Transforming Growth Factor-β1 Represses E-Cadherin Production via Slug Expression in Lens Epithelial Cells

Abstract

purpose. TGFβ is a potent candidate for epithelial–mesenchymal transition (EMT) during the development of anterior polar cataracts in the human lens. The Snail superfamily is involved in EMT through the repression of E-cadherin production. This study was conducted to determine whether the Snail gene family is activated in the process of TGFβ1-induced EMT and how TGFβ1 regulates the expression of this gene family.

methods. Total RNA extracted from human cataract samples was subjected to the real-time PCR quantification of Slug mRNA. Induction of Slug expression by TGFβ1 (10 ng/mL) in lens epithelial cells was determined by RT-PCR, immunostaining, immunoblot analysis, and Slug promoter analysis. A series of Slug promoter deletion constructs was used to identify the putative regulatory element responsive to TGF signaling. Chromatin immunoprecipitation was performed to determine whether Sp1 associates with the endogenous Slug promoter. Inhibition of Slug expression with Slug siRNA was used to investigate the role of Slug in TGFβ-mediated EMT.

results. Slug levels were highly upregulated in lens epithelial cells obtained from patients with anterior polar cataracts. Treatment of TGFβ1 induced the expression of Slug in both lens and other epithelial cells in vitro. TGFβ1-induced Slug expression was significantly inhibited by the MEK- and JNK/SAPK-specific inhibitors, but not by transfection with dominant-negative forms of Smads or small GTPase proteins, indicating that MAPK pathways are involved in the regulation of Slug expression by TGFβ1. The Slug promoter analysis revealed that the Sp1 binding site in the Slug promoter is responsible for TGFβ1-induced Slug expression. In addition, the TGFβ1-mediated repression of E-cadherin was significantly inhibited by Slug siRNA.

conclusions. These data suggest that TGFβ1 induces Slug expression and that the repression of E-cadherin production by TGFβ1 is mediated by the induction of Slug in lens epithelial cells.

Epithelial–mesenchymal transition (EMT), a highly conserved and fundamental process governing morphogenesis, occurs during a critical phase of embryonic development in multicellular organisms. In mammals, EMT first occurs at the blastula stage, during formation of the parietal endoderm, which later contributes to the extra embryonic tissues. ¹ EMT is also reactivated in a variety of diseases, including renal fibrosis, ² liver cirrhosis, ³ breast cancer, ⁴ and anterior polar cataracts. ⁵ During the formation of anterior polar cataracts and posterior capsular cataracts, lens epithelial cells transdifferentiate and proliferate into plagues of large spindle-shaped cells, or myofibroblasts through EMT. ^{6 7} These myofibroblasts promote lens capsule wrinkling and an accumulation of abnormal extracellular materials, including type I collagen. ^{8 9}  

It is likely that loss of E-cadherin production is heavily involved in EMT. Downregulation of E-cadherin in TGFβ-induced EMT has been demonstrated in several systems and also in the lens. ¹⁰ In vitro, there is a direct correlation between the lack of E-cadherin production and loss of the epithelial phenotype. ¹¹ Several transcriptional repressors of E-cadherin have now been identified, including the Snail superfamily of the zinc-finger transcription factors Snail ^{12 13} and Slug, ^{14 15} the two-handed zinc factors ZEB-1 and Sip1, ^{16 17} and the basic helix–loop–helix transcription factor E12/E47. ¹⁸ Snail and Slug control E-cadherin expression in epithelial cells, ^{12 15} as well as in embryonic development. ¹⁹ Studies using either Slug antisense treatment or the expression of dominant-negative Slug constructs have shown that Slug is involved in neural crest specification and mesoderm delamination in chick and Xenopus embryos. ^{20 21} Moreover, Slug gain of function leads to an increase in neural crest production in the chick embryo. ²² These studies indicate that the Snail gene family plays an important role in EMT. 

Several signal-transduction pathways, including the activation of several receptor and nonreceptor tyrosine kinase receptors, are involved in the process of EMT both in epithelial cells and in embryonic development. ^{23 24 25} Among these, TGFβ is an important molecular player during EMT. ^{26 27} TGFβ2 has been proposed to be a signal for EMT and Slug induction in heart development. ²⁸ Signaling by other members of the TGFβ superfamily, the bone morphogenetic proteins, participates in the induction of the neural crest ²¹ by upregulating Slug. ^{29 30} TGFβ1 also induces EMT and Snail expression in hepatocytes. ³¹ However, the mechanisms that regulate the expression of Snail gene family members are still poorly understood. 

Smad proteins have been considered to be important mediators in the regulation of target gene expression induced by TGFβ. In response to TGFβ binding to the type II receptor, the TGFβ type I receptor is recruited to form complexes that lead to phosphorylation of receptor-regulated Smad2 and Smad3 (R-Smads). ³² The phosphorylated R-Smads bind to Smad4 (Co-Smad) to form a stable hetero-oligomeric complex, and the Smad complexes then translocate to the nucleus where they regulate target gene expression in collaboration with other transcription factors. ³³ Several reports suggest, however, that alternate pathways may also be involved in TGFβ-signal transduction. These include the small GTPase RhoA, PI3K/Akt, and MAPKs pathways. ³³ Previous studies have demonstrated the functional interaction between TGFβ and extracellular signal-regulated kinase (ERK). MAP kinase ERK is rapidly activated by TGFβ in culture models of EMT, and the specific inhibitor of MEK upstream of ERK blocks key morphologic features, such as the disassembly of E-cadherin-mediated adherens junction, in these models. ^{34 35 36}  

TGFβ is a strong candidate inducer of the EMT that characterizes anterior polar cataracts in the human lens. In this process, epithelial cells from the lens delaminate and undergo transdifferentiation toward a myofibroblastic phenotype, which is characterized by an increase in α-smooth muscle actin, the accumulation of abnormal ECM molecules (e.g., type I collagen, tenascin, fibronectin), and the loss of E-cadherin production. ^{5 37} We therefore attempted to determine whether the Snail gene family is induced in the process of TGFβ-induced EMT in lens epithelial cells and how TGFβ might regulate the expression of this gene family. 

Slug, but not Snail, was highly expressed in lens epithelial cells obtained from patients with anterior polar cataracts. Stimulation of lens epithelial cells by TGFβ1 in vitro induced the expression of Slug, and MAPK pathways were involved in this process. Analysis of the human Slug promoter indicated that TGFβ1 induced the expression of Slug through an Sp1 binding site. Our results also demonstrate that the repression of E-cadherin production by TGFβ1 was mediated by the induction of Slug. These findings suggest a novel mechanism for the regulation of Slug expression, and the consequent production of E-cadherin after TGFβ1 treatment of lens epithelial cells in vitro. 

Materials and Methods

Antibodies and Reagents

Anti-phosphospecific JNK/SAPK (Thr183/Tyr185) antibody was purchased from Cell Signaling (Beverly, MA). Anti-phosphospecific ERK1/2, anti-ERK1/2, JNK/SAPK, and Sp1 antibodies were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-Myc antibody was purchased from Calbiochem (La Jolla, CA). Anti-fibronectin, α-SMA, Flag, and actin antibodies were obtained from Sigma-Aldrich (St. Louis, MO). The protease inhibitors aprotinin, leupeptin, and pepstatin, the MAPK inhibitors U0126, SP600125, and SB202190, the PI3K inhibitor wortmannin, the Src inhibitor PP2, and the RhoA inhibitor Y27632 were obtained from Calbiochem. The phosphatase inhibitors sodium orthovanadate (Na3VO4) and sodium fluoride (NaF), cyclohexamide and mithramycin A were purchased from Sigma-Aldrich. 

Human Lens Capsules and Ex Vivo Rat Lens Epithelial Explants

Lens capsules with attached lens epithelial cells were obtained during cataract surgery from patients with the clinical diagnosis of nuclear or anterior polar cataracts, and specimens were obtained with informed consent, in accordance with the Declaration of Helsinki. The ages of patients ranged from 38 to 91 years. The lens capsules were immediately placed in transcription reagent (TRIzol; Invitrogen-Gibco, Grand Island, NY) for RNA preparation, or frozen in liquid nitrogen and stored at −70°C for further experiments. Rat lens epithelial explants were prepared from Sprague-Dawley (3 weeks old) rats, as previously described. ³⁸ The lens explants were incubated in Medium 199 with Earle’s salts (Invitrogen) supplemented with 0.1% bovine serum albumin (BSA; Invitrogen). 

Cells and Cell Culture

Human lens epithelial cell lines (HLE B3 and SRA 01/04) were kindly provided by Usha P. Andley (Washington University, St. Louis, MO) and Venkat N. Reddy (University of Michigan, Ann Arbor, MI), respectively. HLE B3 cells were cultured in Eagle’s minimum essential medium (MEM; Invitrogen) supplemented with 20% fetal bovine serum (FBS; Invitrogen). SRA 01/04 cells were grown in Dulbecco’s modified Eagles medium (DMEM; Invitrogen) with 10% FBS. To obtain quiescent cells, the cells were incubated in growth medium containing 1% FBS for 18 hours and then further cultured in serum-free medium for 24 hours. Serum-starved cells were stimulated with 10 ng/mL TGFβ1 (R&D systems, Minneapolis, MN) for the indicated times. 

Transfection and Luciferase assay

Transient transfections of HLE B3 cells with reporter and internal control (pRL-TK) plasmids were performed (Lipofectamine 2000; Invitrogen) and the reporter assay system (Dual-Luciferase; Promega, Madison, WI) was conducted, according to the manufacturers’ protocols. The amount of DNA in each transfection was kept constant by the addition of an appropriate amount of empty expression vector, and the internal control plasmid was used to control for transfection efficiency. Luciferase activities were determined 24 hours after transfection. 

Plasmids and Gene Silencing

The full-length cDNA for human Slug was amplified by PCR using primers specific for hSlug (Table 1 , c1F and c1R) from HLE B3 cells and subcloned into pEGFP-N3 (BD Biosciences-Clontech, Palo Alto, CA). The human Slug promoter fragment (full-length FL, 0.7 kb) was amplified by using specific primers (Table 1 , p1F and p1R) from human genomic DNA and cloned into pGL3-Basic (Promega). Deletion mutants M1 to M5 were generated by sequential restriction digests of FL. Site-directed mutations were introduced by standard PCR techniques by using two-step PCR with four primers: two of which are flanking primers containing XhoI and HindIII (Table 1 , p1F and p1R), respectively, and the other two primers (Table 1 , p2F, p2R, p3F, p3R, p4F, and p4R) include mutation site for the desired elements. Two PCR reactions were performed to obtain the two halves of the final product, which were combined in a third reaction using flanking primers to generate a chimerical product. The PCR products were cut with XhoI and HindIII and cloned into pGL3-basic, and the presence of the mutations was confirmed by sequencing. The constitutively active form of TGFβ receptor I, TβRI (T204D), was generously provided by Joan Massague (Memorial Sloan-Kettering Cancer Center, New York, NY). Wild-type Smad 2 and Smad 3, dominant negative form of Smad 2 and Smad 3 were kindly provided by Rik Derynck (University of California, San Francisco). The dominant negative form of Rac1 (Rac1 T17N), RhoA (RhoA T19N), and CDC42 (CDC42 T17N) were obtained from the UMR cDNA Resource Center (www.cdna.org). For siRNA knockdown experiments, two complementary oligonucleotides (Table 1 , si1F and si1R) were designed to contain nucleotides specific for Slug (AAAGACTACAGTCCAAGCTTT). After annealing, the DNA was cloned into a vector (pSilencer 2.1-U6 neo; Ambion, Inc., Austin, TX) according to the manufacturer’s recommendation, and the identity of the insert was confirmed by sequencing. Plasmids expressing the Slug siRNA (Slug siRNA) or the nonspecific negative control (pSilencer neo Negative Control; Ambion, Inc.) were transfected into HLE B3 cells (Lifectamine 2000; Invitrogen) according to the manufacturer’s protocol. These cells were treated with G418 (500 μg/mL) to create a population of cells stably expressing the slug siRNA for ∼10 days. The selected population of cells was stimulated with TGFβ1 for 16 hours. 

Real-Time PCR Quantification

Total RNA was extracted from clinic samples and HLE B3 cells (TRIzol; Invitrogen-Gibco) according to the manufacturer’s protocol. Real-time PCR quantification was performed (iCycler iQ Real-Time Detection System; Bio-Rad Laboratories, Inc, Hercules, CA) as follows: Two micrograms of total RNA was reverse-transcribed with AMV reverse transcriptase (Promega), and PCR was performed with specific primers (Table 1 , rt1F, rt1R, rt2F, rt2R, rt3F and rt3R). PCR amplification reactions included nucleic acid stain (iQ SYBR Green Supermix; Bio-Rad Laboratories, Inc). All samples were run in triplicate. The specificity of the amplification reactions was confirmed by melting curve analysis and subsequently by agarose gel electrophoresis. The threshold cycle (C _t) value for each gene was normalized to the C _t value for β-actin. 

Immunofluorescent Staining

Rat lens epithelial explants were exposed to 10 ng/mL of TGFβ1 for the indicated time, fixed, permeabilized, and incubated with primary antibodies against Slug (1:200; Santa Cruz Biotechnology, Inc.) or E-cadherin (Transduction Laboratories, Lexington, KY) for 1 hour at 37°C. The cells were then incubated with rhodamine-conjugated horse anti-rabbit immunoglobulin (1:200; Jackson ImmunoResearch, West Grove, PA) for 1 hour at 37°C in the dark. The nuclei were counterstained with Hoechst 33258 (Invitrogen-Molecular Probes, Inc., Eugene, OR) or propidium iodide (Sigma-Aldrich). Images were acquired with a confocal laser scanning microscope (MRC1024; Bio-Rad Laboratories, Inc.) or fluorescence microscope (Axiovert S100; Carl Zeiss Meditec, Inc., Oberkochen, Germany). 

Western Blot Analysis

Cells were lysed in radioimmunoprecipitation assay (RIPA) buffer (0.05 M Tris-buffer [pH 7.2], 0.15 M NaCl, 1% Triton X-100, 1% deoxycholate, 0.1% SDS) supplemented with protease inhibitor (2 μg/mL aprotinin, 2 μg/mL leupeptin, and 2 μg/mL pepstatin), and phosphatase inhibitor (1 mM Na₃VO₄, and 1 mM NaF). The samples containing 10 to 30 μg proteins were boiled in Laemmli sample buffer, separated on SDS polyacrylamide gels, electrophoretically transferred to nitrocellulose membranes (GE Healthcare, Arlington Heights, IL), and blotted with the indicated primary antibodies. Proteins were visualized with the horseradish peroxidase-conjugated secondary antibodies (Zymed Laboratory, Inc., South San Francisco, CA) followed by chemiluminescence (ECL-Plus; Santa Cruz Biotechnology) detection. 

Quantitative Chromatin Immunoprecipitation

The cells were treated with TGFβ1 for the indicated time, rinsed with PBS, and cross-linked with 1% formaldehyde. Crude cell lysates were sonicated to generate 300-to 3000-bp DNA fragments (Sonifier 250; Danbury, CT) followed by centrifugation for 10 minutes. The lysates were diluted 1:5 in chromatin immunoprecipitation dilution buffer (15 mM Tris [pH 8.0]), 1% Triton X-100, 0.01% SDS, 1 mM EDTA, 150 mM NaCl and protease inhibitors) and aliquots of the lysates (1% of total volume) were used for input controls for each immunoprecipitation. Immunoprecipitation was performed with anti-Sp1 or rabbit IgG antibodies (Zymed Laboratory), followed by incubation with salmon sperm DNA/protein G agarose (Upstate, Charlottesville, VA). The protein-DNA complexes were extracted and incubated overnight at 65°C to reverse the formaldehyde cross-linking. DNA fragments were purified (QIAquick Spin Kit; Qiagen Inc.), and the enrichment of specific promoter DNA was measured by real-time PCR with the promoter-specific primers (Table 1 , ch1F and ch1R) and adjusted relative to amplification of target sequences in the initial sonicated chromatin lysate (input control). 

Statistical Analysis

Results are expressed as the mean ± SD. Student’s t-test was used for statistical analysis and differences at P < 0.05 were considered significant. 

Results

Slug in Lens Epithelial Cells from Anterior Polar Cataracts

In anterior polar cataract, the lens epithelial cells undergo an EMT and adopt a mesenchymal phenotype, whereas in nuclear cataract, they retain an epithelial phenotype. We compared the expression of the Snail family members by real-time PCR in lens epithelial biopsy specimens from patients with polar and nuclear cataracts. The level of Slug mRNA was increased markedly in lens epithelial cells from anterior polar cataracts, compared with that in lens epithelial cells from patients with nuclear cataracts (Fig. 1A) . Reverse transcription polymerase chain reaction (RT–PCR) products corresponding to the mRNA of Snail were not detected. 

Consistent with the real-time PCR data, immunofluorescence experiments showed prominent cytoplasmic and nuclear Slug immunoreactivity in anterior polar, but not in nuclear cataract samples (Fig. 1B) . These results indicate that slug expression is upregulated in anterior polar cataract, which involved EMT of lens epithelial cells, but not in nuclear cataract. 

TGFβ1 Stimulation of Levels of Slug Expression

To determine whether Slug is a downstream target of TGFβ, we examined the expression and intracellular localization of Slug expression in lens epithelial cells. The expression and intracellular localization of Slug protein were analyzed by immunofluorescence assay in rat lens epithelial explants. In the absence of TGFβ1, the explants showed very weak immunoreactivity for the protein in both the nucleus and cytosol. However, in presence of TGFβ1, Slug protein production was greatly increased, and immunostaining was detected mainly in the nucleus (Fig. 2A) . Then we examined the level of Slug expression in lens epithelial cells. As expected, the mRNA level of fibronectin, which is a well-known marker of EMT, was increased by the treatment with TGFβ1. The level of Slug mRNA was also significantly elevated in human lens epithelial B3 cells stimulated with TGFβ1 (Fig. 2B) . 

Consistent with this mRNA expression, Slug protein synthesis was induced on stimulation by TGFβ1. An immunoblot assay with anti-Slug antibodies showed that the level of this protein was increased as early as 4 hours after treatment with TGFβ1; this increase was maintained for 24 hours and returned to baseline by 48 hours (Fig. 2C) . 

The induction of Slug gene expression by TGFβ1 was further confirmed by the Slug promoter assay. We isolated a 0.7-kb fragment of the 5′ flanking region of human Slug and tested its response to TGFβ1 in transiently transfected human lens epithelial B3 (HLE B3) cells, by using a luciferase reporter construct. We observed an increase in luciferase activity on TGFβ1 treatment (Fig. 2D) . The Slug promoter activity was increased by cotransfection with the expression plasmids for the constitutively activated form of the TGFβ receptor I, TβRI (T204D; Fig. 2E ). 

To determine whether the induction of Slug expression by TGFβ1 was restricted to lens epithelial cells, we analyzed the levels of Slug mRNA and protein in other cell lines: human lens epithelial SRA01/04 cells; human epidermal keratinocyte line HaCaT, breast cancer cell line MCF7, and the fibroblast line NIH 3T3. Treatment with TGFβ1 increased the levels of Slug mRNA and protein in SRA 01/04, HaCaT, and MCF7 cells (Fig. 2Fand data not shown). However, the levels of Slug mRNA were not changed in NIH 3T3 cells (data not shown). 

Effect of Activation of MAPK Pathways on Slug Expression by TGFβ1

To determine whether TGFβ1 directly regulates Slug transcription, the cells were pretreated with cycloheximide to block new protein synthesis before the addition of TGFβ1. RNA was prepared and analyzed by RT-PCR. To our surprise, Slug mRNA levels failed to respond to the treatment with TGFβ1 (Fig. 3) , suggesting TGFβ-induced slug expression requires translation of a labile protein. 

To identify the mechanism by which TGFβ1 regulates the induction of Slug. First, we analyzed the role of Smad in the induction of Slug by TGFβ1 in lens epithelial cells, because Smad is a well-known mediator of TGFβ signaling. ³² Transfection with dominant-negative forms of Smad2 or Smad3, which block the classic TGFβ–Smad signaling pathway, did not change the TGFβ1-stimulated induction of Slug (Fig. 4A) . 

We then examined the involvement of small GTPase proteins in the induction of Slug expression by TGFβ1, as these are important mediators in the TGFβ signaling pathway. ^{39 40 41} HLE B3 cells were transfected with dominant negative forms of RhoA, Rac1, or CDC42, followed by treatment with TGFβ1, and the levels of Slug expression were then analyzed. None of the constructs inhibited the induction of Slug by TGFβ1 (Fig. 4B) . 

Next, we examined whether MAPKs are activated by the treatment with TGFβ1 of lens epithelial cells. When HLE B3 cells were stimulated with TGFβ1, the phosphorylated forms of ERK were increased. Western blot analysis showed that the maximum activation of ERK occurred 10 minutes after treatment and that this activation was rapidly deactivated over 45 minutes. Stimulation with TGFβ1 also resulted in a rapid increase in the phosphorylated forms of c-Jun N-terminal kinase/stress–activated protein kinase (JNK/SAPK) with an activation profile similar to that of ERK, whereas there was no change in total ERK and JNK/SAPK expression (Fig. 4C) . 

We also investigated whether the TGFβ1-stimulated induction of Slug in HLE B3 cells involves the MAPK cascade, because this has been implicated in such signaling in other contexts. ⁴² When cells were pretreated with the MAPK kinase (MEK)-specific inhibitor U0216 and the JNK/SAPK-specific inhibitor SP600125 for 30 minutes followed by stimulation with TGFβ1, Slug expression levels were dramatically inhibited by 83.9% and 74.5%, respectively, compared with dimethylsulfoxide (DMSO) treatment. However, there were no noticeable effects on Slug expression when cells were treated with the p38-specific inhibitor SB202190, the phosphatidylinositol 3-kinase (PI3K)-specific inhibitor wortmannin, and the Src inhibitor PP2. Consistent with data in Figure 4B , treatment with the RhoA-specific inhibitor Y27632 did not block the induction of Slug by TGFβ1 (Fig. 4D) . 

Role of Sp1 Binding Site within the Slug Promoter in TGFβ1 Responsiveness

To gain further insights into the regulation of Slug expression by TGFβ1, we examined the putative regulatory elements. We used a series of Slug promoter deletion constructs to identify the region of Slug responsible for its induction by TGFβ1. Several mutant constructs of the Slug promoter were cotransfected in HLE B3 cells, together with TβRI (T704D), and the luciferase activity was analyzed. The results shown in Figure 5Ademonstrate that 5′-deletion analysis down to −296/+126 slightly increased Slug promoter activity. In contrast, removal of the Slug promoter region between positions −296 and −141 decreased Slug promoter activity by approximately 90.7%. We further deleted the region between positions −179 and −141 from the −296/+126 construct, to determine a narrowly defined region. Although the promoter activity was slightly increased, significant decreases of activity in TGFβ1 responsiveness were observed. 

We therefore performed sequence analyses of the Slug promoter. Analysis of the region between positions −179 and −141 revealed a sequence that contains two E-box elements, which are potential consensus binding sites for the transcription factor AP-4, and one GC-box element that is a potential consensus binding site for Sp1 (Fig. 5B) . To elucidate which of these transcription factors participates in TGFβ1 responsiveness, mutational analysis of the potential binding sites was performed. To determine whether both potential AP-4 binding sites are required for the induction of the Slug promoter by TGFβ1, different AP-4 binding site mutations were constructed: ΔEb1 and ΔEb2. Both constructs displayed slightly increased promoter activity compared to FL. By contrast, the mutant ΔGC, which contains a three-base substitution in the Sp1 site, resulted in a significant decrease of promoter activity (Fig. 5C) . 

Effect of TGFβ1 on the Expression of Slug through the Activation of Sp1

We examined whether TGFβ1 induces the phosphorylation of Sp1 in lens epithelial cells. When HLE B3 cells were treated with TGFβ, phosphorylation of Sp1 increased steadily with time. Immunoblot analysis showed that the maximum phosphorylation of Sp1 occurred 60 minutes after treatment and the levels of phosphorylated Sp1 then decreased over 4 hours (Fig. 6A) . To determine whether Sp1 physically associates with the endogenous Slug promoter, we performed a chromatin immunoprecipitation assay with anti-Sp1 antibodies in untreated and TGFβ1-treated HLE B3 cells, followed by quantitative real-time PCR analysis. TGFβ1 induced rapid and transient in vivo binding of endogenous Sp1 protein with a GC-box element in the Slug promoter, showing the highest association of Sp1 with Slug promoter at 1 hour after treatment with TGFβ (Fig. 6B) . We therefore tested whether the induction of Slug by TGFβ1 is mediated via Sp1 by interrupting Sp1 binding with the specific pharmacologic inhibitor, mithramycin A. HLE B3 cells were pretreated for 30 minutes with mithramycin A and then stimulated with TGFβ1. As expected, the TGFβ1-induced increase of Slug expression was significantly inhibited by treatment with mithramycin A in a dose-dependent manner. Also, the repression of E-cadherin by TGFβ1 was almost completely blocked by mithramycin A (Fig. 6C) . 

Effect of TGFβ1 in E-cadherin Production through an Increase in Slug Gene Expression

We investigated the role of Slug protein in TGFβ1 signaling in the EMT of lens epithelial cells. To examine whether TGFβ1 induces EMT, we performed immunofluorescence analysis in rat lens epithelial explants. Rat lens epithelial explants treated with TGFβ1 for 72 hours showed significant morphologic changes with a redistribution of E-cadherin from cell–cell contacts to the cytoplasm. Confocal immunofluorescence analysis demonstrated that E-cadherin at cell–cell contact points almost completely disappeared, compared with the untreated control cells, which displayed intense E-cadherin staining confined to cell–cell interfaces (Fig. 7A) . We also analyzed the expression of E-cadherin, α-SMA, and fibronectin as prototypic markers of EMT. Treatment of HLE B3 cells with TGFβ1 induced marked increases in fibronectin and α-SMA levels, whereas the production of E-cadherin decreased steadily with time (Fig. 7B) . 

We then examined whether Slug is necessary for the regulation of E-cadherin production by TGFβ1 in lens epithelial cells. To address this, we used an RNA interference method. We designed siRNAs specific for Slug and demonstrated that transfection with these siRNAs reduced levels of the Slug mRNA and protein. Analysis of the levels of E-cadherin in this experiment revealed that the TGFβ1-induced repression of E-cadherin production was significantly abolished by depletion of Slug, but the increase of α-SMA and fibronectin by TGFβ1 was not affected (Fig. 7C) . 

Discussion

Transcriptional repression of E-cadherin and the associated morphologic changes in cells occur during EMT in embryonic development and in tumor cell invasion. Several transcriptional repressors, including the Snail family, are implicated in such repression by interacting with the E-box sequence in the proximal E-cadherin promoter. ¹ TGFβ is a potent candidate for EMT of lens epithelial cells, which mainly causes anterior polar cataracts. ⁵ In this study, we demonstrated that the levels of Slug mRNA and proteins were highly increased in clinical samples obtained from patients with anterior polar cataracts. Although previous studies have shown that Snail was detected at an early stage of EMT in the injured lens epithelium of mice, ^{43 44} we did not detect Snail transcripts in the clinical samples. However, as the biopsy samples used in this study were obtained from anterior polar cataracts diagnosed clinically, it is likely that they represent late stages of EMT. It is possible that Snail is expressed at earlier stages of polar cataract formation. The observation that Snail expression was detected in both human lens epithelial B3 cells and rat explants stimulated by TGFβ (data not shown) supports this conclusion. We showed om the current study that TGFβ1 induces Slug expression in lens epithelial cells and other epithelial cells, indicating that Slug expression is controlled by signals downstream of TGFβ1. Immunofluorescence assays in rat lens epithelial explants showed that the production of Slug protein was greatly increased on stimulation with TGFβ1, and that its intracellular localization was mainly in the nucleus. This subcellular distribution of Slug gene expression in rat lens explants was very similar to that of ectopically expressed Slug–GFP in human embryonic kidney 293T cells, which showed complete nuclear staining (data not shown). TGFβ1-mediated induction of Slug was further confirmed by our analysis of the human Slug promoter, showing that its activity was induced either by TGFβ1 treatment or by cotransfection with the constitutively active form of the TGFβ receptor I, TbR I (T704D), indicating that the induction of Slug by TGFβ1 is mediated through the activation of TGFβ receptors. Taken together, these expression data from clinical samples and in vitro experiments, suggest that there is a direct link between the increase of Slug expression and anterior polar cataracts through the EMT of lens epithelial cells and that Slug is a downstream target of the TGFβ-signaling pathway. 

We investigated the regulatory mechanism by which TGFβ1 stimulates the induction of Slug expression. Previous studies have demonstrated that MAPKs, PI3K, the small GTPases RhoA and Rac1, and the Smads have all been implicated as mediators of some or all phenotypic aspects of TGFβ-induced EMT. ^{26 35 41 45 46 47} Our data show that Slug induction by TGFβ1 in lens epithelial cells appears to be primarily dependent on the MAPK pathways, but not on other pathways, including the Smad pathway, the small GTPase, Src, or PI3K pathways. In lens epithelial cells, activation of ERK was induced by early stimulation with TGFβ1, and treatment with the MEK inhibitor U0126 significantly inhibited the TGFβ1-mediated induction of Slug. These data are consistent with those in a previous study showing that the ERK pathway is involved in the E-cadherin regulation and induction of Slug expression. ⁴⁸ Among the most interesting findings in our study is that Slug induction by TGFβ1 was also dependent on the activation of JNK/SAPK. This is supported by our observation that stimulation of TGFβ1 in lens epithelial cells induced activation of JNK/SAPK and that the TGFβ1-mediated induction of Slug was significantly inhibited by treatment with the JNK/SAPK inhibitor SP600126. This finding suggests that the activation of both ERK and JNK/SAPK may be involved in the TGFβ1-mediated induction of Slug expression. 

We have further identified the putative regulatory elements in the promoter region between positions −179 and −141, which contains an Sp1 binding site that is required for TGFβ1-induced transcriptional activation of the Slug promoter. Previously, Sp1 was considered to be a factor responsible for basal expression of numerous genes. ⁴⁹ However, several recent studies indicate that it can also play a key role in the regulation of certain genes in response to specific signals. ^{50 51 52 53} There are several possible mechanisms by which TGFβ might activate gene expression through this transcription factor. In the current study, we show that TGFβ1 increased in vivo binding of Sp1 with the endogenous Slug promoter and that the treatment of lens epithelial cells with mithramycin, an inhibitor of Sp1–DNA binding, dramatically inhibited the expression of endogenous Slug, as well as the TGFβ1-mediated repression of E-cadherin. This result suggests that TGFβ1 regulates Slug expression and repression of E-cadherin by altering the DNA binding activities of Sp1. Previous studies have shown that increased activation of the ERK pathway enhances the phosphorylation of Sp1 and the binding of Sp1 to a target sequence. ^{54 55 56} In addition, the phosphorylation of Sp1 is achieved through either ERK or JNK/SAPK, or through both of them. ⁵⁷ Therefore, in lens epithelial cells, activation of the ERK and JNK/SAPK pathways by stimulation of TGFβ1 may augment the phosphorylation of Sp1, which in turn affects Slug expression. Taken together, these results suggest that Sp1 plays a role in the induction of Slug expression and the repression of E-cadherin by TGFβ1. We have demonstrated here that TGFβ1-induced repression of E-cadherin was almost completely abolished by depletion of Slug using its siRNA, suggesting that TGFβ1 may regulate E-cadherin expression through the induction of Slug. However, the levels of fibronectin and α-SMA, which were increased during EMT of lens epithelial cells induced by TGFβ1, were not changed by overexpression of Slug, suggesting that TGFβ may use different regulation pathways for a complete EMT in lens epithelial cells. Previous studies have demonstrated that overproduction of Smad2, Smad3, and Smad4 induces stress fiber formation in murine mammary gland epithelial NMuMG cells. ⁴¹ In addition, TGFβ regulates actin cytoskeleton and adhesion junctions through RhoA-dependent signaling pathways in these cells. Thus, the repression of E-cadherin may be necessary, but is not sufficient to induce EMT of lens epithelial cells by TGFβ. 

In summary, TGFβ1 induces phosphorylation of the ERK and JNK/SAPK pathways through the activation of TGFβ receptors. Subsequently, these activated MAPKs may enhance the phosphorylation of Sp1 and its DNA binding activity, leading to the induction of Slug expression. We cannot, however, rule out the possible involvement of additional factors in the pathway between TGFβ1 and Slug. The cycloheximide experiments that we performed on cultured cells further support the existence of such a factor(s). The increased level of Slug in response of TGFβ1 could result in the repression of E-cadherin production through binding to the E-box element of its promoter (Fig. 8) . The signaling cascade described in this work provides a novel mechanism by which TGFβ1 may induce Slug expression. Furthermore, we suggest that the induction of Slug expression may play an important role in the EMT of lens epithelial cells via TGFβ1 and in the formation of anterior polar cataracts. 

²  Present affiliation: Department of Biochemistry and Molecular Biology, Keck School of Medicine, University of Southern California, Los Angeles, California.

 Supported by Grant 03-PJ1-PG3-20700-0019 of the Korea Health 21 R&D Project, Ministry of Health and Welfare, Republic of Korea.

 Submitted for publication June 13, 2006; revised November 13, 2006, and March 5, 2007; accepted April 23, 2007.

 Disclosure: J. Choi, None; S.Y. Park, None; C.-K. Joo, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Primers Used for PCR Amplification

Code	Primers	Strand	Sequence (5′ → 3′)
c1F	hSlug FL	Sense	TCTGCGGTACCATGCCGCGCTCCT
c1R		Antisense	GGATCCGTGTGCTACACAGCAGCCAGATTC
p1F	hSlug-promoter	Sense	GGGCTCGAGTTTTCAAAAGCCAAGAGG
p1R		Antisense	AAAAGCTTAAGGAGCGCGGGATCTTGCC
p2F	hSlug-promoter ΔEb1	Sense	ATTCTTCCTCTGTTCACGGTTGTCCCAGAGGGA
p2R		Antisense	AGCTCCTCCCTCTGGGACAACCGTGAACAGAGGA
p3F	hSlug-promoter ΔEb2	Sense	TGAAATCTGAACCTCTCGGTTGTGATTGGATCT
p3R		Antisense	AGAAAGATCCAATCACAACCGAGAGGTTCAGA
p4F	hSlug-promoter ΔGC	Sense	TTCACAGCTGTCCCAGAGTTAGAAGCTGAAAT
p4R		Antisense	AGAGGTTCAGATTTCAGCTTCTAACTCTGGGACA
si1F	Slug siRNA	Sense	GATCCAGACTACAGTCCAAGCTTTCTCAAGAGAAAAGCTTGGACTGTAGTCTTTTTTTGGAAA
si1R		Antisense	AGCTTTTCCAAAAAAAGACTACAGTCCAAGCTTTTCTCTTGAGAAAGCTTGGACTGTAGTCTG
rt1F	hSlug	Sense	ATAAGCAGCTGCACTGCGAT
rt1R		Antisense	ATTCCTCATGTTTGTGCAGGA
rt2F	Fibronectin	Sense	CAAGTGTGACCCTCATGAGG
rt2R		Antisense	CTTGGCAGAGAGACATGCTT
rt3F	β-Actin	Sense	AGGCCAACCGCGAGAAGATGACC
rt3R		Antisense	GAAGTCCAGGGCGACGTAGCAC
ch1F	Sp1	Sense	AGTCTGACAACAGATTC
ch1R		Antisense	TCCTTTACGAACTGAGC

Figure 1.

 

Slug was upregulated in lens epithelial cells from patients with anterior polar cataracts. (A) Total RNAs from lens epithelial cells from patients with anterior polar (AP) and nuclear (NU) cataracts were purified, and real-time PCR quantification was performed for the analysis of Slug, Snail, fibronectin, and β-actin. The normalized levels of Slug, Snail, and fibronectin mRNAs based on the level of β-actin mRNA show that the level of Slug was upregulated, whereas Snail was not detected in the clinical samples of anterior polar and nuclear cataracts. Fibronectin (FN) was used as a positive marker of the anterior polar cataracts. (B) Lens capsules with attached lens epithelial cells obtained from clinical samples of anterior polar and nuclear cataracts were carefully flatmounted on a slide glass. After fixation, specimens were subjected to indirect immunofluorescence (IF) staining with anti-Slug antibody and counterstained with propidium iodide (PI). Images of wholemounts of lens cells were taken by confocal microscopy; merged and phase-contrast images are also shown. Higher magnifications of boxed regions (B a–B c) show nuclear localization of Slug. Arrowheads: Slug-positive cells. Bar, 100 μm.

Figure 2.

 

TGFβ1 induced the expression of Slug. (A) Rat lens epithelial explants were incubated in the absence or presence of 10 ng/mL TGFβ1. After 24 hours of culture, the explants were immunostained with anti-Slug antibodies and counterstained with propidium iodide (PI). Bar, 50 μm. (B) Total RNA was purified from HLE B3 cells treated with 10 ng/mL of TGFβ1 for 24 hours. RT-PCR was performed for the analysis of Slug and fibronectin. β-Actin was used as an internal control. (C) Serum-starved HLE B3 cells were stimulated with TGFβ1 for the indicated time, and then cell lysates were immunoblotted with anti-Slug antibody. Actin was used as the loading control. (D) Serum-starved HLE B3 cells were transfected with human Slug promoter reporters. Cells were stimulated with TGFβ1 for 16 hours and luciferase activity determined. The error bars represent the mean ± SD of results in triplicate plates. *P < 0.05 compared with FL+TGFβ (−). (E) Human Slug promoter reporters were cotransfected into the HLE B3 cells with empty expression vector or constitutively activated form of the TGFβ receptor I, TβRI (T204D). Cells were incubated for 24 hours and luciferase activity was determined. *P < 0.05 compared with FL+TβRI (T204D) (−). (F) Analysis of Slug mRNA and protein in HLE B3 cells, stimulated with TGFβ1 for 16 hours, by RT-PCR and immunoblot. Expression of β-actin was used as the loading control.

Figure 3.

 

TGFβ1 indirectly regulated Slug transcription. (A) Expression of Slug mRNA and protein was analyzed by RT-PCR and Western blot in quiescent HLE B3 cells that were pretreated with cycloheximide (CHX, 3 μM) and incubated in absence or presence of TGFβ for 16 hours. (B) Quantification of RT-PCR data reveals that TGFβ-induced expression of Slug is significantly inhibited by CHX. The error bars represent the mean ± SD of results in triplicate experiments. *P < 0.05 compared with DMSO+TGFβ (+).

Figure 4.

 

TGFβ1 regulates Slug expression through the activation of ERK and JNK/SAPK. (A) HLE B3 cells were transiently transfected with Flag-tagged wild-type Smad 2, Smad 3, dominant-negative Smad 2 (dn-Smad2), or Smad 3 (dn-Smad3) constructs. After a 24-hour incubation, cells were treated with or without TGFβ1 for 16 hours before lysis. Whole-cell lysates were probed with anti-Slug and anti-Flag antibodies. Actin was used as a loading control. (B) Myc-tagged dominant-negative Rac1 (Rac1 T17N), RhoA (RhoA T19N), or CDC42 (CDC42 T17N) constructs were transiently transfected into HLE B3 cells. The transfected cells were incubated for 24 hours, followed by stimulation with or without TGFβ1 for 16 hours. Cell lysates were subjected to Western blot analysis using anti-Slug, Myc, and actin. (C) Serum-starved HLE B3 cells were stimulated with 10 ng/mL of TGFβ1 for the indicated time. Whole-cell lysates were subjected to Western blot analysis using either specific antibody against the active phosphorylated forms of ERK, and JNK/SAPK, or anti-ERK, and JNK/SAPK antibodies that react with both active and inactive forms. (D) The quiescent cells were treated with the specific MEK inhibitor U0126 (10 nM), JNK/SAPK inhibitor SP600125 (10 nM), p38 inhibitor SB202190 (10 nM), PI3K inhibitor wortmannin (10 nM), Src inhibitor PP2 (4 nM), or RhoA inhibitor Y27632 (10 nM) for 30 minutes, and then incubated in absence or presence of TGFβ1 for additional 16 hours. Cell lysates were subjected to Western blot analysis using anti-Slug and actin. The blots were scanned and analyzed by densitometry then represented in graphs as average ± SD of three independent experiments. *P < 0.05 compared with DM+TGFβ (+). DM, DMSO; U0, U0126; SB2, SB202190; SP, SP600125; WO, wortmannin; Y27, Y27632.

Figure 5.

 

The p1 binding site within the Slug promoter was responsible for slug induction by TGFβ1. (A) Diagram of sequential deletions (M1–M5) from the 5′ end of the Slug promoter (full-length FL). The Slug promoter deletion constructs were transfected into HLE B3 cells with TβRI (T740D), and luciferase activity was determined. The error bars represent the mean ± SD from triplicate plates. *P < 0.05 compared with FL. (B) Sequence of the cis-acting region in the Slug promoter involved in the response to TGFβ1 stimulation showing positions of two E-box elements and a GC-box element. (C) Schematic diagram of substitution mutations of E-box element (⋄) or GC-box element (▿) within the Slug promoter are depicted. These mutation constructs were transfected into HLE B3 cells with TβRI (T740D), and luciferase activity was determined. Error bars: the mean ± SD of results in triplicate plates. *P < 0.05 compared with FL.

Figure 6.

 

TGFβ1 induced the expression of Slug through Sp1. (A) Serum-starved HLE B3 cells were stimulated with 10 ng/mL of TGFβ1 for the indicated time. Whole-cell lysates were subjected to Western blot analysis using anti-Sp1 antibodies. Actin was used as a loading control. (B) Cells were treated with TGFβ1 for the indicated time and then subjected to the quantitative chromatin immunoprecipitations (ChIP) assay. A ChIP assay with anti-Sp1 antibody showed that a product amplified for the Sp1-binding site was increased rapidly and transiently by TGFβ1, indicating that Sp1 is recruited to the Slug promoter in response to TGFβ1. Input DNA was used as a positive control, and rabbit IgG was used as a negative control for antibodies (top). The error bars represent the mean ± SD from three independent experiments. *P < 0.05 compared with TGFβ (−). (C) The quiescent HLE B3 cells were pretreated with different doses of mithramycin A as indicated for 30 minutes, and then incubated in the absence or presence of TGFβ1 for an additional 16 hours. Cell lysates were probed with anti-Slug, E-cadherin, and actin antibodies.

Figure 7.

 

TGFβ1 induced the downregulation of E-cadherin through alterations in Slug levels. (A) Rat lens epithelial explants were incubated in the absence or presence of 10 ng/mL TGFβ1. After 72 hours of culture, the explants were immunostained with anti-E-cadherin antibodies. Bar, 50 μm. (B) The quiescent HLE B3 cells were stimulated with TGFβ1 for the indicated time. Immunoblot analysis of whole-cell lysates were probed for fibronectin, E-cadherin, and α-SMA. Actin was used as loading control. (C) RNAi was performed in HLE B3 cells using siRNA for Slug. Total RNA was analyzed by RT-PCR for Slug and β-actin. The ethidium-stained gel is depicted alongside quantitation by real-time PCR. Whole-cell lysates from HLE B3 cell transfected with Slug siRNA were analyzed by immunoblot with anti-Slug, E-cadherin, fibronectin, α-SMA, and actin antibodies. Error bars: the mean ± SD of results from three independent experiments. *P < 0.05 compared with NC+TGFβ (+).

Figure 8.

 

Model of induction of Slug expression by TGFβ1. A model proposed to illustrate that TGFβ1 induces the expression of Slug through activation of Sp1 by MAPK pathway and thus represses E-cadherin production, which lead to initiate EMT of lens epithelial cells.