The potassium chloride cotransporter (KCC) is a major determinant of osmotic homeostasis and plays an emerging role in tumor biology. Here, we investigate if KCC is involved in the regulation of epithelial-mesenchymal transition (EMT), a critical cellular event of malignancy. E-cadherin and β-catenin colocalize in the cell-cell junctions, which becomes more obvious in a time-dependent manner by blockade of KCC activity in cervical cancer SiHa and CaSki cells. Real-time reverse transcription-PCR on the samples collected from the laser microdissection indicates that KCC3 is the most abundant KCC isoform in cervical carcinoma. The characteristics of EMT appear in KCC3-overexpressed, but not in KCC1- or KCC4-overexpressed cervical cancer cells, including the elongated cell shape, increased scattering, down-regulated epithelial markers (E-cadherin and β-catenin), and up-regulated mesenchymal marker (vimentin). Some cellular functions are enhanced by KCC3 overexpression, such as increased invasiveness and proliferation, and weakened cell-cell association. KCC3 overexpression decreases mRNA level of E-cadherin. The promoter activity assays of various regulatory sequences confirm that KCC3 expression is a potent negative regulator for human E-cadherin gene expression. The proteosome inhibitor restores the decreased protein abundance of β-catenin by KCC3 overexpression. In the surgical specimens of cervical carcinoma, the decreased E-cadherin amount was accompanied by the increased KCC3 abundance. Vimentin begins to appear at the invasive front and becomes significantly expressed in the tumor nest. In conclusion, KCC3 down-regulates E-cadherin/β-catenin complex formation by inhibiting transcription of E-cadherin gene and accelerating proteosome-dependent degradation of β-catenin protein. The disruption of E-cadherin/β-catenin complex formation promotes EMT, thereby stimulating tumor progression. [Cancer Res 2007;67(22):11064–73]
- KCl cotransporter
- epithelial-mesenchymal transition
The activity of potassium chloride cotransporter (KCC) plays an important role in several cellular functions, such as cell volume regulation, epithelial ion transport, and osmotic homeostasis ( 1). Functionally, KCC is defined as the Cl−-dependent bidirectional K+ transport measured in the presence of ouabain and bumetanide to inhibit Na+-K+ pump and Na+K+2Cl− cotransporter, respectively ( 2). The KCC family includes four isoforms (KCC1-KCC4), which differ at amino acid residues within key transmembrane domains and in the distribution of putative phosphorylation sites within the amino- and carboxyl-terminal cytoplasmic domains. The activities of KCC1, KCC3, and KCC4 are osmotically sensitive and involved in cell volume regulation ( 3). The neuron-specific KCC2 is critical for the maturation of inhibitory γ-aminobutyric acid responses in the central nervous system by the control of intracellular Cl− concentration ( 4, 5). KCC activity in RBC was undiminished in KCC1 knockout mice, decreased in KCC3 knockout mice, and almost completely abolished in mice lacking both isoforms, indicating that KCC activity of mouse RBC is mediated largely by KCC3 ( 6). Loss of KCC3 caused deafness, neurodegeneration, and reduced seizure threshold ( 7). KCC3 also plays an important role in the regulation of cell proliferation ( 8). Deafness and renal tubular acidosis were noted in mice lacking KCC4 ( 9).
The important role of KCC in tumor behaviors emerged from our previous studies. The malignant transformation of cervical epithelial cells is associated with the differential expression of volume-sensitive KCC activities ( 10). Cervical carcinogenesis is also accompanied by the up-regulation of mRNA transcripts in KCC1, KCC3, and KCC4 ( 11). The loss-of-function KCC mutant cervical cancer cells exhibit inhibited cell growth accompanied by decreased activity of the cell cycle gene products retinoblastoma and cdc2 kinase. Reduced cellular invasiveness in loss-of-function KCC mutant cervical cancer cells is in parallel, by reduced expression of αvβ3 and α6β4 integrins, accompanied by decreased activity of matrix metalloproteinase 2 and 9. Inhibition of tumor growth in severe combined immunodeficient mice confirms the crucial role of KCC in promoting cervical cancer growth and invasion ( 10). In addition, KCC activation by insulin-like growth factor-I (IGF-I) stimulation plays an important role in IGF-I signaling to promote the growth and spread of gynecologic cancers ( 12). Furthermore, IGF-I up-regulates KCl cotransporter KCC3 and KCC4, which are differentially required for breast cancer cell proliferation and invasiveness ( 13). Hence, KCC may aid the invasive biology of cancer cells, a feature that poses a serious problem to the successful treatment of neoplastic disease.
Studying on the conversion of epithelial cells to mesenchymal cells [epithelial-mesenchymal transition (EMT)] in different developmental and pathologic stages provides a better understanding of tumor biology. Some cellular functions of tumor would be enhanced by EMT, such as increased invasive migration and resistance to anoikis ( 14, 15). EMT has been characterized as a hallmark for cancer metastasis in several types of cancer, including colorectal cancer ( 16), breast cancer ( 17), and lung cancer ( 18, 19). The alterations in the abundance of several molecules are used to detect the events of EMT, for example, the decreased abundances of epithelial markers such as E-cadherin and β-catenin, and the increased abundances of mesenchymal markers such as vimentin and fibronectin ( 20, 21).
Little information is available on the roles of cellular ion transport systems in the regulation of EMT to promote tumor progression. The literature on the important role of EMT in cervical carcinogenesis is scarce. Because EMT has been characterized as a hallmark of cellular event for cancer invasion and metastasis, we investigate if KCC activity is involved in the regulation of EMT by the model of cervical carcinoma. The data indicate that KCC3 down-regulates E-cadherin/β-catenin complex formation by inhibiting transcription of E-cadherin gene and accelerating degradation of β-catenin protein.
Materials and Methods
Cell culture and transfection. Cultures of cervical cancer SiHa and CaSki cell lines were prepared as previously described ( 11). The full-length cDNA of human KCC1 ( 22), human KCC3 ( 8), and mouse KCC4 ( 22) construct were subcloned into the eukaryotic expression vector pCDNA3.1 (Invitrogen). Plasmids were transfected into SiHa cells by LipofectAMINE (Invitrogen) and stable clones were selected.
Reverse transcription-PCR. Total RNAs were isolated by RNeasy Mini Kit (Qiagen) and cDNA was prepared with Superscript II (Invitrogen). For E-cadherin, forward (5′-GGGTGACTACAAAATCAATC-3′) and reverse (5′-GGGGGCAGTAAGGGCTCTTT-3′) primers were used to amplify a 251-bp fragment from human E-cadherin. For β-catenin, forward (5′-ACTCTAGGAATGAAGGTGTGGC-3′) and reverse (5′-AGTGTGTCAGGCACTTTCTGAG-3′) primers were used to amplify a 822-bp fragment from human β-catenin. For KCC1, forward (5′-TGGGACCATTTTCCTGACC-3′) and reverse (5′-CATGCTTCTCCAC GATGTCAC-3′) primers were used to amplify a 421-bp fragment from human KCC1. For KCC3, forward (5′-CTATCCTTGCCATCCTGACC-3′) and reverse (5′-GCAGCAGTTGTCACTCGAAC-3′) primers were used to amplify a 1,099-bp fragment from human KCC3. For KCC4, forward (5′-GACTCGTTTCCGC AAAACC-3′) and reverse (5′-AGAGTGCCGTGATGCTGTTGG-3′) primers were used to amplify a 783-bp fragment from human KCC4. PCR was carried out as described in detail elsewhere ( 11).
Surgical specimens, laser microdissection, and real-time reverse transcription-PCR. From January 2006 to April 2006, 36 consecutive patients with cervical carcinoma of stage Ib to IIa were scheduled for radical hysterectomy and pelvic lymphadenectomy at National Cheng Kung University Hospital, Taiwan. Patients who had undergone the loop electrosurgical excision for the transformation zone of uterine cervix before radical hysterectomy and who had unusual histopathology such as clear cell adenocarcinoma, adenosquamous carcinoma, and small-cell carcinoma were excluded from this study. The tumor volume of cervical carcinoma was measured as previously described ( 23). The tumor and normal tissues were collected from patients immediately after surgical removal. The frozen tissue specimens were serially cut into pieces of 8 μm. The sections were stained with H&E according to NIH laser capture microdissection protocols, 8 and was subjected to laser microdissection by Leica AS LMD system (Leica Microsystems). Five thousand laser pulses per specimen were used to collect ∼10,000 morphologically normal epithelial cells and primary carcinoma cells from each case. Total RNA of these cells was extracted by the PicoPure RNA Isolation Kit (Arcturus) according to the manufacturer's instructions, and cDNA was prepared with Superscript II (Invitrogen). The RNA levels of individual KCC isoforms in the surgical specimens were examined by real-time quantitative reverse transcription-PCR (RT-PCR) with the use of ABI Prism 7900 Analyzer and TaqMan probes (Applied Biosystems). cDNA was normalized against glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Amplification primers and probes for real-time RT-PCR are listed as follows. For KCC1, forward GTGATCATCTCCATCCTCTCCATCT; reverse GTTGCC CAGCATGCATACC; probe CCCTCCCGTGTTTCC. For KCC3, forward ACATTGACGTTTGCTCTAAGACCAA; reverse ACTCGAGTTACAG AAGAATCCCCATA; probe CTGTCATGTTGTTAATTTC. For KCC4, forward TGCGCCTGACGTGGAT; reverse GGCCACGATGAGGAAGGA; probe CCAGGACACCAGCCACC.
Functional K+ (86Rb+) efflux assays. By using 86Rb+ as a congener of K+, unidirectional K+ efflux was carried out at 37°C as described in detail elsewhere ( 8). KCC activity is defined as the Cl−-dependent K+ transport measured in the presence of 0.1 mmol/L ouabain and 0.01 mmol/L bumetanide to inhibit Na+-K+ pump and Na+K+2Cl− cotransporter, respectively ( 2). Release of 86Rb+ from preloaded cells was measured in the efflux medium at indicated time intervals within a 15-min duration. 86Rb+ efflux rate constants were estimated from the negative slope of the graph of ln[Xi(t)/Xi(t = 0)] versus time (t), where Xi(t = 0) denotes the total amount of 86Rb+ inside the cells at the beginning of the efflux time course and Xi(t) denotes the amount of 86Rb+ inside the cells at the time point of t. The isotonic efflux medium contained (in mmol/L) NaCl 110, KCl 5, MgCl2 1, CaCl2 1.5, glucose 10, HEPES 10 and mannitol 60, titrated to pH 7.4 with NaOH (300 ± 3 mOsm/L). The components of the hypotonic medium (240 ± 2 mOsm/L) are the same as those of the isotonic medium except that mannitol was omitted. To study the KCC activity, the Cl− dependence of K+ (86Rb+) efflux was examined by substituting NO3− for Cl− in the efflux medium. The Cl−-dependent K+ (86Rb+) flux was defined as the efflux difference between Cl− and NO3− media.
Measurements of cell volume. Cell volume was measured as described in detail elsewhere ( 24). The relative volume change (V/Vo) was calculated from the cross-sectional surface area at the beginning (So) of experiment and during (S) the experiments from the relation V/Vo = (S/So)3/2 ( 10, 24). Data were presented as the percentage of starting volume (V/Vo), as a function of time. Fluorescence-activated cell sorting was also used to measure cell size. Cells were suspended in the isotonic solution containing 20 μg/mL propidium iodide. A flow cytometer (Becton Dickinson FACSCalibur System) was used to measure the light-scattering properties of the cells. An argon laser (488 nm) was used as the probing beam, with red light emission being used to exclude the dead, propidium iodide–stained cells during later analysis. Because forward-angle light scatter originates in particles with diameters that are larger than the wavelength of the probing light, the forward-angle light scatter serves as an indirect measure of overall cell size. Forward-angle light scatter distribution histograms for viable cells were analyzed using Cell Quest software.
Proliferation, migration, and invasion assay. To assess proliferation, cells were plated at the density of 1 × 105 per dish on 60-mm dishes and the medium was changed every 2 days. Cell counts were done with the aid of a hemocytometer by using trypan blue exclusion (0.08%) to monitor cell viability. Invasive migration was done in the BD Matrigel invasion chamber (BD Biosciences) for 6 h in serum-free culture medium at 37°C, as an index of invasive activity of tumor cells ( 25). Cells were then fixed with paraformaldehyde, stained with crystal violet, and counted immediately after staining. For the migration assays ( 26), cells were allowed to migrate across a membrane (with 8-μm pore) toward the medium containing 10 μg/mL fibronectin or vitronectin for 6 h at 37°C. Cells that migrated through the membrane were then fixed with methanol, stained with GIEMSA stain, and counted immediately.
Cell dissociation assay. Cell dissociation assay was done following a previously published method ( 27). Briefly, cells were detached from culture plates with a rubber policeman and passed through glass Pasteur pipettes 10 times. The cells were then fixed in 1% glutaraldehyde and pictures were taken under an Olympus IX70 inverted microscope equipped with a CoolSnap-Pro color digital camera (Roper Scientific). The extent of cell dissociation was represented by the ratio of the number of particles (Np) over the number of cells (Nc; Np/Nc). The values of Np/Nc were calculated by analyzing at least 150 cells from each sample.
E-cadherin promoter reporter assay. Serial deletions of the regulatory sequences of the human E-cadherin gene were constructed and attached with pGL3 luciferase reporter plasmid ( 28). These constructs or control vector were transfected into wild-type or KCC3-overexpressed SiHa cells. The cells were transfected in six-well plastic dishes using LipofectAMINE 2000 (Invitrogen). Reporter assays were determined using the Dual-Luciferase Reporter Assay System (Promega). At 48 h after transfection, the cells were lysed and the soluble fraction was used for luciferase assays following the manufacturer's instructions. The luciferase activities were expressed as folds over that of the control vector.
Immunoblotting, immunofluorescence, and immunoprecipitation. For Western immunoblotting, cells were harvested with ice-cold protein lysis solution containing a protease inhibitor mixture (Roche Diagnostics), 100 mmol/L KCl, 80 mmol/L NaF, 10 mmol/L EGTA, 50 mmol/L β-glycerophosphate, 10 mmol/L p-nitrophenyl phosphate, 1 mmol/L vanadate, 0.5% sodium deoxycholate, and 1% NP40. Protein concentrations were determined with the use of a Bio-Rad protein assay. Equal protein loads (50 μg in each column) were separated by 7.5% SDS-PAGE, and then transferred to polyvinylidene difluoride membranes (Stratagene). The primary antibodies used were monoclonal anti–E-cadherin (BD Transduction Laboratories, 1:3,000 for immunoblotting; 1:150 for immunofluorescence), monoclonal anti-tubulin (BD Transduction Laboratories, 1:5,000 for immunoblotting), polyclonal anti–β-catenin (Cell Signaling Technology, 1:3,000 for immunoblotting; 1:150 for immunofluorescence), polyclonal anti–α-catenin (Santa Cruz Biotechnology, 1:5,000 for immunoblotting), polyclonal anti–γ-catenin (Santa Cruz Biotechnology, 1:5,000 for immunoblotting), monoclonal anti-vimentin (BD Transduction Laboratories, 1:150 for immunofluorescence), and monoclonal anti–α-actin (BD Transduction Laboratories, 1:5,000 for immunoblotting). KCC3 antibody was kindly provided by Dr. David B. Mount (Harvard Medical School, Boston, MA). The horseradish peroxidase–conjugated secondary antibody against mouse source (BD Transduction Laboratories, 1:5,000 for immunoblotting), rabbit source (BD Transduction Laboratories, 1:5,000 for immunoblotting), or goat source (Santa Cruz Biotechnology, 1:2,000 for immunoblotting) was purchased. The antibodies against KCC1 (1:1,000 for immunoblotting) and KCC4 (1:1000 for immunoblotting) were purchased from Santa Cruz Biotechnology and Chemicon, respectively. For immunofluorescent staining, the fixed cells were permeabilized with 0.05% Triton X-100 in PBS at room temperature for 20 min. These cells were blocked with commercial blocking serum (Vector Laboratories) and incubated with primary antibody at 4°C overnight, washed, and then incubated with fluorescence-conjugated secondary antibody for 1 h at room temperature. After that, cells were washed and mounted, and examined by a scanning confocal microscope (FV-1000, Olympus). For immunoprecipitation, agarose-conjugated antibody was prepared by adding 50 μL of protein G-agarose bead (Amersham Biosciences) slurry to 3 μg of antibody. Immunoprecipitation was done by adding cell lysates (1 mg/sample) to the agarose-conjugated antibody complex and incubating at 4°C for 1 h to overnight on a rotator. After extensive washing with radioimmunoprecipitation assay lysis buffer, samples were resuspended in reducing sample buffer, boiled for 3 min, centrifuged to pellet the agarose beads, and subjected to immunoblotting.
Statistics. All values were reported as mean ± SE. Student's paired or unpaired t test was used for statistical analyses. Differences between values were considered significant when P < 0.05.
KCC activity regulates E-cadherin/β-catenin complex formation in cell-cell junctions. First, we investigated whether KCC activity is involved in the regulation of EMT. In cervical cancer SiHa cells, KCC activity was quiescent in the isotonic solution and significantly increased under hypotonic conditions ( Fig. 1A ). The hypotonicity-activated KCC activity was almost abolished by 50 μmol/L [(dihydroindenyl)oxy]alkanoic acid (DIOA). N-ethylmaleimide potently activated KCC activity, which was also inhibited by 50 μmol/L DIOA. These data indicate that DIOA is a potent KCC inhibitor. Blockade of KCC activity by 50 μmol/L DIOA increased both E-cadherin and β-catenin abundances in a time-dependent manner ( Fig. 1B). The immunofluorescent images of two different cervical cancer cell lines show that E-cadherin and β-catenin colocalized in the cell-cell junctions, which became more obvious upon the inhibition of KCC activity ( Fig. 1C and D).
KCC3 is the most abundant KCC isoform in tumor tissues. To address the roles of individual KCC isoforms in tumor biology, we analyzed the expression levels of KCC isoforms in the surgical specimens of cervical carcinoma. The laser microdissection with microscopic observation was used to precisely sample the targeted tissues ( Fig. 2A ). Compared with normal or noncancerous squamous epithelia, mRNA levels of KCC3 and KCC4 were significantly increased 25-fold and 14-fold in the primary cervical carcinoma (n = 12), respectively ( Fig. 2A). In addition, the expression level of KCC3 mRNA in tumor tissues was closely correlated with tumor size (Supplementary Fig. S1; R2 = 0.86), an important indicator of human cervical carcinoma progression in vivo ( 23). This is consistent with our previous study showing that KCC3 may play an important role in cell growth regulation ( 8).
Increased KCC3 abundance induces EMT. To study KCC-mediated tumor behavior, we established various clones of cervical cancer SiHa cell lines with KCC1, KCC3, or KCC4 overexpression. The parental SiHa cells exhibited well-organized cell-cell association and islet-like structure, which are the characteristics of squamous cell carcinoma ( Fig. 2B). Overexpression of KCC1 or KCC4 did not change cell morphology or induce cell scattering. In striking contrast, KCC3 overexpression was accompanied by the elongation of cell shape and increased scattering, similar to the morphology of mesenchymal cells ( Fig. 2B). To detect if EMT occurs in KCC3-overexpressed cells, we studied the possible alterations of molecular markers. As shown in Fig. 2C, KCC3 overexpression decreased the abundance of epithelial markers such as E-cadherin and β-catenin in a cell density–independent manner. The abundance of vimentin, a mesenchymal marker, was simultaneously increased ( Fig. 2D). These results indicate that KCC3 overexpression is accompanied by the loss of epithelial markers and the gain of mesenchymal marker.
Increased KCC3 abundance enhances cellular functions. We then tested if the cellular functions of cancer cells benefit from KCC3 overexpression. The immunoblots showed that KCC3 overexpression did not change the protein levels of other KCC isoforms in cervical cancer SiHa cells (Supplementary Fig. S2A). There is no significant difference in cell size between parental wild-type and KCC3-overexpressed cells (Supplementary Fig. S2B). As expected, KCC3-overexpressed cells exhibited a better capability of regulatory volume decrease than did parental wild-type cells (Supplementary Fig. S2C). In addition, the proliferation (Supplementary Fig. S2D), migration ( Fig. 3A ), and invasiveness ( Fig. 3B) of cervical cancer SiHa cells were markedly enhanced by KCC3 overexpression.
E-cadherin and β-catenin could form a complex in cell-cell junctional area, which provides the strength for cell-cell association ( 29). The abundances of E-cadherin and β-catenin were decreased by KCC3 overexpression ( Fig. 2C and D). We then tested whether the physiologic functions of both proteins are affected by KCC3 overexpression. The cell lysate was immunoprecipitated by E-cadherin antibody and then immunoblotted by E-cadherin or β-catenin antibody, respectively, to analyze E-cadherin/β-catenin complex formation. The protein amount of E-cadherin in KCC3-overexpressed SiHa cells was significantly decreased ( Fig. 3C, right). The abundance of E-cadherin–associated β-catenin was remarkably decreased in KCC3-overexpressed SiHa cells ( Fig. 3C, left), indicating that E-cadherin/β-catenin complex formation was reduced. Consistent with the disorganization of β-catenin and E-cadherin, KCC3 overexpression weakened the cell-cell association ( Fig. 3D).
Mechanisms controlling E-cadherin and β-catenin expression in KCC3-overexpressed cells. As shown in Fig. 4A , KCC3 overexpression decreased mRNA expression level of E-cadherin but did not affect that of β-catenin, compared with wild-type cells. The promoter activity of E-cadherin gene was further examined. We have previously studied the structure and functions of human E-cadherin gene regulatory sequences ( 28). Serial deletions of the regulatory sequences of human E-cadherin gene were constructed and attached with pGL3 luciferase reporter plasmid ( Fig. 4B, left). These constructs and control vector were transfected into wild-type or KCC3-overexpressed SiHa cells. The luciferase activities were measured and expressed as folds over that of the control vector. These different constructs displayed differential reporter activities in wild-type SiHa cells, which were similar to our previous study in breast cancer cell lines ( 28). The reporter activities of these constructs were all significantly decreased in KCC3-overexpressed cells ( Fig. 4B, right). This implies that KCC3 expression is likely a potent negative regulation for human E-cadherin gene expression.
Recent studies have shown that β-catenin was translocated into nucleus or degraded through proteosome-mediated pathway while EMT happened ( 29). Because there was no obvious nuclear translocation of β-catenin in KCC3-overexpressed SiHa cells ( Fig. 2D), we studied if KCC3 affects the proteosome-dependent degradation pathway. The wild-type and KCC3-overexpressed cells displayed the similar protein abundances of β-catenin after the treatment of 26S-proteosome–specific inhibitor MG132 for 12 h ( Fig. 4C). On the other hand, the decreased abundance of E-cadherin in KCC3-overexpressed SiHa cells was not affected by treatment of MG132. The protein levels of β-catenin and E-cadherin were not affected by the treatment of NH4Cl, an inhibitor of lysosome-dependent protein degradation pathway either in wild-type or KCC3-overexpressed SiHa cells ( Fig. 4D). These data indicate that KCC3 down-regulates E-cadherin/β-catenin complex formation by inhibiting transcription of E-cadherin gene and accelerating β-catenin protein degradation through the proteosome-dependent pathway.
Association between KCC3 abundance and E-cadherin/β-catenin complex in vivo. Studies in cell culture system have revealed that KCC3 plays an important role in the regulation of EMT. To evaluate the in vivo condition, we examined the association of KCC3 expression and EMT in the surgical specimens of cervical carcinoma by immunofluorescent staining of E-cadherin, vimentin, and KCC3. In the normal or noncancerous cervix, vimentin was only present in stromal tissues ( Fig. 5A, top ). KCC3 expression was weak in both stromal and normal/noncancerous squamous epithelial tissues. In striking contrast, KCC3 protein was abundant in cervical carcinoma. Vimentin appeared at the invasive front of cervical carcinoma, an area where squamous cell carcinoma just broke through the basal layers of squamous epithelia ( Fig. 5A, middle). Tumor nest was formed when cervical carcinoma invaded deeply into stromal tissues, where KCC3 protein was abundant and vimentin was significantly expressed at the periphery of tumor nest ( Fig. 5A, bottom).
E-cadherin protein was obvious in the normal or noncancerous cervical squamous epithelial tissues but was absent in the underlying stromal tissues ( Fig. 5B, top). The loss of cell integrity was noted in the adjacent cervical carcinoma tissues, where the decreased abundance of E-cadherin was accompanied by the increased expression of KCC3 protein ( Fig. 5B, bottom). To confirm if there is a cause and effect, KCC3-overexpressed SiHa cells were treated with KCC inhibitor DIOA and then assayed the expression of E-cadherin and β-catenin. As shown in Fig. 5C, the decrease in both E-cadherin and β-catenin in KCC3-overexpressed cells was rescued by DIOA treatment. These results suggested that the association between KCC3 abundance and E-cadherin/β-catenin complex exists both in vitro and in vivo.
EMT is a key event occurring in the normal development and pathologic processes during which epithelial cells assume a mesenchymal phenotype. EMT may play a gatekeeper role and regulate the frequency of cancer invasiveness and metastasis ( 30). Although multiple extracellular stimuli and transcriptional regulators can trigger EMT, this study is the first report to show that a cellular ion transport system, KCC3, regulates EMT to promote tumor progression. Here, we show the following findings: (a) The characteristics of EMT appear in the KCC3-overexpressed tumor cells, including the elongation of cell shape, increased scattering, down-regulation of epithelial markers (E-cadherin and β-catenin), and up-regulation of a mesenchymal marker (vimentin). (b) Blockade of KCC activity in wild-type and KCC3-overexpressed tumor cells is accompanied by the increased abundances of E-cadherin and β-catenin in cell-cell junctions. The cellular KCC activity is mediated largely from KCC3 ( 3, 6, 31). Therefore, we suggest that there is a tight link between the expression and activity of KCC3 and EMT in cervical cancer cells.
E-cadherin plays a critical role in establishing cell polarity and in maintaining normal tissue morphology ( 32, 33). Down-regulation or loss of E-cadherin has been implicated in the gain of invasive potential by carcinomas ( 34– 37). β-Catenin also functions as a component of adherent cell-cell junctions that promote cell adhesion by binding to the intracellular domain of E-cadherin and linking E-cadherin to the cytoskeleton through the adaptor protein α-catenin ( 38, 39). Several studies have shown that posttranscriptional regulations, including nuclear translocalization and protein degradation, are mainly responsible for regulation of β-catenin while EMT happens ( 29, 37, 40). The E-cadherin/β-catenin complex therefore provides the major strength for cell-cell association ( 29). It has been reported that loss of this complex formation would promote the invasive ability of breast cancer cells due to the decreased cell-cell association ( 16). Here, we show how KCC3 orchestrates the E-cadherin/β-catenin complex that facilitates EMT. KCC3 overexpression decreases mRNA expression level of E-cadherin, compared with that of wild-type tumor cells. We have previously cloned a 1.2-kb DNA fragment upstream from the transcription initiation site of the human E-cadherin gene to understand how human E-cadherin gene expression is regulated ( 28). Several transcription factor–binding sites were identified, for example, three AML1-binding sites, four HNF3-binding sites, and four snail-binding sites. By using this system of serial deletions of the regulatory sequences of E-cadherin gene, KCC3 expression down-regulates the reporter activities of these constructs. Because KCC3 expression does not change the cell size and control vector has been used as a negative control, the results of luciferase assays are the promoter-specific effects rather than artifacts of cell shrinkage. We therefore conclude that KCC3 expression is likely a potent negative regulation for human E-cadherin gene expression. However, neither proteosome inhibitor nor lysosome inhibitor change the protein abundance of E-cadherin in wild-type and KCC3-overexpressed tumor cells. On the other hand, KCC3 overexpression decreases β-catenin abundance by posttranslational modification. The proteosome inhibitor, but not the lysosome inhibitor, restores the decreased protein abundance of β-catenin by KCC3 overexpression. Thus, KCC3 down-regulates E-cadherin/β-catenin complex formation by inhibiting transcription of E-cadherin gene and accelerating proteosome-dependent degradation of β-catenin protein.
KCC3 is the most abundant KCC isoform in tumor tissues, detected by laser microdissection. KCC3 overexpression enhances cervical cancer cell proliferation in vitro. In the surgical specimens, the tumor size of cervical carcinoma, an indicator of tumor growth in vivo, correlates well with KCC3 expression level in tumor tissues. This is consistent with our previous work demonstrating a close link between the expression and activity of KCC3 and cell growth by a NIH/3T3 fibroblast expression system ( 8). In the NIH/3T3 fibroblast expression system, KCC3 activity is important for cell cycle progression, thereby regulating cell growth. Our study also shows that KCC3 overexpression is accompanied by the enhanced capability of regulatory volume decrease, increased cellular invasive migration, and decreased cell-cell association. We previously showed that the invasive migration and proliferation of cancer cells require shape and/or volume changes ( 10). KCC3 plays a dominant role in cell volume regulation ( 7) and may contribute to such shape-volume changes by mediating net salt fluxes across cell membranes ( 11). These data suggest that up-regulation of KCC3 expression in the tumor tissues enhances the malignant behaviors of tumor tissues.
The literature on the important role of EMT in human cervical carcinogenesis is scarce. Here, we show the association between KCC3 abundance and EMT in cervical carcinoma. Studies in cervical cancer cell lines indicate that KCC3 plays an important role in the regulation of EMT. To evaluate the in vivo condition, we examined the association of KCC3 expression and EMT molecular markers in the surgical specimens of cervical carcinoma. E-cadherin protein was abundant in the normal or noncancerous squamous epithelia of the portio of the cervix, where KCC3 protein level is scanty. The loss of cell integrity was noted in the adjacent cervical carcinoma tissues, where the decreased abundance of E-cadherin was accompanied by the increased expression of KCC3 protein. Vimentin, the mesenchymal marker, was only present in stromal tissues. Interestingly, it appeared at the invasive front of cervical carcinoma, an area where squamous cell carcinoma just invaded the basal layers of squamous epithelia. The abundance of vimentin became better visible at the periphery of tumor nest, an area where cervical carcinoma invaded deeply into stromal tissues of tumor nest. Moreover, the decrease in both E-cadherin and β-catenin in KCC3-overexpressed cells was rescued by KCC3 inhibitor, DIOA. This suggests that there is a cause and effect between KCC3 expression and EMT.
The KCC3 activity and expression are sensitive to the stimulation of oncogenic growth factors such IGF-I and vascular endothelial growth factor (VEGF) in cervical cancer cells, ovarian cancer cells, breast cancer cells, and endothelial cells ( 12, 13, 41). This and our previous study ( 8) suggest that KCC3 may play an important role in the regulation of cell growth. In addition, KCC3 expression enhances the cell capability of regulatory volume decrease. KCC3 also down-regulates E-cadherin/β-catenin complex formation in the cell-cell junction. More importantly, KCC3 expression stimulates cell invasive migration. The close linkage of KCC3 abundance, cell proliferation, invasiveness, and EMT suggests the possibility that expression and activity of KCC 3 may serve as a selective advantage for cancer cells in malignant behaviors. Further work will be required to determine whether these changes reflect KCC3-associated changes in intracellular Cl− concentration or intracellular K+ concentration or other yet-to-be determined functions of KCC3 perhaps not directly related to ion transport. Taken together, the results of cell culture systems and surgical specimens, and the regulatory mechanisms of KCC3 on EMT, can be summarized as follows ( Fig. 6 ). The expression and activity of KCC3 affect E-cadherin and β-catenin abundances by inhibiting transcription of E-cadherin gene and accelerating proteosome-dependent degradation of β-catenin protein. The disruption of E-cadherin/β-catenin complex formation causes the dysfunction of cell-cell junction system, thereby triggering cancer invasion and metastasis. Therefore, blockade of KCC3 activity or expression may provide a potential target for the treatment of tumor progression.
Grant support: Center of Excellence for Clinical Trial and Research (DOH-TD-B-111-004), Department of Health, Executive Yuan, Taiwan; National Science Council, Taiwan (NSC 96-2314-B-006-001; M-R. Shen); National Health Research Institutes grants NHRI-EX94-9311BS (M-R. Shen) and NHRI-EX94-9422BI (C.Y. Chen); and Center for Gene Regulation and Signal Transduction Research, National Cheng Kung University, Taiwan.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).
- Received June 29, 2007.
- Revision received August 20, 2007.
- Accepted September 12, 2007.
- ©2007 American Association for Cancer Research.