b iframe width height src
3.3. Cofilin-1 is required for switching between epithelial and mesenchymal states and junctional disassembly during EMT
The cortical F-actin belt of intestinal cells anchors at TJs and AJs, driving their assembly and function . To evaluate if cofilin-1 could regulate the disassembly of the cell–cell adhesion system during EMT, we analyzed the expression and subcellular localization of epithelial and mesenchymal proteins in siCofilin-1/TGF-β cells. As expected, in cells treated with TGF-β, the E-cadherin and claudin-3 protein levels were decreased relative to those in control cells; however, cells treated with siCofilin-1/TGF-β exhibited restoration of E-cadherin and claudin-3 at cell-cell contact. In addition, the claudin-3 protein levels but not of E-cadherin were restored in these cells. Consistently, the TGF-β group showed increased vimentin. Nevertheless, with concomitant depletion
Fig. 1. TGF-β activates RhoA/p-LIMK2/p-cofilin-1 signaling and increases F/G-actin ratio in HT-29 cell line. (A) Representative RhoA pull-down assay for detection of RhoA-GTP with Rhotekin RBD. HT-29 cells were grown and treated with TGF-β (10 ng/mL) for 48 h. The precipitate and lysate were used to determine RhoA-GTP and total RhoA levels by western blot using anti-RhoA antibody following a pull-down assay.
(D) Cells were treated with TGF-β (10 ng/mL) for 48 h, and then total fractionation of F-actin and G-actin was performed and quantified as described in Materials and methods. As a control, treatment with 20 nM of cytochalasin D was used. Representative images are shown. GAPDH was used as a loading control in (B) and (C). Bar graphs represent relative quantification of p/LIMK2/LIMK2 and p/cofilin1/cofilin1 ratio (where non-treated cells = 1). Data are presented as the mean ± SEM of three independent experiments. Significance was determined using ANOVA followed by the Bonferroni post-test (*P < 0.05, **P < 0.01, ***P < 0.001).
of cofilin-1, the vimentin protein levels were decreased, and the vi-mentin filament network was reorganized (Fig. 4A and B). Additionally, Dexamethasone aggregates in close proximity to cytoplasmic E-cadherin deposits in the TGF-β group were observed, suggesting that the arrangement and/or dynamics of cortical actin filaments contributed to the
disassembly of junctions, whereas SB43154 inhibitor restored the actin and E-cadherin organization at AJCs (Supplementary Fig. 3). These data suggested that cofilin-1 signaling is required for switching between epithelial and mesenchymal states and for junctional disassembly by regulating the dynamics of the actin cytoskeleton.
Fig. 2. Subcellular localization and organizational pattern of p-cofilin1, cofilin-1, and actin cytoskeleton in TGF-β-treated cells. HT-29 cells were grown on glass coverslips, treated with TGF-β for 48 h, and subjected to double labeling. (A) Total cofilin-1/TRITC-phalloidin labeling; a white line (from the lamellipodia area to inside the cell is shown. (B) p-cofilin-1 (Ser3)/TRITC-phalloidin labeling; a white line from the rear to the inside of the cell is shown.
(C) The same labeling as in (B), with a white line between actin filaments shown. Analysis was performed using super-resolution microscopy though structured illumination microscopy. The fluorescence signal intensity measurement was determined with fluorescence intensity profiles taken along the colored lines indicated in each image. Actin filaments were labeled with TRITC-phalloidin. Scale bar: 10 μm.
To further validate the role of the RhoA-p-LIMK2-p-cofilin-1 pathway in the junctional disassembly during EMT we used cofilin-1 mutants to mimic TGF-β eﬀect. We generated transient and stable mutant HT-29 cells with the following constructs: RFP-tag WT cofilin-1, RFP-tag cofilin-1 S3A, and RFP-tag cofilin-1 S3E. The latter two con-structs were used to mimic dephosphorylated cofilin-1 (active form), in which serine 3 is replaced with alanine (S3A), and phosphorylated cofilin-1 (inactive form), in which serine 3 is replaced by glutamic acid (S3E). Firstly, we evaluated F-actin formation in RFP-positive and RFP-negative cells obtained by transient transfection and treatment with TGF-β. As expected, cells exposed to TGF-β showed membrane pro-jections and increased F-actin similarly to WT-RFP-positive and RFP-negative cells. However, S3A-RFP and S3E-RFP mutants did not show changed cofilin-1 phosphorylation status after treatment with TGF-β because of the inserted mutations. Thus, S3A-RFP-positive cells (active cofilin-1) exhibited actin aggregates and decreased F-actin formation in contrast to RFP-negative cells responsive to TGF-β treatment, which showed F-actin structures typical of EMT cells. Therefore, we confirmed that active S3A-cofilin-1 was eﬀective in intensifying the severing of F-actin and promoting distinct patterns of actin structures in EMT cells. Interestingly, we found that S3E-RFP-positive cells (inactive cofilin-1) and RFP-negative cells displayed the same pattern of F-actin, including