GDF11 supplementation improved retinal structure and function in retinal ischemia injury

Abstract

Background

Retinal ischemia is a major cause of blindness and plays a detrimental role in various diseases, including occlusion of arteries or veins, diabetic retinopathy, and ocular ischemic syndrome, which could lead to the neuronal and vascular dysfunction. As a result, maintaining their activities may help to avoid visual loss. Growth differentiation factor 11 (GDF11) has been implicated that exert neuroprotective effects and promote the angiogenesis after cerebral or cardiac ischemic injury. Here, we demonstrate that GDF11 has a protective effect on ischemia retinal injury.

Methods

To simulate the morphological and functional results following retinal ischemia, a mouse unilateral common carotid artery occlusion (UCCAO) model was employed. Electroretinography (ERG) was conducted to assess the severity of retinal impairment. The effects were evaluated using hematoxylin and eosin (H&E)-staining and immunohistochemistry. In addition, angiogenic activity affected by retinal ischemia was assessed in vivo and in vitro models.

Results

We successfully established the UCCAO model and expanded the pathological understanding of UCCAO-induced retinal ischemia. The treatment with GDF11 attenuated cell death, retinal edema, gliosis and apoptosis processes in the acute phase after UCCAO. In addition, GDF11 further reduced apoptosis and improved angiogenesis in the chronic phase after UCCAO. Mechanistically, GDF11 exerted its protective effects by activating the phosphoinositide 3-kinase (PI3K)/protein kinase B (AKT) pathway, underscoring its potential as a multifaceted therapeutic agent for retinal ischemia.

Conclusions

Our study shows the potential of GDF11 as a novel therapeutic for recovering retinal function following retinal ischemia. The results reveal that the therapeutic benefits of GDF11 are exerted during the acute and chronic phases following retinal ischemic injury.

Graphical abstract

1. Introduction

Retinal ischemia is a leading cause of blindness worldwide. It has been suggested to be involved in the common pathogenesis of various vascular diseases such as diabetic retinopathy (DR), central retinal artery occlusion (CRAO), ischemic optic neuropathies, and ocular ischemic syndrome (OIS).1, 2, 3, 4 The common feature of these conditions is a phase of ischemia or reduced blood supply to the retinal tissue, which eventually leads to neuronal degeneration.⁵

A principal cause of retinal ischemia is common carotid artery occlusion (CCAO) or stenosis (CCAS).⁶ Consequently, a bilateral common carotid artery occlusion (BCCAO) model in rats, which mimics chronic retinal hypoperfusion, has been employed to investigate the pathological mechanisms for retinal ischemia.^5,7, 8, 9 However, the BCCAO model is primarily limited to rats. Adopting this model for mice disrupts cerebral blood circulation due to the lack of posterior communicating arteries (PCA) in their circle of Willis, resulting in elevated mortality.¹⁰ Therefore, the unilateral common carotid artery occlusion (UCCAO) model was developed to delve into the pathological mechanisms underlying retinal ischemia in mice.11, 12, 13, 14 It was first used as an experimental model for retinal ischemia, showing changes in morphology and molecule.¹¹ However, the more precise mechanisms of retinal ischemia in this model have not been widely studied.

Growth differentiation factor 11 (GDF11) pertains to the transforming growth factor-β (TGF-β) superfamily and is pivotal in mammalian development.¹⁵ Recently, GDF11 has garnered significant interest as a potential anti-aging therapeutic candidate. Overexpression of GDF11 can rejuvenate aging hearts and improve cardiac function post-ischemic cardiac injury.¹⁶ Furthermore, there is evidence that GDF11 promotes the proliferation of primary brain capillary endothelial cells leading to vascular restructuring and neuronal revival in aging mice brains.¹⁷

Recent research on GDF11 has been conducted using younger animal models. Su et al. suggested that exogenous GDF11 can mitigate acute myocardial ischemic injury by diminishing inflammation and oxidative stress.¹⁸ Moreover, GDF11 overexpression supports mitochondrial biogenesis and inhibits apoptosis in acute myocardial ischemia.¹⁹ Beyond myocardial concerns, accumulating evidence suggests that GDF11 promotes neurofunctional recovery and angiogenesis in stroke model.20, 21, 22

Despite increasing interest in GDF11 and its therapeutic potential, its role in retinal ischemia models remains uncharted. Accordingly, this study seeks to ascertain the effect of GDF11 on retinal ischemia using the UCCAO mouse model. Our hypothesis posits that exogenous recombinant GDF11 could confer protection against retinal ischemia by enhancing neuronal recovery, reducing apoptosis, and promoting retinal blood supply.

2. Material and methods

2.1. Animals

Male mice (C57BL/6, 6-8 weeks old) were obtained from the Shanghai Laboratory Animal Center of the Chinese Academy of Science (Shanghai, China). The animals were maintained under a 12-h light-dark cycle at a consistent temperature of 24 ± 1 °C with free access to food and water. The Ethics Committee of Xinhua Hospital of the School of Medicine at Shanghai Jiao Tong University granted approval for all animal protocols (XIIEC-KJB-2024-013). Furthermore, all experimental procedures adhered to the Association for Research in Vision and Ophthalmology's guidelines for animal usage in ophthalmic and vision research.

2.2. Cell culture and ischemia model

Human retinal microvascular endothelial cells (HRMECs) were sourced from the AoYin Cell Center (SAc0515, Shanghai, China) and were cultivated under standard conditions in endothelial cell medium (ECM; No. 1001, Sciencell). This medium comprised 5% fetal bovine serum (FBS; No. 0025, Sciencell), 1% endothelial cell growth supplement (ECGS; No.1052, Sciencell), and 1% penicillin/streptomycin solution (P/S; No. 0503, Sciencell). Our team has previously developed a targeted model for hypoxia in HRMECs,²³ a concentration of 400 μmol/L H₂O₂ was chosen to build this model.

2.3. UCCAO-induced retinal ischemia model and GDF11 administration

The procedure for UCCAO was adapted from previous methods.¹¹ Briefly, anesthesia was administered intraperitoneally using a ketamine (60 mg/kg) and xylazine (10 mg/kg) combination. After creating a cervical incision, the right common carotid artery was meticulously isolated. Double 6-0 silk sutures were tied around the artery, which was then severed between the ligations. Postoperative recovery entailed suturing the incision and allowing the mouse to recuperate. Following 1 h post-UCCAO, 1 μL of recombinant GDF11 (50 μg/mL) (120-11, PeproTech, USA) was injected intravitreally. The control procedure mirrored the aforementioned steps, substituting GDF11 with 0.1% bovine serum albumin in phosphate-buffered saline (PBS). The eyes that underwent UCCAO were subsequently excised at 24-h and 1-week post-surgery intervals.

2.4. Western blot analysis

Retinal tissue or HRMECs lysates were prepared with cold radioimmunoprecipitation assay (RIPA) buffer (Beyotime, China). Protein concentrations were quantified using a BCA Protein Assay Kit (Beyotime, Shanghai, China). Proteins were loaded on sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto polyvinylidene fluoride membranes (Immobilon-P; Millipore, Schaffhausen, Switzerland). Membranes were blocked using QuickBlock™ Blocking Buffer (Beyotime, China) for 30 min at room temperature, followed by incubation with primary anti-GDF11 antibody (1:1000; Abcam), anti-hypoxia‑inducible factor 1 (HIF1) antibody (1:1000; Affinity), anti-vascular endothelial growth factor (VEGF)-A antibody (1:1000; Affinity), anti-P-vascular endothelial growth factor receptor 2 (VEGFR2) antibody (1:1000; Affinity), anti-VEGFR2 antibody (1:1000; Affinity), anti-phosphoinositide 3-kinase (PI3K) antibody (1:1000; Affinity), anti-P-PI3K antibody (1:1000; Affinity), anti-protein kinase B (AKT) antibody (1:1000; Affinity), anti-P-AKT antibody (1:1000; Affinity), anti-superoxide dismutase 1 (SOD1) antibody (1:1000; Affinity), anti-SOD2 antibody (1:1000; Affinity), anti-C-CASPASE 3 antibody (1:1000; Affinity), anti-4-hydroxynonenal (HNE) antibody (1:1000; Affinity), anti-Bcl-2 associated X protein (BAX) antibody (1:1000; Affinity), anti-B-cell lymphoma-2 (BCL2) antibody (1:1000; Affinity), anti-glyceraldehyde-3-phosphate dehydrogenase antibody (1:1000; Affinity), overnight at 4 °C. After washing, membranes were incubated with horseradish peroxidase-conjugated secondary antibodies for 1 h at room temperature. Protein bands were visualized and their intensity quantified using ImageJ software.

2.5. Electroretinography (ERG)

ERG measurements were acquired using the Espion Diagnosys system (Diagnosys LLC, Littleton, MA). A minimum of 8 h of dark adaptation preceded the dark-adapted ERG tests. Prior to the evaluation, mice were anesthetized with an intraperitoneal injection of ketamine (60 mg/kg) and xylazine (10 mg/kg) and their body temperature was regulated using a heating pad. Pupil dilation was achieved by topically administering a 0.5% blend of tropicamide and phenylethylamine (Santen Pharmaceutical, Osaka, Japan). Active electrodes were placed in contact with the cornea, a reference electrode was positioned in the mouth, and a ground electrode was inserted at the tail's end. After applying a topical anesthetic (Proparacaine; Alcaine® drop 0.5%), the active electrodes were positioned. A series of light stimulations were used to record a-wave and b-wave amplitudes.

2.6. Histologic and morphometric analysis

Post-UCCAO (24 h), eyes were enucleated, fixed in special fixation solution (Servicebio, China) for 24 h, and embedded in paraffin. Sections (4-μm-thick) were obtained vertically through the optic disc and stained with hematoxylin and eosin (H&E). Cells in the ganglion cell layer (GCL) with pyknotic nuclei were deemed dead. Layer thicknesses for the inner nuclear layer (INL) were gauged from microscopic images (Olympus, Japan).

2.7. Immunohistochemistry

Twenty-four hours and one-week post-UCCAO, ocular globes were fixed in 4% paraformaldehyde for 4 h, subsequently dehydrated in 40% sucrose solution overnight, embedded in Tissue-Tek OCT (Sakura Finetek, Tokyo, Japan), and cryopreserved at -80 °C. Sagittal sections (10 μm) were obtained using a freezing microtome. The sections underwent triple washing in PBS, blockage in a PBS solution with 10% FBS and 0.15% Triton X-100 for an hour at room temperature, followed by an overnight incubation at 4 °C with primary antibodies targeting aquaporin 4 (AQP4, 16473-1-AP, Proteintech, China), GFAP (Z0334, Dako, USA), Iba1 (ab187846, abcam, USA), caspase 1 (22915-1-AP, Proteintech, China), caspase 3 (9664 S, CST, USA), and CD31 (ab182981, abcam, USA). Secondary antibodies conjugated with Alexa Fluor 488 or 592 were applied for 1 h at room temperature, followed by nuclear staining with DAPI. Fluorescent imaging was conducted with an Olympus BX61 microscope.

2.8. TUNEL assay for apoptosis

Cell apoptosis was quantified utilizing the terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) apoptosis assay kit (Beyotime, C1086), executed per manufacturer directives. Apoptotic cells were visualized and quantified by enumerating green fluorescent cells across ten randomly selected fields for each sample, with imaging carried out on an Olympus BX61 fluorescent microscope.

2.9. Angiogenesis assays

For tube formation, Matrigel (Corning Inc.) coated 48-well plates were used to culture HRMECs (1 × 10⁵ cells/well), with tube structures forming post 4-6 h of hypoxic incubation. Morphology was documented using an Olympus light microscope and analyzed with WimTube software (www.wimasis.com). Migration was assessed via a scratch assay on HRMECs cultured in serum-free media for 8 h, with scratches generated using a 200-μL pipette tip. Initial (0 h) and subsequent (24 h) migration were imaged, and cell-free areas quantified using WimScratch software (www.wimasis.com).

2.10. Statistical analysis

Data were presented as mean ± standard error of the mean (SEM) and analyzed using GraphPad Prism 9. Differences were deemed statistically significant at P-values less than 0.05, determined via Student's t-test, One-way or two-way ANOVA with Tukey's or Sidak's post-hoc tests.

3. Results

3.1. UCCAO successfully induced retinal ischemia and decreased GDF11 expression

To establish the retinal ischemia model, we performed UCCAO in mice (Fig. 1Aa). Intravitreal injection of GDF11 or vehicle was administered at 1 h post-surgery, and evaluations were conducted at one day (1 D) and one week (1 W) post-UCCAO (Fig. 1Ab). Successful arterial occlusion was confirmed by ipsilateral ptosis immediately after surgery (Fig. 1Ac). Western blot analysis revealed a progressive decline in GDF11 protein levels following UCCAO, with expression decreasing at 1 D and further declining at 1 W compared to controls (P<0.01, Fig. 1C).

Fig. 1.

UCCAO model establishment, GDF11 expression, and effects on retinal function and structure.

(A) Schematic diagram of the experimental design. (a) Illustration of UCCAO surgical procedure. (b) Timeline of experimental procedures, including UCCAO induction, intravitreal injection (1 h post-UCCAO), and sample collection at 1 day (1 D) and 1 week (1 W). (c) Representative photographs showing ipsilateral ptosis in UCCAO mice compared with control mice. (B) Representative scotopic ERG waveforms and quantification of b-wave amplitudes and b/a wave ratios at 1 D (upper panels) and 1 W (lower panels) post-UCCAO. (C) Western blot analysis of GDF11 protein expression in retinal tissue. (D) Representative H&E-stained retinal sections and quantification of INL thickness. Data are presented as mean ± SEM. n=3 animals per group. ns P>0.05, ∗P<0.05, ∗∗P<0.01, ∗∗∗P<0.001, ∗∗∗∗P<0.0001 vs. indicated group.

3.2. GDF11 improved retinal function after UCCAO

ERG was performed to assess retinal function. Vehicle-treated UCCAO mice showed marked reductions in scotopic b-wave amplitudes at both 1 D and 1 W time points compared to controls (P<0.01). GDF11 treatment significantly restored b-wave amplitudes at 1 D (P<0.01) and 1 W (P<0.01) compared to vehicle-treated groups. Additionally, the b/a wave ratio, an indicator of inner retinal function, was improved by GDF11 treatment at both time points (P<0.01 at 1 D; P<0.01 at 1 W, Fig. 1B).

3.3. GDF11 attenuated UCCAO-induced retinal structural damage

H&E staining revealed notable morphological alterations in vehicle-treated UCCAO retinas. Neuronal cells in the GCL displayed disorganized arrangement with signs of nuclear pyknosis. Quantitative analysis showed that INL thickness was markedly reduced following UCCAO at both 1 D and 1 W, indicating structural disruption of the inner retina. GDF11 treatment partially restored INL thickness compared to vehicle groups (P<0.01 at 1 D; P<0.01 at 1 W, Fig. 1D), suggesting a protective effect on retinal architecture.

3.4. GDF11 reduced retinal edema marker expression

AQP4, a water channel protein associated with retinal edema, was examined by immunofluorescence. UCCAO induced a substantial increase in AQP4 expression at both 1 D and 1 W. GDF11 treatment markedly decreased AQP4 immunoreactivity at both time points (P<0.01, Fig. 2A), suggesting reduced retinal edema in the ischemic retina.

Fig. 2.

GDF11-mediated attenuation of retinal edema and glial activation following UCCAO.

(A) Immunofluorescence staining of AQP4 (green) in retinal sections at 1 D and 1 W post-UCCAO. (B) Immunofluorescence staining of Iba1, a microglial marker, in retinal sections at 1 D and 1 W post-UCCAO. (C) Immunofluorescence staining of GFAP, an astrocyte/Müller cell marker, in retinal sections at 1 D and 1 W post-UCCAO. Zoomed insets show representative morphological features of reactive gliosis. Nuclei were counterstained with DAPI. Data are presented as mean ± SEM. n=3 animals per group. ns P>0.05, ∗P<0.05, ∗∗P<0.01, ∗∗∗P<0.001, ∗∗∗∗P<0.0001 vs. indicated group.

3.5. GDF11 inhibited glial activation following UCCAO

Reactive gliosis was assessed using Iba1 (microglial marker) and GFAP (astrocyte/Müller cell marker) immunostaining. At 1 D post-UCCAO, vehicle-treated retinas exhibited robust Iba1-positive microglial activation, which was attenuated by GDF11 (P<0.01, Fig. 2B). Similarly, GFAP immunoreactivity was elevated in vehicle-treated UCCAO mice, particularly in the GCL and INL regions, with activated Müller cell processes extending toward outer retinal layers. GDF11 treatment reduced GFAP expression at both 1 D and 1 W (P<0.01, Fig. 2C). The inhibitory effects of GDF11 on both microglial and astrocytic activation persisted through the 1 W time point.

3.6. GDF11 suppressed apoptosis in the acute and chronic phases of UCCAO

To evaluate apoptotic cell death, we performed immunofluorescence staining for caspase-1 and caspase-3, along with TUNEL assay. At 1 D, UCCAO induced upregulation of both caspase-1 and caspase-3 expression, which was reduced by GDF11 treatment (P<0.01, Fig. 3A). The anti-apoptotic effect of GDF11 was maintained at 1 W post-UCCAO. TUNEL-positive cells were markedly increased in vehicle-treated UCCAO retinas at both 1 D and 1 W, and GDF11 treatment decreased the number of apoptotic cells at both time points (P<0.01, Fig. 3B).

Fig. 3.

GDF11-mediated inhibition of retinal cell death following UCCAO.

(A) Immunofluorescence co-staining of caspase-1 (green) and caspase-3 (red) in retinal sections at 1 D and 1 W post-UCCAO. Nuclei were counterstained with DAPI (blue). (B) TUNEL assay for detection of apoptotic cells in retinal sections at 1 D and 1 W post-UCCAO. Data are presented as mean ± SEM. n=3 animals per group. ns P>0.05, ∗P<0.05, ∗∗P<0.01, ∗∗∗P<0.001, ∗∗∗∗P<0.0001 vs. indicated group.

3.7. GDF11 promoted angiogenesis during chronic phase recovery

To investigate the pro-angiogenic effects of GDF11, we examined CD31 expression at 1 W post-UCCAO. Immunofluorescence analysis showed that CD31-positive vascular endothelial cells were increased in GDF11-treated retinas compared to the vehicle-treated UCCAO groups (P<0.01, Fig. 4A). Western blot analysis revealed that GDF11 treatment upregulated HIF-1α (P<0.01), VEGF-A (P<0.01 vs control; P<0.01 vs UCCAO), and enhanced VEGFR2 phosphorylation (p-VEGFR2/VEGFR2 ratio, P<0.01), while total VEGFR2 levels remained unchanged (Fig. 4B).

Fig. 4.

GDF11-mediated promotion of retinal angiogenesis via the HIF-1α/VEGF-A/VEGFR2 signaling pathway.

(A) Immunofluorescence staining of CD31 (vascular endothelial cell marker) in retinal sections at 1 W post-UCCAO. (B) Western blot analysis of angiogenesis-related proteins, including HIF-1α, VEGF-A, VEGFR2, and phosphorylated VEGFR2 (p-VEGFR2), in retinal tissue at 1 W post-UCCAO. Data are presented as mean ± SEM. n=3 animals per group. ns P>0.05, ∗P<0.05, ∗∗P<0.01, ∗∗∗P<0.001, ∗∗∗∗P<0.0001 vs. indicated group.

3.8. GDF11 enhanced angiogenic capacity of HRMECs under oxidative stress

The angiogenic potential of GDF11 was further examined in HRMECs subjected to H₂O₂-induced oxidative stress. Tube formation assays demonstrated that H₂O₂ exposure reduced capillary-like network formation, as evidenced by decreased branching points. Co-treatment with GDF11 restored tube formation capacity (P<0.01, Fig. 5A). Scratch wound migration assays showed that H₂O₂ treatment impaired wound closure at 24 h, while GDF11 co-treatment improved cell migration (P<0.01, Fig. 5B).

Fig. 5.

GDF11-mediated promotion of angiogenesis and protection of HRMECs against oxidative stress via the PI3K/AKT pathway.

(A) Tube formation assay in HRMECs, with quantification of total branching points. (B) Scratch wound migration assay in HRMECs at 24 hours, with quantification of wound closure. (C) Western blot analysis of oxidative stress-related proteins, including SOD1, SOD2, 4-HNE, and apoptosis-related proteins, including BCL2, BAX, and cleaved caspase-3. (D) Western blot analysis of PI3K/AKT signaling-related proteins, including PI3K, phosphorylated PI3K (P-PI3K), AKT, and phosphorylated AKT (P-AKT). Data are presented as mean ± SEM. n=3 independent experiments per group. ns P>0.05, ∗P<0.05, ∗∗P<0.01, ∗∗∗P<0.001, ∗∗∗∗P<0.0001 vs. indicated group.

3.9. GDF11 alleviated oxidative stress and apoptosis in HRMECs

Western blot analysis was performed to evaluate GDF11 effects on oxidative stress and apoptosis markers in HRMECs. H₂O₂ exposure decreased the expression of antioxidant enzymes SOD1 and SOD2, while increasing 4-HNE (P<0.01). GDF11 co-treatment restored SOD1 and SOD2 levels and reduced 4-HNE expression (P<0.01, Fig. 5C). For apoptosis markers, H₂O₂ treatment reduced BCL2 while increasing BAX and cleaved caspase-3 expression (P<0.01). GDF11 reversed these changes, restoring the BCL2/BAX balance and reducing caspase-3 cleavage (P<0.01, Fig. 5C).

3.10. GDF11 activated the PI3K/AKT signaling pathway in HRMECs

To explore the molecular mechanisms underlying GDF11's protective effects, we examined the PI3K/AKT signaling pathway. H₂O₂ exposure decreased the phosphorylation levels of PI3K and AKT, while total PI3K and AKT protein levels remained unchanged. GDF11 treatment restored P-PI3K and P-AKT levels, as well as the P-PI3K/PI3K and P-AKT/AKT ratios (P<0.01, Fig. 5D). These findings suggest that the protective effects of GDF11 against oxidative stress-induced damage are mediated through activation of the PI3K/AKT survival pathway.

4. Discussion

This study explored the therapeutic potential of exogenous GDF11 in retinal ischemia and oxidative stress, employing both in vivo and in vitro approaches. Employing a UCCAO model in mice and assays with HRMECs, we demonstrated that GDF11 exhibits neuroprotective, antioxidative, anti-apoptotic, and pro-angiogenic effects at both acute (1 D) and chronic (1 W) phases. These findings target multiple key pathological features of retinal ischemia, providing a comprehensive understanding of GDF11's therapeutic potential.

Retinal ischemia contributes significantly to visual impairment and is central to the progression of many ocular diseases. Historically, BCCAO in rats has been widely used to simulate ischemic injury.^5,7, 8, 9 However, in mice, the incomplete circle of Willis presents significant challenges with BCCAO, including high mortality.¹⁰ The UCCAO model was developed to overcome these limitations, selectively inducing retinal ischemia while avoiding systemic complications, such as brain injury.^11,13 In this study, the UCCAO model effectively replicated retinal ischemic pathologies, including cellular death, edema, gliosis, and impaired angiogenesis. Moreover, the observed time-dependent decline in endogenous GDF11 expression in ischemic retinas suggests that ischemia disrupts GDF11 homeostasis, underscoring the potential therapeutic value of exogenous GDF11 supplementation.

A prominent consequence of retinal ischemia is neuronal dysfunction, as assessed in this study using ERG. GDF11 treatment effectively restored the reduced b-wave amplitudes and b/a wave ratios caused by UCCAO-induced ischemia (Fig. 1B). Notably, the improvements were observed at both 1 D and 1 W post-UCCAO, indicating that GDF11 provides both immediate and sustained functional protection. The b-wave amplitude reflects the physiological function of the inner retinal layers, suggesting that GDF11 may mitigate ischemia-induced damage and preserve retinal function.

Retinal structural integrity was assessed by H&E staining and quantitative measurement of INL thickness. UCCAO induced marked thinning of the INL at both 1 D and 1 W, reflecting structural disruption in the inner retina. GDF11 treatment partially restored INL thickness (Fig. 1D), indicating its protective effect on retinal architecture.

Retinal edema, primarily mediated by Müller cells, is another critical feature of ischemia. The end-foot of Müller cells, which encircles the superficial retinal blood vessels, regulates AQP4 expression to maintain retinal water and ion homeostasis.¹⁰ During ischemia, AQP4 expression is upregulated, contributing to tissue swelling.²⁴ Our findings showed that UCCAO substantially increased AQP4 expression at both 1 D and 1 W, and GDF11 treatment markedly reduced AQP4 immunoreactivity at both time points (Fig. 2A), demonstrating that GDF11 ameliorates edema by modulating Müller cell activity.

Reactive gliosis, characterized by activation of astrocytes and microglia, is a hallmark of retinal ischemia and contributes to inflammation and secondary neuronal damage.²⁵ GDF11 treatment significantly reduced gliosis in the retina, as demonstrated by decreased expression of GFAP in astrocytes and Iba1 in microglia (Fig. 2B and C). The anti-gliotic effect of GDF11 was sustained from the acute phase (1 D) through the chronic phase (1 W), suggesting long-term modulation of neuroinflammatory responses. This reduction in glial activation likely plays a key role in preserving retinal integrity, as chronic gliosis exacerbates tissue damage and impairs recovery.²⁶ The ability of GDF11 to suppress gliosis aligns with its broader neuroprotective effects and emphasizes its potential to mitigate the inflammatory consequences of ischemia.

Significant neuronal cell death was observed in the GCL in retinopathy induced by UCCAO, characterized by pyknosis, an irreversible chromatin condensation process associated with necrosis or apoptosis. GDF11 treatment significantly reduced the number of pyknotic cells, suggesting it mitigates ischemia-induced neuronal loss (Fig. 1D). Most of these neurons underwent apoptosis due to retinal ischemia. Therefore, preventing cell death is crucial for salvaging the penumbra and improving outcomes after ischemic injury.

In vivo experiments showed that GDF11 treatment reduced the expression of caspase-1 and caspase-3, compared to vehicle-treated mice (Fig. 3A). Additionally, TUNEL-positive apoptotic cells were significantly reduced at both 1 D and 1 W post-UCCAO in the GDF11-treated group (Fig. 3B). The role of caspase-1 in neurodegenerative diseases, including cerebral ischemia and Parkinson's disease, is well-documented.^27,28 Caspase-1 activation can also indirectly promote caspase-3 activation, exacerbating apoptosis under stress.²⁶ Our findings highlight GDF11's ability to disrupt these key apoptotic pathways, suggesting it offers long-term neuroprotection.

Further support for GDF11's anti-apoptotic effects comes from in vitro experiments using HRMECs. Under oxidative stress induced by H₂O₂, there was a marked reduction in the anti-apoptotic protein BCL2 and an increase in pro-apoptotic markers BAX and C-CASPASE3 (Fig. 5C). These changes highlight the increased apoptosis caused by oxidative stress. GDF11 treatment reversed these effects, restoring the BCL2/BAX balance and significantly reducing apoptosis (Fig. 4C), providing mechanistic evidence of GDF11's protective role in preventing endothelial cell apoptosis under oxidative stress. This dual action on neuronal and endothelial cells highlights the multifaceted protective effects of GDF11 in ischemic retinopathy.

Oxidative stress is a central driver of ischemic injury, and GDF11's antioxidative effects were clearly demonstrated in HRMECs. Exposure to H₂O₂ significantly reduced the expression of antioxidative enzymes SOD1 and SOD2, while increasing the oxidative stress marker 4-HNE (Fig. 5C). GDF11 treatment restored SOD1 and SOD2 levels and significantly decreased 4-HNE expression, further emphasizing its ability to combat oxidative stress. These results are consistent with findings in other ischemic tissues and underscore the importance of addressing oxidative stress in therapeutic strategies for retinal ischemia.²⁹

Mechanistically, the protective effects of GDF11 were linked to activation of the PI3K/AKT pathway, a critical regulator of cell survival and oxidative stress responses.30, 31, 32 In HRMECs exposed to H₂O₂, phosphorylation levels of PI3K (P-PI3K) and AKT (P-AKT) were significantly reduced, while total PI3K and AKT levels remained unchanged (Fig. 5D). GDF11 treatment restored P-PI3K and P-AKT levels, indicating that its protective effects are mediated by activation of the PI3K/AKT pathway, which plays a crucial role in cell survival and oxidative stress responses.

Another key aspect of retinal recovery is angiogenesis, which is essential for repairing ischemic tissue. Consistent with previous studies showing GDF11's role in vascular remodeling in cerebral and cardiac ischemia 20, 21, 22, our findings revealed increased CD31 expression in the retina following GDF11 treatment at 1 W post-UCCAO (Fig. 4A). Ischemia leads to tissue hypoxia, which stabilizes HIF-1α and subsequently upregulates VEGF-A expression. VEGF-A binds to VEGFR2 on endothelial cells, promoting their survival and proliferation. In our study, GDF11 treatment increased HIF-1α and VEGF-A levels, along with enhanced VEGFR2 phosphorylation at 1 W post-UCCAO (Fig. 4B). Importantly, this molecular activation was accompanied by a clear increase in CD31-positive vascular density in GDF11-treated retinas (Fig. 4A), suggesting that activation of the HIF-1α/VEGF-A/VEGFR2 axis is linked to improved microvascular restoration after ischemic injury. Consistent with the in vivo findings, GDF11 also strengthened angiogenic behaviors of HRMECs in vitro, including migration and tube formation. These results highlight GDF11's ability to facilitate vascular repair in ischemic conditions, aligning with its known pro-angiogenic effects in other tissues.

While these findings highlight GDF11's potential, several limitations should be addressed in future studies. The long-term efficacy and safety of GDF11 require further investigation, particularly in the chronic ischemic conditions. Although we demonstrated the involvement of the PI3K/AKT and HIF-1α/VEGF-A pathways, the specific cell types expressing GDF11 in the retina and how its expression is regulated at the cellular level remain to be determined. Additionally, the interactions of GDF11 with other signaling pathways, such as TGF-β, needed to be explored to fully understand its therapeutic potential.

5. Conclusions

In conclusion, this study demonstrates that GDF11 alleviates retinal ischemia-induced retinal edema, apoptosis, gliosis, and impaired angiogenesis. GDF11 exerts its protective effects through multiple mechanisms, including suppression of AQP4-mediated edema, inhibition of caspase-dependent apoptosis, attenuation of glial activation, and promotion of angiogenesis via the HIF-1α/VEGF-A/VEGFR2 pathway. The activation of PI3K/AKT signaling in endothelial cells further supports its cytoprotective role under oxidative stress. These findings underscore the beneficial effects of GDF11 across both acute and chronic phases following retinal ischemia injury, highlighting its potential as a novel therapeutic avenue for restoring retinal function post-ischemia insult.

Study approval

The authors confirm that any aspect of the work covered in this manuscript that involved human patients or animals was conducted with the ethical approval of all relevant bodies and the study was performed in accordance with the Declaration of Helsinki, and the protocol was approved by the Ethics Committee of Xinhua Hospital of the School of Medicine at Shanghai Jiao Tong University (XIIEC-KJB-2024-013). Furthermore, all experimental procedures adhered to the Association for Research in Vision and Ophthalmology's guidelines for animal usage in ophthalmic and vision research.

Author contributions

The authors confirm contribution to the paper as follows: Conceptualization: YX, HF, PZ; Methodology: YX, HF, PZ; Investigation: XZ, YY, JY, HX, JP, ML, EC; Formal Analysis & Statistical Analysis: YY, JP, HX, ZL, YW, JY, WZ, HL, ML, EC, MW, XZ, MW; Writing (Original Draft): XZ; Writing (Review & Editing): YX, HF, PZ. All authors reviewed the results and approved the final version of the manuscript.

Funding

This study was supported by the National Key R&D Program of China (2024YFC2510900), Discipline Peak-Climbing Plan of Xinhua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine (XKPF2024A5002), Data Sharing and Emulation of Clinical Trials (CCS-DASET) of SHDC (No.SHDC2024CRI011), Clinical Research Unit, Xin Hua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine (24XHCR04A), Science and Technology Commission of Shanghai Municipality (No.24ZR1449900), and Natural Science Foundation of Shandong Province (ZR2023MH054).

Declaration of competing interests

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Abbreviations

GDF11

Growth differentiation factor 11

UCCAO

Unilateral common carotid artery occlusion

ERG

Electroretinography

DR

Diabetic retinopathy

CRAO

Central retinal artery occlusion

OIS

Ocular ischemic syndrome

CCAO

Common carotid artery occlusion

CCAS

Common carotid artery stenosis

PCA

Posterior communicating arteries

TGF-β

Transforming growth factor-β

HRMECs

Human retinal microvascular retinal endothelial cells

ECM

Endothelial cell medium

ECGS

Endothelial cell growth supplement

P/S

Penicillin/streptomycin solution

H&E

Hematoxylin and eosin

PFA

Paraformaldehyde

GCL

Ganglion cell layer

IPL

Inner plexiform layer

INL

Inner nuclear layer

TUNEL

Terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling

RGCs

Retinal ganglion cells

AQP4

Aquaporin 4