β-Catenin and Smad3 regulate the activity and stability of myocardin-related transcription factor during epithelial–myofibroblast transition

E-Cadherin down-regulation facilitates TGFβ-induced SMA expression

Our previous studies showed that disruption or absence of intercellular contacts (as induced by LCM, scratch wounding, or subconfluence) enables TGFβ to provoke SMA expression in tubular epithelial (LLC-PK1) cells (Masszi et al., 2004; Fan et al., 2007). To assess the role of AJs in this EMyT-promoting effect, we specifically targeted their chief component, E-cadherin. TGFβ failed to induce SMA expression in intact, confluent monolayers transfected with a control (nonrelated [NR]) small interfering RNA (siRNA) but provoked robust SMA expression after siRNA-mediated E-cadherin down-regulation (Figure 1, A and B). E-Cadherin silencing alone (without TGFβ stimulation) did not induce (or did so only marginally) SMA expression (Figure 1, A and B). In agreement with our previous data (Masszi et al., 2004), in subconfluent layers, TGFβ did provoke SMA expression, concomitant with ∼60% decrease in E-cadherin level (Figure 1, C and D). However, when E-cadherin was fully knocked down prior to TGFβ treatment, the cytokine triggered a 3.5-fold-higher increase in SMA expression in the subconfluent cultures compared with the NR-treated controls (Figure 1, C–E). Thus the absence of E-cadherin permits TGFβ-induced SMA expression in confluent monolayers and strongly potentiates SMA expression in subconfluent cultures, implying that E-cadherin is an important regulator of the myogenic program in the epithelium.

β-Catenin is a crucial permissive regulator of SMA expression

E-Cadherin has been shown to mitigate Rho activation (Cho et al., 2010), which might contribute to its suppressive effect on SMA expression. In addition, E-cadherin is the major intracellular binding partner of β-catenin, and this molecule has been implicated in SMA expression (Masszi et al., 2004; Onder et al., 2008). Therefore we set out to characterize the role of β-catenin in SMA expression under myogenic (two-hit) conditions. Cells were transfected with NR or β-catenin siRNA and, after reaching confluence, exposed to LCM plus TGFβ for 48 h. β-Catenin knockdown (Figure 2A and Supplementary Figure 1A) resulted in strong suppression of SMA expression (Figure 2, A and B). Similar observations were made on another tubular cell line, NRK-52E, as well, in which the two-hit stimulation also caused marked increase in SMA expression, and β-catenin down-regulation fully blocked this response (Supplementary Figure S2, A and B). To test whether this effect manifested at the transcriptional level, we measured SMA mRNA in control and β-catenin–silenced cells after 48 h of stimulation with the combined treatment. The message for SMA was 50% less in β-catenin siRNA-transfected cells than in the NR siRNA-transfected controls (Figure 2C). To determine whether the presence of β-catenin affects the activity of the SMA promoter, we performed reporter assays using a 765–base pair SMA promoter coupled to firefly luciferase (pSMA-Luc; Masszi et al., 2010). Cells were first transfected with NR or β-catenin siRNA, followed 24 h later with cotransfection of pSMA-Luc and an internal control plasmid (pRL-TK). The combined treatment triggered strong activation of the SMA promoter in control cells, whereas β-catenin silencing suppressed the (already low) basal SMA promoter activity and—of importance—prevented its stimulation-induced rise over the baseline obtained in nonstimulated control cells (Figure 2D). Taken together, the results indicate that the absence of β-catenin dramatically reduces the activation of the SMA promoter and consequently the level of SMA mRNA and protein expression.

Having seen that β-catenin is essential for SMA expression induced by the conventional two-hit scheme, we asked whether it could be involved in the E-cadherin down-regulation–promoted SMA response as well. It is noteworthy that in our cells (similar to many other systems) E-cadherin silencing did not reduce the total β-catenin level (Figures 1C and 2E), implying that β-catenin might bind to other partners or its synthesis may keep up with its degradation. Moreover, elimination of E-cadherin resulted in a greater-than-threefold rise in nuclear β-catenin content, indicating an increase in the mobile pool of this molecule (Figure 2, F and G). Similarly, stimulation of the cells by the two-hit regimen (TGFβ plus LCM) also caused a threefold increase in cytosolic β-catenin, along with the dissipation of the peripheral β-catenin signal (Figure 2H). Of importance, concomitant silencing of β-catenin and E-cadherin strongly reduced the TGFβ-provoked SMA expression compared with the level observed in E-cadherin down-regulated cells (Figure 2E). This finding confirms the key role of β-catenin in the E-cadherin silencing–promoted SMA expression.

Next we sought to determine whether β-catenin plays a permissive or/and inductive role in the regulation of the SMA promoter. Overexpression of β-catenin induced a 10-fold increase in the activity of the β-catenin–responsive reporter TOPFlash, verifying the efficacy of our expression vector (Supplementary Figure 1B). Nonetheless, β-catenin overexpression did not stimulate the SMA promoter, nor did it increase its activation provoked by TGFβ plus LCM (Figure 2I). Thus β-catenin is necessary but not sufficient for the optimal activation of the SMA promoter, in full agreement with the fact that the SMA promoter does not harbor any obvious β-catenin–responsive motifs.

Taken together these results suggest that β-catenin plays a crucial permissive role in the regulation of the SMA promoter, and the cytosolic β-catenin levels obtained after contact injury or E-cadherin down-regulation are sufficient for this permissive effect.

β-Catenin prevents the inhibitory effect of Smad3 on the MRTF-induced activation of the SMA promoter

Next we sought to gain insight into mechanism by which β-catenin facilitates the myogenic program. Because previous observations showed that β-catenin can associate with Smad3 and that Smad3 is a strong inhibitor of MRTF-dependent transcription (Masszi et al., 2010), we considered whether the Smad3/β-catenin interaction might regulate the myogenic program. We initially tested their interaction under the various conditions of the two-hit regimen after short-term stimulation, since the nuclear translocation of MRTF peaks around 30–60 min, and during this time neither β-catenin nor Smad3 levels change (Masszi et al., 2010). Coimmunoprecipitation studies revealed that in resting cells there was only a weak association between Smad3 and β-catenin, whereas stimulation with either TGFβ or LCM induced a substantial (5- to 10-fold) increase in their interaction. The combined treatment exerted an even stronger effect (Figure 3, A and B). Moreover, down-regulation of E-cadherin increased the complex formation between β-catenin and Smad3 in the absence of any stimulus, suggesting that the increased availability of β-catenin is sufficient to promote enhanced interaction between these partners (Figure 3C).

We then asked whether β-catenin could reverse the inhibitory action of Smad3 on MRTF-driven transcription. In introductory experiments we titrated the effect of Smad3 on MRTF. In agreement with our recent findings (Masszi et al., 2010), the MRTF expression–driven SMA promoter response was gradually diminished as the amount of coexpressed Smad3 was increased (Figure 4A). Under our conditions ∼60% inhibition was attained with 1 μg of Smad3 DNA, and this dose was applied in the subsequent studies. As shown in Figure 4B, cotransfection with an increasing amount of β-catenin gradually restored the MRTF-induced SMA promoter activation. A detailed analysis of the β-catenin effect, using cotransfection with MRTF, Smad3, or both (Figure 4C, top) showed that overexpression of β-catenin entirely prevented the Smad3-induced inhibition of the MRTF-triggered promoter activation, and produced only a slight (nonsignificant) decrease in the basal or MRTF-induced promoter activity. In agreement with our previous findings, overexpression of Smad3 without MRTF had no significant effect on the SMA promoter.

To test whether the β-catenin–induced protection against the inhibitory effect of Smad3 required the CArG-boxes or other cis elements in the SMA promoter, including Smad-binding element 1 (SBE1), SBE2, and the TGFβ control element (TCE), we transfected the cells with a triple mutant promoter in which each of these was inactivated (Masszi et al., 2010). Of importance, the inhibition by Smad3 and the protective effect of β-catenin remained the same as observed with the wild-type promoter (Figure 4C, middle). Similar results were obtained with a short (152 base pair) promoter construct, which includes the two CArGs and the TCE but lacks both SBEs and the E-box, the target site for basic helix-loop-helix transcription factors (Figure 4C, bottom). Together these data show that β-catenin reverses the inhibitory effect of Smad3 on MRTF-driven transcription, and this effect is mediated via the CArG boxes.

β-Catenin maintains the SRF/MRTF (myogenic) complex

To investigate whether β-catenin can preserve the interaction between SRF and MRTF, cells were cotransfected with hemagglutinin (HA)-SRF and Myc-MRTF along with FLAG–β-catenin, FLAG-Smad3, or both (Figure 5A). When only MRTF and SRF were coexpressed, immunoprecipitation of MRTF (through the tag) pulled down a sizable amount of SRF (lane 4). Overexpression of Smad3 nearly abolished the association between SRF and MRTF (lane 1). Of importance, β-catenin overexpression fully restored the MRTF/SRF interaction (lanes 2 and 3).

To validate these observations with regard to the endogenous proteins, we investigated the association of SRF with MRTF in control or β-catenin–down-regulated cells with or without stimulation. Suppression of β-catenin had a dramatic effect on the myogenic complex: namely, it disrupted the association between MRTF and SRF under basal conditions and prevented any increase upon TGFβ plus LCM (30 min) stimulation (Figure 5B). Identical results were obtained using NRK-52E rat tubular cells, implying that the need for β-catenin for the SRF/MRTF association is a general phenomenon (Supplementary Figure S2C).

We then tested whether β-catenin down-regulation could interfere with the stimulus-induced translocation of MRTF into the nucleus. β-Catenin silencing did not prevent the fast, LCM-induced nuclear uptake of MRTF (Figure 5C). Thus β-catenin facilitates the integrity of the myogenic complex by a mechanism distal to MRTF translocation, consistent with the prevention of the inhibitory effect of Smad3.

β-Catenin and Smad3 mutually inhibit each other's association with MRTF

We argued that if β-catenin, at least in part, acts through capturing Smad3, then overexpression of β-catenin should result in reduced Smad3–MRTF interaction. To test this, cells were transfected with HA-MRTF, Myc-Smad3, and either an empty vector or FLAG–β-catenin. Immunoprecipitation of MRTF through the HA tag (Figure 6A) brought down a substantial amount of Smad3 in empty vector-transfected cells, whereas overexpression of β-catenin strongly reduced the amount of MRTF-associated Smad3, concomitant with a dramatic increase in the SRF/MRTF interaction (Figure 6A). Intriguingly and unexpectedly, the MRTF immunoprecipitate also contained some β-catenin (see later discussion). Next we investigated the converse situation by asking whether a reduction in endogenous β-catenin could increase the MRTF–Smad3 interaction. Downregulation of β-catenin caused a substantial rise in the amount of MRTF-associated Smad3 (Figure 6B). These experiments indicate that β-catenin is a negative regulator of the Smad3/MRTF interaction.

Having observed that β-catenin itself can be in complex with MRTF, we first surmised that this interaction might be mediated through Smad3. In other words, whereas β-catenin reduces the overall Smad3 binding to MRTF, it was conceivable that the remaining Smad3–MRTF binding was responsible for the presence of β-catenin through the formation of an MRTF–Smad3–β-catenin complex. To assess this, we used an MRTF mutant, ΔB1p, that lacks a seven–amino acid sequence in the B1 region critical for Smad3 binding (Masszi et al., 2010). Indeed, coprecipitation studies revealed that ΔB1p almost entirely lost its capacity to bind Smad3 (Figure 6C). Despite this, ΔB1p retained its capacity to pull down β-catenin; in fact there was more β-catenin in complex with ΔB1p than with wild-type MRTF (Figure 6C). This implies that β-catenin does not bind to MRTF through Smad3, and that the MRTF sequence critical for Smad3 binding is not essential for the complex formation between β-catenin and MRTF.

To substantiate the reciprocal relationship between the binding of β-catenin versus Smad3 to MRTF, we down-regulated Smad3 and found an increased association between endogenous β-catenin and MRTF (Figure 6D). As expected, this was coincident with enhanced interaction between MRTF and SRF. Taken together, the results indicate that β-catenin reduces the binding of Smad3 to MRTF, whereas Smad3 mitigates the binding of β-catenin to MRTF, and these mutually inhibitory effects do not require the same MRTF sequence. These findings imply that the Smad3–β-catenin interaction prevents these partners from accessing MRTF.

The association between β-catenin and MRTF is a novel finding that could also contribute to MRTF regulation. Although the detailed characterization of this interaction will require a separate study, as an initial step we established that this is a stimulus-regulated process, as the two-hit challenge induced a transient increase in MRTF–β-catenin association (Figure 6E).

β-Catenin maintains MRTF stability in stimulated cells by counteracting glycogen synthase kinase-3β–dependent MRTF degradation

During the course of our experiments we noticed that in β-catenin–down-regulated cells the two-hit stimulation appeared to reduce the size of the MRTF immunoreactive band. To investigate this phenomenon, we performed a detailed time course (0–120 min) in NR and β-catenin siRNA-transfected cells. Figure 7A shows that there was no obvious difference in the expression of MRTF in resting control versus β-catenin–silenced cells; in contrast, after stimulation there was a dramatic reduction in the MRTF band obtained from β-catenin–down-regulated cells. The same phenomenon was observed in NRK-52E cells as well (Supplementary Figure S2D). This decrease was apparent at times ≥60 min and became pronounced or near complete by ∼120 min. It is noteworthy that in β-catenin–expressing cells stimulation induces an upward shift in the MRTF band, corresponding to the reported (multiple) phosphorylation of this protein. This often manifests as a widening of the band or the appearance of multiple bands with slightly higher molecular mass. The total density of these usually exceeds the density of the nonstimulated band, which may be due to more efficient or preferential antibody binding to the less condensed epitopes and/or the phosphorylated form(s) of MRTF. Indeed, densitometry showed an early increase in the MRTF-immunoreactive band(s) upon two-hit stimulation (Figure 7A). Because this response was readily apparent after 15 min, it likely reflects posttranslational modification (e.g., phosphorylation) and not a net rise in the MRTF protein. On stimulation, the intensity of the MRTF bands started to increase in the β-catenin–depleted cells as well, but after ∼30–60 min a reversal occurred and the signal eventually dropped well below the nonstimulated level (Figure 7A). The same pattern was observed when probing with a different MRTF antibody (Figure 7A, inset). The decrease of the signal below the basal level might be due to an additional modification that masks the epitope seen by the antibody (e.g., ubiquitination) and/or the degradation of MRTF protein. To determine whether protein degradation might play a role, we followed the fate of heterologously expressed, HA-tagged MRTF using an anti-HA antibody. In NR siRNA-transfected cells the intensity of the HA signal remained constant during stimulation, indicating that the increase seen by the MRTF antibody was indeed due to some posttranslational modification. In contrast, in β-catenin–silenced cells the HA signal still dropped below the level observed in the NR siRNA-transfected cells. However this decrease started later and became substantial (usually ∼80%) only after ≥2 h of stimulation (Figure 7B). From these results we conclude that the early phase of the signal reduction is likely due to epitope masking by modification, whereas the later phase reflects (at least in part) reduction in MRTF protein expression.

Glycogen synthase kinase-3β (GSK-3β) has been reported to phosphorylate and inhibit myocardin (Badorff et al., 2005) and to promote its ubiquitination (Xie et al., 2009). We therefore examined whether GSK-3β might be involved in the degradation of MRTF, an effect that might become apparent in β-catenin–depleted cells. The decrease in MRTF observed upon stimulation in β-catenin–depleted cells was efficiently mitigated by LiCl or SB-216763, two inhibitors of GSK-3β (Figure 7C). Moreover, the elimination of β-catenin strongly facilitated MRTF ubiquitination, and this phenomenon was prevented by LiCl (Figure 7D).

GSK-3β interacts with MRTF through a Smad3-dependent, β-catenin–inhibited mechanism

We then investigated whether GSK-3β might interact with MRTF and whether this process might be modulated by β-catenin and Smad3. Of importance, a recent publication reported that Smad3 and GSK-3β can directly associate (Hua et al., 2010). This raised the intriguing possibility that Smad3 might act as an adaptor recruiting GSK-3β to MRTF and this process might be regulated by β-catenin. Probing MRTF immunoprecipitates with anti–GSK-3β revealed a weak binding of GSK-3β to MRTF that was slightly elevated by the two-hit stimulation (Figure 8A). Downregulation of β-catenin robustly potentiated the stimulus-induced association of GSK-3β with MRTF as early as 30 min after stimulation. In contrast, elimination of Smad3 prevented the basal and stimulus-induced interaction between MRTF and GSK-3β (Figure 8A). Moreover, double silencing of β-catenin and Smad3 revealed that the MRTF–GSK-3β interaction, potentiated by the absence β-catenin, was preempted in the absence of Smad3 (Figure 8B). To substantiate the role of Smad3 as a potential recruiter of GSK-3β to MRTF, we compared the association of GSK-3β with wild-type (WT) and ΔB1p MRTF, the mutant with diminished Smad3-binding capacity. In β-catenin–depleted cells, the stimulus-induced association of ΔB1p MRTF with GSK-3β was strongly reduced compared with the WT (Figure 8C). Finally, knockdown of Smad3 significantly reduced MRTF degradation in β-catenin–depleted and stimulated cells (Supplementary Figure S3). Taken together, these results implicate β-catenin as a key factor in the maintenance of MRTF stability and suggest that the absence of β-catenin facilitates the Smad3-dependent association of GSK-3β to MRTF, which primes the latter for ubiquitination and degradation (see the scheme in Figure 9C).

β-Catenin is a key permissive factor for the CArGome

Having seen the importance of β-catenin in MRTF stability, we considered that β-catenin might be necessary to ensure the expression of a variety of MRTF-dependent genes in addition to SMA. We showed previously (Masszi et al., 2010) that MRTF is required for the basal and/or stimulus-induced expression of several CArG-regulated proteins, including filamin, CapZ, cofilin, and SRF itself. Accordingly, we found that β-catenin depletion prevented the two-hit–induced increase in filamin, CapZ, and cofilin and decreased the level of SRF both under resting conditions and after stimulation (Figure 9, A and B). These results suggest that β-catenin might be an essential regulator of the CArGome and might explain how β-catenin, independent of its transcriptional effect, can affect the expression of an array of cytoskeletal proteins.

Reagents

The GSK-3β inhibitor SB-216763 was purchased from Sigma-Aldrich (St. Louis, MO) and TGFβ from R&D Systems (Minneapolis, MN). Commercially available antibodies were obtained from various sources as follows: MRTF-B (C-19), SRF (G-20), HA (Y-11), c-Myc (9E10), β-catenin (C-18), GSK-3α/β (0011-A), and GAPDH (0411), Santa Cruz Biotechnology (Santa Cruz, CA); FLAG (M2) and SMA (1A4), Sigma-Aldrich; E-cadherin, CapZ α , and filamin A (clone 5/ABP-280), BD Transduction Laboratories (Lexington, KY); SMAD3, Abcam (Cambridge, MA); HA.11 clone 16B12, Covance (Berkeley, CA); histones (clone F152.C25.WJJ) and ubiquitin (P4D1), Millipore (Billerica, MA); and cofilin, Cell Signaling Technology (Beverley, MA). Rabbit polyclonal anti-MRTF (BSAC) was described previously (Sasazuki et al., 2002). All secondary antibodies were from Jackson ImmunoResearch Laboratories (West Grove, PA).

Cell culture and treatment

LLC-PK1 (Cl 4) cells, a porcine proximal tubular epithelial cell line (a kind gift from R. C. Harris, Vanderbilt University School of Medicine, Nashville, TN) were cultured in low-glucose DMEM (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum and 1% streptomycin/penicillin solution (Invitrogen) as in our previous studies (Masszi et al., 2004, 2010). Rat proximal tubular cells (NRK-52E) were purchased from American Type Culture Collection (Manassas, VA) and cultured in high-glucose DMEM containing the same supplements as mentioned. Cells were incubated under serum-free conditions for at least 3 h before various treatments. To induce cell contact disassembly, cells were thoroughly washed with phosphate-buffered saline (PBS; Invitrogen) and cultured in nominally calcium chloride–free DMEM (LCM; Invitrogen). Where indicated, cells were treated with TGFβ (10 ng/ml for luciferase reporter assays and 4–10 ng/ml for other experiments).

Plasmids and transfection

The p765-SMA-Luc reporter construct containing the proximal 765–base pair portion of the rat SMA promoter in a pGL3-basic vector (WT), the constructs harboring an inactivating mutation at the SBE1 or SBE2 sites (SBE1mut and SBE2mut) (Hu et al., 2003), and the p152-SMA-Luc reporter (provided by S. H. Phan, University of Michigan Medical School, Ann Arbor, MI) were described in our previous studies (Masszi et al., 2010). Using SBE1mut as a template, we performed PCR-based mutagenesis to create subsequent mutations in SBE2 and the TCE resulting in the triple-mutant Luc construct, as described (Masszi et al., 2010). Briefly, the mutations (in parentheses) and the corresponding primer pairs were as follows:

• SBE2 (C⁺¹⁵/T, A⁺¹⁶/G, and G⁺¹⁷/C), 5′-CCACCCACCTGCAGTGGAGAAGCCCAGC-3′ and 5′- CTGGGCTTCTCCACTGCAGGTGGGTGGT-3′

• TCE (T⁻⁵³/C, G⁻⁵²/T, and G⁻⁵⁰/C), 5′-TGGGAAGCGAGCTGCAGGGGATCAGACCA-3′ and 5′-TGGTCTGATCCCCTGCAGCTCG­CTTCCCA-3′

The thymidine kinase minimal promoter-driven Renilla luciferase internal control plasmid, pRL-TK, was purchased from Promega (Madison, WI). The LEF/TCF reporter plasmid, TOPFlash, was obtained from Upstate (Millipore). The N-terminally Myc- or FLAG-tagged Smad3 expression constructs (in pCMV5B) were a kind gift from L. Attisano (University of Toronto). The FLAG-tagged β-catenin was provided by E. R. Fearon (University of Michigan, Ann Arbor, MI) (Kolligs et al., 1999). The FLAG-tagged MRTF-B plasmid was provided by E. N. Olson (University of Texas Southwestern Medical Center, Dallas, TX). The coding region of MRTF-B was amplified by PCR and cloned into pcDNA3.1/Myc–His A to obtain the myc-MRTF-B expression vector. pcDNA3.1/HA-MRTF-B was generated by engineering a 2xHA tag at the N-terminus of MRTF-B using standard PCR methodology. The B1 region–deletion mutant of HA-tagged MRTF-B (ΔB1) was then constructed using primer pairs complementary to regions upstream and downstream of the specific deletion, as described previously (Masszi et al., 2010). The final construct contained a seven–amino acid (S279–P285 inclusive) deletion. PCR reactions were performed using PfuTurbo (Agilent Technologies, Santa Clara, CA). All constructs were verified by sequencing. The pCGN-SRF plasmid encoding HA-tagged human SRF generated by the Prywes lab (Johansen and Prywes, 1993) was obtained through Addgene (Cambridge, MA). Depending on the experiment, cells were transfected using FuGENE 6 (Roche Applied Science, Indianapolis, IN), Lipofectamine 2000 (Invitrogen), or jetPRIME (Polyplus-transfection SA, Illkirch, France) reagents as in our previous studies (Masszi et al., 2004, 2010; Fan et al., 2007) and (for jetPRIME) according to the manufacturer's recommendation.

Luciferase reporter assays

Luciferase reporter assays were performed as described in our previous studies (Masszi et al., 2003, 2004) using 0.5–μg/well luciferase construct, 0.05–μg/well pRL-TK, and varying amounts of empty carrier or expression vector. Sixteen hours later, cells were serum starved for 3 h, treated for 24 h (if not indicated otherwise), and lysed and the luciferase activity was determined using the Dual Luciferase Reporter Assay System Kit (Promega). For each condition, treatments were performed in duplicate, and experiments were repeated at least three times. From each sample, the firefly luciferase activity corresponding to a specific promoter construct was normalized to the Renilla luciferase activity of the same sample. Results are expressed as fold changes compared with the mean firefly/Renilla ratio of the untreated controls taken as a unit.

RNA interference

Optimal target sequences were determined using the siRNA Target Finder program (Applied Biosystems, Foster City, CA). The siRNA sequences used in the present experiments were as follows: pig β-catenin siRNA, 5′-AAGUACAUACACCAUACUACG-3′; pig SMAD3 siRNA, 5′-AAGAGTTCACTCCACATTCTC-3′; and pig E-cadherin siRNA, 5′-CTCTGCTGGTGTTTGATTATT-3′. The validated siRNA against rat β-catenin was obtained from Dharmacon (Lafayette, CO). Alternative siRNAs were also designed and used for the down-regulation of each of the aforementioned proteins. The siRNAs directed against different sequences of the corresponding mRNA gave identical experimental results. Silencer Negative Control #2 siRNA was purchased from Applied Biosystems. LLC-PK1 cells were cultured in antibiotic-free growth medium and transfected with siRNA using Lipofectamine RNAiMAX (Invitrogen). DNA/siRNA cotransfections were performed using jetPRIME.

Western blotting and coimmunoprecipitation

Following treatments, cells were lysed with Triton lysis buffer (30 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, pH 7.4, 100 mM NaCl, 1 mM ethylene glycol tetraacetic acid, 20 mM NaF, and 1% Triton X-100) supplemented with 1 mM Na₃VO₄, 1 mM phenylmethylsulfonyl fluoride, and Complete Mini Protease Inhibitor Cocktail (Roche). SDS–PAGE and Western blotting were performed on equivalent protein loads, as determined using BCA Protein Assay Reagents (Thermo Scientific, Waltham, MA). Prior to coimmunoprecipitation studies, cell lysates were spun at 12,000 rpm for 5 min to remove cell debris. Precleared supernatants were incubated with appropriate antibodies, and immunocomplexes were captured on protein G–agarose beads (Thermo Scientific). Bound proteins were eluted from the washed beads and analyzed by Western blotting. Antibody-free and lysate-free controls were routinely included to confirm specificity of the immunoprecipitated proteins. Aliquots of each input were run in parallel to monitor expression levels. Densitometry was performed with a GS800 densitometer using Quantity One software (Bio-Rad Laboratories, Hercules, CA).

Nuclear extraction

Nuclear extracts were prepared from confluent layers of LLC-PK1 cells grown on 6-cm dishes using the NE-PER Nuclear Extraction Kit (Thermo Scientific). Equivalent amounts of protein from each extract were analyzed by Western blotting. Equal loading of nuclear protein was monitored by probing with an anti-histone antibody.

Immunofluorescence microscopy

Cells plated on glass coverslips were fixed with 4% paraformaldehyde (Canemco & Marivac, Gore, Canada) for 30 min, washed with PBS, and quenched with 100 mM glycine/PBS for 10 min. Cells were permeabilized for 20 min in PBS containing 0.1%Triton X-100 and 1% bovine serum albumin (BSA), blocked in 3% BSA for 1 h, and incubated with primary antibody for an additional 1 h. Washed coverslips were incubated with the corresponding fluorescently labeled secondary antibody, which included the addition of 4′,6-diamidino-2-phenylindole (Lonza, Basel, Switzerland) for nuclear labeling. When staining for E-cadherin, cells were fixed with cooled methanol for 5 min, washed with PBS, and blocked with 3% BSA prior to immunostaining. Coverslips were mounted on slides using fluorescent mounting medium (Dako, Glostrup, Denmark). Samples were analyzed using a microscope (IX81; Olympus, Center Valley, PA) with a UPlan S-Apo 60× 1.42 numerical aperture oil objective (Olympus) coupled to a camera (Evolution QEi Monochrome; Media Cybernetics, Bethesda, MD) controlled by imaging software (QED InVivo; Media Cybernetics). Images were processed using ImagePro Plus software (3DS 5.1; Media Cybernetics). Modifications were restricted exclusively to minor adjustments of brightness/contrast. Cytoplasmic β-catenin staining was quantified using the MetaMorph image analysis software (Molecular Devices, Sunnyvale, CA). Briefly, the mean fluorescence intensity was determined in circular regions (6.5 μm diameter) randomly placed in the cytosplasmic (extranuclear) area in 50 cells. Data were normalized to the background fluorescence determined in a cell-free area on the same coverslip. The average ratio of all measurements in untreated controls was taken as 1 and compared with the ratio obtained in cells exposed to the combined treatment (LCM plus TGFβ) for 6 h.

mRNA analysis

LLC-PK1 cells were transfected with pig specific beta-catenin or NR siRNA using Lipofectamine RNAiMAX. After 48 h, cells were serum deprived for 3 h and treated with TGFβ and LCM for an additional 48 h. RNA was extracted using an RNeasy Kit (Qiagen, Valencia, CA), and cDNA was synthesized from 1 μg of total RNA using iScript reverse transcriptase (Bio-Rad Laboratories). SYBR green–based real-time PCR was used to evaluate gene expression of SMA, using glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as the reference standard. Primer pairs designed against known pig sequences were as follows:

• SMA, 5′-TGTGACAATGGTTCTGGGCTCTGT-3′ and 5′-TTCGTCACCCACGTAGCTGTCTTT-3′

• GAPDH, 5′-GCAAAGTGGACATGGTCGCCATCA-3′ and 5′-AG­CTTCCCATTCTCAGCCTTGACT-3′

Statistical analysis

Data are presented as representative blots or images from at least three similar experiments or as the means ± SEM for the number of experiments indicated. Statistical significance was determined by one-way analysis of variance (Tukey or Dunn post hoc testing for parametric and nonparametric analysis of variance, as appropriate), using Prism and the InStat software GraphPad, La Jolla, CA). p < 0.05 was accepted as significant; *p < 0.05 and **p < 0.01.