Histone Deacetylase Inhibitors Trichostatin A and Vorinostat Inhibit TGFβ2-Induced Lens Epithelial-to-Mesenchymal Cell Transition

Abstract

Purpose.: Posterior capsule opacification (PCO) after cataract surgery is due in part to proliferation of the adhering lens epithelial cells and transdifferentiation into mesenchymal cells. The histone deacetylase (HDAC) inhibitors, trichostatin A (TSA) and vorinostat (suberoylanilidehydroxamic acid [SAHA]) are known to modulate cell proliferation and epithelial-mesenchymal transition (EMT). Studies have shown that TGFβ2 can induce EMT similar to that seen during PCO. This study evaluated the effects of TSA and SAHA on TGFβ2-induced EMT in lens epithelial explants.

Methods.: Epithelial cells adherent to lens capsules were isolated from fresh pig lenses and human donor lenses and cultured for 12 hours. Explants were pretreated with TSA or SAHA for 1 hour and then treated with TGFβ2 for up to 3 days. Scratch wound healing assay was used to determine epithelial cell proliferation and migration in the samples. The effects of TSA and SAHA on histone acetylation and HDAC 1 to 6 levels were analyzed by Western blotting.

Results.: Western blotting and immunocytochemistry demonstrated high expression of α-SMA in lens epithelial cells treated with TGFβ2. The HDAC inhibitors exerted dose-dependent inhibition of α-SMA expression, with complete inhibition occurring with 0.5 μM of TSA and 2.5 μM of SAHA. Transforming growth factor β2–induced EMT was suppressed by TSA and SAHA. Histone deacetylase inhibition in pig lens epithelia led to increased acetylation of histone 3 and 4 at multiple sites.

Conclusions.: Histone deacetylase inhibitors, TSA, and SAHA prevent EMT in lens epithelial explants. The results also suggest that the epigenetic modifiers are the potential targets to control PCO after cataract surgery.

Introduction

Cataracts are the most common cause of blindness in the world and are usually removed by extracapsular cataract extraction (ECCE), which leaves behind the lens capsule with adherent epithelial cells. The capsular bag provides support and proper orientation for the intraocular lens (IOL) inserted within the capsular bag. In a few months to a few years after the cataract surgery and IOL implantation, the epithelial cells that were adherent to the capsule begin to divide and proliferate and, in some instances, cover the entire posterior surface of the capsule, imparting a cloudy appearance to the posterior capsule and interfering with clear vision, the same way in which the cataract did. This phenomenon is called “secondary cataract” or posterior capsule opacification (PCO). ¹ Despite advances in the treatment and prevention of PCO, it still remains a major cause of vision problems after cataract surgery. ² The development of PCO is, to some extent, age-related, with the incidence higher in younger patients than elderly patients, but PCO remains a significant problem in all age groups after cataract surgery. ³ Estimates are that PCO occurs in 30% to 50% of elderly patients, but the rate climbs to nearly 100% in younger patients. ⁴ Consequently, the prevention of PCO remains extremely important for vision preservation and for reducing the cost of postcataract treatment. ⁵ Posterior capsulotomy with the Nd:YAG laser, creating a central opening in the opacified posterior capsule, is currently the treatment of choice for PCO, but this procedure is associated with potential problems, such as IOL/optic damage, postoperative IOP elevation, cystoid macular edema, retinal detachment, and IOL subluxation. ^6,7 In addition, the cost of this treatment is considerable, adding significant costs to the US health care system. ⁸ Moreover, in the developing world, in places with limited or no access to Nd:YAG, the postoperative risk of PCO can be an obstacle to successful expansion of ECCE-IOL surgery. 

The mechanisms that underlie PCO include cellular hyperproliferation after injury, myofibroblast formation, and matrix deposition and contraction. ^1,3 Epithelial-to-mesenchymal transition (EMT) during PCO is characterized by the expression of alpha smooth muscle actin (α-SMA) and other cytoskeletal components in the transdifferentiating epithelial cells. ^9,10 Different experimental approaches, including both transformed and primary lens epithelial cell cultures, human capsular bag cultures, and in vivo animal models, have been used to study the molecular mechanisms of lens epithelial cell proliferation and differentiation and to identify and characterize pharmacological agents that can prevent these cellular processes. Epithelial-to-mesenchymal transition during development of PCO is similar to that observed during tissue regeneration, organ fibrosis, and cancer progression. ¹¹ Agents, such as EDTA, thapsigargin, matrix metalloproteinase (MMP) inhibitors, and proteasome inhibitors, because of their ability to block cell proliferation and differentiation, have been investigated as potential therapies for PCO, but none of them have advanced beyond phase I trials. ^12,13  

Clearly, there is a need to identify new agents with the potential for preventing or treating PCO. We explored an approach to preventing PCO with the histone deacetylase (HDAC) inhibitors vorinostat (suberoylanilidehydroxamic acid [SAHA]) ^14,15 and trichostatin A (TSA). ¹⁶ These agents are known to exert both antiproliferative and anti-EMT actions. Histone deacetylase inhibitors also arrest the cell cycle at the G1 phase, ¹⁷ suggesting that these agents may work by exerting a combination of actions to inhibit PCO after cataract surgery. Vorinostat was approved in 2006 by the US Food and Drug Administration for the treatment of cutaneous T-cell lymphoma, but is being widely investigated for use in other forms of cancer. Ophthalmic studies of HDAC inhibitors in animal models and cell culture systems have shown that deacetylation by HDACs drives the expression of several genes in the developing retina and that HDAC inhibitors are nontoxic to the eye, ^18,19 suggesting that HDAC inhibitors may have potential for the treatment of various eye diseases and the use of these inhibitors to control PCO may not adversely affect other cells in the eye. Additionally, it has been suggested and experimentally demonstrated that hyperacetylation may have a protective role on ganglion cells. ¹⁹  

This study was undertaken to determine whether SAHA and TSA, the two representative HDAC inhibitors that are proven effective in certain cancers, are also effective in preventing EMT formation in a lens epithelial explant model of PCO. Studies have shown that treatment of lens epithelial explants with TGFβ2 mimics PCO formation. ^20,21 Therefore, we used TGFβ2 to induce PCO-like conditions in cultured pig lens epithelial explants and tested different doses of SAHA and TSA to determine the effectiveness of each one in inhibiting TGFβ2-induced EMT. We used immunohistochemistry and Western blot methods and found that both of the HDAC inhibitors suppressed EMT formation, as evidenced by the abolition of SMA expression. We also found that the HDAC inhibitors decreased the levels of HDACs in the cultured lens epithelial explants, which may signify additional beneficial effects of HDAC inhibitors on impeding the development of PCO after cataract surgery. 

Materials and Methods

Trichostatin A and SAHA were purchased from Sigma-Aldrich Corp. (St. Louis, MO, USA). Dulbecco's modified Eagle's medium with low glucose (DMEM-LG) and fetal bovine serum were obtained from Invitrogen (Carlsbad, CA, USA). This study was carried out in adherence to the tenets of the Declaration of Helsinki. Lenses were obtained from pig eyes (6–8 months old), supplied by Pel Freez (Rogers, AR, USA) in wet ice. The globe was sprayed with 70% ethanol for 5 seconds, followed by two washes with sterile PBS. The lens was then dissected from the globe and placed on a 35-mm culture dish with the anterior pole of the lens face down. Each epithelial explant was prepared by peeling the posterior pole of the lens capsule and pinning the anterior capsule to the culture dish so that the epithelial cells faced upward. The lens fiber mass was separated from the explant and used for other studies. This approach provided a primary lens epithelial culture (LEC) system with growth occurring on the lens capsule, thus mimicking the human lens epithelium left behind on the capsule after cataract surgery. The capsule-LEC was rinsed once with DMEM-LG medium (Invitrogen) and then incubated with DMEM-LG at 37°C, 5% CO_2, for 24 hours before proceeding with the experiment, to avoid the use of any cultures with damaged LEC explants. The TGFβ2 was used to induce cell proliferation, migration, and transdifferentiation into cells expressing α-SMA, as occurs during events leading to PCO. 

Human donor lenses (54–56 years old) were procured from the Heartland Lions Eye Bank (Columbia, MO, USA). The donor lenses were obtained by the eye bank from diseased individuals at the time of cornea recovery and transported in media to the laboratory. The lens capsule explants were prepared from these donor lenses, using the procedure similar to that used to prepare the pig lens explants, as described above. 

Cytotoxicity Studies

The cultures were subjected to TUNEL assay to detect evidence of apoptosis mediated by TSA and SAHA. The cells were fixed at room temperature in 4% paraformaldehyde for 15 to 20 minutes and washed with PBS. Apop Tag red in situ apoptosis detection kit (EMD Millipore, Billerica, MA, USA) was used to detect apoptosis in epithelial cells. The processed slides were photographed after observation under the fluorescence microscope (Leica, Allendale, NJ, USA). 

Immunocytochemistry

At the end of the incubation period, epithelial explants were washed and fixed in cold methanol (4°C) for at least 20 minutes, flat-mounted on glass slides with the cellular side facing up, and air dried at room temperature. The tissues were rehydrated and washed in PBS, treated with 0.5% Triton X-100 in PBS for 10 minutes, and blocked in 5% horse serum in PBS for 1 hour at room temperature. Slides were incubated with primary antibodies at 4°C overnight. Subsequently, the slides were washed three times with PBS and stained using fluorescent-labeled secondary antibodies and 4′,6-diamidino-2-phenylinodole (DAPI). The sections were examined under fluorescent microscope. Antibodies used in immunocytochemistry were tested for nonspecific binding using negative tissue controls and by the omission of primary antibody. 

Western Blot Analysis

Lens epithelial explants were collected from each treatment group for Western blot analysis. The lens capsules with adhering epithelia were lysed in ×2 RIPA buffer (30 μL per capsule) with tris(2-carboxyethyl)phosphine (Thermo Scientific, Waltham, MA, USA), protease inhibitor, and phosphatase inhibitor. Total protein in the lysate was measured by the Bio-Rad (Hercules, CA, USA) protein assay method and equalized by adding ×6 sample loading buffer and water before loading on the gel. An equal amount of total protein from each group was subjected to SDS-PAGE in 4% to approximately 20% gels (Bio-Rad). The resolved proteins were electrotransferred onto polyvinylidene fluoride (PVDF) membrane. The membranes were blocked with 5% nonfat milk in Tris-buffered saline, pH 7.4, for 1 hour at room temperature and incubated overnight at 4°C with anti–α-SMA antibody (1:10,000; Sigma-Aldrich Corp.) or histone-3 (1:3000; Cell Signaling Technology, Danvers, MA, USA). Following this incubation, membranes were probed with appropriate horseradish peroxidase (HRP)-conjugated secondary antibodies (1:20,000) and enhanced chemiluminescent detection reagents (Pierce, Rockford, IL, USA). The Western blots were visualized by x-ray film exposure. Standardization of Western blots was done by omission of primary antibody and by using negative control samples. 

Expression of HDAC 1, 2, 3, 4, 5, and 6 in pig lens epithelia was investigated by Western blot by using an HDAC antibody sampler kit (Cell Signaling Technology). Extracts were prepared from cultured unscratched samples and from samples scratched and cultured with SAHA before and after scratching. The samples were treated with 2.5 μM of SAHA for 1 hour before and 3 hours after the scratch. The cultured cells were harvested and lysed as described previously and subjected to SDS-PAGE and transferred to PVDF membrane at 20 V for 30 minutes and probed with antibodies diluted at 1:1000 to 1:10,000 times. The gel loading was normalized using glyceraldehyde 3-phosphate dehydrogenase (GAPDH). The amount of HDACs 1 to 6 in the blots was estimated by densitometric analysis with the help of Scion Image software (shareware developed by Scion Corporation downloaded from www.NIST.gov). 

To assess the effect of SAHA on the overall activity of HDACs, histone-3 acetylation in cultured pig lens epithelial explant cell extracts was investigated by Western blot analysis of equivalent amounts of proteins. The explants were subjected to treatment with SAHA (2.5 μM) for 1 hour, scratched, and then treated again with SAHA (2.5 μM) for an additional 3 hours. Unscratched explants and scratched explants cultured for the same duration without SAHA served as controls. Protein acetylation was detected by overnight incubation of membranes at 4°C with primary acetyl-lysine histone H3 antibodies (Lys 9, 14, 18, 27 and 56; Cell Signaling) as well as acetyl-lysine histone-H4 (Ser1/Lys 5,8, and 12) antibody (Santa Cruz Biotechnology, Dallas, TX, USA), followed by washing and incubation with HRP-conjugated secondary antibody. Band intensities were determined using Scion image software. Expression relative to GAPDH was calculated using densitometric values. 

Wound Healing Cell Proliferation and Migration Assay

Scratch and cell proliferation assay was used to assess the effects of TGFβ2, TSA, and SAHA on migration of epithelial cells on explants. Briefly, after 24-hour culture of the epithelial explants, the cells were scratched with a sterile microspatula tip (1.7 mm). The detached cells were removed by washing the capsule with medium. Fresh medium with and without TGFβ2 was added. Representative capsule cultures received either TSA or SAHA and TGFβ2. The capsule cultures were photographed before and after the addition of TGFβ2 and the HDAC inhibitor and at various time points, for up to 24 hours. The wound gap between the two edges was recorded and averaged for three independent experiments. 

Human Lens Epithelial Explant Study

Because only a small number of human donor lenses were available for this study, we were limited in the extent of our studies on human lens explants. Vorinostat was the only HDAC inhibitor studied in human explants. Human lens epithelial explants were pinned to culture plates. Explants were cultured in media for 24 hours and then treated with TGFβ2 (5 ng/mL). The effect of SAHA (2.5 μM) on TGFβ2-treated explants was determined by immunocytochemistry using α-SMA antibody. 

Results

Culturing of Lens Epithelial Explants

Explants were discarded if they showed evidence of liquefaction and loss of epithelial cells, which occur when there is a delay in receiving the lens after the donor's death. After removal of adhering tissues, sterilized globes were immersed in ice-cold cell culture media, which reduces the cell loss that might occur during lengthy dissection. As observed previously with bovine and rabbit explants, ²² pig lens epithelial explants showed α-SMA expression within 1 week in cells that filled the cell-free zones and were found immediately after dissection. Therefore, to avoid sample variation, we selected lens capsules of similar cell confluence for a particular set of experiments. For example, the initial testing of TGFβ2 and the HDAC inhibitors was done on fully confluent capsules (no cell-free region). All experiments with pig explants described in this study were carried out multiple times. We did the experiments in batches, as we were not able to obtain sufficient explants with similar cell confluence to include all treatment replicates on the same day. In total, we used two to five explants per sample in multiple batches and the results were consistent. The data (Figs.) shown are representative of one such treatment. 

Effects of TGFβ2, TSA, and SAHA on α-SMA and E-Cadherin Expression

Transforming growth factor β2 is known to be involved in initiating the fibrotic changes in the lens after cataract surgery. ¹ To mimic the implicated role of TGFβ2 in PCO, we treated lens epithelial explant cultures with 5 ng/mL of TGFβ2 for 3 days and looked for α-SMA and E-cadherin expression (the markers for EMT) using immunocytochemistry (Fig. 1). During EMT, the cells show increased α-SMA and decreased E-cadherin expression. ²³ Untreated explants exhibited normal epithelial morphology with a cobblestone-like appearance of the cells, intense E-cadherin staining at cell boundaries, and no staining for α-SMA. Transforming growth factor β2–treated cells showed intense α-SMA staining with diminished E-cadherin intensity, suggesting that the cells lost their epithelial characteristics and adopted myofibroblastic phenotype. Western blot analyses of the dose-response effects of TSA and SAHA on α-SMA expression in pig epithelial explants revealed a negligible amount of α-SMA expression in explant cultures not treated with TGFβ2, whereas a strong band was seen in TGFβ2-treated explants (Fig. 2). Trichostatin A or SAHA pretreatment 1 hour before the addition of TGFβ2 to the culture was found to inhibit α-SMA expression in a dose-dependent manner. Nearly complete inhibition of α-SMA expression occurred at 0.5 μM and 2.5 μM concentrations of TSA and SAHA, respectively. These TSA and SAHA concentrations were used in all subsequent experiments. The optimum dosage of HDAC inhibitors required in completely preventing α-SMA expression was initially determined from the dose-response studies using single explant per dose. Multiple times by Western blot analysis, complete suppression of α-SMA expression was observed when HDAC inhibitors were used at the concentrations used for other experiments. The changes in epithelial cell morphology in explants pretreated with TSA before the addition of TGFβ2 were observed under phase-contrast microscopy and by rhodamine-phalloidin staining of F-actin (Figs. 3A–H). Transforming growth factor β2–treated cells appeared elongated and showed abnormal morphology under phase-contrast microscopy (Fig. 3B), whereas explants pretreated with TSA before the addition of TGFβ2 had normal phenotype at the end of the assay (Fig. 3C). Trichostatin A, as such, had no discernable effect on pig epithelial cells (Fig. 3D) and the morphology was comparable to that of explants not exposed to TSA (Fig. 3A). Transforming growth factor β2–treated explants showed an increase in DAPI-stained cells (Fig. 3F), suggesting an increase in the number of epithelial cells. The cells appeared crowded and multilayered in TGFβ2-treated explants (Fig. 3F), unlike the explants not treated with TGFβ2. Explants pretreated with TSA before the addition of TGFβ2 (Fig. 3G) or explants treated with TSA by itself (Fig. 3H) showed normal cell phenotype, suggesting that TSA suppressed the TGFβ2-induced effect on lens epithelial cells. 

Figure 1

 

Effect of TGFβ2 on fully confluent pig lens epithelial explants. The pig lens capsules with adhering epithelial cells were cultured as described in the Materials and Methods section. Fully confluent explants were treated without TGFβ2 (A–C) or with TGFβ2 (5 ng/mL) (D–F), and the expression of E-cadherin and α-SMA was determined after 3 days by immunocytochemistry by using respective antibody. The α-SMA staining was seen only in TGFβ2-treated cells, suggesting EMT. The experiments were repeated three times and the results were consistent. The data shown are representative of one such experiment.

Figure 2

 

Western blot analysis of extracts from lens explants cultured with TGFβ2 and different concentrations of TSA and SAHA. The explants were pretreated with one of the HDAC inhibitors for 1 hour before treating with TGFβ2 for 3 days. Trichostatin A and SAHA suppressed α-SMA expression in TGFβ2-treated explants in a dose-dependent manner.

Figure 3

 

Epithelial cell morphological changes in fully confluent explants treated with TGFβ2 in the presence and absence of TSA. (A–D) Phase-contrast images. (E–H) Rhodamine-phalloidin staining of F-actin. TGFβ2-treated culture showed abnormal cell morphology (B) with increased cell proliferation and multiple cells piling on one another (F). The morphology of TSA-treated cells (D, H) on the lens capsule appeared comparable to that of the control group (A, E). The morphology of the cultured cells treated with TGFβ2 and the inhibitor were similar to that of controls (C, G). The experiments were repeated multiple times and the results were consistent.

Effects of SAHA and TGFβ2 on α-SMA Expression in Human Epithelial Cells

To determine whether HDAC inhibitors block TGFβ2-induced EMT in human epithelial cells, epithelial explants from a 55-year-old human donor lens were treated with SAHA for 1 hour before the addition of TGFβ2. The effect of SAHA pretreatment on TGFβ2-induced α-SMA expression in the human lens epithelial explant is shown in Figure 4. Although TGFβ2 treatment led to α-SMA expression, indicative of EMT, α-SMA expression was completely suppressed in the explants pretreated for 1 hour with SAHA (2.5 μM). Furthermore, SAHA at the same concentration did not affect the epithelial cell morphology in explants. Due to limited availability of age-matched human donor lenses, we performed experiments with only a single explant per treatment group. In spite of sample number limitation, the results show that the effect of SAHA on human and pig lens epithelial explants are comparable. 

Figure 4

 

Effect of SAHA on human epithelial explants treated with TGFβ2. SAHA pretreatment almost completely suppressed α-SMA expression in TGFβ2-treated explants.

Effects of TSA, SAHA, and TGFβ2 on Scratched Pig Epithelial Explants

Remnant epithelial cells migrate, proliferate, and transdifferentiate into myofibroblasts, resulting in PCO after cataract surgery. To determine if HDAC inhibitors block epithelial cell EMT, a single scratch was made on the explant using a 1.7-mm microspatula and the migration of cells into the cell-free scratched area and transdifferentiation to myofibroblasts were observed by α-SMA immunocytochemistry. Experiments were performed on TGFβ2-treated and -untreated explants that were pretreated with or without TSA and SAHA (Fig. 5). In control explants (i.e., scratched, no HDAC inhibitor, no TGFβ2), the epithelial cells migrated and filled the cell-free region, with some of the migrating cells expressing α-SMA. Migration was much more rapid and α-SMA expression was much more severe in TGFβ2-treated explants not pretreated with TSA or SAHA. Trichostatin A and SAHA pretreatment suppressed the migration of cells into the cell-free region and led to a significant reduction in the number of α-SMA–expressing cells even in presence of TGFβ2. Explants treated with an HDAC inhibitor only, and not exposed to TGFβ2, did not show α-SMA expression in migrating cells. 

Figure 5

 

Trichostatin A and SAHA suppressed migration and transdifferentiation of cells in scratched pig lens explants treated with or without TGFβ2. The explants (one per group) were pretreated with TSA (0.5 μM) or SAHA (2.5 μM) for 1 hour before scratching and addition of TGFβ2. After 24 hours, the explants were fixed and processed for immunocytochemistry. The migration of cells into the scratched area was visualized by DAPI staining and the cells that transdifferentiate into myofibroblasts were observed by α-SMA immunostaining. The experiments were repeated at least two additional times and the results were similar.

To evaluate the effect of HDAC inhibition on the extent of cell migration in the wound-healing assay, the distance of cell migration was measured in the absence and presence of TGFβ2 at various time points, for up to 24 hours (Figs. 6A, 6B). As shown in Figure 6A, the scratch wounds in explants treated with TGFβ2 were fully covered by the proliferating and migrating cells in 24 hours, unlike the samples that did not receive TGFβ2. Pretreatment of the cultures with TSA or SAHA before the addition of TGFβ2 suppressed both cell proliferation and cell migration (Fig. 6A). Figure 6B shows the rate of cell migration in the assay. In the absence of an HDAC inhibitor, migration of cells in TGFβ2-treated cultures was comparable, for up to 9 hours, with that in cultures not treated with TGFβ2, after which a sudden increase in the migration rate occurred in cells treated with TGFβ2 and by 24 hours the cells completely covered the scratched area. In comparison, cells in untreated explants covered 60% of the scratched area by 24 hours. Trichostatin A– or SAHA-treated cells showed less than 30% coverage of the wound gap in the presence or absence of TGFβ2. Furthermore, on a molar basis, TSA was more effective than SAHA in controlling cell migration, with 0.5 μM of TSA as effective as 2.5 μM of SAHA. 

Figure 6

 

Wound-healing assay showing increased migration of epithelial cells treated with TGFβ2 and delay in cell migration by TSA and SAHA. (A) Phase-contrast images of the scratched explants at 0- and 24-hour time points. (B) Graphical representation of wound closure in control and test explants. The distance of cell migration was estimated from the phase-contrast images taken at different time points, as described in the Materials and Methods section. Trichostatin A and SAHA showed similar inhibition of cell migration when the experiments were repeated two additional times.

To confirm that apoptosis was not a factor in the decreased migration and proliferation in TSA- or SAHA-treated cells, TUNEL assay was performed on scratched pig epithelial explants treated and not treated with TGFβ2. Apoptosis of few migrating cells was seen in explants treated with TSA or TGFβ2 alone, and in combination the apoptotic cells doubled (Fig. 7). No apoptosis was seen in unscratched areas of the cultured explant. Explants treated with SAHA by itself or in combination with TGFβ2 did not show apoptotic migrating cells, suggesting that SAHA neutralizes the apoptotic action of TGFβ2. In spite of increased apoptotic cells in TSA+TGFβ2-treated explants, the migration and wound healing was complete in 48 hours, suggesting that apoptosis in TSA-treated samples is not the cause of reduced migration. 

Figure 7

 

The TUNEL assay for apoptosis on epithelial explants showed migration of cells. The TUNEL assay was done on multiscratched pig epithelial explants 48 hours after TGFβ2 treatment. Apoptosis of a few migrating cells was seen in explants treated with TGFβ2 and TSA, each by itself or in combination. Vorinostat prevented TGFβ2-induced apoptosis of migrating cells. Arrows show the direction of cell migration. Blue, DAPI; red, TUNEL-positive cell. The TUNEL-positive cells are also highlighted by an asterisk. The experiments were repeated one additional time and the results were consistent.

Effect of SAHA on Acetylation of Histone H3 and H4

To assess whether HDAC inhibitors exert their effects on epithelial cell proliferation and migration by hyperacetylation of histones, acetylated histone H3 and H4 levels were determined by Western blot analysis of explants cultured in SAHA (2.5 μM). Relatively low levels of acetylated histone H3 were found in both unscratched and scratched samples when the antibodies specific to Lys 9, 14, 18, 27, and 56 were used (Fig. 8A). However, after 3 hours of culture in SAHA, a 1.5- to 5.0-fold increase in histone H3 acetylation occurred, depending on Lys position (Fig. 8A). Further, the highest acetylation was observed at Lys 9 of H3, as compared with only a 1.5-fold increase in acetylation at Lys 56 of H3. No increase in acetylation of histone H4 was seen in explants scratched and untreated with SAHA when compared with unscratched controls, whereas scratched SAHA-treated explants showed an approximately 6-fold increase in acetylated forms. Increased acetylation of histone H3 and H4 also was seen with TSA (data not shown). 

Figure 8

 

Effect of SAHA on (A) acetylation of histone H3 and histone H4 and (B) on the expression of HDAC family members in the lens epithelial explants. Multiscratched pig lens epithelial explants (3 numbers) were treated with SAHA. Unscratched (2 numbers) and multiscratched explants (3 numbers) without SAHA served as controls. The samples were pooled and analyzed by Western blot analysis by using respective antibodies.

HDAC 1 to 6 Expression in Lens Epithelia and Effect of SAHA

We examined the expression of representative members from both class I and II HDAC proteins in the lens epithelia by using antibodies against HDACs 1 to 3 (class I) and HDACs 4 to 6 (class II) (Fig. 8B). The epithelial extracts prepared from both wounded (scratched) and unwounded samples showed similar immunoreactivity for HDACs 1 to 5, whereas no immunoreactivity was seen for HDAC 6. Surprisingly, SAHA treatment of the scratched explants decreased immunoreactivity of HDACs 1, 4, and 5, suggesting that the levels of these HDACs can be modulated by core histone acetylation. Further, SAHA-induced inhibition was highest with HDAC 5 expression, followed by HDAC 4 and HDAC 1 expression. 

Discussion

The wound-healing response of the lens capsule and adherent epithelial cells after the cataract surgery underlies PCO, the most common postoperative complication of cataract surgery, affecting nearly 50% of adults undergoing cataract surgery and IOL implantation. ⁴ Although the introduction of square-edge lenses has reduced the risk of PCO in adults, ²⁴ PCO remains a big problem in children, with PCO occurring in nearly 100% of children undergoing cataract surgery. ⁴ Proliferation of the residual epithelial cells, from the anterior capsule and equator region to the posterior capsule region, is the prime cause for PCO. Many of the migrated epithelial cells undergo EMT, resulting in the formation of spindle-like myofibroblasts that impart opacity to the capsule. ²⁵ Studies have shown that increases in cytokine levels, such as TGFβ , IL-1 and -2, and hepatocyte growth factor, in the tissue and surrounding media after the cataract surgery might be responsible for the epithelial cell proliferation that culminates in PCO. ⁹  

Although Nd:YAG laser capsulotomy is effective treatment of PCO, it is not without the risk of complications, such as increased IOP, IOL damage, cystoid macular edema, and retinal detachment. ^6,7 The development of pharmacologic approaches to prevent or treat PCO is appealing both to avoid the risks of laser-related complications and to eliminate the expense of laser capsulotomy. Therapeutic approaches are used during cataract surgery to prevent PCO, such as irrigating the surgical site with agents in an attempt to neutralize the effect of cytokines to control cell proliferation, EMT, and PCO formation, ³ yet the incidence of PCO has not been fully eliminated. 

In vitro studies have shown that treatment of lens epithelial cells with TGFβ2 mimics the molecular changes that occur during lens epithelial proliferation. ²⁶ Epithelial explant studies involving cultures with added TGFβ2 have become a standard approach to characterize inhibitors of PCO formation. ²⁷ Posterior capsule opacification studies have been carried out with animal models, ^28–33 capsular bag models, ^34–38 epithelial explants, ^27,39–41 and cell cultures ⁴² to identify potential therapeutic targets and to screen candidate drug molecules. Although primary or transformed epithelial cells can be used for proliferation and EMT studies, this method does not replicate the proliferation of epithelial cells on the capsule that occurs after cataract surgery. Pig lens capsule explants offer an advantage over rat, mouse, or chicken capsular models because of the close similarity of the pig lens to the human lens. Previous studies have shown that the cellular changes occurring in the pig lens capsular model after TGFβ2 stimulation are similar to those observed in human lens capsular model studies. ⁴³ In our study, the conditions of TGFβ2-induced EMT were replicated in the pig lens epithelial explant to characterize therapeutic potential of TSA and SAHA. 

The array of pharmacological agents that have been tested for their ability to prevent lens epithelial proliferation and EMT include 5-fluorouracil, methotrexate, colchicine, mitomycin, ^31,44 protease inhibitors, ^13,27,45 cytoskeletal drugs, ⁴⁶ aldose reductase inhibitors, ⁴³ chelators, ⁴⁷ small interfering RNAs, ⁴⁸ and signal transduction inhibitors. ⁴⁹ Gene transfer techniques also have been investigated. ^50,51 All of these approaches are intended to modulate gene expression to control cell proliferation and prevent EMT. Gene expression control by reversible acetylation of histones has been established by a number of studies. ^52,53 Studies have shown that acetylation of conserved lysine residues of histone tail by histone acetyltransferase enhances gene expression by relaxing the histone-histone as well as histone-DNA interactions by providing the access to transcription factors. Deacetylation, on the other hand, is accompanied by suppression of gene expression by limiting the access of transcription factors to DNA by promoting chromatin condensation. The ability to control the net acetylation level by using small molecules, such as HDAC inhibitors, provides a window to control cellular events selectively. 

Eighteen HDACs have been identified to date and they are grouped into four general classes. ⁵⁴ Those in class I (HDAC 1, 2, 3, and 8) are ubiquitous, whereas those in class II (HDAC 4, 5, 6, 7, 9, and 10) have restricted expression. ⁵⁵ Class III HDACs include Sirtuin 1 to 7 (SIRT1–SIRT7) and are homologous to yeast silent information regulator 2 (Sir2). The lone member of class IV is HDAC 11. Both class I and II HDACs are susceptible to inhibition by TSA and SAHA and are found in the nucleus and cytosol. ⁵⁴ Recently we reported the expression of both class I and II HDACs in pig lens epithelial explants (Sharma KK, et al. IOVS 2013;54:ARVO E-Abstract 2953). Chen et al. ⁴² evaluated mRNA levels of HDACs 1 to 11 and SIRT1 to 7 in cultured human epithelial cell lines, HLEB3 and SRA01/04 cells treated with TGFβ2. They found that mRNA levels of HDACs 1, 2, 3, 5, 8, and 10 increased in SRA01/04 cells after TGFβ treatment, whereas in HLEB3 cells only HDACs 2, 5, and 10 were increased. Our studies show that HDAC proteins in class I (HDACs 1, 2, and 3) and class II (HDACs 4 and 5) are present in lens epithelia and that HDAC 3 and 5 levels in these cells can be suppressed by SAHA. Why SAHA treatment leads to a decrease in the HDAC 3 and 4 levels is not known at this time, but it is possible that the acetylation of histones 3 and 4 might be affecting the normal level of transcription of the candidate genes because it is known that HDAC inhibitors can modulate gene expression. ⁵³ Because both SRA01/04 and HLEB3 cell lines are not primary cell cultures and come from different species, it is likely that the HDAC expression profile in these cell lines is different from that of epithelial explants used in the present study. Furthermore, at this time it is unclear whether the transformed cell lines used in the earlier study ⁴² displayed the same HDAC gene expression as that of a primary cell line. As expected, SAHA treatment of lens epithelial explants led to increased acetylation of histone 3 and 4 at multiple sites (Fig. 8A). However, the acetylation of histone 3 at usually acetylated lysine residues was not elevated to the same extent. The acetylation of lysine 56 was least affected by SAHA, whereas lysine 9 was maximally modified. 

Static cells are relatively resistant to HDAC inhibitors, whereas rapidly dividing and proliferating cells are susceptible to their actions, including induction of growth arrest, activation of the apoptotic pathway, autophagic cell death, reactive oxygen species–induced cell death, mitotic cell death, and senescence. Therefore, we reasoned that HDAC inhibitors also may have the capacity to prevent the lens epithelial cell proliferation, migration, and EMT formation that culminates in the development of PCO after cataract surgery. Histone D acetylase inhibitors induce expression of cyclin-dependent kinase inhibitor p21, the expression of which correlates with cell cycle arrest. Trichostatin A has been shown to induce P21 in both SRA01/04 and HLEB3 cells. ⁴² In addition, the same study also reported that TSA arrests the HLEB3 cells at G1 phase through inhibition of cyclin/cyclin-dependent kinases (CDKs) complexes. We hypothesize that the effects of TSA and SAHA during our wound-healing assays are from enhanced expression of CDK 21. Other studies have shown that CDK 21 is involved in the control of lens epithelial growth and migration. ^27,56 Although we have not investigated the effect of SAHA on canonical TGFβ-Smad2 and Jagged/Notch signaling pathway, we expect that SAHA also inhibits those two pathways, based on data from studies of the effects of TSA on HLEB cell lines. ⁴²  

Both ex vivo and in vivo studies have shown that TSA and SAHA, at concentrations similar to those used in this study, are not toxic to eye tissues. ^57,58 In a rabbit model of corneal haze, a single corneal application of SAHA after photorefractive keratectomy (PRK) was shown to reduce haze formation due to SMA expression, ⁵⁸ a general response to TGFβ1 cytokine expressed during wound healing. Corneal studies of PRK-induced haze also showed that TSA (2.5 μM) was not toxic to the cells in the cornea. ⁵⁷ In retinal studies, TSA was found to protect the retina from ischemic injury, ¹⁸ which was mediated by suppression of TNF-α expression and signaling and blocking of secretion of MMP1 and MMP3 from astrocytes. In a study of the protective effects of HDAC inhibition on retinal ganglion cells, inhibition of HDACs 1 and 3 in mice led to increased survival of retinal ganglion cells following experimental optic nerve injury. ¹⁹ Together, the data suggest that HDAC inhibition is not harmful to cells in the visual system. Therefore, the use of HDAC inhibitors to control epithelial cell proliferation and migration is unlikely to impart toxicity to surrounding tissues. The results of our study suggest that gene expression regulation by manipulating the epigenetic modification of histones can be exploited to control EMT formation and, in turn, prevent PCO formation after cataract surgery. 

Acknowledgments

Supported by an unrestricted grant-in-aid from Research to Prevent Blindness to the Department of Ophthalmology. The authors alone are responsible for the content and writing of the paper. 

Disclosure: L. Xie, None; P. Santhoshkumar, None; L.W. Reneker, None; K.K. Sharma, None 

Footnotes

 LX and PS contributed equally to the work presented here and therefore should be regarded as equivalent first authors.