• MKK4;
  • EMT;
  • ovarian cancer;
  • twist;
  • NF-κB;
  • E-cadherin


  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

In the current study, we investigated the mechanism relating downregulation of mitogen-activated protein kinase kinase 4 (MKK4) expression to development of ovarian cancer. Over-expression of the MKK4 gene in TOV-21 G cells, a line with homozygous deletion of MKK4, resulted in morphologic changes in which cells growing in a scattered, fibroblast-like pattern formed tightly packed colonies. Based on a wound healing assay and a Matrigel invasion assay, we determined that both motility and invasiveness of MKK4-transfected TOV-21G cells were significantly reduced compared to control vector-transfected cells. To confirm that MKK4 expression related to tumor invasion resulted from an epithelial to mesenchymal transition (EMT)-like morphological change, we used 2 independent but complementary approaches. MKK4 gene knockdown in MDAH 2774 cells over-expressing MKK4 increased invasion activity. Additionally, engineered expression of MKK4 in SKOV3 cells, a line with low endogenous MKK4 expression, produced a phenotype similar to that of TOVG-21G. Interestingly, we found that MKK4 upregulation caused downregulation of phosphorylated NF-κB and Twist, as well as upregulation of E-cadherin, in TOVG-21G and SKOV3 cells. Reciprocal results were obtained in MDAH 2774 cells with MKK4 knockdown. Our results suggest that MKK4 downregulation causes increased phosphorylation NF-κB. This promotes Twist over-expression, resulting in E-cadherin downregulation that induces EMT in ovarian cancer.

Ovarian cancer is the most lethal gynecological malignancy in the world.1 As there is no method of early detection, the majority of patients present with peritoneal dissemination and distant metastasis at the time of diagnosis. With disseminated disease, treatment is often unsuccessful and overall survival is low. The identification of an invasion-related molecule associated with early and rapid spread of ovarian cancer is the current focus of many investigators. Recently, inactivation or downregulation of metastasis suppressor genes has been associated with ovarian cancer progression. Mitogen-activated protein kinase kinase 4 (MKK4), a member of the stress-activated protein kinase signaling cascade, has been identified as a metastasis-suppressor gene.2 Progressive loss of expression occurs in prostate,3 pancreatic4 and ovarian cancers.5 The consequences of downregulation of MKK4 could include development of a more aggressive phenotype, one prone to invasion and metastasis that is more difficult to optimally debulk and is more chemoresistant.6 The role of MKK4 in cancer is incompletely studied4–6 and further studies will be required to clarify the function. Recently, we identified homozygous deletion of MKK4, suggesting a possible mechanism for the downregulation of MKK4 in ovarian cancer.7 Cells with a homozygous deletion of MKK4 are an ideal model to analyze its function using gene transfection methods. In the current study, we successfully identified an ovarian cancer cell line carrying a homozygous MKK4 deletion. We utilized this line in a gene transfection assay in order to clearly demonstrate the involvement of MKK4 in tumor growth and invasion both in vitro and in vivo.

Invasion and metastasis are the biologic hallmarks of malignancy. The molecular mechanism responsible for invasion and metastasis is a key area to investigate. A number of molecules related to tumor invasion and spread in ovarian cancer have been reported and include the following: Twist,8 E-cadherin,9 NF-κB10 and Snail.11 However, the molecular changes associated with acquisition of metastatic ability in ovarian cancer progression are poorly understood. Recently, the epithelial to mesenchymal transition (EMT) has been described as an important mechanism promoting invasion and causing metastasis of cancer.12 The EMT is basically an embryonic trait through which cells adopt a phenotype more amenable to migration and invasion.13 The phenotypic change after forced expression of MKK4 in our present model raised the possibility that the molecular mechanism by which MKK4 suppressed metastasis involved the EMT. In light of this, we also measured expression levels of other invasion-related molecules promoting EMT changes in MKK4 deleted cells. Our findings are the first evidence linking EMT to the metastasis suppressive function of MKK4.

Material and Methods

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Tissue samples

A total of 89 paraffin-embedded tumor tissue and normal ovarian tissue samples were obtained from the Department of Obstetrics and Gynecology at Shimane University Hospital: 54 ovarian carcinomas, for example, 24 serous carcinomas, 3 mucinous carcinomas, 6 clear cell carcinomas and 8 endometrioid carcinomas, 28 ovarian borderline tumors (8 serous and 20 mucinous cases), 11 ovarian benign cystadenomas and 9 normal ovarian tissue samples. Diagnosis was based on conventional morphological examination of H&E-stained sections. Tumors were classified according to the WHO classification. Tumor staging was performed according to the International Federation of Gynecology and Obstetrics (FIGO) classification. Acquisition of tissue specimens and clinical information was approved by an institutional review board (Shimane University). The paraffin tissues were organized into tissue microarrays, which were made by removing 3-mm diameter cores of tissue from each block. The selection of the area to core was made by a gynecologic oncologist (K.N) and pathological technician (K.I) based on review of the H&E slides.


Formalin-fixed and paraffin-embedded sections were dewaxed in xylene and hydrated in graded alcohol. After antigen retrieval in a sodium citrate buffer, slides were incubated overnight at 4°C with antibodies to MKK4 (Novocastra, Newcastle, United Kingdom), Twist (Santa Cruz Biotechnology, Santa Cruz, CA, catalog no. sc-15393), E-cadherin (Upstate Cell Signaling, Lake Placid, NY, catalog no. 51-9001922) and phospho-IκB-α (Upstate Cell Signaling, catalog no. 4814) at dilutions of 1:40,1:100,1:100, and 1:1,000, respectively. This was followed by incubation with a biotinylated linker and streptavidin-horseradish peroxidase (LSAB2 system-HRP, DAKO Cytomation, Carpinteria, CA). The signals were visualized using ABC+ (DAKO Cytomation) as the substrate-chromagen at room temperature for 10 min. Sections were counterstained with hematoxylin and mounted. Immunoreactivity was scored by 2 investigators as follows: 0: undetectable, 1+: weakly positive, 2+: moderately positive and 3+: intensely positive.

Cell lines

SKOV3, MDAH2774 and TOV-21G human ovarian cancer cell lines were obtained from the American Tissue Culture Center (Rockville, MD). The human ovarian carcinoma cell lines KF28, KF28TX and KFr13TX were a kind gift of Dr. Yoshihiro Kikuchi (Ohki Memorial Kikuchi Cancer Clinic for Women, Saitama, Japan).14 OVK#18, a human ovarian cancer cell line, was obtained from Tohoku University (Sendai, Japan). JHOC5, another human ovarian cancer cell line, was obtained from Riken Bioresource Center (Ibaragi, Japan). All cell lines were maintained in DMEM (Life Technologies, Gaithersburg, MD) supplemented with 5% fetal bovine serum (FBS), 100 units/mL penicillin, and 100 μg/mL streptomycin at 37°C in an atmosphere of 5% CO2.

Western blot analysis

Western blot analysis of the above mentioned ovarian carcinoma cell lines was performed using antibodies to MKK4 (Upstate cell signaling, catalog no. 9152), Twist (Santa Cruz Biotechnology, catalog no. sc-15393), E-cadherin (Upstate Cell Signaling, catalog no. 51-9001922), phospho-IκB-α (Upstate Cell Signaling, catalog no. 92465) total-IκB-α (Upstate Cell Signaling, catalog no. 4814), β-catenin (Transduction Laboratories, Palo Alto, CA catalog no. 51-900192), and vimentin (Chemicon International, Temecula, CA, catalog no. sc-53464) at dilutions of 1:500,1:200,1:200.1:1,000, 1:1,000, 1:2,000, and 1: 2,000, respectively. Similar amounts of total protein (50 μg) from each lysate were loaded and separated on 10% Tris-Glycine-SDS polyacrylamide gels and electroblotted to Millipore Immobilon-P polyvinylidene difluoride membranes (Millipore, Billerica, MA). Western blots were developed by chemiluminescence (Pierce, Rockford, IL).

PCR analysis for MKK4

Genomic DNA was purified from TOV-21G and KF28 cell lines using a DNA purification kit (Qiagen, Valencia, CA). Polymerase chain reaction (PCR) was then performed with an iCycler (Bio-Rad, Hercules, CA). The primers for MKK4 and LINE1 were GACCAGATGAGGATCCGAAA (Forward), AGGCCCAAGCAAAACCTAAT (Reverse), TCTCCCCGAGTTTCATCAAC (Forward), GCTTCCTCTCAGCCCTTTTT (Reverse), and AAAGCCGCTCAACTACATGG (Forward), TGCTTTGAATGCGTCCCAGAG (Reverse). PCR products were 212, 239 and 172 bp, respectively. The genomic locations of PCR primers are listed in Supporting Information Table 1.

Fluorescence in situ hybridization

BAC clones (CTD-2331E5) containing the genomic sequences harboring the MKK4 gene located at 17p12 were labeled with biotin by nick translation. Bac clones (RP11-127K18 and RP11-629A22) located at Chr2q11.2 were used to generate the reference probes. FISH methods and scoring criteria have been detailed in previous reports.7, 15

Plasmid constructs and generation of stable clones

TOV-21G and SKOV3 cells were transfected with plasmid vector pcDNA3 (Invitrogen, Carlsbad, CA) that contained cDNA for MKK4 (construct provided by R. J. Davis, Howard Hughes Medical Institute, University of Massachusetts, Worcester, MA), using LipofectAMINE reagent (Life Technologies). The MKK4 siRNA expression vector was obtained from Bioscience Corporation (SABioscience Corporation, Frederick, MD). Two siRNA-expressing vectors targeting MKK4 were designed with the following sense sequences: GCCTTACGAAGGATGAATCCA and CGCATCACGACAAGGATATGA. A control siRNA-expressing vector was also purchased from Bioscience Corporation (Frederick, MD). Briefly, cells were grown to ∼60% confluence in T75 culture dishes, rinsed with serum-free medium, overlaid with a mixture of 7 μg of DNA and 54 μL of LipofectAMINE reagent diluted in serum-free medium, and incubated at 37°C in 5% CO2/95% air for 12 hr. The transfection medium was then replaced with fresh medium; 36 hr later, cells were harvested, diluted in growth medium containing 600 μg/mL G418 for TOV-21G cells, 500 μg/mL G418 for SKOV3 and 800 μg/mL G418 for MADH cells, and split to select and establish clonal cell lines. Selection of PCMV/MKK4-stable clones and the MKK4 knockdown clone was performed by minimal dilution in selection medium containing 600, 500 or 800 μg/mL G418 for TOV-21G, SKOV3 or MDAH 2774 cells, respectively (Sigma, St. Louis, MO).

Wound assay to assess cell motility

Cells were seeded in 6-well plates and grown to a confluent monolayer. An acellular area was created by scraping the cell surface using a 200-μL pipette tip (time 0). Floating cells were removed by 2 gentle washes with culture media. The speed of defect closure was monitored for 24 hr. The numbers of cells invading the acellular area were counted at 8, 12 and 24 hr postscraping. The individual cells in the monolayer defect were quantified as an average from multiple fields (at least 5) at 200× magnification for each experiment.

Matrigel invasion assay

The invasion study was performed using chemotaxic cell culture chambers containing a membrane with a 0.8-μm pore size (Kurabo, Osaka, Japan). 30 μL of serum-free DMEM-diluted Matrigel (1 mg/mL) (BD Bioscience, Bedford, MA) was added to the membrane and incubated at room temperature for 4 hr to form matrix gels. Chambers without the Matrigel coating were used as control chambers. Control vector-transfected TOV-21G, SKOV3 and MDAH 2774 cells, or the MKK4-stable transfected or MKK4-stable knockdown clone, were seeded individually in wells at a density of 25,000 cells/250 μL in serum-free medium in the upper chamber. In the lower compartment, 750 μL of DMEM media containing 5% FBS was added. After 22 hr of incubation at 37°C under 5% CO2, the Matrigel was carefully removed using a cotton swab. The membranes were fixed with 4% paraformaldehyde and stained with crystal violet. Cells migrating through the membrane and cells invading the Matrigel were counted in 5 nonoverlapping ×200 fields under a light microscope.

In vitro kinase assay

Serum-starved TOV-21G-MKK4 clones were stimulated with 50 ng/mL anisomycin (Sigma) for 20 min and protein lysates were subjected to Western blotting using anti-phospho-MKK4antibody (Abcam, Cambridge, MA, MEK4 (phospho S80) antibody (ab39403)) with 1:100 dilution.

Tumor xenograft in nude mice

To confirm the above findings of an MKK4 effect in vitro, we injected either 3 × 106 TOV-21G, SKOV3 or MKK4 cells, all of which had been stably vector transfected, or control shRNA- or MKK4 shRNA-stable transfected MDAH2774 cells into the intraperitoneal tissue of 4-week-old athymic BALB/c nu/nu mice (Charles River Japan, Kanagawa, Japan). At the start of the experiment, the end point was set for validity and ethics to the day when the mice first developed ascites.

Three mice were used for each experimental group. Two to four weeks after injection, necropsy was performed on all mice to assess intraperitoneal tumor growth and quantity of ascites. Tumors were excised and weighed. Animal experiments were performed in accordance with the regulations of the institutional ethical commission (Shimane University) and that of the United Kingdom Coordinating Committee on Cancer Research Guidelines.16

Statistical analysis

Results are expressed as the mean ± SD. A value of p < 0.05 was considered statistically significant. To generate the p value, Student's t-test was used. The Pearson correlation coefficient test was used to examine statistical significance in immunohistochemical analysis values.


  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

MKK4 expression is lower in invasive ovarian carcinomas than in borderline tumors, benign cystadenomas and normal ovarian surface epithelium

To assess the biological roles of MKK4 in tumor progression, we first performed immunohistochemistry on surgically removed ovarian tumor tissue and benign counterpart tissue, as well as tissue from normal ovaries. The number of samples of normal ovarian surface epithelium, ovarian cystadenoma, borderline tumor and invasive carcinoma is listed in Figure 1e.

thumbnail image

Figure 1. MKK4 immunoreactivity in normal ovarian surface epithelium and ovarian tumor tissues. MKK4 immunointensity is (a) strong in normal surface epithelium, (b) strong in a benign cystadenoma, and (c) strong in a borderline tumor. (d) MKK4 immunostaining is undetectable in an invasive ovarian cancer sample. (e) Immunostaining was scored as a 4-tiered system based on the intensity of immunoreactive tumor cells. The number of samples of normal ovarian surface epithelium, ovarian cystadenoma, borderline tumor and invasive carcinoma are listed. [Color figure can be viewed in the online issue, which is available at]

Download figure to PowerPoint

Cytoplasmic staining of MKK4 was uneven in ovarian tissues (Figs. 1a1d). In contrast to normal ovarian surface epithelium, 29% of benign cystadenomas and borderline tumors and 66% of invasive ovarian carcinomas demonstrated lower MKK4 immunoreactivity (e.g., 0 and 1+ positivity; chi-square test, p < 0.05) (Fig. 1e). In other words, positive MKK4 staining intensity (2+ and 3+) was more frequent in cells from ovarian surface epithelium, benign cystadenomas and borderline tumors than in cells from invasive ovarian carcinomas (p < 0.05, chi-square test) (Fig. 1e).

Loss of expression of MKK4 protein and a homozygous deletion of MKK4 in TOV-21G cell lines

We analyzed the MKK4 protein expression level in several ovarian cancer cell lines and found reduced expression in SKOV3 (serous carcinoma) cells and absolutely no expression in TOV-21G (clear cell carcinoma) cells (Fig. 2a). MDAH2774 (serous carcinoma) cells over-expressed MKK4 and the other cell lines, including JHOC5 (clear cell carcinoma), OVK#18 (serous carcinoma), KF28TX (serous carcinoma), KF28 (serous carcinoma) and KFr13TX (serous carcinoma), expressed MKK4 moderately. To identify the mechanism underlying loss of MKK4 protein expression in the TOV-21G cell line, genomic DNA copy number was analyzed by PCR. Interestingly, we found a homozygous deletion of MKK4 in this line (Fig. 2b). To confirm the PCR results, we employed a dual color FISH analysis. The FISH analysis is based on directly counting probe signals on slide cells and provides a highly sensitive and specific approach of assessing gene copy number.17, 18 TOV-21G cells showed a signal corresponding to the reference probe, but no signal was seen for the MKK4 probe from the BAC clone, CTD-231E5, indicating a homozygous deletion (Supporting Information Fig. 1).

thumbnail image

Figure 2. (a) Western blot analysis of ovarian cancer cell lines shows absolutely no expression of MKK4 protein in TOV-21G cells, weak expression in SKOV3 cells, and high expression in MDAH cells. (b) Genomic-PCR using MKK4 primers show homozygous deletion of MKK4 in TOV-21G cells and expression in the KF28 ovarian carcinoma line. (c) Functional analysis of MKK4 expression: Western blot analysis showing forced expression of MKK4 protein in MKK4-transfected TOV-21G cells and no expression of MKK4 in control vector-transfected TOV-21G cells. MKK4 expression was somewhat variable in the 3 clones tested. [Color figure can be viewed in the online issue, which is available at]

Download figure to PowerPoint

Constitutive expression of MKK4 leads to morphological mesenchymal to epithelial transition (MET)-like changes

TOV-21G cells were transfected with a vector expressing MKK4; 3 independent clones were randomly selected for functional analysis. Western blot analysis confirmed MKK4 expression in these clones (Fig. 2c). Over-expression of the MKK4 gene in TOV-21G cells resulted in morphologic changes in which cells growing in a scattered, fibroblast-like pattern formed tightly packed colonies (Fig. 3a). SKOV3, another line with low endogenous MKK4 expression, demonstrated similar morphological changes when MKK4 was over-expressed (Supporting Information Fig. 2A).

thumbnail image

Figure 3. Ectopic expression of MKK4 in TOV-21G cells. (a) Morphologic changes were induced by over-expression of MKK4. Over-expression of MKK4 led to morphological changes consistent with a mesenchymal to epithelial transition (MET). (b) Effect of MKK4 transfection on cell migration in a wound healing assay. Photographs were taken immediately after incision (0 hr) (left panel) and 24 hr postincision (right panel). Phase contrast images show a significant decrease in cell migration after MKK4 transfection. (c) The numbers of migrated cells were significantly lower in MKK4-transfected TOV-21G clones than in control vector-transfected cells in a time course assay. *p < 0.01. (d) Matrigel invasion assay. The invasiveness of TOV-21G control vector and TOV-21G MKK4 clones was evaluated by their ability to invade and penetrate through Matrigel-coated pores on membrane inserts of the trans-wells. TOV-21G MKK4 clones showed a significantly lower invasion capacity than TOV-21G control vector clones. (e) TOV-21G MKK4 clones showed a significantly lower invasion capacity than TOV-21G control vector-transfected cells. * p < 0.05. The error bars were standard deviation. [Color figure can be viewed in the online issue, which is available at]

Download figure to PowerPoint

Constitutive expression of MKK4 leads to decreased motility and invasive capabilities in vitro

The EMT induces cell motility, an essential component for proper gastrulation.19, 20 Similar phenomena involving cell motility are also necessary for tumor invasion.21 To understand the phenotypic characteristics of the EMT, we analyzed the motility and invasive abilities of TOV-21G cells after MKK4 transfection. Cell motility was investigated with a wound-healing assay. Transfected cells showed 90% reduced cell motility compared with the vector control (p < 0.01) (Figs. 3b and 3c). Following the Matrigel invasion assay, a 45% decrease in cell invasion was observed in MKK4-transfected TOV-21G cells in comparison with the empty vector control (p < 0.05) (Figs. 3d and 3e). Similar results were also obtained with MKK4-transfected SKOV3 cells (p < 0.01) (Supporting Information Figs. 2B, C). These results indicate that MKK4 inactivation might lead to increased invasiveness of TOV-21G and SKOV3 cells through regulation of EMT changes.

Constitutive knockdown of MKK4 leads to progression of invasion ability through induction of EMT-like changes

To confirm that MKK4 over-expression led to suppression of invasion ability through induction of MET-like changes, we utilized complementary approaches using a gene knockdown system. MDAH 2774 cells were transfected with a vector expressing MKK4 siRNA, and 2 independent, stable MKK4 knockdown clones were randomly selected for functional analysis. Western blot analysis confirmed decreased MKK4 expression in both clones (Fig. 4a). The colony morphology of the MKK4-knockdown clones changed from tightly packed colonies to a scattered, fibroblast-like growth pattern (Fig. 4b). Cell motility was investigated with a wound-healing assay. MKK4-knockdown in MDAH 2774 cells resulted in a 60% increase in cell motility in comparison with cells transfected with the vector expressing control siRNA (p < 0.01) (Fig. 4c). In the Matrigel invasion assay, a 60% increase in cell invasion was observed in MKK4-knockdown MDAH 2774 cells in comparison with control siRNA vector-transfected cells (p < 0.01) (Fig. 4d).

thumbnail image

Figure 4. (a) Silencing RNA knockdown of MKK4 gene expression: Western blot analysis showed reduced expression of MKK4 protein in MKK4-siRNA-expressing vector transfected MDAH cells and no change in MKK4 expression in control siRNA-expressing vector-transfected MDAH cells. (b) Effect of MKK4 knockdown on cell migration in a wound healing assay. The number of migrated cells was significantly higher in MKK4 siRNA-transfected MDAH clones than in control siRNA-transfected cells in a time course assay. *p < 0.01. (c) MDAH MKK4 siRNA-transfected clones showed significantly higher invasion capacity than MDAH control siRNA-transfected cells. *p < 0.01. The error bars were standard deviation. [Color figure can be viewed in the online issue, which is available at]

Download figure to PowerPoint

Constitutive expression of MKK4 results in tumor suppression and constitutive knockdown of MKK4 results in tumor progression in vivo

Having established the molecular and phenotypic changes in MKK4-transfected cells, we next examined whether or not these cells had altered tumor growth and peritoneal dissemination in athymic nu/nu mice. Intraperitoneal injections of MKK4-expressing TOV-21G clones in athymic nu/nu mice produced significantly smaller tumors and a lower volume of hemato-ascites than control cells transfected with control vector (p < 0.01) (Figs. 5a and 5b) (Supporting Information Fig. 3). Furthermore, knockdown of MKK4 in MDAH 2774 cells resulted in production of disseminated tumors and greater hemato-ascites when these cells were injected into athymic nu/nu mice (p < 0.01) (Figs. 5c and 5d).

thumbnail image

Figure 5. (a) TOV-21G MKK4 cells produced very small tumor nodules in the peritoneal cavity; however, the TOV-21G control vector cells grew much larger intraperitoneal tumors. *p < 0.01. (b) TOV-21G MKK4 cells produced almost no hemato-ascites in the peritoneal cavity; however, mice injected with TOV-21G control vector-transfected cells had a much larger volume of hemato-ascites. *p < 0.01. The error bars were standard deviation. (c) MDAH control siRNA-transfected cells produced very small tumor nodules in the peritoneal cavity; however, MDAH MKK4 siRNA-transfected cells grew disseminated intraperitoneal tumors. *p < 0.01. (d) MDAH control siRNA-transfected cells produced almost no hemato-ascites in the peritoneal cavity; however, mice injected with MDAH MKK4 siRNA-transfected cells had a much larger volume of hemato-ascites. *p < 0.01.

Download figure to PowerPoint

Constitutive expression of MKK4 leads to EMT inactivation via downregulation of phosphorylated NF-κB, Twist and upregulation of E-cadherin. Reciprocal changes were seen following constitutive knockdown of MKK4

Given the negative correlation between MKK4 expression and the EMT in TOV-21G and SKOV3 cell lines, we then investigated the mechanism by which downregulation of MKK4 stimulated EMT changes. NF-κB and Twist are important promoters of the EMT in cancer. In contrast, E-cadherin is a negative regulator.12 When epithelial cells are undergoing EMT, their morphology changes well organized cell–cell adhesion and cell polarity to loss of cell–cell contacts and cell scattering that is often associated with the loss of epithelial makers such as E-cadherin and catenins, and gain of mesenchymal markers such as vimentin.22 Therefore, we measured expression levels of Twist, E-cadherin, β-catenin, vimentin and phosphorylated IkB-α using a Western blot. Because phosphorylation of IκB-α is an essential step for the release of active NF-κB, phosphorylated IκB-α (phospho-Ik Bα) is an excellent marker of NF-κB activation.23 Decreased expression of Twist and phospho-IkB-α and increased expression of E-cadherin were noted in MKK4-transfected TOV-21G and SKOV3 clones. Over-expression of Twist and phosphorylated IkB-α accompanied by absent expression of E-cadherin were found in control vector-transfected cells (Fig. 6a). These results suggest that MKK4 may prevent the EMT through inhibition of Twist-mediated NF-κB-dependent prosurvival activity. To confirm this, we used the same approach for MKK4-knockdown in MDAH 2774 cells. Reciprocal changes were obtained following MKK4-knockdown in MDAH 2774 cells (Fig. 6a).

thumbnail image

Figure 6. (a) Effect of MKK4 transfection on proteins related to EMT changes. We examined expression of Twist, E-cadherin, β-catenin, vimentin, total-IκB-α, and phospho-IκB-α by Western blot. Decreased Twist, vimentin and phospho-IκB-α, along with increased E-cadherin and β-catenin, were noted in MKK4-transfected TOV-21G and SKOV3 cells. In contrast, over-expression of Twist, vimentin and phospho-IκB-α, along with no expression of E-cadherin and β-catenin, were found in parental counterparts. Increased Twist, vimentin and phospho-IκB-α, as well as decreased E-cadherin and β-catenin, were noted in MKK4-siRNA-transfected MDAH cells, whereas low or no expression of Twist, vimentin and phospho-IκB-α, along with high expression of E-cadherin and β-catenin, were found in the parental counterpart. (b) Representative samples showed the positive correlations among MKK4, phospho-IκBα, Twist and E-cadherin expression. [Color figure can be viewed in the online issue, which is available at]

Download figure to PowerPoint

Next, we analyzed Twist and E-cadherin expression by immunostaining the ovarian carcinoma samples. Of the 54 samples, 41 had sufficient tissue for MKK4 and all other immunostaining procedures. Twist expression was more pronounced in invasive ovarian carcinomas compared with benign ovarian tissues. This finding suggests a negative correlation between MKK4 and Twist in tumor progression. This negative correlation was statistically significant (Y = 2.08–2.37X, N = 54, R = 0.301, p = 0.025, Supporting Information Fig. 4). Therefore, we tested for a correlation between the loss of MKK4 expression and downregulation of phosphorylated NF-κB, upregulation of Twist, and downregulation of E-cadherin in invasive ovarian cancer tissues. Interestingly, 24% (10/41) of samples had findings that matched those obtained in our in vitro experiments. Immunostaining of MKK4, phosphorylated NF-κB, Twist and E-cadherin is summarized in Figure 6b and Supporting Information Table 2.

Ectopic expression of MKK4 had kinase-independent activities in vitro

To confirm that ectopically expressed MKK4 proteins were biochemically functional, serum-starved cells were treated with anisomycin, a known activator of the SAPK pathway, and protein lysates were immunobloted using ant- phospho-MKK4 antibody.

Western blot analysis using phospho-MKK4 was performed after treating anisomycin. However, TOV-21G MKK4 clones did not have phosphorylated form of MKK4 (Supporting Information Fig. 5).


  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Although MKK4 is the most extensively studied MKK, it may have different functions depending on cell context. Initially, MKK4 was considered a candidate tumor suppressor gene.24 Subsequently, MKK4 was shown to be a metastasis suppressor gene in ovarian cancer with a higher expression level in normal ovarian epithelium compared to metastatic ovarian cancer. Transfection of the MKK4 gene into SKOV3 ovarian cancer cells inhibited formation of peritoneal metastases by 90%.5 Recently, we found that about 5% of ovarian cancer cases had a homozygous deletion of MKK4.7 Despite substantial evidence arguing in favor of an antimetastatic function of MKK4, one group recently reported a pro-oncogenic role for MKK4.25 To better understand this gene, we utilized an ovarian cancer cell line containing a homozygous deletion of MKK4. This enabled us to precisely control MKK4 expression artificially to elucidate the underlying mechanism of its action more reliably. Current results from Western blot analysis showing SKOV3 has weak MKK4 expression are consistent with those of previous reports.5 TOV-21G, however, had no detectable MKK4 expression. Over-expression of MKK4 resulted in a marked reduction of invasiveness in vitro and smaller tumors with little peritoneal dissemination in vivo. Reciprocal results were obtained following downregulation of MKK4 expression. The results of our in vitro migration and invasion assay investigating the metastasis suppressor functions of MKK4 are consistent with those of a recent prior report.26

In the current study, we demonstrated that normal ovarian surface epithelium, benign adenomas and ovarian tumors of low malignant potential express moderate to high levels of MKK4. In contrast, the majority of ovarian cancer specimens show significantly decreased expression, particularly if metastatic to the peritoneum. This suggests that MKK4 is downregulated in clinical disease when cells acquire the ability to invade and disseminate. This prompted the hypothesis that MKK4 regulates other mediators responsible for triggering the EMT, thereby blocking acquisition of the ability of the tumor cells to migrate, invade and eventually metastasize.

Tumor progression and invasion are complex biological processes that involve the remodeling of stromal tissue by invading cells. Twist expression activates dormant developmental pathways in invading tumor cells.12 Twist has recently been associated with metastasis in ovarian cancer,27 liver cancer28 and breast cancer.29 Exogenous expression of Twist promotes colony formation in anchorage-independent assays.30 Our current results, showing that constitutive expression of MKK4 leads to suppression of invasion through induction of the EMT, as well as the reciprocal results with gene knockdown, are consistent with these previous studies. Therefore, suppression of invasion-related molecules such as Twist may be one mechanism by which MKK4 suppresses metastasis. In this study, increased levels of MKK4 corresponded to decreased levels of Twist and concomitant decrease in reduction in cell motility and invasiveness in vitro. Conversely, depletion of endogenous MKK4 corresponded in increased levels of Twist and increased cell motility and invasiveness in vitro. Therefore, Twist is an important downstream participant in MKK4 signaling, with MKK4 expression being inversely correlated with Twist level in regulating metastasis in the current model.

Once the primary tumor is established and proliferating, dissolution of cell–cell junctions is the next logical step in initiating EMT-mediated metastasis. In this regard, loss of E-cadherin expression is considered a central event in tumor metastasis.31 Twist plays a crucial role by downregulating E-cadherin, promoting the EMT, and mediating cell motility and invasiveness.32 Loss of tight or adherens junction proteins is essential for EMT downregulation of E-cadherin and has been associated with disruption of epithelial integrity and consequent acquisition of invasiveness in many advanced tumor types. As expected from these observations, TOV-21G and SKOV3 cells transfected with MKK4 were less aggressive, and MDAH 2774 cells in which MKK4 had been knocked down were more aggressive, as evidenced by performance in in vivo and in vitro migration, invasion and mouse xenograft assays. The gain of E-cadherin and the loss of Twist indicate that over-expression of MKK4 in TOV-21G and SKOV3 cells causes the loss of the EMT trait. In this manner, acquisition of increased motility and invasiveness required for metastasis is prevented. In contrast, loss of E-cadherin and gain of Twist indicate that knockdown of MKK4 in MDAH 2774 cells induces the EMT trait, thus facilitating development of increased motility and invasiveness leading to metastasis.

The development of a highly invasive ovarian cancer phenotype requires coordinated upregulation or downregulation of many signaling pathways. A key question is how MKK4 regulates Twist. In recent reports, NF-κB was identified as a central mediator of the EMT in a model of breast cancer progression.33 The E-cadherin repressors Twist and Snail are candidate downstream targets of NF-κB.12 Furthermore, MKK4 has recently been shown to promote cell survival through an NF-κB–dependent pathway.34 NF-κB normally resides in the cytosol in an inactive state, complexed with the inhibitory IκB protein. Activation occurs via phosphorylation of IκBα at Ser32 and Ser36 resulting in the release and nuclear translocation of active NF-κB. Nuclear translocation of NF-κB induces target genes related to oncogenesis, thereby influencing key oncogenic processes including cell proliferation, angiogenesis, metastasis and antiapoptosis.23 As expected from these observations, expression of phosphorylated IκB-α in MKK4-transfected cells was higher than in control vector-transfected cells. This suggests that NF-κB is an important downstream target of MKK4 and acts as a key regulator of the EMT. Finally, we hypothesized a possible association between MKK4, NF-κB, Twist, E-cadherin and the EMT. Our in vitro and in vivo findings together suggest that MKK4 downregulation causes upregulation of phosphorylated NF-κB. This might promote Twist over-expression resulting in E-cadherin downregulation, which induces the EMT in ovarian cancer. In contrast to primary site disease, the possible association among MKK4, NF-κB, Twist, E-cadherin and the MET (mesenchymal epithelial transition) may occur in metastatic disease. Further study is necessary to confirm this possible model of MKK4-related EMT using kinetics assays, direct DNA binding assay and to analyze the MET at metastatic sites using a clinical sample obtained from a metastatic site. In our results, 24% (10/41) of samples had findings that completely matched those obtained in our in vitro experiments. This raises the possibility that the mechanisms of ovarian cancer metastasis are multifactorial. Our current result could be one of the mechanisms underlying ovarian cancer metastasis and other mechanisms may exist.

MKK4 is a kinase and its known activities are via its kinase activity; however, the protein is not generally active in cells that are grown on tissue culture plastic. For this reason, most studies that examine MKK4 signaling use an artificial stimulus such as anisomycin to activate MKK4.26 Therefore, we also analyzed the kinase assay using anisomycin, but activated MKK4 did not show the same result as our in vitro gene transfection/silencing assay. In contrast, nearly 70% of the cases of invasive ovarian cancer lost total MKK4 protein in this study. Taken together, previous reports and current results suggest that MKK4 may have both kinase-dependent and kinase-independent activities.

Angiogenesis is vital to the growth and metastatic dissemination of solid tumors in that it facilitates passage of tumor cells into the circulation. Interestingly, our in vivo mouse study showed that knockdown of MKK4 in MDAH 2774 cells allowed them to produce hemato-ascites when transplanted. This effect was completely prevented by over-expression of MKK4, implicating a factor downstream of MKK4 in tumor angiogenesis. Twist and NF-κB both promote angiogenesis, as recently demonstrated by a breast cancer model using Twist29 and an in vitro model of angiogenesis using NF-κB.35 Our in vivo study demonstrated that transfection of TOV-21G cells with MKK4 completely blocked their ability to form hemato-ascites. Further studies are needed to confirm whether suppression of angiogenesis is another potential mechanism by which MKK4 prevents tumor spread.

This is the first study demonstrating that forced expression of MKK4 in ovarian cancer cells reduces phosphorylation of NF-κB with subsequent downregulation of Twist and over-expression of E-cadherin. Our results show that downregulation of MKK4 expression induces several molecular changes important for the EMT in ovarian cancer. MKK4 may therefore represent a potential new target for preventing tumor invasion and metastasis.


  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information
  • 1
    Wingo PA, Tong T, Bolden S. Cancer statistics, 1995. CA Cancer J Clin 1995; 45: 830.
  • 2
    Robinson VL, Hickson JA, Vander Griend DJ, Dubauskas Z, Rinker-Schaeffer CW. MKK4 and metastasis suppression: a marriage of signal transduction and metastasis research. Clin Exp Metastasis 2003; 20: 2530.
  • 3
    Kim HL, Vander Griend DJ, Yang X, Benson DA, Dubauskas Z, Yoshida BA, Chekmareva MA, Ichikawa Y, Sokoloff MH, Zhan P, Karrison T, Lin A, et al. Mitogen-activated protein kinase kinase 4 metastasis suppressor gene expression is inversely related to histological pattern in advancing human prostatic cancers. Cancer Res 2001; 61: 28337.
  • 4
    Xin W, Yun KJ, Ricci F, Zahurak M, Qiu W, Su GH, Yeo CJ, Hruban RH, Kern SE, Iacobuzio-Donahue CA. MAP2K4/MKK4 expression in pancreatic cancer: genetic validation of immunohistochemistry and relationship to disease course. Clin Cancer Res 2004; 10: 851620.
  • 5
    Yamada SD, Hickson JA, Hrobowski Y, Vander Griend DJ, Benson D, Montag A, Karrison T, Huo D, Rutgers J, Adams S, Rinker-Schaeffer CW. Mitogen-activated protein kinase kinase 4 (MKK4) acts as a metastasis suppressor gene in human ovarian carcinoma. Cancer Res 2002; 62: 671723.
  • 6
    Spillman MA, Lacy J, Murphy SK, Whitaker RS, Grace L, Teaberry V, Marks JR, Berchuck A. Regulation of the metastasis suppressor gene MKK4 in ovarian cancer. Gynecol Oncol 2007; 105: 31220.
  • 7
    Nakayama K, Nakayama N, Davidson B, Katabuchi H, Kurman RJ, Velculescu VE, Shih Ie M, Wang TL. Homozygous deletion of MKK4 in ovarian serous carcinoma. Cancer Biol Ther 2006; 5: 6304.
  • 8
    Kajiyama H, Hosono S, Terauchi M, Shibata K, Ino K, Yamamoto E, Nomura S, Nawa A, Kikkawa F. Twist expression predicts poor clinical outcome of patients with clear cell carcinoma of the ovary. Oncology 2006; 71: 394401.
  • 9
    Hirohashi S. Inactivation of the E-cadherin-mediated cell adhesion system in human cancers. Am J Pathol 1998; 153: 3339.
  • 10
    Chua HL, Bhat-Nakshatri P, Clare SE, Morimiya A, Badve S, Nakshatri H. NF-κB represses E-cadherin expression and enhances epithelial to mesenchymal transition of mammary epithelial cells: potential involvement of ZEB-1 and ZEB-2. Oncogene 2007; 26: 71124.
  • 11
    Peinado H, Portillo F, Cano A. Transcriptional regulation of cadherins during development and carcinogenesis. Int J Dev Biol 2004; 48: 36575.
  • 12
    Kang Y, Massague J. Epithelial-mesenchymal transitions: twist in development and metastasis. Cell 2004; 118: 2779.
  • 13
    Bates RC, Mercurio AM. The epithelial-mesenchymal transition (EMT) and colorectal cancer progression. Cancer Biol Ther 2005; 4: 36570.
  • 14
    Yamamoto K, Kikuchi Y, Kudoh K, Nagata I. Modulation of cisplatin sensitivity by taxol in cisplatin-sensitive and -resistant human ovarian carcinoma cell lines. J Cancer Res Clin Oncol 2000; 126: 16872.
  • 15
    Wang TL, Diaz LA, Jr, Romans K, Bardelli A, Saha S, Galizia G, Choti M, Donehower R, Parmigiani G, Shih Ie M, Iacobuzio-Donahue C, Kinzler KW, et al. Digital karyotyping identifies thymidylate synthase amplification as a mechanism of resistance to 5-fluorouracil in metastatic colorectal cancer patients. Proc Natl Acad Sci USA 2004; 101: 308994.
  • 16
    UKCCCR. United Kingdom Co-ordinating Committee on Cancer Research (UKCCCR) Guidelines for the Welfare of Animals in Experimental Neoplasia (Second Edition). Br J Cancer 1998; 77: 110.
  • 17
    Cox MC, Maffei L, Buffolino S, Del Poeta G, Venditti A, Cantonetti M, Aronica G, Aquilina P, Masi M, Amadori S. A comparative analysis of FISH, RT-PCR, and cytogenetics for the diagnosis of bcr-abl-positive leukemias. Am J Clin Pathol 1998; 109: 2431.
  • 18
    Pauletti G, Godolphin W, Press MF, Slamon DJ. Detection and quantitation of HER-2/neu gene amplification in human breast cancer archival material using fluorescence in situ hybridization. Oncogene 1996; 13: 6372.
  • 19
    Thiery JP. Epithelial-mesenchymal transitions in development and pathologies. Curr Opin Cell Biol 2003; 15: 7406.
  • 20
    Hay ED. An overview of epithelio-mesenchymal transformation. Acta Anat (Basel) 1995; 154: 820.
  • 21
    Thiery JP. Epithelial-mesenchymal transitions in tumour progression. Nat Rev Cancer 2002; 2: 44254.
  • 22
    Yang J, Mani SA, Donaher JL, Ramaswamy S, Itzykson RA, Come C, Savagner P, Gitelman I, Richardson A, Weinberg RA. Twist, a master regulator of morphogenesis, plays an essential role in tumor metastasis. Cell 2004; 117: 92739.
  • 23
    Mabuchi S, Ohmichi M, Nishio Y, Hayasaka T, Kimura A, Ohta T, Kawagoe J, Takahashi K, Yada-Hashimoto N, Seino-Noda H, Sakata M, Motoyama T, et al. Inhibition of inhibitor of nuclear factor-κB phosphorylation increases the efficacy of paclitaxel in in vitro and in vivo ovarian cancer models. Clin Cancer Res 2004; 10: 764554.
  • 24
    Teng DH, Perry WL, III, Hogan JK, Baumgard M, Bell R, Berry S, Davis T, Frank D, Frye C, Hattier T, Hu R, Jammulapati S, et al. Human mitogen-activated protein kinase kinase 4 as a candidate tumor suppressor. Cancer Res 1997; 57: 417782.
  • 25
    Wang L, Pan Y, Dai JL. Evidence of MKK4 pro-oncogenic activity in breast and pancreatic tumors. Oncogene 2004; 23: 597885.
  • 26
    Hickson JA, Huo D, Vander Griend DJ, Lin A, Rinker-Schaeffer CW, Yamada SD. The p38 kinases MKK4 and MKK6 suppress metastatic colonization in human ovarian carcinoma. Cancer Res 2006; 66: 226470.
  • 27
    Terauchi M, Kajiyama H, Yamashita M, Kato M, Tsukamoto H, Umezu T, Hosono S, Yamamoto E, Shibata K, Ino K, Nawa A, Nagasaka T, et al. Possible involvement of TWIST in enhanced peritoneal metastasis of epithelial ovarian carcinoma. Clin Exp Metastasis 2007; 24: 32939.
  • 28
    Niu RF, Zhang L, Xi GM, Wei XY, Yang Y, Shi YR, Hao XS. Up-regulation of Twist induces angiogenesis and correlates with metastasis in hepatocellular carcinoma. J Exp Clin Cancer Res 2007; 26: 38594.
  • 29
    Mironchik Y, Winnard PT, Jr, Vesuna F, Kato Y, Wildes F, Pathak AP, Kominsky S, Artemov D, Bhujwalla Z, Van Diest P, Burger H, Glackin C, et al. Twist overexpression induces in vivo angiogenesis and correlates with chromosomal instability in breast cancer. Cancer Res 2005; 65: 108019.
  • 30
    Maestro R, Dei Tos AP, Hamamori Y, Krasnokutsky S, Sartorelli V, Kedes L, Doglioni C, Beach DH, Hannon GJ. Twist is a potential oncogene that inhibits apoptosis. Genes Dev 1999; 13: 220717.
  • 31
    Shiozaki H, Oka H, Inoue M, Tamura S, Monden M. E-cadherin mediated adhesion system in cancer cells. Cancer 1996; 77: 160513.
  • 32
    Kwok WK, Ling MT, Lee TW, Lau TC, Zhou C, Zhang X, Chua CW, Chan KW, Chan FL, Glackin C, Wong YC, Wang X. Up-regulation of TWIST in prostate cancer and its implication as a therapeutic target. Cancer Res 2005; 65: 515362.
  • 33
    Huber MA, Azoitei N, Baumann B, Grunert S, Sommer A, Pehamberger H, Kraut N, Beug H, Wirth T. NF-κB is essential for epithelial-mesenchymal transition and metastasis in a model of breast cancer progression. J Clin Invest 2004; 114: 56981.
  • 34
    Xia D, Srinivas H, Ahn YH, Sethi G, Sheng X, Yung WK, Xia Q, Chiao PJ, Kim H, Brown PH, Wistuba II, Aggarwal BB, et al. Mitogen-activated protein kinase kinase-4 promotes cell survival by decreasing PTEN4 expression through an NF κB-dependent pathway. J Biol Chem 2007; 282: 350719.
  • 35
    Yang LH, Wang KF, Zhou YP, Li ML. [Role and mechanism of nuclear factor κ B in angiogenesis of human ovarian carcinoma]. Ai Zheng 2004; 23: 5314.

Supporting Information

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Additional Supporting Information may be found in the online version of this article.

IJC_25332_sm_suppfig1.tif131KSupporting Figure 1
IJC_25332_sm_suppfig2.tif254KSupporting Figure 2
IJC_25332_sm_suppfig2b-c.tif62KSupporting Figure 2 B–C
IJC_25332_sm_suppfig3.tif759KSupporting Figure 3
IJC_25332_sm_suppfig4.tif51KSupporting Figure 4
IJC_25332_sm_suppfig5.tif119KSupporting Figure 5
IJC_25332_sm_SuppTable-1.tif50KSupporting Table 1
IJC_25332_sm_SuppTable-2.tif48KSupporting Table 2

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.