Introduction
Age-related macular degeneration (AMD) is characterized by the progressive degeneration of the neuronal retina, retinal pigment epithelium (RPE) and choriocapillaris complex, leading to severe and permanent central visual impairment among aging population in developed countries [
1]. Neovascular AMD (nAMD), accounting for up to 90% of AMD-associated vision loss and blindness [
2], is primarily characterized by the presence of choroidal or retinal neovascularization (CNV or RNV) in the macular region, ultimately resulting in the formation of fibrotic scar [
2]. The prevalence of fibrosis at baseline, 12, 24, and 60 months in patients with nAMD was reported as follow: 13%, 32%, 36%, and 56%, respectively, according to a recent epidemiological report [
3]. Despite the efficacy of anti-vascular endothelial growth factor (anti-VEGF) therapy as the cornerstone in nAMD management, capable of stabilizing or even enhancing the visual function [
4], approximately 1/3 of the patients with subretinal fibrosis still experience poor prognosis, and even deterioration of the visual acuity [
5]. Therefore, it is imperative and urgent to explore the molecular mechanisms underlying the pathogenesis of subretinal fibrosis to develop effective therapies for subretinal fibrosis secondary to nAMD.
The RPE, a highly specialized monolayer of polarized, pigmented epithelial cells, is located between the outer segment of photoreceptors and the choriocapillaris endothelial cells, playing an essential role in maintaining visual function [
6]. The primary functions of RPE encompass light energy absorption focused by the lens onto the retina [
7]; transportation of nutrients and ions between photoreceptors and the choriocapillaris [
8]; secretion of various factors crucial for maintaining structural integrity in both retina and choriocapillaris [
9]; facilitation of phagocytosis in photoreceptor outer segments; and preservation of the blood-retinal barrier [
10]. Dysfunction of the RPE, particularly its transition from an epithelial to a mesenchymal phenotype, is implicated in the pathogenesis of subretinal fibrosis, a hallmark of the end-stage complications of nAMD. Recently, several signaling pathways regulating the epithelial–mesenchymal transition (EMT) of RPE cells have been identified, among which the canonical Wnt/β-catenin pathway has been extensively studied and well-characterized [
11]. Jung Woo Han et al. demonstrated a crucial role of Wnt/β-catenin signaling in RPE proliferation and EMT using a laser-induced mouse model, while also observing a significant upregulation of the non-canonical Wnt ligand Wnt5a expression in laser-treated RPE compared to control RPE [
12]. With accumulating evidence indicating a substantial role for the non-canonical Wnt ligand Wnt5a in the processes of EMT and fibrosis across diverse tissues [
13‐
16], the precise molecular mechanisms underlying Wnt5a-triggered signaling and its interplay with canonical Wnt/β-catenin in EMT, as well as its contribution to the development of subretinal fibrosis in the context of nAMD remain poorly elucidated.
The Wnt signaling pathway plays a crucial role in both development and various diseases. Overall, two primary Wnt signaling pathways can be distinguished: the canonical Wnt/β-catenin pathway, which functions through β-catenin as a transcriptional cofactor, and the non-canonical, β-catenin-independent Wnt pathway, encompassing the Wnt/Ca
2+ pathway and the planar cell polarity (PCP) pathway [
17]. Wnt5a, a highly conserved non-canonical Wnt ligand, functions as a secreted signaling molecule that plays pivotal roles in regulating PCP, convergent extension, and epithelial–mesenchymal interaction during embryonic morphogenesis [
18]. Although Wnt5a primarily signals through the non-canonical Wnt signaling pathways by binding to various members of the Frizzled (FZD)- and receptor tyrosine kinase-like orphan receptor (ROR)-family receptors, thereby initiating intracellular signaling cascades, under certain circumstances, it also possessed the capacity to activate the canonical Wnt signaling pathway [
19]. This activation led to stabilization of β-catenin and modulation of gene transcription regarding cell proliferation, survival, differentiation, and migration [
20]. In the field of cancer biology, the role of Wnt5a exhibits intricate and context-dependent characteristics. Emerging evidence suggests that the dysregulated activation or inhibition of Wnt5a signaling remains an important event in cancer progression, exerting both oncogenic and tumor suppressive effects depending on the availability of key receptors, thereby highlighting the paradoxical role of Wnt5a across different cancers [
18]. The involvement of Wnt5a in regulation of cancer cell invasion, metastasis, metabolism and inflammation renders it a subject of intense research in oncology [
21]. Furthermore, Wnt5a has been widely recognized as an essential factor in tissue repair and regeneration. For example, in some instances, Wnt5a acting as a particularly attractive growth factor stimulates tissue regeneration (such as colonic crypt [
22], osteochondral regeneration [
23] and cartilage interface integration [
23]) and wound healing [
24]. While in other cases, the upregulation of Wnt5a has been implicated in aggravating a variety of tissue fibrosis and scar formation, such as cardiac fibrosis under pressure overload [
25], myocardial fibrosis following myocardial infarction [
26], atrial fibrosis [
25], keloid scarring caused by aberrant genetic activation [
13], idiopathic pulmonary fibrosis [
27], renal fibrosis [
28], and liver fibrosis [
29]. However, the specific involvement of Wnt5a in the EMT process of RPE cells and subretinal fibrosis formation in nAMD remains elusive.
Therefore, the objective of this study to elucidate the molecular mechanisms of Wnt5a, a non-canonical Wnt ligand, and its crosstalk with the canonical Wnt signaling pathway involved in EMT of RPE cells, as well as its contribution to subretinal fibrosis progression in nAMD. Additionally, we aim to identify potential therapeutic targets for preventing or halting fibrotic progression in nAMD patients, thereby preserving visual function and improving the quality of life of individuals affected by this devastating disease.
Materials and methods
Animals
All mice were housed under specific pathogen-free (SPF) conditions and exposed to a 12-h light/dark cycle with free access to food and water. All procedures were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and the Guides for the Care and Use of Animals (National Research Council and Tongji University; Permit Number: TJHBLAC-2020–06).
Scotopic and photopic electroretinography (ERG)
Mice were dark-adapted for a minimum of 12 h before experiments to ensure maximal sensitivity. Prior to ERG recordings, mice were anesthetized by intraperitoneal injection of 0.2 mL 2% pentobarbital sodium. After pupil dilation with compound tropicamide eye drops. ERG recordings were performed in a dark room to ensure minimal ambient light. Gold wire electrodes were carefully placed on the corneas, with the reference electrode positioned under the skin behind the ear. Additionally, a ground electrode was subcutaneously inserted in the upper part of the tail.
For the evaluation of rod photoreceptor function (scotopic ERG), mice were exposed to a dim red light to maintain dark adaptation during electrode placement and recording. Scotopic ERG responses were recorded in response to light stimuli of 0.01 and 3.00 cds/m2.
For the assessment of cone function (photopic ERG), mice were exposed to a background light of 25 cds/m2 for 5 min, and responses were elicited with flashes of light of 3.00 cds/m2. Recorded ERG signals from 6 or 7 eyes were analyzed for amplitude of a-wave and b-wave components.
Laser-induced CNV mouse model and intravitreal injection
In this study, 7-week-old male C57BL/6J mice were purchased from SLAC Laboratory Animal Co., LTD., Shanghai. The laser-induced CNV model was carried out following the previously established protocol [
30]. Briefly, mice were anesthetized by intraperitoneal injection of 0.2 ml 2% pentobarbital sodium and the pupils were dilated with 1% tropicamide (Santen, Osaka, Japan). A 532-nm laser photocoagulation was used to induce CNV in mice (laser power: 120 mW, duration: 100 ms, spot size: 50 μm, Carl Zeiss Meditec; Dublin, Ireland). Four or fifteen laser-induced lesions, resulting in rupture of the RPE and underlying Bruch's membrane, were generated around the optic nerve head in a standardized fashion using a slit lamp delivery system for subsequent immunofluorescence or Western blot analysis on both eyes of each animal.
On the same day or 14 days following laser induction, intravitreal injection of FH535 or Box5 (1 μL) dissolved in dimethyl sulfoxide (DMSO) was administered according to the previous method [
31] to reach the target site in the posterior segment of the eye, achieving a final concentration of 0.5/3.0 μmol/L or 90 μmol/L in the vitreous cavity. The contralateral eye received an injection of 1 μL phosphate-buffered saline (PBS) buffer containing 0.03%/0.18% or 5.4% DMSO as a Vehicle control.
Quantification of the sizes of CNV and subretinal fibrosis
Seven days after laser induction, cardiac perfusion was performed with 4% paraformaldehyde in mice, followed by enucleation of the eyeballs and continued fixation for 30 min. The RPE-Bruch's membrane choriocapillaris complex (RBCC) tissues in each group were isolated and subjected to immunofluorescence staining for fibrosis-related molecules, following the procedure outlined in section “
Immunofluorescence”. The immunofluorescence staining results for isolectin B4 (IB4), collagen I, fibronectin, alpha-smooth muscle actin (α-SMA), active β-catenin, and RPE65 were quantified for the measurement of CNV and subretinal fibrosis areas using ImageJ software (National Institutes of Health, Bethesda, MD, USA). Briefly, the images were converted to an 8-bit format. By employing scale calibration and establishing appropriate thresholds, valid fluorescence signals were included for quantitative area measurement, followed by standardized processing.
Cell culture and treatments
The human RPE cell line (ARPE-19) was obtained from American Type Culture Collection (ATCC, Manassas, VA). ARPE-19 cell line was cultured in Dulbecco's Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/F-12, Hyclone, UT, USA) supplemented with 10% fetal bovine serum (FBS, 40130ES76, Yeasen Biotechnology, Shanghai), 5 U/mL penicillin/streptomycin (60162ES76, Yeasen Biotechnology, Shanghai) and incubated in a humidified atmosphere with 5% CO2 at a constant temperature of 37 °C. The ARPE-19 cell line was frequently screened for mycoplasma contamination using an EZ-PCR kit (HaEmek, Israel). Cell culture dishes, palates, and centrifuge tubes were obtained from NEST Biotechnology Co. Ltd. (Wuxi, China).
Recombinant transforming growth factor beta 1 (TGFβ1) was purchased from ThermoFisher (PHG9214, Shanghai, China) and was used at the final concentration of 10 ng/mL. FH535 was obtained from MedChemExpress (MCE, HY-15721, Shanghai) and was used at the final concentration of 0.1, 0.5, 1, and 2 μmol/L respectively. Foxy-5 (Wnt5a agonist) was purchased from MedChemExpress (MCE, HY-P1416A, Shanghai) and was used at the final concentration of 50, 100, and 200 μmol/L, respectively. Box5 (Wnt5a antagonist) was purchased from MedChemExpress (MCE, HY-123071, Shanghai) and was used at the final concentration of 10, 45, and 90 μmol/L. The peptides were purified by reverse-phase high performance liquid chromatography and the purity of Box5 was 91%, and the purity of FH535 and Foxy-5 was more than 99%. The human ARPE-19 cell line was subjected to serum-free starvation for 12 h, followed by treatment with TGFβ1 or combination with above small molecules in fresh serum-free medium (SFM) for an additional 48 h to investigate the impact of Wnt signaling on EMT.
The short hairpin RNA (shRNA) was employed to knockdown Wnt5a to explore the effect of Wnt5a on the EMT of ARPE-19 cells in vitro. Briefly, ARPE-19 cells were seeded in appropriate culture dishes or plates and allowed to reach 50–60% confluency. The plasmids containing Wnt5a shRNA1 or shRNA2 was introduced into the cells using a transfection reagent Lipofectamine 3000 according to the manufacturer's instructions (Beijing Tsingke Biotech Co., Ltd.). Optimal transfection conditions, including plasmid-to-reagent ratio (1:1) and incubation time (36 h), were determined through optimization experiments. Following transfection, the effectiveness of Wnt5a knockdown was assessed using quantitative real-time polymerase chain reaction (qRT-PCR) and Western blot analysis. The sequences targeting Wnt5a knockdown were 5′- CCGGGCTGGAAGTGCAATGTCTTCCCTCGAGGGAAGACATTGCACTTCCAGCTTTTTT-3′ (Wnt5a shRNA1), and 5′-CCGGGGTCGCTAGGTATGAATAACCCTCGAGGGTTATTCATACCTAGCGACCTTTTTT-3′ (Wnt5a shRNA2, Tsingke, Beijing, China). The ARPE-19 cells exhibiting successful knockdown of Wnt5a gene will be selected for subsequent experiments.
Cell viability assay
Cell viability was measured by CCK-8 kit (40203ES76, Yeasen, Shanghai, China) according to the manufacturer's protocol. A total of 5 × 103/well cells were seeded uniformly and cultured in five replicate wells in a 96-well microplate (Corning, USA) with medium containing 10% FBS, and then were incubated at 37 °C, 5% CO2 incubator for 2 days. Afterwards, the culture medium was replaced with SFM for 12 h of starvation treatment. Then, the cells were exposed to varying concentrations of FH535 or Box5, dissolved in DMSO, or Foxy-5 dissolved in PBS, all within SFM. The Vehicle Control groups for FH535 and Box5 contained DMSO concentrations of 0.2% and 2%, respectively. After treatment for 48 h, the medium was discarded. The CCK-8 reagent (10 μL) was added to 90 μL SFM to generate a working solution (total 100 μL) and then added to each well and incubated for 1–2 h. A microplate reader (Bio‐Rad, Hercules, CA, USA) was employed to test the absorbance of each experimental well at 450 nm, and to detect changes of cell viability among each group.
Scratch assay
ARPE-19 cells were uniformed seeded in 6-well plates (1 × 106 cells per well) and cultured for 2 days till the desired confluency. ARPE-19 cells were then deprived of serum for 12 h. Migration was evaluated by straight scratching a confluent layer of ARPE-19 cells using a P200 pipette tip. After gently washing with serum-free media to remove cell debris, 2 mL of SFM with or without TGFβ1 [10 ng/mL], FH535 [0.5 μmol/L], Box5 [90 μmol/L], or Foxy-5 [200 μmol/L] was added, followed by incubation at 37 °C. Images were obtained at 0, 24 and 48 h under a fluorescence microscope (Leica, DMI3000, Germany), after which the reduction in the wound area was determined using Image‐Pro Plus software (Media Cybernetics, Rockville, MD, USA).
Transwell migration assay
1 × 106 cells/well of ARPE-19 cells were uniformly seeded in 6-well plates. After 48 h of TGFβ1 treatment with or without FH535 co-incubation. ARPE-19 cells were trypsinized and then resuspended in SFM. 1 × 104 cells/well of ARPE-19 cells were seeded into the upper compartment of transwell cell culture chambers (8-µm pores; Falcon; Corning Life Sciences, Corning, NY, USA). 500 µL of complete DMEM/F12 medium was added to the lower inserts of the transwell. After incubation for 48 h at 37 °C, cells on the upper compartment of the transwell were removed by using a swab and cells across pores were fixed with cold methanol for 10 min and stained with crystal violet stain solution (0.5%, Beyotime Biotechnology, Haimen, China)) for 1 h. Images of migrated RPE cells were taken under the Leica microscope (DMI3000, Germany) and were quantified with the help of Image J software. The number of migrated cells in more than 12 random fields (magnification × 20) was counted.
qRT-qPCR
Total RNA was extracted from ARPE-19 cells using TRIzol Reagent (19201ES60, Yeasen; Shanghai, China) according to the manufacturer’s instructions. Complementary DNA was synthesized using a reverse transcription SuperMix (11120ES60, Yeasen; Shanghai, China). The SYBR Green master mix (11201ES03, Yeasen; Shanghai, China) was employed to conduct RT-qPCR. The conditions were set as follows: 95 °C for 5 min, 40 cycles of 95 °C for 10 s, 60 °C for 20 s and extension at 72 °C for 20 s. Relative expression of mRNA was calculated using the 2
−ΔΔCt method [
32] and normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH). Primer sequences are listed in Table
1.
Table 1
The information of human primers
α-SMA | CAGAAGGAGATCACGGCCCTAG | CGGCTTCATCGTATTCCTGTTTG |
collagen I | AATGTGGTTCGTGACCGTGA | AGCCTTGGTTGGGGTCAATC |
Dvl2 | TGGGGCTTCAGACCAGGATA | GCCCCCACCATAGGTGTAAG |
Dvl3 | TGCTGATAACCCATCGGAGC | TGGGCAGACACCAAAGAGTC |
fibronectin | AAGACCATACCCGCCGAATG | GGCATTTGGATTGAGTCCCG |
FZD1 | GAAAGTGCAGTGTTCCGCTG | TCACACTTGAGCGTGTCTGG |
FZD2 | CTATCCGCTGGTGAAGGTGC | CTCACAGATAGAGCGGCACGG |
FZD3 | TGATGGCTCTCATAGTTGGCA | ACCTGTCGGCTCTCATTCAC |
FZD4 | GTCTCAGTCTGGGGTTGCTC | ACGTTGTAGCCGAGGTTCTG |
GAPDH | CAAATTCCATGGCACCGTCA | GACTCCACGACGTACTCAGC |
ITGA5 | GGTCGGGGGCTTCAACTTA | AGCACACTGACCCCGTCTG |
MMP2 | TGATGGCATCGCTCAGATCC | GGCCTCGTATACCGCATCAA |
NAKED1 | AGGGGGAATAGGTGAGACCC | GGAATCCCATTGGCTGGTCA |
NAKED2 | CTTTCCGGGAGGACCAGTGT | GCACTGGAGTGCGTCAATGT |
Snail1 | TGCAGGACTCTAATCCAGAGTTT | GGACAGAGTCCCAGATGAGC |
Transgelin | AAGCGCAGGAGCATAAGAGG | ACTGATGATCTGCCGAGGTC |
Wnt1 | CAAGATCGTCAACCGAGGCT | TCACACGTGCAGGATTCGAT |
Wnt3a | ACAAAGCTACCAGGGAGTCG | TCCCACCAAACTCGATGTCC |
Wnt5a | CGGTGTACAACCTGGCTGAT | GCGCTGTCGTACTTCTCCTT |
Western blot
Total cell lysates and samples of mouse RBCC tissues were lysed with radio immunoprecipitation assay lysis buffer (RIPA, 20101ES60, Yeason, Shanghai, China) on ice and sonicated for 10 s immediately. The cytoplasmic and nuclear extracts of ARPE-19 cells were prepared using the ProteinExt® Mammalian Nuclear and Cytoplasmic Protein Extraction Kit (DE20101, transgen, Beijing, China), according to the manufacturer’s instructions. Lamin B1 and GAPDH were employed as nuclear and cytoplasmic loading controls, respectively.
The samples were centrifuged at 4 °C at 12,000 rpm for 30 min and the supernatant was collected for bicinchoninic acid assay (20201ES76, Yeason, Shanghai, China) and Western blot as previously described [
31]. Briefly, protein was denatured and resolved in 7.5–12.5% SDS-PAGE gels and transferred onto nitrocellulose membranes. The blots were blocked for 30 min at room temperature with 5% bovine serum albumin (BSA) in tris-buffered saline with Tween-20 (TBST) and then probed with fibronectin, α-SMA, zonula occludens-1 (ZO-1), Wnt3a, Wnt5a, ROR1, dishevelled 2 (Dvl2), Naked1, non-phospho (active) β-catenin (Ser33/37/Thr41), lamin B1 and GAPDH overnight at 4 °C separately. After thorough washing with TBST solution four times for 6 min each time, the corresponding secondary antibodies (goat anti-mouse/rabbit) conjugated to HRP were employed to incubate the blots for 1 h at room temperature. The bands were washed by TBST for 4 times and then visualized by chemiluminescence. Quantification was performed by measuring the relative intensity of interested signals which normalized by GAPDH or lamin B1 with the aid of Quantity One software (Bio-Rad). The primary antibodies are listed in Table
2.
Table 2
Antibody information
α-SMA | Ab7817 | WB (1:2,000); IF (1: 100) | Abcam |
collagen I | Ab34170 | WB (1:1,000); IF (1: 100) | Abcam |
Dvl2 | 3224 | WB (1:1,000) | CST |
fibronectin | 15,613-1-AP | WB (1:500); IF (1: 100) | Proteintech |
GAPDH | 10,494-1-AP | WB (1: 5,000) | Proteintech |
IB4 | I21411/I32450 | IF (1:2000) | Invitrogen |
lamin B1 | AF5161 | WB (1:500) | Affinity |
Naked1 | 2262 | WB (1:1,000) | CST |
Non-phospho (Active) β-Catenin (Ser33/37/Thr41) | 8814 | WB (1:1,000); IF (1:800) | CST |
ROR1 | 4102 | WB (1:1,000) | CST |
RPE65 | Ab13826 | IF (1: 100) | Abcam |
vimentin | 10,366-1-AP | IF (1: 100) | Proteintech |
Wnt3a | 2721 | WB (1:1,000) | CST |
Wnt5a | 2392 | WB (1:1,000) | CST |
ZO-1 | 61–7300 | WB (1:1,000); IF (1: 100) | Invitrogen |
Immunofluorescence
The ARPE-19 cells and mouse RBCC were fixed with 4% paraformaldehyde (PFA) (Servicebio, Wuhan, China) for 20 min, permeabilized with 0.05% Triton X-100 for 20 min, and then blocked with 1% BSA for 1 h. The slides were incubated overnight with primary antibodies against collagen I, fibronectin, α-SMA, vimentin, RPE65, ZO-1, or active β-catenin overnight, followed by IB4 probe, Alexa Fluor 488/555 goat anti-mouse or Alexa Fluor 555/488 goat anti-rabbit conjugated secondary antibodies, respectively, at room temperature in the dark for 1 h. The nucleus was visualized with 4′,6-diamidino-2-phenylindole (DAPI). The slides were mounted with coverslips. Images were obtained with a confocal microscope from Carl Zeiss (LSM 710; Königsallee, Germany).
Statistical analysis
Data generated in this study are presented as means ± SEM from at least three independent experiments. Statistical analysis was performed using GraphPad Prism 8.0 software (San Diego, CA, USA). For the comparison between the two experimental groups and the intergroup comparison, we applied two-tailed unpaired Student's t-test or one-way analysis of variance (ANOVA) as appropriate. Differences were considered statistically significant when P-values are less than 0.05.
Discussion
Subretinal fibrosis, as the most prevalent natural sequelae of nAMD, causally impacts the photoreceptors, RPE, and choriocapillaris, thereby leading to irreversible central vision loss [
38]. Despite extensive research on the pathogenesis of nAMD, the molecular mechanisms underlying subretinal fibrosis remain insufficiently characterized. Although the involvement of canonical Wnt/β-catenin activation in subretinal fibrosis has been reported [
12,
39], the role of the non-canonical ligand Wnt5a in EMT and subretinal fibrosis remains obscure.
In the present investigation, we observed a significant upregulation of Wnt5a expression in both in vivo and in vitro models simulating subretinal fibrosis or EMT (Fig.
3A and Fig.
4B). Above molecular and phenotypic changes were evidently inhibited by inactivating β-catenin using FH535 (Figs.
2,
3,
5 and
6, and Additional file
1: Fig. S3), thereby corroborating the indispensable role of the canonical Wnt/β-catenin signaling in facilitating EMT [
12], as well as its transcriptional effect on the non-canonical ligand Wnt5a. The action of the small molecule FH535 is achieved by interfering with the nuclear translocation of β-catenin in the canonical Wnt signaling pathway, thereby impeding its entry into the cell nucleus and subsequent binding with transcription factors, ultimately resulting in the reduction of downstream target gene expression [
40], such as Wnt5a [
41]. Furthermore, the inhibition of Wnt5a through Box5 or shRNA could notably reverse the activation of β-catenin, EMT, and subretinal fibrosis both in vivo and in vitro (Figs.
3,
7 and
8). The causative role of Wnt5a in promoting EMT and its positive regulation on β-catenin could be further reinforced in a Wnt5a agonist assay, demonstrating that Wnt5a alone is sufficient to activate β-catenin and initiate EMT (Fig.
6). Based on the aforementioned data, activation of Wnt5a and β-catenin establishes a mutually reinforcing interaction loop involving "Wnt5a/β-catenin" significantly promoting the EMT process and subsequent subretinal fibrosis (F
ig.
8I). However, the activation sequence of these two pathogenic factors in the development of EMT or fibrosis remains unclear.
The expression of Wnt5a in myofibroblasts was reported to be induced by the profibrotic factor TGFβ, playing a crucial role in the regulation of fibrotic matrix proteins induced by TGFβ in the context of liver fibrosis. This effect was reversed after silencing Wnt5a [
42]. It was also reported that TGFβ-mediated regulation of Wnt5a was confirmed in primary cells, with Smad binding sites identified within the Wnt5a promoter [
43,
44]. Thus, it is possible to propose that TGFβ1 mediates the upregulation of Wnt5a transcription, promoting the activation of β-catenin, which ultimately contributes to subsequent EMT and subretinal fibrosis. The precise mechanism underlying TGFβ1-induced upregulation of Wnt5a requires further exploration in the context of subretinal fibrosis secondary to nAMD.
The intricate Wnt signaling system plays a pivotal role in regulating the processes implicated in the pathogenesis of EMT as well as various fibrotic diseases. Although the canonical and non-canonical Wnt signaling pathway are two independent cellular signaling pathways, they can interact and cross-regulate each other in cell development, tissue repair, and disease processes [
45]. Wnt5a is recognized to stimulate non-canonical signaling independent of β-catenin [
46]. Nonetheless, this non-canonical Wnt ligand can trigger the activation of the canonical Wnt signaling pathway [
47,
48]. For example, Wnt5a overexpression in certain cases resulted in an enhancement of TCF/ lymphoid enhancer-binding factor (LEF) transcriptional activity and direct activation or cooperation with the Wnt/β-catenin pathway [
49]. Additionally, one study demonstrated that myometrial cells, acting as niche components, regulate the self-renewal activity of endometrial mesenchymal stem-like cells (eMSCs) via Wnt5a-dependent activation of the Wnt/β-catenin signaling pathway [
49]. Meanwhile, in specific circumstances, the activation of the canonical Wnt/β-catenin signaling pathway can induce the binding of the transcription factor TCF/LEF to the promoter region of Wnt5a gene, thereby enhancing its transcription [
50]. One of the limitations of this study is that we have not yet determined the direct transcriptional impact of β-catenin on Wnt5a, which may necessitate the use of a ChIP-qPCR technique in our future work.
After proposing and validating the concept of the "Wnt5a/β-catenin" interaction loop, we also discovered that FH535, Box5, and Wnt5a shRNA displayed comparable inhibitory effects on subretinal fibrosis and EMT (Figs.
3 and
5–
8). In other words, either the blockage of Wnt5a or β-catenin could disrupt this interaction loop (Fig.
8). Notably, although no apparent changes were observed in the expression of Wnt3a and Wnt1 both in vivo and in vitro (Figs.
3A and
4A), it is important to acknowledge that the canonical Wnt/β-catenin signaling pathway is not solely regulated by these ligands alone. Therefore, we cannot disregard the potential involvement of additional Wnt ligands in β-catenin activation within the scope of this study. One research validated our findings by demonstrating elevated level of Wnt5a and unaltered expression of wnt3a in laser-treated RPE compared to control RPE, as determined through RT-qPCR analysis. However, this study also identified significant increases in other Wnt ligands, such as Wnt1, Wnt2b, Wnt3, Wnt7a, Wnt7b, and Wnt10b, which might also contribute to the activation of β-catenin [
12]. Furthermore, we cannot exclude the potential involvement of the Wnt5a-mediated non-canonical Wnt signaling activation in the pathogenesis of EMT [
51,
52] and subretinal fibrosis, which also merits further study.
The underlying mechanisms of subretinal fibrosis secondary to nAMD are complicated and remain incompletely elucidated. Extensive literature corroborated the involvement of additional factors, such as pericyte-myofibroblast transition (PMT) [
53], EndMT [
54], activated microglia [
55], macrophages [
56] and Müller glia [
57] in the laser-induced CNV mouse model, which contribute to the accumulation of differentiated myofibroblasts and deposition of ECM, ultimately leading to fibrosis formation in nAMD. Therefore, the contribution of Wnt5a or Wnt/β-catenin in the MT of other cell types during subretinal fibrosis remains to be elucidated. Notably, the critical role of EMT undergone by RPE cells during subretinal fibrosis has been extensively recognized and corroborated [
58,
59]. The intricate molecular networks governing EMT exhibit interconnectedness and reciprocal interactions. This process involves the activation of not only the Wnt signaling but also the canonical TGF-β-Smad signaling, as well as the non-canonical TGF-β signaling pathways, including phosphatidylinositol-3-kinase/Akt (PI3K/Akt), Rho/Rho kinase (ROCK), mitogen-activated protein kinases (MAPKs), Jagged/Notch, Hedgehog (Hh), and other signaling pathways [
60]. It is proposed that patients with nAMD who exhibit an inadequate response to anti-VEGF therapy could potentially benefit from a combined therapeutic approach that targets both VEGF and the EMT process of RPE cells within the fibrovascular scar, thereby achieving a more favorable prognosis.
Subretinal fibrosis often occurs concurrently with neovascularization in nAMD, as exemplified by the co-localization of collagen I and IB4 in the CNV lesions visualized by confocal microscope in Fig.
2C and
D. The early intervention of pathological neovascularization potentially mitigates the infiltration of immune cells and inflammatory response, thereby facilitating the prevention of subsequent subretinal fibrosis [
61]. Despite its seemingly dual-functionality in physiological angiogenesis, Wnt5a signaling is essential for pathological angiogenesis [
62]. Experimental studies on various vascular eye diseases, including nAMD, diabetic retinopathy (DR), retinopathy of prematurity (ROP), and corneal neovascularization, indicate that an aberrantly heightened Wnt signaling pathway is among the potential causative factors for pathological ocular neovascularization [
17]. The activation of the Wnt pathway has been reported in laser-induced CNV models, indicating its pathogenic role in CNV. Inhibition of Wnt signaling through the use of an anti-LRP6 antibody demonstrates therapeutic potential for treating CNV [
63]. Therefore, it is difficult to assert that the attenuation of subretinal fibrosis through inhibition of Wnt5a/β-catenin signaling pathway partially stemmed from its inhibitory impact on CNV, which merits further investigation. In order to mitigate the confounding effect of CNV and subretinal fibrosis, we administered the inhibitor injection at a later time point (Day 14) after laser induction in mice. This decision was based on the report that neovascularization in this mouse model reaches its peak stage at day 7 and progresses gradually [
64]. Remarkably, our findings demonstrate that inhibition of Wnt5a/β-catenin pathway remained highly effective in combating fibrosis and EMT during the late stage (Day 21) of this mouse model (Fig.
3).
These findings and discussion should be taken into account when developing specific drugs for the treatment of subretinal fibrosis, as only targeting the mesenchymal transition and activation of multiple retinal cells is insufficient. Instead, a combination approach with inhibition of pathological neovascularization should be employed to effectively inhibit both the neovascularization and subretinal fibrosis. In developed countries, a significant proportion of vision impairment arises from abnormalities in the retinal or choroidal vasculature. These conditions, encompassing ailments such as nAMD [
2], DR [
65], ROP [
66], and neovascular glaucoma [
67], share the hallmark features of macular edema, retinal and vitreous hemorrhage, or the formation of fibrovascular scars. A shared underlying mechanism in these diverse conditions lies within the retina's response to damage, eventually setting off an ongoing cycle of chronic wound healing that may culminate in fibrotic tissue formation and, in the worst cases, irreversible vision impairment [
68]. Therefore, the mechanistic insights gleaned from the nAMD study and the combined inhibition of subretinal fibrosis and neovascularization mentioned above has important clinical significance in developing a treatment for the prevention of the subretinal fibrosis and neovascularization in other ocular diseases.
In summary, our research findings unequivocally demonstrated the pivotal role of the positive interaction loop of Wnt5a/β-catenin in inducing EMT of RPE cells, facilitating cell migration, and contributing to subretinal fibrosis in nAMD, as illustrated in the schematic diagram presented in F
ig.
8I. Furthermore, exploring the involvement of Wnt5a in the pathogenesis of AMD-related conditions beyond EMT, such as EndMT, PMT, and glial cell activation, holds promising potential for further exploration. Moreover, considering the pivotal role of the Wnt signaling pathway in pathological neovascularization across diverse diseases, it is plausible to speculate that Wnt5a/β-catenin might serve as a key role in driving angiogenesis of RNV and CNV secondary to nAMD. Our findings shed light on one of the intricate molecular pathways underlying subretinal fibrosis and offer novel insights into potential therapeutic strategies for the management of the fibrovascular diseases by targeting Wnt5a/β-catenin-mediated EMT. In future, when developing novel therapies for treating nAMD, these findings should be taken into careful consideration, aiming to employ minimal drug combinations to address the pathological neovascularization, subretinal fibrosis, inflammation, and the various cellular processes involved in EMT or activation within the retina. This approach seeks to optimize the therapeutic strategy for the management of nAMD and other fibrovascular diseases.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.