BMP-4 and BMP-7 Inhibit EMT in a Model of Anterior Subcapsular Cataract in Part by Regulating the Notch Signaling Pathway

Abstract

Purpose: The proliferation, migration, and epithelial-mesenchymal transition (EMT) of lens epithelial cells (LECs) are believed to be the pathological mechanisms underlying anterior subcapsular cataract (ASC). Bone morphogenetic proteins (BMPs) inhibit transforming growth factor-beta (TGF-β)-induced fibrosis in the lens. Herein, we aimed to further clarify the roles of BMP-4/BMP-7 in the progression and the underlying mechanisms of fibrotic cataract.

Methods: BMP-4/BMP-7, TGF-β₂, jagged-1 peptide, or DAPT were applied in a mouse injury-induced ASC model and in the human LEC cell line SRA01/04. The volume of opacity was examined by a slit lamp and determined by lens anterior capsule whole-mount immunofluorescence. Global gene expression changes were assessed by RNA sequencing, and the levels of individual mRNAs were validated by real-time PCR. Protein expression was determined by the Simple Western sample dilution buffer. Cell proliferation was examined by CCK8 and EdU assays, and cell migration was measured by Transwell and wound healing assays.

Results: Anterior chamber injection of BMP-4/BMP-7 significantly suppressed subcapsular opacification formation. RNA sequencing of the mouse ASC model identified the Notch pathway as a potential mechanism involved in BMP-mediated inhibition of ASC. Consistently, BMP-4/BMP-7 selectively suppressed Notch1 and Notch3 and their downstream genes, including Hes and Hey. BMP-4/BMP-7 or DAPT suppressed cell proliferation by inducing G1 cell cycle arrest. BMP-4/BMP-7 also inhibited TGF-β₂-induced cell migration and EMT by modulating the Notch pathway.

Conclusions: BMP-4/BMP-7 attenuated ASC by inhibiting proliferation, migration, and EMT of LECs via modulation of the Notch pathway, thereby providing a new avenue for ASC treatment.

Anterior subcapsular cataract (ASC) is associated with aberrant conditions, such as surgical trauma^1,2 and ultraviolet radiation,³ or other diseases, such as retinitis pigmentosa^4,5 and atopic dermatitis.⁶ The abnormal proliferation, migration, and epithelial-mesenchymal transition (EMT) of lens epithelial cells (LECs) lead to the formation of subcapsular fibrotic plaques.⁷ Transforming growth factor-beta (TGF-β) is essential in the formation of ASC by promoting the phenotypic transition of LECs into myofibroblasts.^8–10 The TGF superfamily also contains bone morphogenetic proteins (BMPs), which were originally found to trigger the formation of bone and cartilage and later found to be involved in diverse biological processes and disease states.^11–13 Various BMPs were found to play distinct roles in lens development, and BMP-4 and BMP-7 appear to be particularly essential due to their involvement in multiple stages.^14–18 Notably, BMPs synergize with the FGF signaling to promote the differentiation of LECs into primary lens fiber cells and their cell cycle exit.¹⁸ Similarly, studies have revealed that BMP-4 and BMP-7 attenuate TGF-β-mediated or injury-induced EMT, a process leading to the generation of more proliferative mesenchymal cells in the lens.^19–21 As a generally acknowledged canonical downstream signaling pathway of TGF-β/BMP, the Smad signaling pathway has important functions in the antagonism between TGF-β and BMPs by balancing the TGF-β/Smad2/3 and BMP/Smad1/5/8 pathways.^22–24 Specifically, by utilizing the mouse model of injury-induced EMT of LECs and lens explants, researchers have shown that BMP-4 and BMP-7 upregulate their downstream targets, inhibitors of differentiation 2 and 3 (Id2/3), by inhibiting the activation of Smad2 and enhancing the activation of Smad1/5/8.^19–21 

In addition to the canonical TGF-β/BMP/Smad pathway, growing evidence indicates that TGF-β/Notch signaling is essential in EMT progression.²⁵ Furthermore, the Notch signaling pathway has been proven to be a key intercellular regulator of cellular functions.²⁶ In mammals, ligand binding that promotes proteolysis of the four receptor paralogs by γ-secretase triggers the Notch signaling pathway. The cleaved Notch intracellular domains (NICDs) are then translocated into the nucleus and form complexes with DNA binding proteins that initiate transcription of downstream target genes.^27,28 Notch signaling acts between physically adjacent cells to exert different effects on biological behaviors, including cell proliferation, death, migration, and differentiation.^29,30 Moreover, Notch signaling has emerged as a therapeutic target for different kinds of fibrotic diseases, such as liver fibrosis, skin fibrosis, retinal fibrosis, and chronic kidney disease.^31–36 

Interestingly, BMP signaling was also found to crosstalk with the Notch pathway, which modulated the EMT phenotype in colorectal cancer via a γ‐secretase‐independent mechanism.³⁷ Furthermore, Hey1, a Notch downstream gene, suppressed BMP-2 signaling by inhibiting Smad phosphorylation.³⁸ Additionally, the Notch signaling pathway was suppressed by the Smad2/3 inhibitor SB431542,²⁵ which indicated a link between the Smad and Notch signaling pathways. Moreover, Notch activation abrogated Id2 expression in innate lymphoid cell development,³⁹ and Id2 also antagonized Hey1 in neural stem cells.⁴⁰ These findings strongly indicate an interaction among TGF-β, BMPs, and the Notch signaling pathway. 

Despite these studies, the relationship among TGF-βs, BMPs, and Notch signaling in LECs has not been fully elaborated. In this study, we further dissected the function of BMP-4 and BMP-7 in the progression of ASC and reported that the crosstalk between BMPs and the Notch pathway played an important role in this process. 

Materials and Methods

Model of Injury-Induced Anterior Subcapsular Cataract

All animal procedures followed the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and their approval was granted by the Ethics Committee of Zhongshan Ophthalmic Center (Z2022008). Six- to 8-week-old C57BL/6 mice were used for the model of injury-induced ASC, as described previously.⁴¹ Briefly, an intraperitoneal injection of pentobarbital sodium (50 mg/kg) was used for general anesthesia of mice. After pupillary dilation with compound tropicamide eye drops and topical anesthesia with proparacaine hydrochloride eye drops, for induction of injury, a small incision was made in the center of the anterior capsule by vertically inserting a 26-gauge hypodermic needle through the cornea, reaching a depth of a quarter of the blade (approximately 300 µm). For the BMP treatments, 1.5 µL of phosphate-buffered saline (PBS), BMP-4 (1 ng/µL), or BMP-7 (2 ng/µL) was injected into the anterior chamber with a Hamilton micro syringe. After the surgery, tobramycin eye ointment was applied for postoperative anti-infection. In every experiment, the depth and angle of needling were kept consistent to ensure incisions of the same size. The mouse eyes were photographed by a slit lamp followed by lens harvest under a stereomicroscope at day 7 for subsequent experiments. 

Immunofluorescence

Lens anterior capsules or cell slides were fixed in methanol for 1 hour at room temperature followed by 3 rinses in PBS. Bovine serum albumin (BSA; 5%, BioFroxx, Shanghai, China) diluted in PBS with 0.5% Triton X-100 was used for blocking and permeabilization. The tissues and cells were incubated with primary antibodies overnight at 4°C, followed by an FITC-conjugated antibody (Supplementary Tables S1, S2) and 1 µg/mL DAPI for 1 hour at room temperature. The whole opacity was scanned continuously from the top to the bottom with an interval of 1 µm by utilizing the Z-stack mode of the ZEISS LSM 980 confocal microscope. The volume of the opacity and the EMT marker were calculated using the LSCM Image Browser software, as described previously.⁴¹ Briefly, the plaque can be regarded as a pile of superposed pyramidal frustums and the volume between every 2 images could be calculated based on the following formula: V₁ = 1/3 × H × [S₁ + S₂ + \(\sqrt {{S_1} \times {S_2}} \)], (V = volume, H = altitude of frustum, S₁ = base area 1, and S₂ = base area 2). Hence, the volume of each subcapsular plaque is the sum of individual volumes between two images (V_total = V₁ + V₂ +……+ V_n, V_total = volume of subcapsular plaque, V_n = individual volumes between 2 images). Twelve lens anterior capsule samples per group were used for statistical analysis. 

Simple Western

RIPA buffer containing protease and phosphatase inhibitors (Roche, Basel, Swiss) was used to lyse LECs or lens capsules. Protein concentrations were measured with the BCA-100 Protein Assay Kit (Biocolor Bioscience & Technology Co., Ltd., Shanghai, China), and the protein expression level was determined using Wes (ProteinSimple, Bio-Techne, Minneapolis, MN, USA). In more detail, protein samples were then mixed with the Simple Western sample dilution buffer provided in the Wes separation module (SM-W004, Bio-Techne), and denatured at 95°C for 5 minutes. The prepared protein samples, biotinylated ladder blocking buffer, primary and secondary antibodies (see Supplementary Tables S1, S2), and chemiluminescent substrate included in the detection module were loaded onto the designated wells of a Wes 12-230 kDa prefilled plate and separated and detected automatically according to the manufacturer's manual. Compass for SW software (version 6.0) was used for data analysis and results presentation. In detail, the quantification of relative protein expression was implemented by calculating the area of the chemiluminescent peak of the protein of interest divided by that of the β-actin loading control, with the high dynamic range (HDR) exposure automatically determined by the system. The electrophoretogram was generated by the software based on the area of each chemiluminescent peak. Standard curves were made according to the manufacturer's instructions to ensure that the final total protein concentration (1 µg/µL) used in the present study was in the linear range (Supplementary Fig. S1). 

RNA Sequencing and Pathway Enrichment Analysis

TRIzol reagent (RNAiso Plus, 9108; TaKaRa Bio, Inc., Shiga, Japan) was used for total RNA extraction of the lens capsules. The concentrations of RNA were assessed using a Nanodrop 2000 (Thermo Fisher Scientific, Waltham, MA, USA) and the integrity and purity were evaluated with an Agilent Fragment Analyzer 5400 (Agilent Technologies, Santa Clara, CA, USA). RNA samples with total amounts greater than 400 ng and RNA integrity numbers larger than 7 were used as the input materials, and each group had 4 biological replicates. The NEBNext UltraTM RNA Library Prep Kit for Illumina (New England Biolabs Inc., Ipswich, MA, USA) was used to generate the sequencing libraries. The prepared libraries were sequenced with Illumina NovaSeq6000 after cluster generation. The construction of the reference genome index and the alignment of the paired-end clean reads to the reference genome were accomplished by HISAT2 (version 2.0.5). Furthermore, the reads of each gene were counted by FeatureCounts (version 1.5.0-p3), and the DESeq2 R package (version 1.20.1) was used for the analysis of differential gene expression. Gene Ontology (GO) enrichment with differentially expressed genes (DEGs; P value < 0.05) was performed by the clusterProfiler R package. 

Cell Culture and Treatment

Dulbecco's modified Eagle's medium (DMEM; Gibco, Life Technologies Corporation, Grand Island, NY, USA) with 10% fetal bovine serum (FBS; Gibco) was used for the culture of the human LEC line SRA01/04. A total of 100 ng/mL recombinant human BMP-4 (PeproTech, Cranbury, NJ, USA) or 500 ng/mL recombinant human BMP-7 (PeproTech), 20 ng/mL recombinant human TGF-β₂ (PeproTech), 10 µM DAPT (a γ-secretase inhibitor; MedChemExpress, Shanghai, China), 10 µM jagged-1 (JAG-1) peptide, and 10 µM scrambled JAG-1 (MedChemExpress) were used for the treatment of LECs for the indicated time period. 

Quantitative Real-Time PCR

Total RNA of the samples was extracted with TRIzol reagent and reverse transcribed into cDNA with the FastKing RT Kit (KR116, Tiangen Biotech Co., Ltd., Beijing, China). Quantitative real-time PCR was performed in a QuantStudio 7 Flex (Thermo Fisher Scientific) with ChamQ SYBR qPCR Master Mix (Q331; Vazyme Biotech Co., Ltd., Nanjing, China). Supplementary Table S3 lists the primer sequences. The relative expression of mRNA was calculated with the 2^−ΔΔCT method. 

Cell Viability and Proliferation Assays

The viability of LECs was assessed by Cell Counting Kit-8 assay (CCK-8; K1018; ApexBio Technology, Houston, TX, USA). In brief, cells cultured in a 96-well plate were starved for 12 hours and then treated with BMPs in medium with 1% FBS for 72 hours. Then, 10% CCK-8 assay reagent in 100 µL DMEM was added to each well, and the cells were incubated for 1 hour at 37°C. The optical density (OD) value at 450 nm was measured by Infinite M200 Pro NanoQuant. Cell proliferation was evaluated with an EdU cell proliferation kit (BeyoClickTM EdU Cell Proliferation Kit, C0078S; Beyotime, Shanghai, China). Specifically, LECs were seeded in a 12-well plate and incubated with EdU for 2 hours after BMP treatment. After fixation and permeabilization, a click reaction was performed, and images were taken using a ZEISS LSM 980 confocal microscope. 

Scratch Wound Healing Assay

When the cells in a 6-well plate reached 90% confluency, scratches were made with a 200-µL pipette tip. After removal of the detached cells with PBS, the attached cells were cultured in serum-free DMEM with TGF-β₂ or BMPs for 24 hours. Images of the wound area were acquired using a phase-contrast microscope at 0 hours and 24 hours. The calculation of the wound area and analysis of the migration rate were accomplished by ImageJ software (Media Cybernetics Inc., Silver Spring, MD, USA). 

Transwell Assay

Then, 800 µL of serum-containing medium with TGF-β₂ or BMPs was added into the lower chamber and 2 × 10⁴ LECs suspended in 100 µL of serum-free medium with TGF-β₂ or BMPs were seeded into each upper chamber of a 24-well Transwell filter plate (Corning Incorporated, Corning, NY, USA). After 24 hours, the cells attached to the upper side of the membrane were removed, and the transmigrated cells were fixed with 4% paraformaldehyde and stained with crystal violet (Beyotime). Three images of each Transwell plate were acquired with an inverted microscope, and quantification was performed via cell counting using ImageJ. 

Statistical Analysis

GraphPad Prism version 8.2.1 (GraphPad Software, Inc., La Jolla, CA, USA) was used for statistical analysis. The data of each group are presented as the mean ± SEM, and a value of P < 0.05 by a two-tailed test or one-way ANOVA was considered statistically significant. 

Results

BMP-4 and BMP-7 Suppressed the Formation of Injury-Induced ASC in Mice

We used an injury-induced ASC model in mice to investigate the effect of BMP-4 and BMP-7 on the formation of subcapsular cataract. Seven days following surgery, an obvious opacity forming under the anterior capsule was observed using a slit lamp (Fig. 1A). The BMP treatment groups displayed a much smaller opacity than the control group. To quantitatively analyze the opacity, we used α-SMA as a marker of EMT and DAPI as a marker of the nucleus. The immunofluorescence results showed that the volume of the opacity and the volume of α-SMA were reduced after BMP treatment (Figs. 1B–D). Consistent with the immunofluorescence results, the protein expression of fibronectin and N-cadherin was suppressed, whereas the expression of E-cadherin was elevated by BMP-4 and BMP-7 (Figs. 1E, 1F). These results suggested that BMP-4 and BMP-7 suppressed the EMT of LECs and capsular opacification formation. 

Figure 1.

 

BMP-4 and BMP-7 suppressed the formation of injury-induced ASC in mice. (A) Representative images of anterior segment with slit-lamp of mice 7 days after surgery. White arrows show subcapsular opacity. (B) Immunofluorescence of lens anterior capsule whole mount with α-SMA (red) for the EMT marker and DAPI (blue) for nuclei. The horizontal pictures show the section with the largest area and the orthogonal view displays the thickness of the opacity. Scale bar = 100 µm. (C, D) Quantification and statistical analysis of the volume of subcapsular plaques and α-SMA-positive cells. Twelve capsules were collected in each group in three independent experiments. (E) Output electrophoretogram by Simple Western with the HDR exposure of the relative protein expression of E-cadherin, N-cadherin, and fibronectin of mouse lenses. (F) Quantification and statistical analysis of relative protein expression. Data from three independent experiments are presented as the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

RNA Sequencing Analysis Identified the Notch Pathway as a Potential Mechanism Underlying BMP-4- and BMP-7-Mediated EMT Inhibition In Vivo

To further demonstrate the mechanism of BMP-4 and BMP-7 in capsular opacification formation, we assessed global gene expression changes by transcriptomic profiling. The transcriptional profiling of the BMP-treated and control groups was visualized in the form of a principal component analysis graph and volcano plots (Supplementary Fig. S2). A GO enrichment analysis showed that DEGs were enriched in biological processes, including unsurprisingly, cell migration and cell differentiation. More interestingly, the Notch pathway was also identified as one of the top signaling pathways (Figs. 2A, 2B). The heatmap of DEGs related to the Notch signaling pathway, cell migration, cell proliferation, and cell differentiation induced by BMPs showed a similar pattern (Fig. 2C), indicating an overall functional redundancy between these ligands in the regulation of these processes. Among the DEGs in the heatmap, Nfkbia,^42–44 Kit,^45,46 Cbfa2t2,^47,48 and Fat4^49,50 were related to the Notch signaling pathway and real-time PCR (RT‒PCR) results consistently showed that these genes were downregulated by BMP-4 and BMP-7 (Supplementary Fig. S3). The crosstalk between BMPs and the Notch pathway has not been extensively investigated in capsular opacification formation, so we decided to focus on this particular signaling pathway and test Notch and its downstream target genes to verify the transcriptomic data. The inhibitory effect of BMPs on Notch was at the protein stability level rather than the mRNA level, because the mRNA level of Notch could not be affected by BMP-4 or BMP-7 (Fig. 2D). In contrast, BMP-4 and BMP-7 suppressed the expression of the Notch1 and Notch3 protein (Figs. 2E, 2F) but had no effect on the Notch2 protein (Supplementary Fig. S4). The most extensively studied and understood targets of Notch are hairy and enhancer of split (Hes) and Hes-related with YRPW motif (Hey),⁵¹ and seven Hes genes and three Hey genes have been identified in mammalian genomes.^52–55 The RT‒PCR results showed that BMP-4 and BMP-7 decreased the mRNA levels of Hes1, Hes5, Hes6, Hes7, Hey1, and Heyl, and increased that of Hey2 (see Fig. 2D). Hes2, Hes3, and Hes4 were expressed at low level in mouse lens (CT value > 35). These findings indicated that BMP-4 and BMP-7 inhibited capsular opacification formation by regulating the Notch signaling pathway. 

Figure 2.

 

RNA sequencing analysis identified the Notch pathway as a possible mechanism underlying BMP-4- and BMP-7-mediated EMT inhibition in vivo. (A) GO enrichment analysis between the BMP-4-treated and control groups shown by a dot plot. (B) GO enrichment analysis between the BMP-7-treated and control groups shown by a dot plot. GeneRatio shows the ratio of the genes found enriched in the specific pathway dataset to the total genes in the dataset. Color grading from blue to red represents the P value. The number of genes enriched in the specific pathway set is shown by the size of the dot (count). (C) Heatmap of DEGs related to the Notch signaling pathway, cell migration, cell proliferation, and cell differentiation. Color gradation from orange to blue indicates relative gene expression. Dots of different colors show DEGs related to Notch signaling pathway (green), cell migration (yellow), cell proliferation (blue), and cell differentiation (red). Each group had four biological replications. (D) Results of RT‒PCR show the relative expression of Notch1, Notch2, Notch3, Hes1, Hes5, Hes6, Hes7, Hey1, Hey2, and Heyl. β-actin was used as a reference gene. (E, F) Images and statistical analysis of Simple Western showed the relative protein expression of Notch1 and Notch3 normalized to β-actin. Data represent the mean ± SEM of three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

BMP-4 and BMP-7 Inhibited the Notch Signaling Pathway In Vitro

Given that the Notch signaling pathway is involved in the regulation of capsular opacification by BMPs, we established an in vitro cell culture model to verify the results we observed in vivo, and we explored the underlying mechanism of the crosstalk between BMPs and the Notch signaling pathway. First, we found that BMP-4 and BMP-7 treatment for 72 hours both inhibited the TGF-β₂-induced upregulation of Notch1 and Notch3 (Figs. 3A, 3B). The results of RT‒PCR showed that BMP-4 and BMP-7 selectively attenuated the TGF-β₂-induced increase in Hes1, Hes3, Hes4, Hes5, Hes7, Hey1, and Heyl, whereas they had no significant effect on Hes2 and Hes6. Consistent with the in vivo results, BMP-4 and BMP-7 increased the abundance of Hey2 transcripts above that of TGF-β₂ alone (Fig. 3C), suggesting that regulation of Hey2 expression may be different from the other members. 

Figure 3.

 

BMP-4 and BMP-7 inhibited the Notch pathway in vitro. (A, B) Images and statistical analysis of the results of Simple Western showed the relative protein expression of Notch1 and Notch3. (C) Real-time PCR results of the relative mRNA expression levels of Hes1, Hes2, Hes3, Hes4, Hes5, Hes6, Hes7, Hey1, Hey2, and Heyl. β-actin was used as a reference gene. (D, E, F, G) Images in traditional Western blot style and statistical analysis of Simple Western displayed the relative protein expression of Notch3 normalized to β-actin after cotreatment with JAG-1 or DAPT. Data from three independent experiments are presented as the mean of the relative fold change ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ns = no statistical significance.

The activation of Notch signaling relies on ligand binding to induce receptor proteolysis via γ-secretase.²⁷ The JAG-1 peptide, a fragment of the JAG-1 protein, is an activator of the Notch pathway, whereas DAPT, a γ-secretase inhibitor, functions as an inhibitor.^56,57 Indeed, we found that JAG-1 enhanced the TGF-β₂-induced elevation of Notch3. Additionally, the upregulation of Notch3 induced by both TGF-β₂ and JAG-1 was suppressed by BMP-4 and BMP-7 (Fig. 3D). Furthermore, DAPT suppressed the TGF-β₂-induced upregulation of Notch3 (Fig. 3E). These findings demonstrate that the Notch signaling pathway is a downstream pathway of TGF-β/BMP. Furthermore, BMPs inhibited Notch signaling by suppressing Notch receptors and downstream target genes. 

BMP-4 and BMP-7 Attenuated the Proliferation of LECs by Inducing Cell Cycle Arrest

To investigate the effect of BMPs and the Notch pathway on the biological behavior of LECs, we first explored the impact of BMP-4 and BMP-7 on cellular proliferation. The CCK-8 assay showed that BMP-4 and BMP-7 both significantly suppressed the viability of LECs (Fig. 4A), and the inhibition of proliferating cell nuclear antigens (PCNAs; Figs. 4B, 4C) and EdU incorporation (Figs. 4D, 4E) further confirmed BMP-mediated inhibition of LEC proliferation. 

Figure 4.

 

BMP-4 and BMP-7 attenuated the proliferation of LECs by inducing cell cycle arrest. (A) OD reading at 450 nm after treatment with BMP-4 or BMP-7 for 72 hours. (B) Relative expression of PCNA normalized to β-actin shown by Simple Western assay. (C) Quantification and statistical analysis of protein expression. (D) Representative immunofluorescence pictures of the EdU assay. DAPI (blue) represents the nuclei. Scale bar = 50 µm. (E) Statistical analysis of the percentage of EdU-positive cells per visual field. (F, G) Simple Western results and quantification of relative protein expression of CDK2, CDK4, p21, and p27 normalized to that of β-actin. (H, I) Simple Western results and quantification and statistical analysis of the protein levels of PCNA, CDK4, p27, and p21 normalized to that of β-actin. Data represent mean ± SEM of three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

Furthermore, we examined the variation in key cell cycle regulatory proteins. In normal cell cycles, cyclin D1 combines with cyclin-dependent kinases (CDKs) 4 and 6 to form a complex, and cyclin E1 combines with CDK2 to form a complex. These complexes jointly phosphorylate the retinoblastoma protein, initiating its dissociation from the transcription factors regulating the G1/S phase transition, and causing the cells to enter the proliferation phase.^58,59 We found that the expression of CDK2 and CDK4 was downregulated by BMP-4 or BMP-7, and the expression of the cyclin-dependent kinase inhibitors (CDKIs) p21 and p27 was upregulated (Figs. 4F, 4G). Moreover, LECs treated with DAPT showed the downregulation of PCNA and CDK4, as well as the upregulation of p21 and p27, consistent with the results of BMPs (Figs. 4H, 4I). These findings suggest that BMP-4 and BMP-7 suppressed cellular proliferation by inhibiting the G1/S transition via modulation of the Notch signaling pathway. 

BMP-4 and BMP-7 Inhibited the Migration and EMT of LECs Induced by TGF-β₂

TGF-β₂ induces ASC by increasing the EMT as well as migration of LECs.^60,61 Thus, we next tested whether BMPs abrogated the effect of TGF-β₂-induced EMT and migration on LECs. Indeed, wound healing (Figs. 5A, 5B) and Transwell (Figs. 5C, 5D) assays showed that TGF-β₂ promoted the migration of LECs, while BMP-4 and BMP-7 inhibited this promoting effect. In addition, we found that TGF-β₂ significantly increased the expression of fibronectin, and BMPs reversed this trend (Figs. 5E–H). Furthermore, JAG-1 strengthened the facilitation of EMT by TGF-β₂, whereas BMPs attenuated the effect of JAG-1. In addition, JAG-1 was able to reverse the inhibitory effect of BMPs on TGF-β₂-induced EMT (Fig. 5I). Concerning the effect of DAPT, our findings demonstrated that DAPT reduced the expression of fibronectin induced by TGF-β₂ (Fig. 5J). In summary, these results demonstrate that via the regulation of the Notch signaling pathway, BMPs suppressed TGF-β₂-induced migration and EMT. 

Figure 5.

 

BMP-4 and BMP-7 inhibited the migration and EMT of LECs induced by TGF-β₂. (A) Representative pictures of scratch width at 0 hours and 24 hours. Straight black lines represent wound edges at 0 hours. (B) Quantification and statistical analysis of relative cell migration into the scratch area. (C) Representative pictures showing cells that migrated through the pore and clung to the bottom membrane after 24 hours. (D) Quantification of the transmigrated cells per field. Pictures were taken by an inverted microscope. Scale bar = 200 µm. (E) Simple Western results showed the relative protein expression of fibronectin. (F) Quantification of relative protein expression. (G) Immunofluorescence of LECs stained with DAPI (blue) representing the nuclei and fibronectin (red). Scale bar = 20 µm. (H) Mean relative cell fluorescence of fibronectin. Three independent experiments were carried out in triplicate. (I, J) Simple Western results and quantification as well as statistical analysis of the relative expression of fibronectin in LECs treated with TGF-β₂, BMPs and JAG-1, or DAPT. Data represent the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001.

Discussion

In the current study, we present in vivo evidence that BMP-4 and BMP-7 inhibit ASC. Furthermore, RNA-sequencing analysis revealed that the role of BMPs might be mediated via the Notch signaling pathway. Our in vitro experiments confirmed that interactions of Notch signaling with BMP-4 and BMP-7 regulated the proliferation, migration, and EMT of LECs. Thus, our results reveal that BMPs inhibit ASC via the suppression of Notch signaling (Fig. 6). 

Figure 6.

 

Schematic illustration of the interaction among TGF-βs, BMPs, and the Notch signaling pathway in the regulation of the proliferation, migration, and EMT of LECs. TGF-β ligands, such as TGF-β₂, bind to the receptors, which initiate the following signaling pathways, including the Notch signaling pathway. The ligand of a cell binds to the receptor of its adjacent cell, which brings about proteolysis of the receptors, leading to the release of an active NICD from a membrane tether and the activation of target genes, such as Hes and Hey. Upon binding to BMP-4 or BMP-7, BMP receptors initiate the attenuation of the Notch signaling pathway to inhibit Hes and Hey and consequently reverse the proliferation, migration, and EMT of LECs.

Previous studies have suggested that overexpression of BMP-7 suppressed fibrogenesis in multiple types of tissues, including the lens.^18,21,62 However, no reports have been published regarding the roles of BMPs in the overall effects of ASC pathogenesis. Herein, we utilized a model that led to the formation of subcapsular plaques⁶³ and further validated previous work highlighting the anti-EMT effect of BMPs. In this study, we evaluated the cataract phenotype using a slit lamp and calculated the volume of ASC with 3D reconstruction, which was more intuitive than the mere assessment of EMT markers. We found that BMP-4 and BMP-7 attenuated subcapsular cataract formation. 

To probe the detailed mechanisms involved in BMP-mediated ASC suppression, we conducted an RNA-sequencing analysis using mouse capsules in the ASC model. A GO enrichment analysis identified modulation of the Notch signaling pathway by BMP-4/BMP-7 in the process of ASC regulation. This finding is consistent with those in other diseases, such as esophageal adenocarcinoma and colorectal cancer, where Notch and BMP signaling are also engaged in crosstalk and functional synergism or antagonism.^64–69 In contrast to these studies, which reveal the crosstalk occurs downstream of Notch, for instance, the interactions between various transcription factor, our in vivo results showed that BMP-4/BMP-7 inhibited Notch1 and Notch3 at the level of protein stability. Notch proteins undergo complicated protein processing to transduce their signals,⁷⁰ and previous studies also demonstrate that the ubiquitin–proteasome system and lysosomes play important roles in the degradation of Notch1^71–75 and Notch3.⁷⁶ Hence, it will be interesting to determine how BMPs regulate Notch protein processing in future studies, which may provide an opportunity to further elucidate Notch signal regulation. For Notch target genes, unsurprisingly, we found that BMPs suppressed most Notch downstream targets, including most of members of the Hes and Hey families, with the exception of Hey2, which displayed an opposite pattern in response to treatment with BMP-4 and BMP-7. This phenomenon reflects very complicated regulation among the Hes and Hey family members. One reason may be that not all members of the Hes and Hey family are regulated by Notch signaling alone. Hes1, Hes2, Hes3, Hes4, Hes5, and Hes7 were reported to be regulated by the Notch pathway,^77–80 whereas Hes6 regulation appeared independent of Notch signaling.⁸¹ In the organ of Corti, expression of Hey2, unlike Hes5 and Hey1 which are tightly regulated by the Notch pathway, depends on the FGF pathway.⁸² Rather than suppressing, signaling through BMP receptor 1 acts on the Hey2 promoter to induce the expression of Hey2.⁸³ Although BMP-4 and BMP-7 may share some common functions in regulating the Notch pathway, each of them may have unique roles by binding to different receptors and regulating different signaling pathways,^84–87 which should be further investigated in future studies. 

The generally acknowledged mechanism of fibrotic cataract includes proliferation, migration, and EMT of residual LECs.^88,89 Our findings have validated that BMP-4 and BMP-7 inhibited the proliferation and TGF-β₂-induced EMT and migration of LECs. Moreover, we identified the inhibitory role of the Notch signaling pathway via DAPT-induced cell cycle arrest by upregulating p21 and p27 and downregulating CDK4, and JAG-1 interacted with BMP-4 and BMP-7 in the regulation of TGF-β₂-induced EMT of LECs. A previous study verified the negative effect of DAPT on the proliferation of smooth muscle cells through the Notch3-Hes1/2/5 pathway,⁹⁰ and Hes1-deficient mice showed reduced proliferation and increased p27 transcripts.⁹¹ Furthermore, Notch activity promoted EMT during cardiac development,⁹² whereas the combined inactivation of Hes1 and Heyl caused impaired EMT in mice.⁹³ 

Taken together, our data from the mouse ASC model and human LECs jointly support the idea that BMP-4 and BMP-7 alleviate ASC by inhibiting the proliferation, migration, and EMT of LECs by regulating the Notch pathway. Presently, the use of recombinant BMP-2 and BMP-7, has been approved by the US Food and Drug Administration (FDA), marketed as InFuse Bone Graft and OP-1. Therefore, BMP-4 and BMP-7 can be developed as promising drugs to prevent ASC. 

Acknowledgments

Supported by grants from the National Natural Science Foundation of China (No. 81970783 and No. 81770909). 

Disclosure: F. Jiang, None; Y. Qin, None; Y. Yang, None; Z. Li, None; B. Cui, None; R. Ju, None; M. Wu, None