A ROCK inhibitor suppresses the transforming growth factor-beta-2-induced endothelial–mesenchymal transition in Schlemm’s canal endothelial cells

In the normal eye, most of the aqueous humor drains through the trabecular meshwork (TM) and Schlemm’s canal (SC). The concentration of transforming growth factor beta 2 (TGF-β2) is increased in the aqueous humor of primary open angle glaucoma patients. TGF-β2 increases outflow resistance by affecting the TM and SC, and endothelial–mesenchymal transition (EndMT) of SC cells is involved in these changes. Here, we investigated the effect of a ROCK inhibitor on TGF-β2-induced EndMT in SC cells. The ROCK inhibitor Y-27632 suppressed the TGF-β2-induced increase in the trans-endothelial electrical resistance (TER) and proliferation of SC cells. Y-27632 suppressed the expression of α-SMA, N-cadherin, and Snail, which are upregulated by TGF-β2. Moreover, TGF-β2 decreased mRNA levels of bone morphogenetic protein (BMP) 4 and increased those of the BMP antagonist gremlin (GREM1), but Y-27632 significantly suppressed these changes. Y-27632 also inhibited TGF-β2-induced phosphorylation of p-38 mitogen-activated protein kinase (MAPK). BMP4 and the p-38 MAPK inhibitor SB203580 suppressed the TGF-β2-induced TER elevation in SC cells. Moreover, SB203580 suppressed TGF-β2-induced upregulation of fibronectin, Snail, and GREM1. These results indicate that a ROCK inhibitor inhibited the TGF-β2-induced EndMT in SC cells, implying the involvement of p38 MAPK and BMP4 signaling.


Effect of a ROCK inhibitor on TGF-β2-induced changes in protein and mRNA expression.
We performed immunostaining of cell-cell adhesion and cytoskeletal proteins in SC cells (Fig. 2). ZO-1, which forms tight junctions, did not show changes in expression upon TGF-β2 stimulation. Expression of N-cadherin and β-catenin increased near the cell membrane upon TGF-β2 stimulation, an effect suppressed by Y-27632. Cytoskeletal proteins such as F-actin and α-SMA were upregulated by TGF-β2 stimulation, and this effect was suppressed by Y-27632. In addition, Y-27632 inhibited TGF-β2-induced cell morphological changes.
Western blotting was performed to compare the protein levels of α-SMA, fibronectin, and N-cadherin as mesenchymal markers, Tie2 as an endothelial marker, and Snail as an EndMT inducer. TGF-β2 significantly increased the levels of α-SMA, fibronectin, and N-cadherin at 72 h ( Fig. 3A-C). Y-27632 significantly inhibited the TGF-β2-induced increase in the expression of α-SMA and N-cadherin. Moreover, Y-27632 partially inhibited the TGF-β2-induced increase in fibronectin expression, albeit not significantly so. Tie2 expression was significantly decreased by TGF-β2, an effect significantly suppressed by Y-27632 (Fig. 3D). The expression of Snail was significantly increased by TGF-β2 treatment for 24 h, and the effect was suppressed by Y-27632 (Fig. 3E). Furthermore, immunostaining showed that TGF-β2 increased the expression extracellular matrix collagen type IV, which was suppressed by the simultaneous addition of Y-27632 (Fig. 3F).

Effect of a p38 MAPK inhibitor on TGF-β2-induced changes in SC cells. Y-27632 significantly
inhibited phosphorylation of p38 MAPK. We investigated the involvement of p38 MAPK in the EndMT induced by TGF-β2. SB203580, a p38 MAPK inhibitor, inhibited TGF-β2-induced TER elevation and cell proliferation (Fig. 7A, B). In addition, SB203580 tended to suppress cell morphological changes induced by TGF-β2 (Fig. 7C). SB203580 did not affect the increase in α-SMA expression (Fig. 7D). By contrast, SB203580 suppressed the elevation of fibronectin expression by TGF-β2 (Fig. 7E). N-cadherin was slightly inhibited by SB203580, but not significantly so (Fig. 7F). SB203580 had no effect on the TGF-β2-induced decrease in Tie2 (Fig. 7G). The increases in Snail and collagen type IV expression mediated by TGF-β2 were significantly inhibited by SB203580 (Fig. 7H, I). Real-time PCR showed that SB203580 caused changes in the expression of ACTA2, CDH2, SNAI1, and TEK, in agreement with the western blotting results ( Supplementary Fig. S3). SB203580 had no effect on BMP4 expression but significantly suppressed GREM1 expression ( Supplementary Fig. S3).

Discussion
SC cells from glaucoma patients show increased expression of the fibrosis markers α-SMA, fibronectin, and collagen, and increased cell proliferation 19,20 . These changes are similar to those induced by TGF-β2 stimulation 17,18 .
Our findings imply induction of EndMT by TGF-β2 in SC cells. Therefore, EndMT may be induced by aqueous humor cytokines such as TGF-β2 in the SC of glaucoma patients, resulting in decreased endothelial function and decreased aqueous humor outflow control. Furthermore, Y-27632 suppressed the TGF-β2-induced EndMT in SC cells. Suppression of EMT or EndMT by ROCK inhibitors has been reported in corneal endothelial cells, retinal pigment epithelial cells, lens epithelial cells, and lung epithelial cells [31][32][33][34][35] . These results indicate that ROCK inhibitors act directly on SC cells in glaucoma patients, suppress the EndMT induced by TGF-β2 in the aqueous humor, and restore endothelial function to normal, thereby improving outflow resistance. However, because Y-27632 does not suppress the decrease in PECAM1 expression and the increase in FN1 and COL4A1 expression induced by TGF-β2 ( Supplementary Fig. S1), ROCK inhibitors do not completely restore endothelial function.
In addition, we did not examine the mesenchymal-to-endothelial transition, the reverse of EndMT, which is a subject for future investigation. BMP4 reportedly affects intraocular pressure, and expression of gremlin, which acts as its endogenous inhibitor, causes intraocular pressure elevation 36,37 . In this study, TGF-β2 stimulation decreased the expression of BMP4 and increased that of GREM1. In addition, a ROCK inhibitor significantly suppressed TGF-β2-induced BMP4 downregulation and GREM1 upregulation. Furthermore, BMP4 stimulation significantly suppressed the fibronectin expression and TER elevation induced by TGF-β2. BMP4 reportedly suppresses EMT in lens epithelial cells and retinal pigment epithelial cells 38,39 . These results implicate the suppression of fibronectin expression in SC cells via BMP4 signaling in the improvement of aqueous outflow resistance by ROCK inhibitors. In addition, the effect of BMP4 in the presence of TGF-β2 was observed only at a high concentration (10 ng/mL), and BMP4 had no effect on TGF-β2-induced mRNA level changes ( Supplementary Fig. S2). These results are thought to be related to the TGF-β2-induced increase in gremlin expression.
Y-27632 significantly inhibited p38 phosphorylation irrespective of the presence or absence of TGF-β2. Activation of Rho/ROCK upstream of p38 MAPK has been reported in orbital fibroblasts and microglia 40,41 . Therefore, p38 MAPK is likely located downstream of the Rho/ROCK signal in SC cells. SB203580, a p38 MAPK inhibitor, significantly suppressed TGF-β2-induced upregulation of fibronectin, collagen type IV, and Snail, implying that p38 MAPK is an important signal in TGF-β2-induced EndMT induction. However, unlike the Y-27632, p38 MAPK inhibition had no significant effect on TGF-β2-induced Tie2 expression reduction and N-cadherin expression elevation. In this study, although there was no statistically significant difference, Y-27632 showed a tendency to suppress the phosphorylation of Akt and Smad3 by TGF-β2. It has been reported in other endothelial cells that suppression of Akt 42,43 and Smad 44-46 signals suppress EndMT. In addition, we previously reported that HDAC inhibitors significantly suppressed Akt signaling and suppressed EndMT in SC cells 17 . These reports suggest that the inhibition of Akt and Smad3 phosphorylation by ROCK inhibitors also contributes to the suppression of EndMT along with the suppression of p38 MAPK signals.

Experimental procedures
Cell culture. Primary monkey SC cells were isolated from enucleated eyes of cynomolgus monkeys as described previously 17,29 . The cells were cultured in low-glucose Dulbecco's modified Eagle's medium (DMEM; 041-29775, FUJIFILM Wako Pure Chemical, Osaka, Japan) in the presence of 10% FBS (SH30910.03, HyClone™, Cell proliferation assay. Cell proliferation was evaluated using the WST-8 assay (Cell Counting Kit-8, CCK-8; CK04, Dojindo, Kumamoto, Japan), as described previously 17 . SC cells were seeded in 96-well plates at 1 × 10 4 cells per well and incubated for 24 h. After serum starvation for 24 h, TGF-β2, Y-27632, BMP4 and SB203580 were added to the cells and incubated for 72 h. CCK-8, which is a detection reagent, was added and the absorbance at 450 nm was measured using a microplate reader (Multiskan FC, Thermo Fisher Scientific) after incubation for 2 h. Cell proliferation is presented as relative change compared to the control.
Immunocytochemistry. Fluorescent immunostaining of SC cells was performed as reported previously 17,47 .
Phase-contrast images were acquired with an inverted microscope (IX71, Olympus, Tokyo, Japan) before fixation. Treated cells were fixed with 4% (v/v) paraformaldehyde in PBS for 15 min at room temperature. After washing with cytoskeletal buffer (10 mM 2-morpholinoethanesulfonic acid potassium salt, 150 mM NaCl, 5 mM EGTA, 5 mM MgCl 2 , and 5 mM glucose, pH 6.1), the cells were treated with 0.5% (v/v) Triton X-100 in PBS for 12 min at room temperature. For blocking, the cells were treated with serum buffer (10% FBS and 0.2 mg/mL sodium azide in PBS) at 4 °C for at least 2 h. The cells were incubated with primary antibodies (see Table 1) overnight at 4 °C. Next, the cells were incubated with an anti-mouse or anti-rabbit IgG secondary antibody labeled with Alexa Fluor 488 at room temperature for 30 min. For visualization of F-actin, phalloidin labeled with Alexa Fluor 546 was incubated with the secondary antibody. After mounting with VECTASHIELD mounting medium containing 4, ' 6-diamidino-2-phenylindole (DAPI; H-1200, Vector Laboratories, Burlingame, CA), the cells were observed under a laser confocal microscope (FV-1200; Olympus) or an all-in-one epifluorescence microscope (BZ-X710; Keyence, Itasca, IL, USA). We performed immunostaining in at least three independent experiments.
Western blotting. Western blotting was performed as described previously 17,30 . Cell lysates were collected from SC cells 24 and 72 h after treatment using LIPA buffer (89900, Thermo Fisher Scientific) with pro-  (Table 2) diluted with 5% BSA or 5% skim milk in TBS-T overnight at 4 °C. After washing with TBS-T, the membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies for 30 min at room temperature. The chemiluminescence signal was detected using ECL Prime (RPN2232, Cytiva) or ECL Select western blotting detection reagent (PRN2235, Cytiva) and a luminescence imager (LAS 4000mini; FUJIFILM, Tokyo, Japan). All membranes were stripped of antibodies using WB stripping solution (05364-55, Nacalai Tesque) and incubated with an anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody followed by an HRP-conjugated rabbit IgG antibody as a loading control. The densitometry of immunoreactive bands was analyzed using Image J software (National Institutes of Health, Bethesda, MD).

Real time RT-PCR.
Real time RT-PCR was performed as described previously 17,48 . RNA samples were prepared from SC cells using a NucleoSpin RNA Kit (U0955, Takara Bio, Shiga, Japan), according to the manufacturer's instructions. The concentration of RNA samples was measured using a DS-11 NanoPad spectrophotometer (DeNovix, Wilmington, DE). Reverse transcription was performed using PrimeScript™ RT Master Mix (RR036A, Takara Bio) according to the manufacturer's protocol. Quantitative PCR was performed using TB Green® EX Taq II (RR820A, Takara Bio) and the StepOnePlus Real-Time PCR System (Thermo Fisher Scientific). Relative expression of target mRNAs was compared to the control samples using the comparative threshold cycle method; GAPDH was used as an endogenous control based on the results of preliminary studies (Supplementary Fig. S4). The primer sequences are listed in Table 3.  3.0; SAS Institute, Cary, NC) was used for statistical analysis. Comparison of multiple groups was conducted using the Tukey-Kramer honestly significant difference (HSD) test. In all analyses, differences were considered statistically significant at p < 0.05.

Data availability
All data generated or analyzed during this study are included in this published article and its Supplementary Information files.