Tetramethylpyrazine Attenuates Transdifferentiation of TGF-β2–Treated Human Tenon's Fibroblasts

Abstract

Purpose: To investigate the pharmacologic effect of tetramethylpyrazine (TMP) on human Tenon's fibroblasts (HTFs), the cells implicated in scarring after filtration surgery.

Methods: Transforming growth factor-β2 (TGF-β2) was used to stimulate a fibrotic phenotype in primary HTFs, and the influence of TMP on the fibrotic phenotype was assessed. Cell proliferation and cell cycle regulation were profiled. Immunofluorescence staining tracked proliferating cell nuclear antigen (PCNA) expression. Transwell assays monitored cell migration. Flow cytometry measured TMP toxicity. In addition, in TGF-β2–treated HTFs, Western blot and immunofluorescence were employed to assess the expression of α-smooth muscle actin (α-SMA). The TMP-mediated activity on cytoskeletal arrangements and extracellular matrix (ECM) accumulation in HTFs was evaluated using actin polymerization and Western blot assays. Moreover, TGF-β–dependent activation of Smad3 and p38 was examined by Western blot analysis.

Results: In TGF-β2–treated HTFs, TMP reduced proliferation and migration but did not induce apoptosis. Moreover, TMP attenuated expression of α-SMA and suppressed stress fiber formation stimulated by profibrotic cytokine; it also counteracted TGF-β2–induced cytoskeletal rearrangements, morphologic changes, and ECM accumulation. Smad3 and p38 mitogen-activated protein kinase (MAPK) signaling were downstream of the TMP-sensitive effect.

Conclusions: Tetramethylpyrazine counteracts TGF-β2–mediated myofibroblast transdifferentiation and attenuates ECM component deposition and cell proliferation in HTFs, implicating TMP as a potential antifibrosis agent in glaucoma filtration surgery.

Trabeculectomy is the most common and effective surgical procedure for relieving glaucoma¹ but when it fails, the complication is usually bleb scarring, caused by proliferating Tenon's fibroblasts that produce a large amount of extracellular matrix (ECM).^2,3 Currently, the most effective treatment for bleb scarring is an antimetabolite, such as mitomycin-C or 5-fluorouracil. This is a drastic and suboptimal approach, associated with severe and potentially blinding complications. Improved therapeutic approaches with milder side effects are an urgent need. 

Transforming growth factor-β (TGF-β) has been identified as the initiator and main propellant in both physiological wound healing and pathologic tissue fibrosis.^5–7 It drives the conversion of fibroblasts to myofibroblasts, a normal part of healthy wound healing. However, the persistence of myofibroblasts is associated with overexpression of anti-smooth muscle actin (α-SMA) and deposition of ECM components, which leads to tissue fibrosis.^8,9 Hence, targeting this profibrotic cytokine and its downstream intracellular signaling pathways might offer a therapeutic strategy for disrupting postoperative fibrosis. Smad-dependent signaling is the main pathway controlling TGF-β–mediated fibrosis.¹⁰ In lens epithelial cells, Smad3 silencing attenuates the TGF-β2–mediated proliferation and aberrant expression of ECM.^11,12 Mitogen-activated protein kinase (MAPK) signaling is also stimulated by TGF-β; and recent findings demonstrated that inhibition of the MAPK p38 attenuated the TGF-β2–induced transition of human Tenon's fibroblasts (HTFs) into myofibroblasts.¹³ 

A widely used clinical treatment for fibrosis and degenerative nerve diseases is tetramethylpyrazine (TMP), a natural, small molecule extracted from the traditional Chinese medicinal herb chuan xiong (Ligusticum wallichii Franchat). Several studies demonstrated that TMP has multiple pharmacologic activities, including suppressing tumor growth, inhibiting neovascularization, and attenuating pathologic fibrosis.^14–17 In vivo and in vitro studies indicate that TMP protects the liver from CCl4-induced fibrogenesis by modulating the NLRP3 inflammasome pathway¹⁸; and our previous studies in rats demonstrated that TMP protects against pulmonary bleomycin-induced fibrosis by modulating the SDF-1/CXCR4 axis.¹⁶ Studies in the liver suggest that TMP likely exerts antifibrotic effects by inhibiting Smad2/3 expression; and in hypoxic myocardial cells, protective doses of TMP downregulated the p38 MAPK pathway.^19,20 

Accumulating evidence suggests that TMP may have therapeutic potential as a complementary therapy for various ocular diseases, including vascular retinopathy, corneal neovascularization, optic neuropathy, and glaucoma.^16,21–23 In particular, our latest investigation implicated TMP in attenuation of TGF-β–induced pathologic changes in trabecular meshwork cells, showing that it inhibits cytoskeleton remodeling and ECM accumulation through the SDF/CXCR4 axis.²⁴ 

Prompted by these previous reports, we sought to characterize the bioactivity of TMP and its underlying mechanism. In the present study, fibrosis of HTFs was stimulated with TGF-β2, transdifferentiating fibroblasts to myofibroblasts. Here, we first demonstrated that TMP might modulate the Smad and p38 MAPK pathways to attenuate the TGF-β2–induced fibrosis, suggesting its potential therapeutic value for glaucoma patients. 

Materials and Methods

In Vitro Culture and TMP Treatment of Primary Human Tenon's Capsule Fibroblasts

Anonymous human Tenon's capsule samples were obtained (eight samples from eight different subjects), under sterile conditions during strabismus surgery at the ZhongShan Ophthalmic Center, Sun Yat-sen University, with the approval of the ethical committee of the ZhongShan Ophthalmic Center and in accordance with the Declaration of Helsinki (20140311). Exclusion criteria included any other ocular pathologies or systemic diseases. Preoperatively, tobramycin 0.3% (Tobrex; Alcon, Puurs, Antwerpen, Belgian) eye drops were administered three times daily for 3 days, which would not affect HTF status and physiology. Samples (∼4 × 4 mm) were taken from an area near the surgical incision that appeared to be free of blood vessels. Samples were immediately transferred into saline solution and sent for lab analysis. Fresh tissue was minced, seeded onto 60-mm culture dishes, and maintained in Dulbecco's modified Eagle's medium (DMEM)/F12 (Gibco, Carlsbad, CA, USA) containing 10% fetal bovine serum (FBS, Gibco) and 1% penicillin/streptomycin (Gibco) at 37°C in 5% CO₂. 

Approximately 7 days after seeding, when primary cultures had reached 50% confluence, they were transferred into culture dishes using 0.25% trypsin-EDTA (Gibco). Human Tenon's fibroblasts were characterized by adherent morphology (phase-contrast microscopy) and expression of vimentin. Cells from passages 3 to 5 were used in TMP experiments. 

For TMP treatment, HTFs were serum starved and randomly divided into four groups: dimethyl sulfoxide (DMSO) alone (DMSO; 2 μL for 1 mL medium, 0.2%; Sinopharm Chemical Reagent Co., Ltd., Shanghai, China) or including TMP (200 μM; Sigma-Aldrich Corp., St. Louis, MO, USA), TGF-β2, (10 ng/mL; CST, Danvers, MA, USA) or TGF-β2 (10 ng/mL) + TMP (200 μM). For serum-induced proliferation assay, HTFs were treated with TMP at different concentrations, 200 μM and 400 μM, in the presence of 10% FBS. 

Western Blotting

Whole-cell lysates were resolved by SDS-PAGE, transferred to polyvinylidene fluoride (PVDF) membrane, and Western blotted using primary antibodies recognizing α-SMA, P-p38, p38, P-Smad3, and Smad3 (1:1000, CST), thrombospondin-1 (TSP-1) (1:1000; Abcam, Cambridge, MA, USA), fibronectin (FN), collagen I, collagen III (1:1000; Proteintech, Chicago, IL, USA), and mouse anti-TSP-1 (1:1000, Abcam), respectively; GAPDH served as a loading control (1:10000, Proteintech). Immunoreactive species were detected with horseradish peroxidase–conjugated goat anti-rabbit and anti-mouse (CST) and visualized by chemiluminescence. 

Immunofluorescence

Human Tenon's fibroblasts were fixed for 15 minutes in 4% paraformaldehyde; permeabilized for 10 minutes in PBS plus 0.1% Triton X-100 (Sigma-Aldrich Corp.); and blocked for 30 minutes in 10% normal goat serum (Boster, Wuhan, China). Cells were then incubated with rabbit anti-FN (1:100, Boster), mouse anti-TSP1 (1:100, Abcam), mouse anti-proliferating cell nuclear antigen (anti-PCNA) (1:100, Abcam), rabbit anti-collagen I (1:100, Proteintech), or rabbit anti-collagen III (1:100, Proteintech) at 4 °C. Cytoskeleton remolding was evaluated by double staining with FITC-phalloidin (Sigma-Aldrich Corp.) and mouse anti-vimentin (1:100, CST). Secondary antibodies were anti-mouse (Alexa Fluor 488 conjugate, green, or Alexa Fluor 555 conjugate, red, CST) and anti-rabbit (Alexa Fluor 555 conjugate, red, CST). Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI) (Sigma-Aldrich Corp.). Photomicrographs were captured by an Axioplan-2 imaging microscope system (Carl Zeiss, Inc., Jena, Germany). Cytoskeleton rearrangement was quantified as the intensity of F-actin fluorescence as measured by LSM 510 Examiner software (Carl Zeiss, Inc.). Ten randomly selected photomicrographs were analyzed, and the results were normalized by cell number. 

Cell Proliferation Assays

For serum-induced proliferation, HTFs were seeded in 96-well plates and then treated with 200 μM TMP or 400 μM TMP or DMSO, in the presence of 10% FBS. For TGF-β2–induced proliferation, serum-starved HTFs were divided into four groups: DMSO, TMP (200 μM), TGF-β2 (10 ng/mL), and TGF-β2 (10 ng/mL) + TMP (200 μM). After 48 hours, HTFs were incubated with CCK8 (Invitrogen, Carlsbad, CA, USA) for 3 hours. Absorbance was measured at 450 nm using a fluorescence plate reader (Power Wave XS; BIO-TEK, Winooski, VT, USA). Cell viability was determined based on the optical density ratio of a treated culture relative to an untreated control. 

Flow Cytometry

Tetramethylpyrazine-induced apoptosis of HTFs cultured with 10% FBS was quantified by flow cytometry using FITC-labeled annexin V and propidium iodide (PI) (Annexin-V-PI Kit; Roche, Mannheim, Germany) per manufacturer protocols. For cell cycle assay, serum-starved HTFs were divided into four groups: DMSO, TMP (200 μM), TGF-β2 (10 ng/mL), and TGF-β2 (10 ng/mL) + TMP (200 μM). After 48 hours, HTFs were fixed with 70% ice-cold ethanol, and their distribution in the cell cycle was determined with PI staining and evaluated by flow cytometry. 

Transwell Invasion Assay

The motility of HTFs was examined by a transwell invasion assay. Human Tenon's fibroblasts were trypsinized and resuspended in serum-free medium and then cultivated to the upper chamber of 24-well inserts with 8-μm pores (Corning Costar, New York, NY, USA). Six hundred microliters DMEM/F12 containing 20% FBS was added to the lower chambers, and then randomly divided into four groups: DMSO, TMP (200 μM), TGF-β2 (10 ng/mL), and TGF-β2 (10 ng/mL) + TMP (200 μM). The cell treatments were added to the upper chamber in the serum-free medium. After 24 hours, cells in lower chambers were fixed and stained for 15 minutes with 0.1% crystal violet (Beyotime, Haimen, China), photographed (×40), and counted in five randomly selected fields of each membrane. 

Statistical Analysis

All in vitro experiments were performed in triplicate, and all data were represented as the mean value ± SD; a P < 0.05 was considered statistically significant (*statistically significant differences between the control and TGF-β; #statistically significant differences between TGF-β2 and TGF-β2 + TMP). The differences between the means were evaluated using a 2-tailed Student's t-test (for two groups) or analysis of variance (ANOVA; for more than two groups). 

Results

TMP Reduces Proliferation of Human Tenon's Capsule Fibroblasts (HTFs)

After 7 days culture, the HTF cells were adhered to the bottom of the culture dish, exhibiting a spindly and elongated phenotype (Fig. 1A). Immunofluorescence staining for vimentin was strongly positive in the cytoplasm (red), consistent with the identification of the cells as HTFs (Fig. 1A). In the presence of serum, TMP treatment had little effect on cell morphology but did affect cell viability (Fig. 1B), in a dose-dependent manner (Fig. 1C, con: 100%, TMP200: 93.19 ± 2.92%, TMP400: 78.26 ± 3.63%, P < 0.05). In contrast, in serum-starved cells, TMP treatment had no effect (data not shown). Trimethylpyrazine also downregulated expression of PCNA, a cellular marker of proliferation, in HTFs cultured with 10% FBS, as evidenced by immunofluorescence (Fig. 1D). Additionally, the reduction in cell number could not be attributed to apoptosis in HTFs cultured with 10% FBS (Fig. 1E, apoptosis: control, 0.943 ± 0.277%; TMP200, 1.047 ± 0.347%; TMP400, 1.09 ± 0.164%; death: control, 1.627 ± 0.481%; TMP200, 1.673 ± 0.505%; TMP400, 2.107 ± 0.206%; P > 0.05), suggesting that TMP may inhibit proliferation in HTFs. 

Figure 1

 

Tetramethylpyrazine inhibits the proliferation of HTFs in a dose-dependent manner. (A) Representative phase-contrast images of human Tenon's capsule fibroblasts (HTFs). Spindle-shaped and elongated cells crawled out of surgically sampled tissue approximately 1 week after seeding. Cells were stained by vimentin (red) and DAPI (blue). (B) The morphologic changes of HTFs treated with different doses of TMP in the presence of 10% FBS for 48 hours. (C) Viability of HTFs in 10% FBS was lowered by TMP. (D) Immunofluorescence staining revealed that PCNA expression of HTFs in the presence of 10% FBS was downregulated by TMP in a dose-dependent manner. (E) Apoptosis of HTFs was quantified by flow cytometry after annexin V and propidium iodide staining demonstrating that TMP had no effect on the incidence of apoptosis in HTFs cultured in 10% FBS. Results were confirmed in at least three independent experiments. Error bars represent the standard deviation of the mean (n = 3). The asterisks indicate statistically significant differences between the control and experimental cells (P < 0.05).

TMP Suppresses TGF-β2–Induced Migration and Proliferation in HTFs

Trimethylpyrazine treatment of conjunctiva has an anti-scarring effect that might be the result of antiproliferative properties.^16,17 To test this, we measured whether TMP (200 μM) affected transit of serum-starved, TGF-β2–exposed HTFs through the cell cycle. We found that, whereas in the presence of TGF-β2 (10 ng/mL), 7.387 ± 1.225% of cells were in S-phase, TMP treatment counteracted the effect of TGF-β2, such that the percentage of cells in S-phase dropped to 4.27 ± 0.517% (Figs. 2A, 2B, P = 0.018). In addition, a cell viability assay confirmed that TMP attenuated TGF-β2–induced proliferation of HTFs (con: 100%, TMP: 93.83 ± 5.86%, TGF-β2: 173.24 ± 23.99% TGF-β2 + TMP: 128.82 ± 13.95%, *P < 0.01, #P < 0.05, Fig. 2C). 

Figure 2

 

Tetramethylpyrazine attenuates TGF-β2–induced migration and proliferation in HTFs. (A) Exogenous TGF-β2 stimulation induced S-phase arrest that was counteracted by TMP treatment. (B) Histograms presenting the percentage of cells in S-phase. (C) Tetramethylpyrazine attenuated proliferation of HTFs induced by TGF-β2. The data are presented as the percent survival compared with controls. (D) Histogram representing the number of migrated cells per ×40 field, demonstrating that TMP suppressed TGF-β2–induced migration in HTFs. (E) The number of migrated HTFs as determined by transwell assay. Results were confirmed in three independent experiments. Error bars represent the standard deviation of the mean (n = 3). *Statistically significant differences between control and TGF-β2 treated; #statistically significant differences between TGF-β2 treated and TGF-β2 + TMP treated (*P < 0.05).

Moreover, transwell invasion assays were employed to evaluate the effect of TMP on cell migration. As shown in Figures 2D and 2E, exogenous TGF-β2 markedly enhanced cell proliferation and migration; however, this effect was significantly weakened by cotreatment with TMP (migration cell numbers: con, 32 ± 4.58; TMP, 29.33 ± 3.51; TGF-β2, 62.33 ± 5.03; TGF-β2 + TMP, 42.67 ± 4.04; *P < 0.01, #P < 0.05). Taken together, our data demonstrated that TMP suppressed TGF-β2–induced proliferation and migration of HTFs. 

TMP Suppresses TGF-β2–Induced α-SMA Expression in HTFs

To track whether TMP affected TGF-β2–induced transdifferentiation of HTFs, we measured changes in the expression level of α-SMA, a myofibroblast marker. Western blotting and immunofluorescence staining showed that culturing HTFs in 10% FBS, as opposed to in serum-free media, upregulated α-SMA expression by 6.62-fold, P = 0.005. Interestingly, TMP treatment reduced this to 3.34-fold (P = 0.03; Figs. 3A, 3B), while in serum-starved cells, TMP treatment had no effect on α-SMA expression. Moreover, 48 hours of TGF-β2 stimulation upregulated α-SMA expression 9.62-fold (P = 0.001) over control treatment (Figs. 3C, 3D), suggesting that the cells were differentiating into myofibroblasts. However, α-SMA upregulation was significantly attenuated by 200 μM TMP treatment (by 6.53-fold, as compared with control treatment, P = 0.01). Consistent with these data, immunofluorescence of untreated cells showed that the cytosol stained only weakly for α-SMA (Fig. 3E), while ∼80% of those treated with TGF-β2 showed robust assembly of α-SMA–positive stress fibers, and this stress fiber assembly was attenuated by treatment with 200 μM TMP. Taken together, these observations suggest that TMP treatment attenuated TGF-β2–induced α-SMA expression and its incorporation into stress fibers. 

Figure 3

 

Tetramethylpyrazine attenuates α-SMA expression in TGF-β2–stimulated HTFs. (A) α-SMA expression in HTFs was induced by 10% FBS, but the upregulation was counteracted by additional TMP. (B) α-SMA expression in HTFs was quantified by densitometry; data are presented graphically. (C) Western blot indicated that TGF-β2 enhanced α-SMA expression; however, this phenomenon was attenuated by TMP. (D) The relative expression of α-SMA in HTFs, quantified by densitometry; data are presented graphically. (E) Immunofluorescence staining revealed that α-SMA–positive stress fibers were induced by TGF-β2 in ∼80% of HTFs. This phenomenon was attenuated by 200 μM TMP. Results were confirmed in three independent experiments. Error bars represent the standard deviation of the mean (n = 3). *Statistically significant differences between the control and TGF-β2 treated; #statistically significant differences between TGF-β2 treated and TGF-β2 + TMP treated (*P < 0.05).

TMP Attenuated TGF-β2–Mediated F-Actin Remodeling in HTFs

Transforming growth factor-β signaling likely elicits major changes in actin cytoskeletal dynamics to drive the pathogenesis of conjunctiva fibrosis after trabeculectomy.⁶ Therefore, cytoskeletal rearrangements induced by TGF-β2 (10 ng/mL) were assessed in actin polymerization assays. Cells were serum starved for 48 hours in the presence of TMP (200 μM) or DMSO (as a negative control), with or without TGF-β2 (10 ng/mL), and then stained with ani-vimentin (red) and F-actin (green). The accumulation of curved filamentous stress fibers traversing the entire cell as well as cellular expansion was observed in HTF cells exposed to TGF-β2 (Fig. 4A). Interestingly, TMP treatment significantly attenuated this phenomenon (F-actin content: con, 23.31 ± 4.73; TMP, 16.51 ± 1.82; TGF-β2, 47.76 ± 10.67; TGF-β2 + TMP, 26.21 ± 6.31, *P < 0.01, #P < 0.01, Figs. 4A, 4B). These findings agree with previous studies of trabecular meshwork cells and ECV 304 cells.^16,24 In addition, it is well established that the actin cytoskeleton is associated with the vimentin filament system, which is involved in regulating cell shape and tissue homeostasis.²⁵ By immunofluorescence staining assay, more pronounced vimentin expression was observed in TGF-β2–treated HTFs, but counteracted by TMP treatment (vimentin content: con, 26.92 ± 6.67; TMP, 19.11 ± 3.48; TGF-β2, 77.52 ± 9.39; TGF-β2 + TMP, 42.67 ± 8.38, *P < 0.01, #P < 0.01, Figs. 4A, 4C). Taken together, these observations strongly suggest that TMP could attenuate the actin cytoskeleton remodeling in TGF-β2–stimulated HTFs. 

Figure 4

 

Tetramethylpyrazine attenuated F-actin remodeling in HTFs stimulated with TGF-β2. (A) The HTFs were double-stained with F-actin (green) and vimentin (red). The nucleus was stained with DAPI (blue). Tetramethylpyrazine attenuated TGF-β2–induced cytoskeleton arrangement and reorientation of the vimentin network in HTFs. (B) Images were captured by a confocal laser scanning microscope (Zeiss LSM 510). F-actin content was quantified using Zeiss LSM 510 Examiner software and normalized by cell number to evaluate stress fiber formation. (C) Vimentin content was quantified using Zeiss LSM 510 Examiner software and normalized by cell number. Results were confirmed in at least three independent experiments. The error bars represent the standard deviation of the mean (n = 3). *Statistically significant differences between control and TGF-β2 treated; #statistically significant differences between TGF-β2 treated and TGF-β2 + TMP treated (*P < 0.05).

TGF-β2–Induced Extracellular Matrix Protein Production Is Reduced by TMP

Western blotting and immunofluorescence were employed to assess the bioactivity of TMP on ECM production in HTFs. As shown in Figures 5A and 5B, our results revealed a dramatically upregulated expression of ECM components, including TSP-1, collagen I, collagen III, and FN (by 3.64-, 3.22-, 2.83-, and 5.09-fold, respectively, P < 0.05), by TGF-β2 treatment compared to untreated controls. Again, treatment with TMP significantly attenuated this phenomenon, resulting in a 1.37-, 1.04-, 0.94-, and 2.08-fold reduction in TSP-1, collagen I, collagen III, and FN, respectively (as compared to TGF-β2–treated cells; P < 0.05). Moreover, in HTFs stimulated with TGF-β2, TMP treatment also decreased the intracellular upregulation of TSP-1, collagen I, and collagen III, and attenuated the extracellular microfibrillar network formation of FN connecting the neighboring cells, as evidenced by immunofluorescence staining (Fig. 5C). Collectively, these results suggest that for HTFs, TMP could reduce the fibrotic effects of TGF-β2. 

Figure 5

 

Tetramethylpyrazine suppressed ECM accumulation in TGF-β2–stimulated HTFs. (A) Western blot analysis indicated that TGF-β2 dramatically elevated expression of ECM proteins, including TSP-1, collagen I, collagen III, and FN. Nevertheless, treatment with TMP significantly attenuated this phenomenon, resulting in a 1.37-, 1.04-, 0.94-, and 2.08-fold reduction in TSP-1, collagen I, collagen III, and FN, respectively, compared to TGF-β2 treatment alone. (B) The relative expression of ECM in HTFs was quantified by densitometry; data are presented as histograms. Results were confirmed in three independent experiments. (C) Immunofluorescence staining revealed that TMP could downregulate the TGF-β2–induced ECM deposition of TSP-1, collagen I, collagen III, and FN in HTFs. Error bars represent the standard deviation of the mean (n = 3). *Statistically significant differences between the control and TGF-β2 treated; #statistically significant differences between TGF-β2 treated and TGF-β2 + TMP treated (*P < 0.05).

TMP Attenuates TGF-β–Mediated HTF Fibrosis Through the Smad/p38 MAPK Axis

To investigate the antifibrotic mechanism of TMP, we studied intracellular signaling via p38 MAPK and Smad3, two pathways that can independently trigger TGF-β–induced fibrosis.^10,13 As shown in Figures 6A and 6B, TGF-β2 stimulation triggered Smad3 phosphorylation by 7.87-fold over control treatment (P = 0.016), and this phenomenon was attenuated by TMP treatment: Smad3 phosphorylation level increased only 4.09-fold, as compared with control treatment (P = 0.028). A similar effect was observed for MAPK p38: TGF-β2 induced a 4.07-fold phosphorylation of p38 (P = 0.005); however, only 2.15-fold phosphorylation of p38 was observed in TGF-β2–stimulated HTFs cocultured with TMP (P = 0.012). Data are presented as levels relative to the housekeeping gene GAPDH and normalized to the ratio in control cells. Although neither signaling pathway was completely blocked by TMP, blocking both simultaneously might have a synergistic effect that could explain its antifibrotic bioactivity. 

Figure 6

 

Tetramethylpyrazine attenuates TGF-β–mediated HTF degeneration by modulating the Smad and p38 axis. (A) Western blot analysis demonstrated increased phosphorylation of both Smad3 and p38 in TGF-β2–stimulated HTFs, which was suppressed by TMP. (B) The relative expression of phosphorylated Smad3 and p38 in HTFs was quantified by densitometry; data are presented as histograms. Results were confirmed in three independent experiments. Error bars represent the standard deviation of the mean (n = 3). *Statistically significant differences between control and TGF-β2 treated; #statistically significant differences between TGF-β2 treated and TGF-β2 + TMP treated (*P < 0.05).

Discussion

When surgical therapy for glaucoma fails, the most common cause is postoperative scarring, and effective anti-scarring treatments with mild side effects are urgently needed. Here, we demonstrated that TMP, a traditional Chinese herb, could attenuate myofibroblast transdifferentiation of HTFs as evidenced by reduced α-SMA expression, modified cytoskeleton arrangement, and decreased ECM protein deposition. 

Tenon's fibroblast plays a pivotal role in the pathogenesis of filtering bleb scarring by virtue of the transdifferentiation of quiescent cells into myofibroblastic cells.²⁶ Controlling the balance of fibroblasts and myofibroblasts is thought to be critical in postoperative healing. Clinical studies observed increased levels of TGF-β in both conjunctival tissue and aqueous humor of postfiltration surgery patients,^26,27 which might stimulate the proliferation and migration capacity of HTFs in vivo, resulting surgical failure. Transforming growth factor-β2 is recognized as the predominant isoform of all three isoforms in humans and is implicated in the pathogenesis of several ocular fibrosis diseases.²⁸ Therefore, we evaluated the bioactivity of TMP on HTFs in an in vitro model mimicking the state of transdifferentiation in Tenon's capsule fibrosis and elucidated the underlying molecular mechanisms. 

Numerous studies have provided solid evidence concerning the multifunction of TMP in inhibiting neovascularization, fibrosis, and thrombosis under pathologic conditions and in suppressing tumor growth.^14–18,29 Clinical studies also suggest that TMP shows promise as complementary therapy for various ocular diseases, including corneal neovascularization and glaucoma.^21–24 In the present study, our results indicate that TMP exhibits an inhibitory effect on the proliferation in profibrotic cytokine-treated HTFs. Moreover, in 10% FBS, HTFs treated with TMP were less viable and expressed less PCNA, but were not apoptotic even at a high concentration (400 μM), indicating that the limiting effects of TMP were unlikely to be due to toxicity. This is consistent with our previous report showing that TMP has no cytotoxic effects on PHTM cells,²⁴ suggesting that it is clinically safe. 

Cell migration is a key facet in wound healing as well as scarring. Evidence suggests that TMP may inhibit proliferation and migration of several cell types, including ECV304 and PHTM cells.^16,24 Trimethylpyrazine also delayed TGF-β2–induced migration of primary HTFs in our study, as evidenced by the transwell invasion assays. Interestingly, under serum-free conditions, TMP treatment alone had no effect on either proliferation or migration. Taken together, these observations strongly suggest that TMP attenuates the pathogenic proliferation of HTFs, raising the possibility that it might offer therapeutic protection from fibrogenic insults. 

The phenotypic transition of HTFs into myofibroblasts is similar to the pathologic processes inherent to fibroproliferative diseases elsewhere in the body: α-SMA is overexpressed; stress fibers form; and ECM is produced.³⁰ In particular, during fibrosis, α-SMA is recruited to actin stress fibers to enhance cell contractility.³¹ The present study illustrated that the additional application of TMP could not only downregulate expression of α-SMA induced by TGF-β2 or serum, but also block the TGF-β–mediated incorporation of α-SMA into actin stress fibers. Moreover, the actin polymerization analysis (visible with double staining of F-actin and vimentin) demonstrated that TGF-β2–treated HTFs assemble curved, filamentous stress fibers and spatially reorient the vimentin network as the cells expand. Trimethylpyrazine attenuated these phenomena too. Similar observations have been made in other cell types.^16,24 In HTFs, TMP also attenuated TGF-β2–induced accumulation of ECM proteins, including TSP-1, collagen I, collagen III, and FN, as evidenced by Western blot and immunofluorescence. These findings collectively establish that TMP has antifibrotic effects on HTFs. 

The signal transduction pathway for TGF-β–dependent fibrosis involves multiple cell surface receptors leading to phosphorylation and gene regulation.^10,32,33 Studies have indicated that TGF-β induced the Smad signaling pathway already in quiescent fibroblast,^11,34 while the p38 MAPK is constitutively active during transdifferentiation stages.^13,35 In HTFs, we analyzed both pathways to explore the underlying mechanism of the antifibrotic activity of TMP. Our data indicated that TMP not only impeded the capacity of TGF-β2 to phosphorylate Smad3, but also suppressed TGF-β2–mediated upregulation of p38 phosphorylation. Transforming growth factor-β receptor signaling involves multiple downstream signaling pathways, which are dependent or independent of other signaling pathways.³⁶ Although TMP did not block Smad3 or MAPK p38 signaling completely, partially inhibiting both might have a synergistic effect that gives the molecule its antifibrotic bioactivity. 

Trimethylpyrazine is suggested to have several beneficial effects on the eyes, including inhibiting angiogenesis and protecting retina neurocytes.^16,22 Our previous studies also demonstrated that TMP attenuates TGF-β–mediated ECM deposition and cytoskeletal rearrangements in PHTM cells.²⁴ By using an in vitro model mimicking the pathophysiological changes of Tenon's capsule fibrosis, the present study demonstrates the antifibrotic role of TMP in HTFs. Taken together, our results indicate that TMP may have potential therapeutic value for glaucoma patients, where it may not only slow degeneration of trabecular meshwork cells and protect retina neurocytes from high IOP, but also attenuate excessive scarring formation after filtration surgery. Clearly, as a Chinese traditional herb, TMP warrants further in-depth evaluation both in vivo and in vitro to investigate its therapeutic potential and optimize its clinical benefits. Moreover, since no TMP eye drops have been developed yet, little is known about the potential side effects of topical administration of TMP, which needs further investigation. Collectively, this study not only gives new insight into the pharmacologic activities of TMP, but also provides a basis for future studies to explore effective therapeutic strategies for antifibrosis. 

Acknowledgments

Supported by Administration of Traditional Chinese Medicine of Guangdong Province (Grant No. 20141053), Pearl River Nova Program of Guangzhou (Grant No. 2016-55), and National Natural Science Foundation of China (Grant No. 81300764). 

Disclosure: X. Cai, None; Y. Yang, None; P. Chen, None; Y. Ye, None; X. Liu, None; K. Wu, None; M. Yu, None