIN was detected mainly in carcinoma cells, in which it was seen in 37 (80%), 28 (61%) and 38 (83%) effusions (Figures 3C–3E). Ovarian carcinoma cells in pleural and peritoneal effusions showed comparable mRNA expression (P > 0.05, data not shown).
626
B. Davidson et al.
Figure 2. Protein expression of pERK1/2, MMP and EMMPRIN in effusions using immunocytochemistry: (A) and (B) demonstrate p-ERK1/2 expression in two of the studied effusions. Tumor cells show frequent expression, while inflammatory and mesothelial cells that are dispersed between them are largely negative. (C–E) Diffuse protein expression of MMP-1, MMP-2 and MMP-9 in ovarian carcinoma cells in effusion. (F) shows an MMP-1-negative specimen. Expression of MMP-9 in reactive mesothelial cells and macrophages is shown in (G). (H) and (I) show two specimens with unequivocal EMMPRIN expression at the cell membrane.
Gelatinolytic assay Gelatinolytic activity of latent MMP-2 and MMP-9 showed a range from 27 to 1800 units using densitometry. A cutoff of 500 was chosen in order to separate absent/low from high activity. Absent/low and high gelatinolytic activity of MMP-2 was found in 19 (35%) and 36 (65%) specimens, respectively (Figure 4). MMP-9 activity was considerably
lower, with 41 (75%) specimens showing absent/low activity and only 14 (25%) cases showing a high one (Figure 4). Interestingly, high gelatinolytic activity of MMP-2 was more frequent in peritoneal effusions (27/37, 73%) compared to pleural specimens (9/18, 50%), although this finding failed to reach significance (P = 0.093).
MAPK, MMP and EMMPRIN in ovarian carcinoma
627
Figure 3. mRNA expression of MMP and EMMPRIN in effusions using mRNA in situ hybridization: (A) Positive d(T) control for mRNA integrity. All cells are labeled. (B) negative control using a sense probe. Section is counter-stained with Nuclear Fast Red. (C–E) mRNA expression of MMP-2, MMP-9 and EMMPRIN in ovarian carcinoma cells. All cells are labeled. (All figures ×200, NBT-BCIP staining, counter-stained with nuclear fast red.)
Figure 4. Zymography results for MMP-2 and MMP-9 in 20 effusions. Co-activation of MMP-2 and MMP-9 is seen in some specimens, but MMP-2 activity is consistently more pronounced.
Correlation between MAPK expression, MMP and EMMPRIN expression, and MMP activity MMP-2 protein expression correlated with pan-ERK expression (P = 0.035), as well as with phosphorylation activity of the low mW ERK isoform (P = 0.046) (Table 5). MMP2 mRNA expression correlated with both the expression of pan-ERK (P = 0.031) and p-ERK activity (P = 0.048) (Table 5). Finally, MMP-2 gelatinolytic activity was significantly higher in specimens showing phosphorylation of the high mW isoforms of ERK (P = 0.039) and p38 (P = 0.017) (Table 6). MMP-9 protein expression correlated with phosphorylation activity of the medium mW p38 isoform (P = 0.046, data not shown), but MMP-9 gelatinolytic activity showed inverse relationship with expression and activity of pan- (P = 0.008) and (marginally) p- (P = 0.05) forms of the low mW ERK, as well as activity of the low mW isoform of p-p38 (P = 0.033) (Table 7). MMP-1 pro-
tein expression showed no association with MAPK level or phosphorylation. EMMPRIN mRNA expression correlated with MMP-1 protein expression (P < 0.001), while EMMPRIN protein was co-expressed with MMP-2 mRNA (P = 0.042) and MMP-9 protein (P = 0.029) expression (Table 8). In addition, EMMPRIN mRNA expression was significantly higher in specimens showing phosphorylation of the high mW ERK isoform (P = 0.001) (Table 8). Discussion The presence of malignant effusion, either at diagnosis or later in the clinical course, is one of the most frequent clinical findings in ovarian carcinoma. Despite the frequent involvement of the peritoneal and pleural cavity in this disease, research efforts have so far been directed mainly
628
B. Davidson et al.
Table 5. The association between MAPK expression and MMP-2 protein (55 effusions) and mRNA (46 effusions) expression. MAPK
Scorea
MMP-2 proteinb 0 1 2
Table 8. The association between EMMPRIN protein (= 55 specimens) and mRNA (= 46 specimens) expression and MMP and MAPK expression. EMMPRIN mRNAa
Total P -value Molecule
Score 0
1
2
Total P -value
1 2 3
2 2 3 5 10 3 4 7 19
7 18 30
P = 0.035
MMP-1 proteina
0 1 2
5 2 1
3 0 1 20 3 11
8 23 15
P < 0.001
p- low mW ERK 1 2
11 14 15 0 5 10
40 15
P = 0.046
p- low mW ERKb
1 2
6 2
0 4 7 27
10 36
P = 0.001
MMP-2 mRNAb 0 1 2
Total P -value
pan-ERK
Scorea
Molecule
p-ERK
1 2
8 1
1 15 1 20
24 22
P = 0.048
pan-ERK
1 2 3
2 5 2
0 3 2 8 0 24
5 15 26
P = 0.031
a Score = expression as compared to the A375SM melanoma cell line
for p-ERK and pan-ERK, present (= 2) vs. absent (= 1) activity for the ERK isoform. b Score = staining extent (2 = 21–100% of cells, 1 = 1–20% of cells, 0 = 0%).
Table 6. The association between MAPK expression and MMP-2 gelatinolytic activity (= 55 effusions). Scorea
MMP-2 activityb 1 2 Total
P -value
p- high mW ERK
1 2
8 6 11 30
14 41
P = 0.039
p- high mW p38
1 2
17 21 2 15
38 17
P = 0.017
a Score = present (= 2) vs. absent (= 1) activity. b Activity = defined as absent/low or high, based on densito-
meter analysis of zymography gels (see text).
Table 7. The inverse correlation between MAPK expression and MMP-9 gelatinolytic activity (= 55 effusions). Scorea
MMP-9 activityb 1 2 Total
P -value
pan- low mW ERK 1 2
26 14 15 0
40 15
P = 0.009
p- low mW ERK
1 2
27 13 14 1
40 15
P = 0.05
p- low mW p38
1 2
22 12 19 2
34 21
P = 0.033
a Score = present (= 2) vs. absent (= 1) activity. b Activity = defined as absent/low or high, based on densitometer
analysis of zymography gels (see text).
EMMPRIN proteina Score 0
1
2
Total P -value
MMP-2 mRNAa
0 1 2
3 5 1 0 1 2 2 16 16
9 3 34
P = 0.042
MMP-9 proteina
0 1 2
6 18 9 1 2 11 0 2 6
33 14 8
P = 0.029
a Score = present (= 2) vs. absent (= 1) expression/activity. b Extent = percentage of stained carcinoma cells (2 = 21–
100% of cells, 1 = 1–20% of cells, 0 = 0%).
towards the study of primary tumors, while the biological characteristics of ovarian carcinoma cells in effusions have been poorly characterized at both the phenotypic and genotypic level. As a result, relevant questions regarding tumor progression in ovarian cancer are still unanswered. We have previously reported on the expression of MMP [13] and EMMPRIN [39] in ovarian carcinoma cells in effusions, as well as the co-expression of MMP-2 and membrane-typeMMP [14]. Other workers have demonstrated the presence of MMPs and TIMPs in both malignant and non-malignant exudates, as well as in transudates from the pleural cavity using ELISA, zymography and immunoblotting [45, 46]. Since MMP synthesis and activity has been shown to be regulated by signals that are mediated by the MAPK system in vitro, we wished to investigate the association between their expression and activity and the activation of MAPK signaling in cancer cells in effusions. The availability of a large number of fully characterized fresh frozen samples optimally suited this purpose. As in our previously studied cohort [13], MMP-2 and MMP-1 expression predominated over that of MMP-9 in the present study. Gelatinolytic assay similarly showed more pronounced activation of MMP-2. These findings are in agreement with reports by other investigators, in which MMP-2 has been found to be the predominant enzyme in cancer cells in effusions [47]. They are also supported by our finding of increased expression of MMP-2 in effusions, compared with paired primary ovarian carcinomas [13]. EMMPRIN expression showed comparable frequency to that of MMP-1 and MMP-2, in agreement with the postulated role of this protein in transcriptional activation of MMP. Expression of all proteins and mRNAs was predominantly seen in cancer cells. This was also true for MAPK
MAPK, MMP and EMMPRIN in ovarian carcinoma expression and activity, most notably for ERK, the exception being JNK phosphorylation. Extensive documentation is available regarding the in vitro association between MMP-1 and MAPK signaling. This enzyme has been shown to be activated by all three MAPK families in a variety of cells, following activation by agents such as Phorbol esters, the oncogene Met, cytokines and conditioned media [15–21]. Despite the co-expression of MMP-1 with EMMPRIN in our material, in agreement with the role of the latter in MMP activation, we failed to see any association between MMP-1 expression and MAPK expression or activity. The most likely explanation for this discrepancy would be the different origins of the cells studied in these models. The above-mentioned studies analyzed fibroblasts, transformed keratinocytes, glioma cells and one non-ovarian cell line. In addition, the stimulation provided for activation of MMP synthesis is not necessarily present in vivo. Our data thus suggest that MMP-1 expression in ovarian carcinoma cells in effusions may be regulated by other intracellular signaling pathways Less experimental data links MMP-2 and MAPK signaling in cancer cells. ERK 1/2 activity has been shown to mediate, together with protein kinase C and ornithine decarboxylase, MT1-MMP mRNA expression, MMP-2 activation and cellular invasion [22]. The activation of MMP-2 and the invasiveness of human melanoma cells have been shown to be mediated through p38, but not ERK signaling [23]. Our results suggest that MMP expression in vivo is regulated by ERK activity, while the gelatinolytic activity of MMP-2 is linked to both ERK and p38 activation. Although ERK and p38/JNK signaling was originally thought to mediate proliferation and apoptosis signals, respectively [8], it is currently accepted that these kinases have many overlapping functions [48]. MMP-2 regulation may be one of these roles in metastatic ovarian carcinoma. The effect of MAPK signaling on MMP-9 expression and activity has been studied in several carcinomas. The role of p38 and JNK in MMP-9 promoter activation via AP-1 has been shown in head and neck carcinoma lines [24, 25]. Further, ERK [26, 27] and p38 [26] have been shown to mediate similar effects in breast carcinoma cell lines. Inhibition of JNK signaling suppressed MMP-9 promoter activation via AP-1 in OVCAR-3 ovarian carcinoma cells [29]. However, MMP-9 induction in response to stimulation by epidermal growth factor (EGF) has been shown to be mediated largely through phosphatidylinositol 3-kinase (PI3K) in OVCA 429 ovarian carcinoma cells, with only minor contribution of ERK and p38 [28]. Our results show an in vivo correlation between p38 activation and MMP-9 protein expression, but inverse correlation between p38 and ERK activation and MMP-9 gelatinolytic activity. These results may point to a more central role of other signaling pathways for MMP-9 activation in effusions. Alternatively, they may reflect the minor role of this enzyme in our material. The generally lower expression and phosphorylation of JNK, a known activator of MMP-9 via AP-1, in our material supports this hypothesis.
629 The finding of carcinoma cells in a pleural effusion is interpreted as evidence of distant metastatic (stage IV) disease. In contrast, ovarian carcinoma cells can be found in the peritoneal cavity of patients with tumors confined to the ovary (stage Ic) and are believed to originate from direct shedding of cells from the tumor surface. Despite this distinction, little is available to show that patients with positive ascites and intraperitoneal spread (stage IIIc) have significantly better outcome than those with stage IV disease manifested as isolated pleural effusion. Moreover, no phenotypic or genotypic advantages have been documented for cancer cells in the pleural cavity. As a matter of fact, the only differences in protein or mRNA expression in a series of 11 studies [13–14, 33–34, 39, 49–54], were actually higher proliferation rate [49] and more frequent expression of the αv and β1 integrin chains [54] in peritoneal effusions. In the present study, MMP-2 activity was higher in peritoneal effusions, though not significantly. True tumor progression is therefore questionable in this case on both biological and clinical grounds. We believe that these results disagree with the current FIGO classification of isolated pleural effusion as stage-IV disease. In conclusion, an association between MMP-1, MMP2, MMP-9 and EMMPRIN expression in ovarian carcinoma is reported. However, based on both expression and activity results, MMP-2 is the enzyme linked to MAPK signaling in this material, and is probably the most significant one at this anatomic site. Although the precise role of the various MAPK isoforms is unknown at present, our results shed some light on the possible differences in biologic function of these enzymes in terms of protease activation. Though correlative rather than experimental, the study of MAPK activity provides the first in vivo biologic evidence linking these molecules and proteases expression and activity in human tumor material in vivo. Our findings may aid in understanding the biology of ovarian carcinoma and possibly aid in defining better treatment strategies against this malignancy.
Acknowledgements This study was supported by grant D-02019 from the Norwegian Cancer Society. The work of Vered Givant-Horwitz is supported by the Yeshaya Horowitz fellowship grant.
References 1. 2.
3.
4.
5.
Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000; 100: 57–70. Denhardt DT. Signal-transducing protein phosphorylation cascades mediated by Ras/Rho proteins in the mammalian cell: The potential for multiplex signalling. Biochem J 1996; 318: 729–47. Garrington TP, Johnson GL. Organization and regulation of mitogenactivated protein kinase signaling pathways. Curr Opin Cell Biol 1999; 11: 211–8. Fanger GR. Regulation of the MAPK family members: Role of subcellular localization and architectural organization. Histol Histopathol 1999; 14: 887–94. Chang L, Karin M. Mammalian MAP kinase signalling cascades. Nature 2001; 410: 37–40.
630 6. Davis RJ. MAPKs: New JNK expands the group. Trends Biochem Sci 1994; 19: 470–3. 7. Obata T, Brown GE, Yaffe MB. MAP kinase pathways activated by stress: The p38 pathway. Crit Care Med 2000; 28 (Suppl): N67–77. 8. Xia Z, Dickens M, Raingeaud J et al. Opposing effects of ERK and JNK-p38 kinases on apoptosis. Science 1995; 270: 1326–31. 9. Sivaraman VS, Wang H, Nuovo GJ, Malbon CC. Hyperexpression of mitogen-activated protein kinase in human breast cancer. J Clin Invest 1997; 99: 1478–83. 10. Xu Q, Liu Y, Gorospe M et al. Acute hypertension activates mitogenactivated protein kinases in arterial wall. J Clin Invest 1996; 97: 508– 14. 11. Engelhardt JF. Redox-mediated gene therapies for environmental injury: Approaches and concepts. Antioxid Redox Signal 1999; 1: 5–27. 12. Egeblad M, Werb Z. New functions for the matrix metalloproteinases in cancer progression. Nature Rev (Cancer) 2002; 2: 161–74. 13. Davidson B, Reich R, Berner A et al. Ovarian carcinoma cells in serous effusions show altered MMP-2 and TIMP-2 mRNA levels. Eur J Cancer 2001; 37: 2040–9. 14. Davidson B, Goldberg I, Berner A et al. Expression of membrane-type 1,2 and 3 matrix metalloproteinases (MT1-MMP, MT2-MMP, MT3MMP) mRNA in ovarian carcinoma cells in serous effusions. Am J Clin Pathol 2001; 115: 517–24. 15. Brauchle M, Glück D, Di Padova F et al. Independent role for p38 and ERK1/2 mitogen-activated kinases in the up-regulation of matrix metalloproteinase-1. Exp Cell Res 2000; 258: 135–44. 16. Lim M, Martinez T, Jablons J et al. Tumor-derived EMMPRIN (extracellular matrix metalloproteinase inducer) stimulates collagenase transcription through MAPK p38. FEBS Lett 1998; 441: 88–92. 17. Westermarck J, Li S, Jaakkola P et al. Activation of fibroblast collagenase-1 expression by tumor cells of squamous cell carcinomas is mediated by p38 mitogen-activated protein kinase and c-Jun NH2-terminal kinase-2. Cancer Res 2000; 60: 7156–62. 18. Westermarck J, Li SP, Kallunki T et al. p38 mitogen-activated protein kinase-dependent activation of protein phosphatases 1 and 2A inhibits MEK1 and MEK2 activity and collagenase 1 (MMP-1) gene expression. Mol Cell Biol 2001; 21: 2373–83. 19. Johansson N, Ala-aho R, Uitto VJ et al. Expression of collagenase-3 (MMP-13) and collagenase-1 (MMP-1) by transformed keratinocytes is dependent on the activity of p38 mitogen-activated protein kinase. J Cell Sci 2000; 113: 227–35. 20. Ieda Y, Waguri-Nagaya Y, Iwahasi T et al. IL-1β-induced expression of matrix metalloproteinases and gliostatin/platelet-derived endothelial cell growth factor (GLS/PD-ECGF) in a chondrosarcoma cell line (OUMS-27). Rheumatol Int 2001; 21: 45–52. 21. Garcia-Guzman M, Dolfi F, Zeh K, Vuori K. Met-induced JNK activation is mediated by the adapter protein Crk and correlates with the Gab1-Crk signaling complex formation. Oncogene 1999; 18: 7775–86. 22. da Rocha AB, Mans DRA, Lenz G et al. Protein kinase C-mediated in vitro invasion of human glioma cells through extracellular-signalregulated kinase and ornithine decarboxylase. Pathobiology 2000; 68: 113–23. 23. Denkert S, Siegert A, Leclere A et al. An inhibitor of stress-activated MAP-kinases reduces invasion and MMP-2 expression of malignant melanoma cells. Clin Exp Metastasis 2002; 19: 79–85. 24. Simon C, Simon M, Vucelic G et al. The p38 SAPK pathway regulates the expression of the MMP-9 collagenase via AP-1-dependent promoter activation. Exp Cell Res 2001; 271: 344–55. 25. Gum R, Wang H, Lengyel E et al. Regulation of 92 kDa type IV collagenase expression by the jun aminoterminal kinase- and the extracellular signal-regulated kinase-dependent signaling cascades. Oncogene 1997; 14: 1481–93. 26. Yao J, Xiong S, Klos K et al. Multiple signaling pathways involved in activation of matrix metalloproteinase-9 (MMP-9) by heregulin-β1 in human breast cancer cells. Oncogene 2001; 20: 8066–74. 27. Liu JF, Crepin M, Liu JM et al. FGF-2 and TPA induce Mtrix metalloproteinase-9 secretion in MCF-7 cells through PKC activation of the Ras/ERK pathway. Biochem Biophys Res Comm 2002; 293: 1174–82. 28. Ellerbroek SM, Halbleib JM, Benavidez M et al. Phosphatidylinositol 3-kinase activity in epidermal growth factor-stimulated matrix
B. Davidson et al.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46. 47.
48. 49.
metalloproteinase-9 production and cell surface association. Cancer Res 2001; 61: 1855–61. Shin M, Yan C, Boyd D. An inhibitor of c-jun aminoterminal kinase (SP600125) represses c-Jun activation, DNA-binding and PMAinducible 92-kDa type IV collagenase expression. Biochem Biophys Acta 2002; 1589: 311–6. Lakka SS, Jasti SL, Kyritsis AP et al. Regulation of MMP-9 (type IV collagenase) production and invasiveness in gliomas by the extracellular signal-regulated kinase and jun amino-terminal kinase signaling cascades. Clin Exp Metastasis 2000; 18: 245–52. Xu S, Khoo S, Dang A et al. Differential regulation of mitogenactivated protein/ERK kinase (MEK)1 and MEK2 and activation by a Ras-independent mechanism. Mol Endocrinol 1997; 11: 1618–25. Davidson B, Risberg B, Kristensen G et al. Detection of cancer cells in effusions from patients diagnosed with gynecological malignancies – evaluation of five epithelial markers. Virchows Arch 1999; 435: 43–9. Davidson B, Berner A, Nesland JM et al. E-cadherin and α-, β- and γ catenin protein expression is up-regulated in ovarian carcinoma cells in serous effusions. J Pathol 2000; 192: 460–9. Davidson B, Berner A, Nesland JM et al. Carbohydrate antigen expression in primary tumors, metastatic lesions, and serous effusions from patients diagnosed with epithelial ovarian carcinoma – evidence of up-regulated Tn and Sialyl Tn antigens expression in effusions. Hum Pathol 2000; 31: 1081–7. Khokhlatchev A, Xu S, English J et al. Reconstitution of mitogenactivated protein kinase phosphorylation cascades in bacteria. Efficient synthesis of active protein kinases. J Biol Chem 1997; 272: 11057–62. Raptis L, Brownell HL, Lu Y et al. v-Ras and v-Raf block differentiation of transformable C3H10T1/2-derived preadipocytes at lower levels than required for neoplastic transformation. Exp Cell Res 1997; 235: 188–97. Brownell HL, Lydon NB, Schaefer E et al. Inhibition of epidermal growth factor-mediated ERK1/2 activation by in situ electroporation of nonpermeant ((alkylamino)methyl)acrylophenone derivatives. DNA Cell Biol 1998; 17: 265–74. Nagata K, Puls A, Futter C et al. The MAP kinase kinase kinase MLK2 co-localizes with activated JNK along microtubules and associates with kinesin superfamily motor KIF3. EMBO J 1998; 7: 149–58. Davidson B, Goldberg I, Berner A et al. EMMPRIN (extracellular matrix metalloproteinase inducer) is a novel marker of poor outcome in serous ovarian carcinoma. Clin Exp Metastasis 2003; 20: 161–9. Greene GF, Kitadai Y, Pettaway CA et al. Correlation of metastasisrelated gene expression with metastatic potential in human prostate carcinoma cells implanted in nude mice using an in situ messenger RNA hybridization technique. Am J Pathol 1997; 150: 1571–82. Guo H, Majmudar G, Jensen TC et al. Characterization of the gene for human EMMPRIN, a tumor cell surface inducer of matrix metalloproteinases. Gene 1998; 220: 99–108. Reed JA, Manahan LJ, Parks CS, Brigati DJ. Complete one-hour immunocytochemistry based on capillary action. Biotechniques 1992; 13: 434–43. Davidson B, Goldberg I, Gotlieb WH et al. High levels of MMP-2, MMP-9, MT1-MMP and TIMP-2 mRNA correlate with poor survival in ovarian carcinoma. Clin Exp Metastasis 1999; 17: 799–808. Reich R, Blumenthal M, Liscovitch M. Laminin-induced production of collagenase IV involves activation of phospholipase D in metastatic tumor cells. Clin Exp Metastasis 1995; 13: 134–40. Eickelberg O, Sommerfeld CO, Wyser C et al. MMP and TIMP expression patterns in pleural effusions of different origins. Am J Resp Crit Care Med 1997; 156: 1987–92. Hurewitz AN, Zucker S, Mancuso P et al. Human pleural effusions are rich in matrix metalloproteinases. Chest 1992; 102: 1808–14. Fishman DA, Bafetti LM, Banionis S et al. Production of extracellular matrix-degrading proteinases by primary cultures of human epithelial ovarian carcinoma cells. Cancer 1997; 80: 1457–63. English J, Pearson G, Wilsbacher J et al. New insights into the control of MAP kinase pathways. Exp Cell Res 1999; 253: 255–70. Davidson B, Risberg B, Berner A et al. Expression of cell cycle proteins in ovarian carcinoma cells in serous effusions - biological and prognostic implications. Gynecol Oncol 2001; 83: 249–56.
MAPK, MMP and EMMPRIN in ovarian carcinoma 50. Berner HS, Davidson B, Berner A et al. Expression of CD44 in effusions of patients diagnosed with serous ovarian carcinoma - diagnostic and prognostic implications. Clin Exp Metastasis 2000; 18: 197–202. 51. Davidson B, Lazarovici P, Ezersky A et al. Expression levels of the NGF receptors TrkA and p75 in effusions and solid tumors of serous ovarian carcinoma patients. Clin Cancer Res 2001; 7: 3457–64. 52. Davidson B, Reich R, Kopolovic J et al. Interleukin-8 and vascular endothelial growth factor mRNA levels are down-regulated in ovarian
631
53.
54.
carcinoma cells in serous effusions. Clin Exp Metastasis 2002; 19: 135–44. Davidson B, Risberg B, Goldberg I et al. Ets-1 mRNA expression in effusions of serous ovarian carcinoma patients is a marker of poor outcome. Am J Surg Pathol 2001; 25: 1493–500. Davidson B, Goldberg I, Reich R et al. The αv and β1 integrin subunits are commonly expressed in malignant effusions from ovarian carcinoma patients. Gynecol Oncol 2003; 90: 248–57.
