TGFβ/Smad4-Dependent and -Independent Regulation of Human Lens Epithelial Cells

Abstract

Purpose.: Transforming growth factor (TGF)-β can mediate fibrotic responses via Smad4-dependent and -independent signaling pathways. TGFβ-induced transdifferentiation of lens epithelial cells to myofibroblasts and matrix contraction contribute to posterior capsule opacification (PCO) after cataract surgery. The primary objective of the study was to determine the importance of Smad4 in TGFβ2-induced transdifferentiation, matrix contraction, and Smad signaling by human lens epithelial cells.

Methods.: The human lens cell line FHL 124 was used. Smad4 knockdown was achieved by using siRNA technology and efficiency tested at the message and protein level by real-time PCR and Western blots, respectively. Smad4 and Smad2/3 nuclear distribution after TGFβ2 stimulation (10 ng/mL) was determined by immunocytochemistry. Gene expression of transdifferentiation markers, α-smooth muscle actin and fibronectin and the inhibitory Smad, Smad7, in the presence and absence of TGFβ2 (10ng/mL) was determined with real-time PCR. TGFβ2-induced contraction was assessed with a patch contraction assay. Suspended bead array technology was used to determine pERK, pP38, and pJNK levels in response to TGFβ2 over a 2-hour time course.

Results.: SiRNA targeted against Smad4, successfully reduced expression of Smad4 message and protein after both the initial transfection period and removal of transfection conditions. These findings were confirmed by immunocytochemistry for Smad4, which showed a significant reduction in nuclear Smad4 after siSmad4 treatment relative to control siRNA in both the presence and absence of 10 ng/mL TGFβ2. Smad4 knockdown cells exhibited TGFβ2-induced Smad2/3 nuclear translocation which did not differ significantly from Smad4-expressing cells. Real-time PCR analysis showed significant induction of αSMA and fibronectin transcript in the Smad4-expressing cell line (control siRNA group). These inductions were suppressed in Smad4-knockdown cells. TGFβ-induced mSmad7 expression in FHL 124 cells was unaffected by Smad4 knockdown. Smad4-expressing cells did not exhibit a significant contractile response after 24 hours of culture in the presence or absence of 10 ng/mL TGFβ2. However, Smad4-knockdown cells demonstrated a significant reduction in patch area (i.e., contraction), after TGFβ2 treatment. Bead array analysis showed that TGFβ2 significantly increased pERK and pP38 levels relative to unstimulated control. No significant change was observed with pJNK.

Conclusions.: Smad4 is critical for the expression of transdifferentiation markers, αSMA, and fibronectin; in contrast, Smad4 signaling is not essential for TGFβ-induced Smad7 expression. Smad4 does not appear to be necessary for TGFβ-induced matrix contraction. TGFβ does activate Smad-independent pathways in human lens epithelial cells. Therefore, TGFβ2 regulation of transdifferentiation and matrix contraction appears to be mediated by distinct TGFβ signaling pathways.

Induction of the transforming growth factor (TGF)β signaling cascade after surgical injury has been implicated as a causative factor in the pathologic fibrosis of the lens termed posterior capsule opacification (PCO). ¹ PCO or “after cataract” is the most common complication of cataract surgery, and results from the resilient growth onto the posterior capsule of residual lens epithelial cells. ¹ Analysis of human postmortem capsular bag tissue from donor eyes that had undergone cataract surgery has identified a nuclear distribution of the classic TGFβ-signaling molecules, Smad proteins, ² TGFβ isoforms, and receptors. ³ TGFβ2 is the major isoform within the eye, most of which is detected in the aqueous humor and exists largely in the latent form ^4,5 ; however, after trauma (e.g., by surgical injury) active levels of all TGFβ isoforms can be elevated. ⁴  

Transdifferentiation of lens epithelial cells to a myofibroblast phenotype and contraction are key markers of fibrosis and play critical roles in PCO. Most important, wrinkling of the posterior capsule, as a result of matrix contraction, can induce formation of light-scattering regions. ¹ TGFβ is a potent inducer of matrix contraction in numerous cell types, including lens epithelial cells. ^6–9 Human capsular bags, derived from a sham cataract operation, when cultured with TGFβ2 reveal wrinkling (matrix contraction) of the posterior capsule along with α-smooth muscle actin (αSMA) expression, the major marker of transdifferentiation. ⁹ Moreover, these molecular and physical changes were also observed in postmortem analysis of lens tissue from a donor who had undergone cataract surgery 1 month before death. ⁹ In a study of the human lens cell line FHL 124 αSMA expression and matrix contraction were proposed to be a consequence of increased Smad4-mediated transcriptional activity in response to TGFβ2. ¹⁰ However, the study did not determine whether Smad4 was critical for the process of TGFβ-induced transdifferentiation and matrix contraction. Recently, a targeted inhibition approach to determine the relationship between transdifferentiation of human lens epithelial cells and matrix contraction was performed. ⁶ In contrast to conventional wisdom, which suggests that transdifferentiation gives rise to contractile events, ¹¹ it was revealed that αSMA expression and fibronectin/fibronectin receptors are not critical in TGFβ-induced matrix contraction. ⁶ Therefore, the TGFβ signaling mechanisms that actually regulate transdifferentiation and matrix contraction of lens epithelial cells must be elucidated. Consequently, determining the importance of TGFβ/Smad4 signaling in these fibrotic responses is fundamental. 

Smad4 is a central signaling component of the Smad pathway that transduces signals from TGFβ, ¹² a key mediator of fibrotic diseases throughout the body. ¹³ On TGFβ ligand stimulation, Smad4 forms a heteromeric complex with receptor regulated Smad2 and -3 (R-Smads), which are phosphorylated by an activated TGFβ receptor complex consisting of two type I and two type II transmembrane serine threonine kinase receptors. ¹⁴ Similarly, phosphorylation-dependent interaction of Smad1, -5, and -8 (R-Smads) with Smad4 occurs in bone morphogenetic protein (BMP) and growth and differentiation factor (GDF) signaling, ¹⁵ although TGFβ-induced phosphorylation of R-Smad1/5 has recently been demonstrated in nonocular epithelial cells. ¹⁶ The R-Smad/Smad4 heteromeric complex is free to translocate from the cell cytoplasm to the nucleus and is competent to associate with transcriptional co-activators and co-repressors to direct specific transcriptional responses. ¹² TGFβ can terminate the induction of its own target genes by recruiting Smad7, an inhibitory Smad that prevents R-Smad phosphorylation. ¹⁷  

Smad signaling pathways are widely considered to be central to the transduction of TGFβ signals and have been implicated in numerous fibrotic diseases (reviewed in Ref. 18); however, increasing evidence suggests that TGFβ can activate Smad-independent pathways (reviewed in Ref. 19). In an elegant study, Levy and Hill ²⁰ used a gene knockdown technique, small interfering RNA (siRNA), against Smad4, to determine the extent to which Smad4 is necessary in TGFβ's regulating of target gene transcription in human keratinocyte and pancreatic tumor cell lines. Of 114 TGFβ target genes investigated, 49 genes were Smad4 dependent and 65 were independent. Evidence obtained in recent years suggests TGFβ may signal via alternate (non-Smad) pathways which can either interact with the Smad signaling cascade or be independent from it (reviewed in Ref. 19). Moreover, R-Smads may mediate transcriptional responses independent of Smad4 recruitment. ^21,22 Therefore, the importance of Smad4 in TGFβ signaling events that underpin fibrosis is an area of recent interest. For instance, human dermal fibroblasts that express dominant negative Smad4 constructs demonstrate impaired TGFβ-induced contraction of collagen gels. ²³ In addition, siRNA against Smad4 inhibits TGFβ-induced epithelial-to-mesenchymal transition of breast cancer cells. ²⁴ In contrast, TGFβ-induced expression of extracellular matrix components is observed in Smad4-deficient murine fibroblasts ²⁵ ; moreover, Smad4 was not found to be necessary for TGFβ-induced fibronectin synthesis in a human fibrosarcoma cell line. ²⁶  

A body of evidence has been obtained regarding the involvement of TGFβ signaling in the pathogenesis of fibrotic disorders of the lens. ^7,9,27 For example, TGFβ has been shown to induce anterior subcapsular cataract (ASC) in a rat lens culture model. ^7,27 TGFβ is a potent inducer of the transdifferentiation of lens epithelial cells to myofibroblasts, and analysis of human ASC tissue has revealed elevated levels of the myofibroblast markers αSMA and fibronectin. ²⁸ To understand the role of TGFβ1 signaling in ASC cells, a TGFβ1/Smad3-knockout mouse was recently used. ²⁹ It was found that TGFβ1-induced ASC plaques and transdifferentiation in the absence of Smad3. An additional study was performed in Smad3-knockout mice to determine wound-healing events after a puncture injury of the lens epithelium. ³⁰ In that study, impaired transdifferentiation of lens epithelial cells was observed. Although the results of these studies do not directly link Smad4 to ASC formation they suggest that Smad4 through its interaction with the Smad2/3 heteromeric complex plays a role. 

We used a gene expression knockdown technique, siRNA against Smad4, to determine whether Smad4 has a role in TGFβ-induced transdifferentiation and matrix contraction of human lens epithelial cells (FHL 124). Disruption of Smad4 by siRNA perturbed TGFβ-induced αSMA and fibronectin expression, whereas TGFβ-induced matrix contraction and Smad2/3 nuclear translocation were unaffected. Moreover, we demonstrate that TGFβ can stimulate Smad-independent signaling pathways in human lens epithelial cells. We therefore provide evidence that TGFβ2 regulates transdifferentiation and matrix contraction by distinct signaling pathways. These findings have great relevance to posterior capsule opacification, a fibrotic condition that develops after cataract surgery. 

Materials and Methods

SiRNA Transfection

Custom-made Smad4 siRNA: sense 5′-GGUGGAGAGAGUGAAACAUtt-3′, antisense 5′-AUGUUUCACUCUCUCCACCtt-3′ and control siRNA (custom-made sequence with no detectable targets) were used. Both siRNAs were purchased from Ambion (Huntingdon, UK). FHL 124 cells were seeded onto 35-mm culture dishes at either (1) 25,000 cells in 1.5 mL for RNA and protein extraction, as four patches of 5,000 cells, for patch-contraction assay analysis, or (2) onto sterile glass coverslips at 7,000 cells in 100 μL of 5% FCS-EMEM (for immunocytochemistry). The cells were maintained in EMEM supplemented with 5% FCS for 3 days and then serum starved for 1 day. Transfections were performed with 100 nM siRNAs according to the manufacturer's instructions. Briefly, 1 μL Smad4 siRNA or control siRNA (final concentration, 100 nM) was added to 184 μL reduced-serum medium (Optimem; Invitrogen, Ltd., Paisley, UK). In addition, 5 μL oligofectamine (Invitrogen) was added to 10 μL of the medium. The two solutions were incubated at room temperature for 5 minutes and then mixed by gentle agitation with incubation at room temperature for a further 15 to 20 minutes. Meanwhile, the serum-containing medium was aspirated from the cell preparations and replaced with 2 mL of the reduced-serum medium. This solution was then aspirated and replaced with 800 μL of fresh reduced-serum medium. After the incubation period, 200 μL of siRNA transfection mix was added to the cell preparations. The cells were incubated at 35°C in a 5% CO₂ atmosphere for 4 hours, to initiate transfection, followed by the addition of 500 μL of 6% FCS-EMEM for up to 48 hours to “knockdown” Smad4 gene and protein expression. After this period, the transfection medium was removed and replaced with 1.5 mL of nonsupplemented EMEM, and the cells were cultured for a further 24 hours, followed by the addition of TGFβ2 at 10 ng/mL and incubation in either 1.5 mL of nonsupplemented EMEM or 2% FCS-EMEM. The cells were lysed after 24 hours (for RNA extraction). Glass coverslip cultures (for immunocytochemistry) and patch assays were terminated after 2 or 24 hours, respectively, in experimental conditions. 

Quantitative Real-Time PCR

The human lens cell line FHL 124 was seeded onto 35-mm dishes at ∼30,000 cells in 400 μL of 5% FCS-EMEM (Invitrogen Ltd.) and were maintained in 1.5 mL of 5% FCS-EMEM for 3 days. The medium was replaced with nonsupplemented EMEM and cultured for a further 24 hours before experimental conditions were applied. After 24 hours in experimental conditions, the RNA was collected from the cells (RNeasy mini kit; Qiagen Ltd., Crawley, UK). RNA (FHL 124 cells; 500 ng) was reverse transcribed in a 20-μL reaction mixture (Superscript II RT; Invitrogen). Quantitative RT-PCR (QRT-PCR) was performed (Opticon 2 DNA Engine; MJ Research Inc, Reno, NV). Primer oligonucleotide sequences specific for the genes examined are shown in Table 1. The level of product was determined by SYBR green (Finnzymes Oy, Espoo, Finland). A 50-μL reaction mixture was prepared for each cDNA sample containing: 50 ng cDNA, 2× SYBR green, 2 μM forward and reverse primers (Invitrogen), and double-distilled water to the total final volume. Serial dilutions of cDNA known to express the gene of interest were prepared, to permit relative levels between test samples to be determined. QRT-PCR was performed with the following program: step 1, initial denaturation for 94°C 4 minutes; step 2, denaturation for 94°C for 20 seconds; step 3, annealing at 55°C for 30 seconds; step 4, extension at 72°C for 20 seconds; step 5, “cut off” for 10 seconds at either 80°C (GAPDH, αSMA, Smad7), 77°C (fibronectin) or 75°C (Smad4) to denature the primer dimers; the fluorescence was then measured. Steps 2 to 5 were repeated for 35 cycles. Melting curve analysis was performed to determine the properties of the amplicon. 

Table 1.

 

Primer Sequences Employed for Real-Time PCR

Primer Name	Forward Sequence	Reverse Sequence	Reference
GAPDH	ACCACAGTCCATGCCATCAC	TCCACCACCCTGTTGCTGTA	Wormstone et al. 10
Smad 4	GACTGAGGTCTTTTACCGTTGG	CTTCAAGCTCTGAGCCATGC	e-PCR:NM_005359.3*
Alpha SMA	CCCAGCCAAGCACTGTCA	TCCAGAGTCCAGCACGATG	Lee and Joo 28
Fibronectin	CAGGATCACTTACGGAGAAACAG	GCCAGTGACAGCATACACAGTG	Spirin et al. 31
Smad7	AAAGTGTTCCCTGGTTTCTCCATCAAGGC	CTACCGGCTGTTGAAGATGACCTCCAGCCAGCAC	Kitamura et al. 32

*  e-PCR sequences obtained from http://www.ncbi.nlm.nih.gov.

Western Immunoblot Analysis

FHL-124 cells were washed with 1.5 mL PBS then lysed on ice in 0.5 mL Daub's lysis buffer; (50 mM HEPES [pH 7.5], 150 mM NaCl, 1% Triton X-100, 1 mM EDTA, 10% glycerol, 10 mM sodium pyrophosphate, 2 mM sodium orthovanadate, 10 mM sodium fluoride, 250 mL ddH₂O, 1 mM phenylmethylsulfonylfluoride, and 10 μg/mL aprotinin). Lysates were precleared by centrifuging at 13,000 rpm at 4°C for 10 minutes, and the protein content of the soluble fraction was assayed by bicinchoninic acid (BCA) protein assay (PerBio, Cramlington, UK). Equal amounts of protein per sample were loaded onto 10% SDS-PAGE gels for electrophoresis and transferred onto a polyvinylidene fluoride (PVDF) membrane (Perkin Elmer, Waltham, MA; with a Trans-Blot semidry Transfer Cell; Bio-Rad, Hercules, CA). Proteins were detected using the ECL+ blot analysis system (GE Healthcare, Buckinghamshire, UK) with anti-Smad4 (Santa Cruz Biotechnology, Santa Cruz, CA), anti-Smad2/3 (BD Transduction Laboratories, , Lexington, KY) and anti-β-actin (Cell Signaling Technology, Herts, UK). Gels were scanned (Scanjet 5470c; Hewlett Packard Development Co., Palo Alto, CA) and band intensity was determined (1D 3.5 software; Eastman Kodak, Rochester, NY). 

Immunocytochemistry

FHL 124 cells were seeded onto sterile glass coverslips (placed in 35-mm tissue culture dishes) at 7000 cells in 100 μL of 5% FCS-EMEM and maintained in 1.5 mL of 5% FCS-EMEM until they were 50% to 70% confluent, in a 35°C 5% CO₂ incubator. The medium was then removed, replaced with nonsupplemented EMEM, and cultured for a further 24 hours, followed by exposure to experimental transfection conditions: Smad4 siRNA and control siRNA for 48 hours. The transfection medium was replaced with 1.5 mL of nonsupplemented EMEM, and culture continued for a further 24 hours. After incubation, the cells were stimulated with TGFβ2 at 10 ng/mL for 2 hours, followed by fixation with 1.5 mL of 4% formaldehyde, and three washes in quick succession with PBS. The FHL 124 cell membranes were permeabilized with the addition of 0.5% Triton X-100/PBS for 30 minutes, followed by three washes at 10-minute intervals, with shaking, in 0.02% BSA and 0.05% Igepal in PBS. Nonspecific cellular sites were blocked with the addition of normal goat serum (1:50) in 1% BSA in PBS, with incubation for 1 hour at 35°C. Anti- Smad2/3 (BD-Transduction Laboratories, Dorset, UK) and anti-Smad4 (BD-Transduction Laboratories) antibodies were diluted 1:100 in 1% BSA in PBS and applied for 1 hour at 35°C, followed by washing three times at 15-minute intervals, with shaking, in 0.02% BSA and 0.05% Igepal in PBS. Smad2/3 and -4 were visualized with Alexa 488-conjugated secondary antibodies (Molecular Probes, Leiden, Netherlands), which were applied at 1:200 in 1% BSA in PBS for 1 hour in the dark at 35°C. The F-actin cytoskeleton was stained with Texas Red-X-phalloidin, and chromatin was stained with DAPI (diamidino-2-phenylindole hydrochloride; Molecular Probes), both stains were diluted in 1% BSA in PBS and applied at 2 U/mL and 10 μg/mL, respectively, for 10 minutes in the dark at room temperature. The coverslips were washed three times at 15-minute intervals in the dark with shaking in 0.02% BSA/0.05% Igepal in PBS, floated onto microscope slides, and mounted (Hydromount; National Diagnostics, Atlanta, GA). 

Quantification of Smad2/3 and -4 Nuclear Levels

To quantify nuclear Smad levels, we analyzed fluorescent micrographs of Smad2/3 and -4 with their corresponding chromatin images. Separate Smad2/3 and -4 and chromatin fluorescent micrographs were transformed into binary images by thresholding (Photoshop ver. 8.0; Adobe, San Jose, CA). The images were then transferred to Scion Image (Frederick, MD), and the black areas (fluorescent regions) within the image calculated. The nuclear area, detected by chromatin, was used to calibrate for differences in the number of cells. 

Patch Contraction Assay

FHL 124 cells were seeded at four sites on a tissue culture dish at 5000 cells in 25 μL and maintained in EMEM supplemented with 5% FCS until confluent regions spanning approximately 5 mm developed. ^6,10 The medium was then replaced with nonsupplemented EMEM, and the cells were cultured for a further 24 hours, followed by fixation with 4% formaldehyde for 30 minutes at room temperature and then washing in PBS. The patches were used as a t = 0 reference control. All remaining cell cultures were exposed to transfection conditions (i.e., Smad4 siRNA and control siRNA for 48 hours), followed by removal of transfection medium, replacement with 1.5 mL of nonsupplemented EMEM, and culture for a further 24 hours. Patch cultures were placed into experimental conditions of TGFβ2 at 10 ng/mL and maintained in 2% FCS for 24 hours. Experiments were terminated on the appearance of cell-free regions (holes) within the central region of the patch, a consequence of matrix contraction, ⁶ by fixation for 30 minutes with 4% formaldehyde at room temperature. The cells were washed in PBS (Sigma, Poole, UK) and stained with Coomassie brilliant blue (a total protein dye) for 30 minutes to enable patches to be visualized and measured. The cells were washed several times in PBS to remove excess dye. Images of patches were captured on a CCD camera (Grabber software; Synoptics, Cambridge UK) and analyzed (PC Image; Foster Findlay Associates, Newcastle, UK). After the patches had been measured, PBS medium was aspirated from the culture dish and replaced with 1 mL of 70% ethanol allowing Coomassie blue dye within the cells to be dissolved. The culture dishes were placed on a rotary shaker for 1 hour until all the dye had been extracted from the cells. A 200-μL sample of dye from each dish was placed in a clear plastic 96-well microtiter plate and the absorbance read at 550 nm on a multilabel counter (Wallac Victor 2 1420; with Workout, ver. 15 software; Perkin Elmer, Boston, MA). Total protein, as determined by dye content, has been shown to be proportional to the number of cells. ³³  

Suspended Bead Array Analysis of Smad-Independent Signaling Proteins

FHL 124 cells were seeded onto 35-mm dishes at ∼30,000 cells in 400 μL of 5% FCS-EMEM (Invitrogen Ltd.) and were maintained in 1.5 mL of 5% FCS-EMEM for 3 days. After a 24-hour incubation period in nonsupplemented EMEM, the cells were treated with TGFβ2 at 10 ng/mL for 0,10, 30, 60, and 120 minutes, respectively. The cells were lysed (Bio-Plex Cell Lysis Kit; Bio-Rad, Hemmel Hempstead, UK), according to the manufacturer's instructions. Lysates were precleared by centrifuging at 4500 rpm at 4°C for 20 minutes. Equal amounts of protein per sample were analyzed for phosphorylated and total levels of ERK1/2, P38 and JNK using a commercially available kit (Bio-Plex Suspended Multiplex Bead Array Assay kit; Bio-Rad). Data from the reaction were acquired with a flow cytometry system (X Map-100; Luminex, Austin, TX) and accompanying software (Bio-Plex Manager software; Bio-Rad). The median fluorescence intensity was used as a measure of detection of total and phosphorylated protein. Results are presented as the level of phosphorylated protein normalized to its corresponding total protein level. 

Statistical Analysis

A t-test analysis (Excel software; Microsoft, Redmond, WA) and one-way ANOVA (with the Tukey post hoc analysis; SPSS 12.0 for Windows; SPSS Inc., Chicago, IL) were performed to determine any statistical differences between experimental groups, set at P ≤ 0.05. 

Results

Effect of siRNA Targeted against Smad4 on Smad4 Gene and Protein Expression

QRT-PCR was used to validate siRNA against Smad4 in FHL 124 cells after a 24-hour transfection period (Fig. 1A). Smad4 gene expression was significantly inhibited, such that levels were reduced to 20% of the control siRNA. Western blot analysis was performed to determine that siRNA targeted against Smad4 could effectively inhibit Smad4 protein expression. After 48 hours in siSmad4 transfection conditions, Smad4 protein expression was inhibited by 88% relative to the control (Figs. 1B, 1C). Smad2 and -3 protein expression was not impaired by Smad4 inhibition (Figs. 1B, 1C) suggesting siRNA directed against Smad4 has no off-target effects. After removal of transfection conditions, the cells were maintained in serum-free conditions for 24 hours. The expression of Smad4 and its ability to translocate to the nucleus in response to TGFβ was evaluated by immunocytochemistry. 

Figure 1.

 

Validation of siRNA directed against Smad4. (A) QRT-PCR detection of Smad4 gene expression in FHL 124 cells after 24 hours of transfection with siSmad4 and control siRNA. Data were normalized with mGAPDH control and are expressed as mean ± SEM (n = 4). *Significant difference between siSmad4-treated and control siRNA-treated groups (P ≤ 0.05, one tailed t-test). (B) The effect of siSmad4 on Smad2, -3, and -4 protein expression was assessed by Western immunoblot analysis. FHL 124 cells were transfected with either siSmad4 or control siRNA for 48 hours, after which protein was extracted. Data were normalized with β-actin protein control and are expressed as mean ± SEM (n = 4). *Significant difference between siSmad4-treated and control siRNA-treated groups (P ≤ 0.05, one-tailed t-test). (C) Smad2, -3, and -4 protein bands from a representative Western blot are presented along with their corresponding β-actin profile shown as a control for equal protein loading.

In unstimulated control siRNA cells, Smad4 was distributed largely in the cytoplasm, with relatively low expression in the nucleus (Fig. 2A). Smad4 nuclear accumulation was observed in control siRNA cells after exposure to TGFβ2 (Fig. 2A), which was significantly elevated compared with the control siRNA group (Fig. 2B). The addition of TGFβ2 to Smad4-knockdown cells showed a significant reduction in nuclear Smad4 accumulation compared with the TGFβ2-treated control siRNA group; therefore, Smad4 translocation in response to TGFβ2 has been effectively blocked by siSmad4 transfection. It should be noted that F-actin organization was unaffected by Smad4 knockdown relative to the control siRNA cells. 

Figure 2.

 

Smad4 knockdown impairs TGFβ-induced Smad4 nuclear translocation. (A) Fluorescent micrographs showing Smad4 (green) and F-actin (red) staining in siSmad4 and negative siRNA-transfected FHL 124 cells after 2 hours exposure to TGFβ2 (10 ng/mL). (B) The effects of siSmad4 on TGFβ2-induced Smad4 translocation, quantified by imaging and thresholding techniques. Data represent the mean ± SEM (n =4). *Significant difference between TGFβ2-treated and untreated control siRNA groups; ^πsignificant difference between siSmad4 treated and control siRNA treated groups (P ≤ 0.05, ANOVA with Tukey's test).

Smad4-Dependent TGFβ-Induced αSMA and Fibronectin Gene Expression

siRNA targeted against Smad4 was successful in suppressing Smad4 protein expression and function. This transfection technique was applied to FHL 124 cells to determine the role of Smad4 in TGFβ-induced transdifferentiation, a common cellular characteristic of fibrotic diseases. ¹³ The impact of Smad4 knockdown on the gene expression of αSMA, the major marker of transdifferentiation and fibronectin a determinant of transdifferentiation and fibrosis was evaluated by QRT-PCR. Transfected siSmad4 and control siRNA treated cells were maintained in the presence and absence of TGFβ2 (10ng/mL) for 24 hours. Addition of TGFβ2 to control siRNA cells induced a significant increase in αSMA gene expression (197% ± 6%) compared with the unstimulated control siRNA (Fig. 3A). After transfection with siSmad4 and culture in the absence of TGFβ2, αSMA gene expression was reduced relative to control siRNA. αSMA gene expression was significantly decreased in siSmad4-transfected cells treated with TGFβ2 (43% ± 3%) compared with the TGFβ2-treated control siRNA. The addition of 10 ng/mL TGFβ2 to control siRNA cells induced a significant increase in fibronectin gene expression compared with the unstimulated control siRNA group (Fig. 3B). After transfection with siSmad4 and culture in the absence of TGFβ2, fibronectin gene expression was significantly reduced compared with the control siRNA group. Similarly, fibronectin gene expression was significantly reduced in siSmad4-transfected cells treated with TGFβ2 compared with the TGFβ2-treated control siRNA. However, the addition of TGFβ2 to siSmad4-transfected cells induced a significant increase in fibronectin gene expression compared with that in the unstimulated siSmad4 experimental group. 

Figure 3.

 

TGFβ-induced αSMA and fibronectin gene expression are Smad4 dependent. siSmad4 and negative siRNA transfected FHL 124 cells were maintained in either serum-free EMEM or treated with TGFβ2 at 10 ng/mL for 24 hours. Quantitative RT-PCR was used to analyze (A) αSMA and (B) fibronectin gene expression. Data were normalized with mGAPDH control. Data represent the mean ± SEM (n = 4).*Significant difference between TGFβ2-treated and untreated control siRNA groups; ^πsignificant difference between siSmad4+TGFβ2-treated and control siRNA+TGFβ2-treated groups; ^σsignificant difference between siSmad4 treated and control siRNA treated groups (P ≤ 0.05, ANOVA with Tukey's test).

Smad4-Independent TGFβ-Induced Smad7 Gene Expression

The impact of Smad4 knockdown on Smad7 expression, an inhibitory Smad and major determinant for TGFβ transcriptional responses, ¹² was evaluated by QRT-PCR. Addition of 10 ng/mL TGFβ2 to control siRNA cells induced a significant increase in Smad7 gene expression (230% ± 7%) compared with the unstimulated control siRNA cells (Fig. 4). FHL 124 cells transfected with siSmad4 and cultured in the absence of TGFβ2 showed no significant difference in Smad7 gene expression compared with the control siRNA group. The addition of TGFβ2 to siSmad4-transfected cells induced a significant increase in Smad7 gene expression (210% ± 12%) compared with that in the unstimulated siSmad4 group. The increase in Smad7 gene expression observed with the TGFβ2-treated siSmad4 group was not significantly different when compared with that in the TGFβ2-treated control siRNA group. 

Figure 4.

 

TGFβ-induced Smad7 gene expression is Smad4 independent. siSmad4 and control siRNA-transfected FHL 124 cells were maintained in either serum-free EMEM or treated with TGFβ2 at 10 ng/mL for 24 hours. Quantitative RT-PCR was used to analyze Smad7 gene expression. Data were normalized with mGAPDH control. Data represent the mean ± SEM (n = 4).*Significant difference between TGFβ2-treated and untreated siRNA-transfected groups (P ≤ 0.05, ANOVA with Tukey's test).

Smad4-Independent TGFβ-Induced Matrix Contraction

Matrix contraction is a detrimental characteristic of PCO ¹ and also in fibrotic diseases throughout the body. ¹³ siRNA targeted against Smad4 was applied to FHL 124 cells, seeded to form patches, to determine the role of Smad4 in TGFβ2-induced matrix contraction. Patches transfected with siRNA against Smad4 did not show a significant difference in patch area from that in the control siRNA-treated cells when maintained in nonstimulated conditions (i.e., no TGFβ2; Fig. 5). This result suggests that Smad4 had a negligible effect on patch growth. After the addition of TGFβ2 (10 ng/mL) for 24 hours, cell-free regions within the patches were observed (Fig 5B), a consequence of matrix contraction. ⁶ The patch area was reduced when compared with that in the unstimulated control siRNA group, but the reduction was not significant (Fig 5A). Patches that had been transfected with siSmad4 and then treated with TGFβ2 showed a significant reduction in the patch area, because of matrix contraction, compared with that in the unstimulated siSmad4-treated patches (Fig. 5). Moreover, the siSmad4 group treated with TGFβ2 was also significantly different from the control siRNA group treated with TGFβ2 (Fig. 5A). Analysis of total protein Coomassie blue dye from these patches showed no significant changes in any experimental group. This finding indicates the there was no significant change in cell population between any groups (data not shown). 

Figure 5.

 

Smad4 was not critical for TGFβ-induced matrix contraction. FHL 124 cells were seeded to form patches, then transfected with siRNA targeted to Smad4 or negative siRNA control and maintained in EMEM supplemented with 2% FCS. Patches were measured after 24 hours in culture with 10 ng/mL TGFβ2 (A). Data represent the mean ± SEM (n = 4).*Significant difference between TGFβ2-treated and untreated siSmad4 groups; ^πsignificant difference between siSmad4+TGFβ2-treated and control siRNA+TGFβ2-treated groups (P ≤ 0.05, ANOVA with Tukey's test). Coomassie blue dye was extracted and quantified from the same experiment as presented in (A). No significant difference between groups was observed (data not shown). (B) Representative images of dishes for each experimental group.

Smad4-Independent TGFβ-Induced Smad2/3 Nuclear Translocation

The impact of Smad4 knockdown on Smad2/3 function on TGFβ stimulation was investigated using immunocytochemistry. In both unstimulated control siRNA cells and siSmad4-transfected cells, Smad2/3 was distributed largely in the cytoplasm, with a relatively low expression in the nucleus (Fig. 6A). After a 2-hour exposure to TGFβ2 (10 ng/mL) Smad2/3 nuclear accumulation was observed in control siRNA cells (Fig. 6A) which was significantly greater than the unstimulated control siRNA group (Fig. 6B). Similarly, Smad2/3 nuclear accumulation was also observed in siSmad4-transfected cells after exposure to TGFβ2 (Fig. 6A), such that the extent of Smad2/3 nuclear accumulation in TGFβ2-treated siSmad4 cells was similar to that in the TGFβ2-treated control siRNA cells (Fig. 6B). 

Figure 6.

 

Smad4 was not critical for TGFβ-induced Smad2/3 nuclear translocation. (A) Fluorescent micrographs showing Smad2/3 (green) and F-actin (red) staining in siSmad4 and negative siRNA-transfected FHL 124 cells after 2 hours' exposure to TGFβ2 (10 ng/mL). (B) The effects of siSmad4 on TGFβ2-induced Smad2/3 translocation, quantified by imaging and thresholding techniques. Data represent the mean ± SEM (n = 4). *Significant difference between TGFβ2-treated and untreated control siRNA groups (P ≤ 0.05, ANOVA with Tukey's test).

Activation of Smad-Independent Signaling Pathways in Response to TGFβ

To identify alternate pathways to the conventional Smad-signaling system, we investigated the activation (phosphorylation) of three possible Smad-independent pathways over a 2-hour period after exposure to 10 ng/mL TGFβ2. The pathways were determined with a suspended bead array system (Bio-Plex; Bio-Rad) to detect pERK1/2, pP38 and pJNK simultaneously (Figs. 7A–C). Using this method, we found that TGFβ induced a significant increase in pERK and pP38 relative to unstimulated control. The peak response in both cases was observed at the 60-minute time point. At this time point, pERK1/2 and pP38 level was 151.9 ± 14.0 and 216.2 ± 25.8, respectively (Fig. 7D). Of interest, changes in ERK1/2 phosphorylation were observed sooner than p38 phosphorylation, such that the EC₅₀ was 21 and 46 minutes, respectively (Figs. 7A, 7B). With respect to phosphorylation of JNK, no significant difference relative to unstimulated control was observed at any time point studied (Figs. 7C, 7D). 

Figure 7.

 

TGFβ-induced Smad-independent signaling detected by suspended bead array assay. Changes in phosphorylation levels of (A) ERK1/2, (B) P38, and (C) JNK in response to 10 ng/mL TGFβ2 were detected over a 2-hour period. The data represent mean ± SEM, n = 3 normalized to the corresponding total of ERK1/2, P38, and JNK for each sample respectively. (D) Mean phosphorylation levels at the 1-hour time point for the three target proteins. *Significant difference between TGFβ2-treated and unstimulated control (P ≤ 0.05, 1 tailed t-test).\.

Discussion

The present study determined the relative importance of Smad4 in TGFβ/Smad signaling, gene expression, and matrix contraction. A successful experimental protocol was established using siRNA targeted against Smad4, whereby Smad4 protein levels were significantly knocked down in FHL 124 cells for extended periods after cessation of siSmad4 transfection. Applying this method revealed that Smad4 is important in the TGFβ signaling cascade that regulates the gene expression of transdifferentiation markers in human lens epithelial cells, but is not critical for matrix contraction to proceed. These findings have great relevance to posterior capsule opacification, a fibrotic condition that develops after cataract surgery. 

The present study revealed that in FHL 124 cells, TGFβ-induced gene expression of the transdifferentiation marker αSMA was perturbed by Smad4 knockdown. Previous evidence in the lens has supported the critical role of Smad signaling in the upregulation of αSMA by TGFβ. ^30,34,35 In particular, Saika et al. ³⁰ observed that both TGFβ2-induced and injury-induced αSMA expression was inhibited in the lens epithelium of Smad3-knockout mice. Furthermore, in a recent investigation, Banh et al. ²⁹ observed αSMA expression in a TGFβ1/Smad3-knockout mouse; however, it should be noted that expression was at a reduced level compared with that in wild-type lenses. An interesting observation in the present study was that R-Smads can still translocate to the nucleus independent of Smad4. However, the translocation is not associated with increased levels of αSMA. In conjunction with results in previous studies, ³⁰ the evidence suggests that a complex involving both Smad3 and Smad4 strongly promotes TGFβ-induced transdifferentiation of lens epithelial cells to a myofibroblast phenotype. The current investigation reveals TGFβ2 signaling via the Smad2/3-Smad4 pathway to be important for the induction of αSMA gene expression in FHL 124 cells. 

In the current work, the regulation of fibronectin gene expression appeared to be Smad4-dependent, as fibronectin expression is significantly suppressed in Smad4 knockdown cells in both the presence and absence of TGFβ treatment. Of interest, TGFβ is still capable of upregulating fibronectin expression relative to unstimulated Smad4 knockdown cells, which suggests Smad4 independent regulation of fibronectin expression is also occurring. TGFβ-Smad4 independent regulation of fibronectin expression has been previously observed in a human fibrosarcoma cell line ²⁶ and cross-talk between the Smad- and mitogen-activated protein kinase pathways can regulate TGFβ transcriptional responses. ³⁶ Therefore, the interplay of Smad4 dependent and -independent pathways in regulating fibronectin expression in FHL 124 cells remains an intriguing possibility. 

The induction of Smad7 gene expression by TGFβ was found to be independent of Smad4 regulation in the present study; Smad7 is a major determinant for TGFβ transcriptional responses that regulate the intensity and/or duration of TGFβ signals. A previous investigation by Levy and Hill, ²⁰ also confirmed Smad7 to be a Smad4-independent gene in human keratinocyte and pancreatic tumor cell lines. In contrast, the binding of Smad3 and Smad4, but not Smad2, to the Smad-binding element (SBE) of the Smad7 gene promoter was suggested to be essential for its activation in response to TGFβ. ¹⁷ To date, the signaling mechanisms that regulate Smad7 gene expression in the lens have yet to be discerned. However, the overexpression of Smad7 in the mouse lens prevents injury-induced αSMA expression and Smad2/3 nuclear translocation, ³⁵ confirming the role of Smad7 as an inhibitory regulator of TGFβ signal transduction via a negative feedback mechanism. ³⁷ Smad7 associates with the activated TGFβ receptor and interferes with the activation of Smad2/3 by preventing their receptor interaction and subsequent phosphorylation. ³⁷ Smad7 expression can be induced by the Jak1/Stat1 pathway after stimulation with IFNγ and by activated NF-κB ³⁸ ; therefore, the interplay of several non-Smad pathways may also regulate Smad7 gene expression in the absence of TGFβ-Smad4 signaling. Although TGFβ can induce gene expression by Smad-independent pathways, ¹⁹ there have been no published reports to date of those mechanisms promoting the induction of Smad7 by TGFβ. Therefore, future investigations into the Smad4 independent mechanism(s) that regulate TGFβ-induced Smad7 gene expression in the lens are of interest with respect to determining TGFβ transcriptional responses. 

An important finding of the present study is that TGFβ2-induced matrix contraction is independent of Smad4 regulation. Smad4 knockdown did not affect the induction of matrix contraction by TGFβ2, where in contrast, there was a significant inhibition of αSMA expression by TGFβ2 in these Smad4 knockdown FHL 124 cells. We have previously reported that the expression of αSMA does not correlate with TGFβ-induced matrix contraction of FHL 124 cells ⁶ ; data from the present study support these findings. The evidence presented consequently counters results in previous studies that propose αSMA, the major marker of transdifferentiation, as being essential to TGFβ matrix contraction. ^8,39 The findings in the current investigation indicate that TGFβ-induced αSMA expression and matrix contraction are regulated by two separate pathways, rather than a TGFβ/Smad-dependent pathway where αSMA expression (transdifferentiation) is a precursor to matrix contraction. After treatment with TGFβ2, matrix contraction was promoted in the Smad4 knockdown group, which showed reduced levels of αSMA, compared with the control siRNA group. The promotion of TGFβ2-induced matrix contraction of αSMA-knockdown cells has also been observed, ⁶ indicating a role for αSMA in suppressing TGFβ-promoted matrix contraction. The current body of data suggest that alternate mechanisms of regulation, independent of Smad4 should be considered with respect to TGFβ-induced matrix contraction. 

An interesting finding in the present study is that Smad2/3 can translocate to the nucleus in Smad4-knockdown cells in response to TGFβ2, thus indicating that a TGFβ-promoted Smad2/3 nuclear translocation mechanism independent of Smad4 binding is in place in FHL 124 cells. Although it is generally reported that binding of R-Smads to Smad4 is necessary for nuclear translocation and the formation of high-order transcriptional complexes, ¹² other investigations have demonstrated that R-Smad nuclear translocation can occur independent of Smad4. ^21,22,40 The rationale for this phenomenon is based on evidence that shows R-Smads to have their own NLS (lysine-rich sequence in the MH2 domain), which when phosphorylated allows for importin binding and subsequent nuclear translocation. ¹⁸ The phosphorylation and resulting nuclear translocation of Smad2 and -3 has been observed in Smad4-defective cell lines, ⁴¹ and during mouse embryogenesis, the induction of the lateral plate mesoderm requires Smad2 and -3 but is independent of Smad4. ²¹ There is evidence to suggest that after translocation, the interaction of Smad2/3 with nuclear proteins can enable the induction of transcription independent of Smad4. ^42,43 For instance, TIF1γ was found to interact with Smad2/3 to mediate TGFβ-induced erythroid differentiation independent of Smad4, ⁴² and the tumor suppressor cofactor IKKα can control keratinocyte differentiation via an Smad4-independent TGFβ-Smad2/3 signaling pathway. ⁴³ Therefore, the potential for Smad2/3 to bind to similar nuclear co-regulators in response to TGFβ and induce target gene transcription in FHL 124 cells may be likely. In the present study we could not distinguish whether both Smad2 and -3 were present in equal ratios in the nucleus or whether one R-Smad dominated, as a Smad2/3 (TGFβ-RSmad) antibody was applied to the Smad4-knockdown FHL 124 cells. However, we observed that the level of Smad3 protein expression was comparatively greater than Smad2 in unstimulated FHL 124 cells. Therefore, it is likely that after TGFβ stimulation, Smad3 nuclear translocation is predominant over Smad2. There is much evidence to suggest that in the presence of TGFβ, Smad3 rather than Smad2 is more likely to induce gene transcription independent of Smad4. For instance, only nucleoporins can import Smad2 into the nucleus, ⁴⁴ whereas β-importin and nucleoporins enable Smad3 nuclear import. ^44,45 When in the nucleus Smad2 cannot bind DNA directly, because of a sequence insert in the β hairpin. ⁴⁶ Furthermore, Smad3 has a higher affinity than Smad2 for the cofactors E2F4/5, ATF3, and FoxO that target TGFβ-regulated genes. ⁴⁷ Smad3 dependent signaling has been widely implicated in fibrotic conditions throughout the body (reviewed in Ref. 18). Therefore, it is more likely that Smad3, rather than Smad2, can induce TGFβ-promoted gene expression independent of Smad4 in FHL 124 cells. Most recently, TGFβ-induced phosphorylation of Smad1/5 was observed in nonocular epithelial cells, although this was not necessary for epithelial-to-mesenchymal transition. ¹⁶ However, the potential for Smad1/5 to regulate TGFβ responses in human lens epithelial cells independent of Smad4 binding remains an intriguing possibility. 

In addition to Smad signaling, TGFβ can signal independent of Smads via pathways that include: the ras/MEK/ERK MAP kinase cascade, Rho kinase, JNK, and p38 signaling pathways. ^12,19,36 In the present study, we show that the Smad independent signaling pathways ERK and p38 MAP kinase can be activated by TGFβ in human lens epithelial cells. Understanding the importance of these alternate signaling pathways could reveal mechanisms underlying TGFβ-regulated diseases of the lens. There is much evidence to suggest that Smad-independent signaling pathways can regulate matrix contraction. For example Rho, myosin light chain phosphatase (MLCPPase), and myosin light chain kinase (MLCK) ⁴⁸ have been implicated in matrix contraction. The ERK signaling pathway can promote matrix contraction by its activation of MLCK ⁴⁹ and the JNK and p38 signaling pathways also promote matrix contraction. ^50,51 It has been shown that MLCK gene expression is elevated in FHL 124 cells in response to TGFβ. ⁵² Therefore after TGFβ stimulation, it is likely that the availability of MLCK is increased in human lens epithelial cells and through activation of ERK signaling, increased levels of active MLCK could promote myosin activity, thus resulting in greater contractile force. Therefore, in human lens epithelial cells, the role of Smad-independent signaling pathways in TGFβ-induced matrix contraction needs further investigation to underpin the regulatory mechanism controlling this detrimental fibrotic characteristic. 

In summary, in the present study Smad4 was critical for the expression of TGFβ-induced transdifferentiation markers, but was not essential to the induction of matrix contraction. It therefore appears that alternate TGFβ pathways should be investigated to discern the regulation of matrix contraction; this information consequently will contribute to our understanding of the signaling mechanisms that influence the fibrotic response. 

Footnotes

 Supported by the Humane Research Trust; BBSRC; MedImmune; the National Eye Institute; and John and Pamela Salter Trust.

Footnotes

 Disclosure: L.J. Dawes, Cambridge Antibody Technology (now MedImmune) (F); M.A. Sleeman, MedImmune (E); I.K. Anderson, MedImmune (E); J.R. Reddan, None; I.M. Wormstone, Cambridge Antibody Technology (now MedImmune) (F).

Footnotes

 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.

The authors thank Diane Alden for technical assistance and George Duncan, whose helpful contributions, support and enthusiasm toward this study will always be appreciated.