Hakai reduces cell-substratum adhesion and increases epithelial cell invasion

The rabbit polyclonal anti-Hakai antibody (Hakai-2498) was kindly provided by Yasuyuki Fujita [8]. Anti-FAK antibody was from Upstate (Charlottesville, VA), anti-Paxillin antibody was from Transduction Laboratories, anti-Vinculin was from Chemicon International (Hampshire, UK), antibody to the extracellular portion of E-cadherin (ECCD-2) was from Zymed (South San Francisco, CA), anti-α-Tubulin antibody was from Immunologicals Direct (Oxford Biotechnology Ltd., UK), HRP-rabbit and mouse polyclonal antibodies was from GE Healthcare (UK) and Alexa-Fluor 488 mouse antibody and Alexa-mouse 568 were from Molecular Probes (Paisley, UK). All primary antibodies were used at dilutions of 1:1000 for Western blotting and 1:100 for immunofluorescence, except for anti-FAK antibody that was used at a dilution of 1:5000 for Western blotting and 1:500 for immunofluorescence. Secondary antibodies were used at dilutions 1:10000 for HRP-rabbit, 1:5000 for HRP-mouse antibody, and 1:200 for Alexa-Fluor antibodies. pcDNA-HA-Hakai was kindly provided by Walter Birchmeier [3]. MG132 and lactacystin were obtained from Calbiochem (Darmstadt, Germany).

MDCK cells were cultured in Dulbecco's modified Eagle's medium (DMEM) containing penicillin/streptomycin and 10% FCS at 37°C and ambient air supplemented with 5% CO₂. MDCK cells stably expressing Hakai were kindly provided by Yasuyuki Fujita [8]. MDCK cells were transfected with pcDNA-HA-Hakai using Lipofectamine-Plus™ reagent (Invitrogen) and were selected in a medium containing 800 μg ml^-1 of G418 (Calbiochem). More than ten stable clones were obtained from two independent transfections. All clones, including 4, 6, 7 and 8 clones, represented comparable phenotypes and the results shown were obtained using clone 4, unless otherwise indicated. MDCK cells stably expressing E-cadherin shRNA in a tetracycline-inducible manner were kindly provided by Yasuyuki Fujita [12].

For the adhesion assay, 1 × 10⁵ cells were plated in a 6-well plate using human fibronectin-coated coverslips (BD Bioscience, Belgium). After 24 hours, the cells were vigorously washed five times with phosphate-buffered saline (PBS), and the remaining cells attached to the plate were analyzed by phase contrast images using a Leica DMIRB microscope with a Hamamatsu C4742-95 Orca camera.

Cells cultured in 6-well plate dishes were lysed for 30 min in 0.3 ml of 1% Triton X-100 lysis buffer (20 mM Tris-HCL [pH 7.5], 150 mM NaCl, and 1% Triton X-100) containing 5 μg ml^-1 lupeptin, 50 mM phenylmethylsulfonyl fluoride, and 7.2 trypsin inhibitor units for aprotinin. After centrifugation at 14, 000 × g for 10 min, thirty micrograms of the supernatants were loaded in 10% polyacrilamide SDS-PAGE. Western blotting was performed as previously described [13].

For immunofluorescence, human fibronectin-coated coverslips (BD Bioscience, Belgium) were used. Cells on coverslips were washed twice in PBS and incubated in 3% parafolmaldehyde-PBS for 15 min. After washing twice in PBS, cells were permeabilized in 0.5% Triton X-100-PBS for 15 min, followed by blocking in DMEM containing 10% fetal bovine serum for 1 h. They were further incubated in primary antibodies for 1 h, washed in PBS, incubated in Alexa-Fluor 488 or Alexa-Fluor 568-conjugated secondary antibody solution for 1 h, and washed four times in PBS. The coverslip was the mounted onto Mowiol on a glass slide. Immunofluorescent images were analyzed by epifluorescence microscopy. To obtain epifluorescence images, we used a Zeiss Axioskop 1 with a Roper Scientific Coolsnap camera.

For invasion assays, 5 × 10⁴ cells were plated in DMEM/1% FCS in a cell invasion chamber (Cell invasion assay kit, Chemicon International) in a 24-well plate, which contains an 8-μm pore size polycarbonate membrane covered with a thin layer of collagen matrix. Invasive cells migrate through a membrane according to the gradient of FCS to the lower chamber that contains DMEM/10% FCS. The invaded cells were stained with crystal violet, and the absorbance of stained cells was measured in a 565-nm wavelength.

In a recent study, we reported that MDCK cell lines overexpressing Hakai (4-6-fold higher than endogenous Hakai) exhibited reduced cell-cell contacts and higher cell proliferation than non-transfected parental MDCK cells [8]. Here, we examined the effect of Hakai overexpression on cell-substratum adhesions. When we incubated several clones of Hakai-overexpressing cells, we realized that they had reduced adhesions to the substratum. Parental MDCK cells attached tightly to the substratum and washes with PSB did not detach the cells from the plate (Figure 1, upper panel). In these cells, trypsin was required to detach them from the plate. In contrast, Hakai-overexpressing MDCK cells were easily detached from the plate by washing with PBS (Figure 1, lower panel). This observation suggests that overexpression of Hakai leads to a reduction of cell-substratum adhesions.

Then, we examined key protein components of cell adhesion, such as FAK, Paxillin and Vinculin, which localize to focal adhesions and focal contacts, mediate adhesion to the substratum, and participate in initiating signal transduction pathways [14]. By Western blotting, we found that Paxillin expression was strongly reduced in Hakai-overexpressing cells, whereas the levels of FAK and Vinculin were not affected by Hakai expression (Figure 2A). Further analysis of the adhesions by immunofluorescence revealed that FAK protein localized at focal adhesions and focal contacts in parental MDCK cells, while in Hakai-overexpressing cells, FAK was diffusely distributed in the cytoplasm, enriched at the sites of focal contacts, and absent from focal adhesions (Figure 2B). As expected, in Hakai-overexpressing cells, Paxillin was undetectable (Figure 2B). These data suggest that cell-substratum adhesions and the formation of focal adhesions are highly dependent on overexpression of Hakai, probably due to the specific downregulation of Paxillin. Although Paxillin primarily functions as a molecular adapter or scaffold protein that provides multiple docking sites at the plasma membrane, it is also important for the transduction of external signals to elicit changes in cell motility and modulate gene expression by the various MAP kinase cascades. Paxillin binds to many proteins that are involved in effecting changes in the organization of the actin cytoskeleton, and are thus necessary for cell motility associated with tumor metastasis [15]. Therefore, the complete absence of Paxillin at focal adhesions may be one of the mechanisms that explain the loss of cell adhesion to the substrate.

To investigate whether the reduced Paxillin levels found in Hakai-overexpressing cells was the result of increased degradation of Paxillin protein, we analyzed the effect of proteasome inhibitors MG132 and lactacystin in Hakai-overexpressing and parental MDCK cells. None of these inhibitors affected Paxillin abundance in Hakai-overexpressing cells (Figure 3), indicating that the expression of Paxillin is controlled in a proteasome-independent manner. We therefore ask whether the loss of E-cadherin was sufficient to reduce Paxillin expression levels. Using MDCK cells in which E-cadhering was knocked down by shRNA in a tetracycline-regulated manner, Paxilllin expression was strongly reduced where E-cadherin was silenced (Figure 4). These findings suggest that the loss of Paxillin in Hakai-overexpressing cells may, at least partially, result from the decreased expression level of E-cadherin.

However, several lines of evidence suggest that Hakai can also affect cellular phenotypes in an E-cadherin-independent manner. First, MDCK cells expressing E-cadherin shRNA do not extend spiky protrusions that are seen in Hakai-overexpressing cells [12]. Second, in Hakai-overexpressing MDCK cells, Hakai is localized at the end of the protrusion where FAK is also enriched [8], suggesting that Hakai may be involved in regulating the dynamic extension and retraction of these structures and in influencing cell motility and cell attachment to the ECM. Finally, Hakai is also located in the nucleus, where it can interact with nuclear proteins such as PSF and modulate cancer-related gene expression [8] and it may act as a correpresor of the estrogen receptor in breast cancer cells [16]. These findings suggest that Hakai can have ubiquitin-independent functions that may influence the cell phenotype, in addition to its influence on known substrates (like E-cadherin) and as-yet unreported substrates in the nucleus or the cytoplasm, as previously reported for other E3-ubiquitin ligases [17, 18].

In closing, we examined whether the overexpression of Hakai also affect the invasiveness of cells. We used a polycarbonate membrane over which a thin layer of extracellular matrix (ECM) was applied to serve as an in vitro basement membrane. Since only invasive cells are capable of migrating through the ECM layer, we analyzed the ability of parental and Hakai-overexpressing MDCK cells to invade this in vitro matrix. Staining of the cells that passed through the membrane revealed that, whereas parental cells did not invade the matrix, Hakai-overexpressing cells readily passed through the membrane (Figure 5), indicating that overexpression of Hakai also enhances cell invasion. Taken all together, these observations suggest that Hakai may be involved in the regulation of invasion and cell-substratum adhesion. However, further investigations of Hakai in in vitro and in vivo model systems would lead us to validate its role during tumorogensis. Therefore, Hakai is implicated in several processes that are often occurring during epithelial-mesenchymal transition (EMT). In the last few years, it has been proposed that EMT mRNAs markers can be detected in circulating tumour cells (CTC) in the blood from cancer patients, helping to select new therapeutic targets with important clinical implications [19] Indeed, our group has recently proposed Plakophillin-3 mRNA, involved in epithelial cell-cell adhesions, as a biomarker for detection of CTCs in gastrointestinal cancer patients [20]. Further investigation is warranted to elucidate the molecular mechanism whereby Hakai influences these processes under physiological and pathological conditions, and to assess the potential usefulness of Hakai as therapeutic target for cancer.