MiR‐22‐3p inhibits fibrotic cataract through inactivation of HDAC6 and increase of α‐tubulin acetylation

Abstract Objectives Fibrotic cataract, including posterior capsule opacification (PCO) and anterior subcapsular cataract (ASC), renders millions of people visually impaired worldwide. However, the underlying mechanism remains poorly understood. Here, we report a miRNA‐based regulatory pathway that controls pathological fibrosis of lens epithelium. Materials and methods Expression of miR‐22‐3p and histone deacetylase 6 (HDAC6) in normal and PCO patient samples were measured by qPCR. Human lens epithelial explants were treated with TGF‐β2 in the presence or absence of miR‐22‐3p mimics or inhibitor. Cell proliferation was determined by MTS assay, and migration was tested by transwell assay. Expression of HDAC6 and EMT‐related molecules were analysed by Western blot, qPCR and immunocytochemical experiments. Results We identify miR‐22‐3p as a downregulated miRNA targeting HDAC6 in LECs during lens fibrosis and TGF‐β2 treatment. Mechanistically, gain‐ and loss‐of‐function experiments in human LECs and lens epithelial explants reveal that miR‐22‐3p prevents proliferation, migration and TGF‐β2 induced EMT of LECs via targeting HDAC6 and thereby promoting α‐tubulin acetylation. Moreover, pharmacological targeting of HDAC6 deacetylase with Tubacin prevents fibrotic opaque formation through increasing α‐tubulin acetylation under TGF‐β2 stimulated conditions in both human lens epithelial explants and the whole rat lenses. Conclusions These findings suggest that miR‐22‐3p prevents lens fibrotic progression by targeting HDAC6 thereby promoting α‐tubulin acetylation. The ‘miR‐22‐HDAC6‐α‐tubulin (de)acetylation’ signalling axis may be therapeutic targets for the treatment of fibrotic cataract.


| INTRODUC TI ON
The crystalline lens, composed of only mono-layered lens epithelium and orderly arranged lens fibres with a basement membrane surrounded, is a remarkable structure that contributes to focusing of images on the retina. Opacification in lens, the cataract, is the leading cause of blindness worldwide. 1 Surgery is currently the only effective treatment available. However, the most common complication, known as posterior capsule opacification (PCO), could result in visual axis opacification once again due to the migration of the remaining lens epithelial cells (LECs) towards the posterior lens capsule with excessive proliferation and mesenchymal transformation in response to the surgical injury. 2 In paediatric cataract patients, particularly, the incidence of PCO is almost 100%, which could disrupt their visual development and may result in irreversible blindness. 3 Apart from PCO, fibrotic opacities could also accumulate underneath the anterior capsular induced by ocular trauma, inflammation or radiation, namely anterior subcapsular cataract (ASC). 4 In fibrotic cataracts, lens epithelium loss their integrity and function while disorganized mesenchymal-like cells accumulate with excessive fibronectin (FN) and collagen (Col) secreted into the ECM, thereby damaging the architecture of the lens and leading to eventual visual loss. 5 TGF-βs are considered as the most prominent pro-fibrotic factors, and TGF-β2 is the major isoform in the aqueous humour. 6 Therefore, deciphering the underlying mechanism of proliferation, migration and TGF-β2-induced EMT in lens is of great value for the prevention and treatment of fibrotic cataract.
MicroRNAs (miRNAs), a group of small non-coding RNA strands that regulate their targets via translational repression or mRNA degradation, are involved in diverse biological cellular processes. 7 MiRNAs level is controlled by the TGF-β signalling pathway in both direct and indirect manners in fibrotic diseases. SMADs, the key TGF-β signal transducer, could directly regulate miRNAs expression through several mechanisms, including the inhibition of pri-miRNA transcription, Drosha-mediated pri-miRNA processing and the maturation of miRNAs. 8,9 Moreover, TGF-β could indirectly affect miRNAs, such as miR-203 and miR-200 family, through downstream transcriptional factors, like ZEB1, ZEB2, SNAIL and SLUG. 10,11 Increasing evidence has demonstrated that dysregulation of miR-NAs, such as let-7a-5p, miR-34a, miR-181a and miR-204-5p, [12][13][14][15] is associated with EMT process during cataract. However, the specific effect of miR-22-3p in lens fibrosis remains unexplored. In this study, miR-22-3p was found to decrease significantly in both fibrotic cataract tissue and TGF-β2 induced fibrosis model of human lens epithelial explants, implying miR-22-3p as a repressor during the pathogenesis of fibrotic cataracts.
Our previous study has identified HDAC6 as a target gene of miR-22-3p in mesenchymal stem cells. 16 Histone deacetylases (HDACs) are a group of enzymes that control chromatin condensing and gene transcription through deacetylation on lysine residues mainly in histones. 17,18 Studies have shown that HDAC inhibition prevents EMT of LECs and suggested the epigenetic modifiers as potential targets to control lens fibrosis. 19,20

| In vitro transfection with miRNA mimics, miRNA inhibitors and siRNA
The synthetic miR-22-3p mimics, inhibitors and interfering RNAs of HDAC6 (siHDAC6) were from Life Technologies. About 50 nmol/L miRNA mimics, inhibitors or 40 nmol/L siHDAC6 was transfected to human lens epithelial explants or human LECs for gain-and lossof-function studies by using siPORT™ NeoFX™ Transfection Agent (AM-4511; Applied Biosystems) according to the manufacturer's instructions. The detailed sequences are listed in Table S1.

| Whole-mount staining with immunofluorescence staining analysis
The whole-mount lens epithelial explants were flattened on a glass slide and fixed with 4% paraformaldehyde for 20 minutes after TGF- and rabbit-acetylated-α-tubulin monoclonal antibody (5335; CST).
Images were obtained using a Leica DM3000 microscope system.

| RNA isolation and qPCR
Total RNA from fibrotic PCO tissues or lens epithelial explants were purified using Arcturus PicoPure RNA isolation kit (KIT0204; Applied Biosystems) according to the manufacturer's protocol. cDNAs of miRNAs were synthesized by miScript II RT Kit (218160; Qiagen), and qPCR analysis of miR-22-3p was performed using miScript SYBR Green PCR Kit (218073; Qiagen). cDNAs of mRNAs were synthesized with Maxima First strand cDNA synthesis kit (K1641; Thermo Fisher Scientific) and qPCR was performed using FastStart Universal SYBR Green Master (Rox) (4913914001-1; Roche). The amplification was performed on a StepOnePlus real-time PCR system (Applied Biosystems) on standard settings. The expression levels were normalized to that of U6 or GAPDH, respectively. The sequences of oligonucleotide primers used in this study are listed in Table S1.

| Proliferation assay
We used MTS Cell Proliferation Colorimetric Assay Kit (G3580; Promega) to determine the proliferation abilities of cells as instructions. The cells were seeded into 96-well plates at a density of 2 × 10 4 cells/ well. MTS reagent was added and incubated at 37°C for 1 hour from day 0 to day 5, and the absorbance values of each well were measured by using a spectrophotometer at 490 nm.

| Detection of HDAC6 deacetylase activity
The activity of HDAC6 was determined using the HDAC6 Activity Assay Kit (K466-100; Bio Vision) according to the manufacturer's protocol. Briefly, extract (2 μg total protein) was incubated at 37°C with 2 μL HDAC6 Substrate in HDAC6 Assay Buffer to 100 μL for 30 minutes. The reaction was stopped by adding 10 μL Developer to each well at 37°C for 10 minutes. Fluorescence was measured by using a Synergy H1 Hybrid Reader (BioTek) with excitation at 380 nm and emission at 490 nm. HDAC6 activity was calculated by Relative Fluorescence Unites per milligram of the sample.

| Model of TGF-β2-induced ASC in whole lens culture
All animal procedures were carried out in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Animal Use and Care Committee of Zhongshan Ophthalmic Center at Sun Yat-Sen University. Whole lenses of 21-day-old Wistar rats were cultured in 4 mL serum-free M199 medium with 0.1% BSA, 0.1 µg/mL l-glutamine, 100 IU/mL penicillin and 100 µg/mL streptomycin. Different concentrations of Tubacin (SML0065; Sigma-Aldrich) were added to the medium with or without 5 ng/mL TGF-β2 for up to five days. On Day 5, all lenses were fixed with 10% formaldehyde and embedded in paraffin for further immunofluorescence analysis.

| Statistical analysis
Comparisons between two groups were examined using Student's t test while the differences among multiple groups were analysed using One-way ANOVA. A value of P < .05 was considered statistically significant.

| MiR-22-3p is downregulated in PCO tissues and explants
Transparent lens capsules and fibrotic lens capsules from healthy donors and PCO patients were collected. Expression levels of mesenchymal marker α-SMA, ECM markers FN and Col I as well as proliferative marker Ki67 were markedly increased in PCO tissues relative to normal lens epithelium ( Figure 1A,B), which indicates that lens fibrosis involves a combination of processes including hyperproliferation and EMT. Interestingly, miR-22-3p was significantly downregulated in PCO tissues than that of normal tissues ( Figure 1C). To further explore the regulatory role of miR- 22-3p in lens fibrosis, human lens explants were stimulated with TGF-β2, the main type of pro-fibrotic cytokines in aqueous humour. As shown in Figure 1D,E, lens epithelial explants treated with TGF-

| HDAC6 inhibition suppresses proliferation, migration and TGF-β2-dependent EMT
To investigate whether HDAC6 mediate the biological functions of miR-22-3p during lens fibrosis, we conducted a loss-of-function approach by knocking down HDAC6 in LECs. Introduction of a siRNA targeting HDAC6 mRNA (siHDAC6) effectively inhibited the expression of HDAC6 in SRA01/04 cells ( Figure 4A,B) and blocked proliferation ( Figure 4C) and migration of LECs ( Figure 4D,E). Moreover, lens epithelial explants treated by TGF-β2 exhibited restoration of epithelial state and diminished mesenchymal markers with HDAC6 knockdown ( Figure 4F-H). Collectively, these results demonstrate that HDAC6 is required for LECs transforming into mesenchymallike cells and strengthen the finding that miR-22-3p controls lens fibrosis by targeting HDAC6.

| Involvement of miR-22-3p-HDAC6-α-tubulin (de)acetylation axis in lens fibrosis
HDAC6 can shuttle between cytoplasm and nucleus to deacetylase both non-histone and histone substrates under certain contexts. 23 In this study, HDAC6 was shown to accumulate exclusively in the cytoplasm of the elongated and spindle-shaped mesenchymal cells upon TGF-β2-induced fibrosis of lens ( Figure 6B), excluding the possibility of its modification on nuclear histones and emphasizing its cytoplasmic deacetylation effect during lens fibrosis. Microtubule α-tubulin, the primary component of the cytoskeleton, represents the most important substrate of HDAC6. 27 It is worth noting that cytoskeleton, the basic cellular structure, affects nearly all the cellular processes 28 and that EMT is a process by which epithelial cells reorganize their cytoskeleton to obtain mesenchymal morphology. 29 However, the regulatory network and the role of α-tubulin acetylation on lens fibrosis have not been explored.

| Inhibition of HDAC6 prevents lens fibrosis through α-tubulin acetylation
Based on the mechanistic rationale, we evaluated whether pharmacological targeting of HDAC6 deacetylase with specific inhibitor, Tubacin, could phenocopy the impact of HDAC6 genetic knockdown on lens fibrosis. Tubacin promoted α-tubulin acetylation significantly in lens epithelial explants ( Figure 6A,B). Both

| α-tubulin acetylation with Tubacin, an HDAC6 deacetylase inhibitor, prevents TGF-β2-induced ASC
To study whether blockade of HDAC6 deacetylase abrogates lens fibrosis in a more sophisticated system, we delivered different concentrations of Tubacin to the semi-in vivo ASC model. 30 Whole rat lenses were cultured with 5 ng/mL of TGF-β2 for 5 days and developed obvious cloudy opacities beneath the lens capsule ( Figure 7A) with disorganized proliferation and aberrant fibrosis evidenced by F I G U R E 4 HDAC6 inhibition suppresses proliferation, migration and TGF-β2-dependent EMT. A and B, qPCR and Western blot analysis of HDAC6 expression levels in non-transfected (NT), siRNA negative control (siNC) and siRNA of HDAC6 (siHDAC6)-transfected human LECs after 48 h transfection. C, Proliferative activity of LECs after transfected with siHDAC6 and siNC. D and E, Representative images (D) and quantification (E) of lower chambers following Transwell migration assays after HDAC6 knockdown by siRNA. F, HDAC6 mRNA level in lens epithelial explants by qPCR after transfected with siHDAC6 for 12 h and further treated with or without TGF-β2 (5 ng/mL) for 24 h. G and H, Changes of EMT-related markers after transfected with siHDAC6 in TGF-β2 induced fibrosis model of lens epithelial explants detected by immunofluorescent staining (G) and qPCR (H). All scale bars: 50 μm. **P < .01 and ***P < .001 by Student's t test, all n = 3 per group. Data are shown as mean ± SD increased Ki67, α-SMA and FN ( Figure 7B-D). In contrast, Tubacin reversed TGF-β2-induced ASC as concentration increased and the lenses treated with 5 and 10 μmol/L Tubacin remained transparent as normal lens ( Figure 7A). Additionally, lenses treated by Tubacin not only retained mono-layered lens epithelium with decreased proliferation but also maintained epithelial state by preventing mesenchymal cells accumulation and ECM secretion with decreased α-SMA and FN ( Figure 7B-D). Collectively, these findings reveal that α-tubulin acetylation via HDAC6 inhibition can suppress ASC formation.

| D ISCUSS I ON
Investigating the mechanism underlying lens fibrosis is important for improving strategies for the prevention and treatment of fibrotic cataract. Here, we show that 'miR-22-3p -HDAC6-α-tubulin (de)acetylation' signalling axis plays a pivotal role in hyperproliferation, migration EMT of LECs. MiR-22-3p, which is downregulated during PCO and TGF-β2-induced EMT, is responsible for the maintenance of epithelial markers and prevention of mesenchymal F I G U R E 5 Involvement of miR-22-3p-HDAC6-α-tubulin (de)acetylation axis in lens fibrosis. A, Comparison of HDAC6 deacetylase activity in normal lens epithelium and PCO tissues. B, Western blot analysis of acetylated α-tubulin in PCO tissues vs normal lens epithelium. C and D, Activity of HDAC6 deacetylase (C) and acetylated α-tubulin levels (D) in lens epithelial explants in response to TGF-β2 with different concentrations (2.5, 5 and 10 ng/mL). E and F, Activity of HDAC6 deacetylase (E) and acetylated α-tubulin levels (F) in lens epithelial explants in response to TGF-β2 (5 ng/mL) at different time points (48 and 72 h). G, HDAC6 deacetylase activity in lens epithelium after 12 h transfection by siHDAC6 or siNC and further treatment with or without TGF-β2 for 24 h. H, Western blot analysis of acetylated α-tubulin in lens epithelium after HDAC6 knockdown under TGF-β2 treated conditions. I and J, Effect of miR-22-3p overexpression on HDAC6 deacetylase activity (I) and acetylated α-tubulin levels (J) in lens with or without TGF-β2 treatment for 24 h. K and L, Effect of miR-22-3p inhibitor on HDAC6 deacetylase activity (K) and acetylated α-tubulin levels (L) in lens with or without TGF-β2 treatment for 24 h.  Hsp70 and Hsp90 are shown to prevent EMT and enhance cell survival F I G U R E 7 α-tubulin acetylation with Tubacin, an HDAC6 deacetylase inhibitor, prevents TGF-β2-induced ASC. A, Representative images of lenses treated by different concentrations of Tubacin (0, 2, 5 and 10 μmol/L) with or without TGF-β2 (5 ng/mL) for 5 d. B-D, Immunofluorescence staining for Ki67 (B), α-SMA (C) and FN (D) under indicated conditions. All scale bars: 50 μmol/L. *P < .05 by Student's t test, all n = 3 per group. Data are shown as mean ± SD of rat lens epithelium , 40 and tyrosine phosphorylated cortactin is required for E-cadherin stability. 41 Additionally, HDAC6 deacetylates β-catenin and thus represses its phosphorylation, leading to its nuclear translocation and promoting proliferation of cancer cells. 42,43 Hence, further studies are still needed to reveal whether other cytoplasmic substrates of HDAC6 are involved in miR-22-3p regulated lens fibrosis.
In addition, HDAC6 has functions beyond its deacetylase activity due to the ubiquitin-binding motif at its C-terminus by which HDAC6 binds to poly-ubiquitinated proteins and transports them into autophagosomes through microtubules for clearance by autophagy. 44 Autophagy is an intracellular homoeostatic process asso- also be considered as potential mechanism during lens fibrosis, and further investigation focusing on autophagy will be performed in our future work.
A key feature of EMT is the loss of cell-cell junctions. More than adhesive functions, E-cadherin/β-catenin complex plays a crucial role in controlling EMT whereby loss of E-cadherin promotes the release of β-catenin from the cell membrane, thus stabilizing β-catenin in the cytosol for nuclear import to promote the expression of EMT-related genes. 51,52 Moreover, β-catenin is a key mediator in the Wnt signalling implicated in lens development and cataract formation. [53][54][55] In this study, both TGF-β2 treatment and miR-22-3p inhibition to lens epithelial explants decreased E-cadherin expression and promoted β-catenin translocating into cytoplasm.
In contrast, miR-22-3p overexpression, HDAC6 inhibition and α-tubulin acetylation retained β-catenin to the cell membrane and restored E-cadherin expression levels, thereby maintaining the lens epithelial state. Additionally, HDAC6 was responsible for Wnt-and EGF-induced nuclear localization of β-catenin through deacetylating β-catenin at different lysine sites in cancer cells. 42,43 Together, these results indicated that miR-22-3p and HDAC6 might modulate lens fibrosis through controlling the stability of the E-cadherin/β-catenin complex and related pathways. Hence, it would be interesting to explore whether 'miR-22-3p-HDAC6-(de) acetylated α-tubulin' axis induced by TGF-β2 modulates lens fibrosis via regulation of the E-cadherin/β-catenin complex through crosstalk with Wnt signalling or other pathways.
In summary, we identify miR-22-3p as an anti-fibrotic factor during lens fibrosis. We have established a post-translational mechanism by which miR-22-3p directly targets HDAC6 and thereby promotes α-tubulin acetylation to control lens fibrosis by maintaining the homoeostasis of the lens epithelium, providing a novel mechanistic insight for prevention and treatment of fibrotic cataract and other fibrotic diseases. Yat-sen University.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data generated or analysed during this study are included in this article.