Effect of ranibizumab on diabetic retinopathy via the vascular endothelial growth factor/STAT3/glial fibrillary acidic protein pathway

Abstract

BACKGROUND

Diabetic retinopathy (DR) is the leading cause of vision loss in patients with diabetes. The vascular endothelial growth factor (VEGF) pathway plays a critical role in the pathogenesis of DR, and ranibizumab, an anti-VEGF agent, has shown promise in its treatment. Signal transducer and activator of transcription 3 (STAT3) is involved in inflammatory processes and cellular signaling, while glial fibrillary acidic protein (GFAP) is a marker of glial cell activation, both contributing to retinal damage in DR. However, the mechanisms by which ranibizumab affect early-stage DR through the VEGF/STAT3/GFAP pathway are not fully understood.

AIM

To investigate the role of ranibizumab in early DR via the VEGF/STAT3/GFAP pathway.

METHODS

Adult retinal pigment epithelial 19 (ARPE-19) cells and human retinal microvascular endothelial cells (HRMECs) were cultured under high-glucose conditions to simulate a diabetic environment. The effects of ranibizumab on cytokine mRNA and protein expression were analyzed by quantitative polymerase chain reaction and Western blot analysis. A diabetic rat model was induced with streptozotocin (60 mg/kg). Retinal changes, including retinal ganglion cell (RGC) apoptosis, vascular alterations, and cytokine expression, were evaluated using fundus fluorescein angiography, hematoxylin and eosin and periodic acid Schiff staining, immunofluorescence, confocal imaging, and Western blot analysis.

RESULTS

High-glucose conditions significantly increased the mRNA and protein levels of VEGF, STAT3, GFAP, and other cytokines in ARPE-19 and HRMECs. However, these levels were partially suppressed by ranibizumab. RGC apoptosis, vascular leakage, and elevated cytokine expression were observed during early-stage DR in diabetic rats. Ranibizumab treatment in diabetic rats reduced cytokine expression, restored RGCs, and repaired vascular networks.

CONCLUSION

Intravitreal ranibizumab modulates the VEGF/STAT3/GFAP pathway, suppresses cytokine expression, and promotes retinal repair, effectively delaying or preventing early DR progression.

Key Words: Diabetic retinopathy; Ranibizumab; Early stage; Vascular endothelial growth factor; Signal transducer and activator of transcription 3; Glial fibrillary acidic protein

Core Tip: Diabetic retinopathy (DR) is the leading cause of vision loss in patients with diabetes; however, the mechanisms behind its early stages remain unclear. This study explored the therapeutic effects of intravitreal ranibizumab on early DR through its effect on the vascular endothelial growth factor/STAT3/glial fibrillary acidic protein signaling pathway. Using high-glucose retinal cells and diabetic rat models, ranibizumab suppressed cytokine expression, reduced retinal ganglion cell apoptosis, and repaired vascular networks. These findings highlight the potential of ranibizumab in delaying or preventing early DR progression and provide a foundation for its clinical application.

INTRODUCTION

Diabetic retinopathy (DR) is a microvascular complication of diabetes mellitus (DM) and a leading cause of global vision loss. It is characterized by retinal microvascular abnormalities, including increased vascular permeability, microaneurysms, and capillary occlusion. In severe cases, DR can progress to include retinal neovascularization and detachment. Depending on the severity of its clinical signs and symptoms, DR can be classified as non-proliferative and proliferative.

A 2024 Lancet Global Research report estimated the global prevalence of diabetes at 14%, affecting approximately 828 million people worldwide. Urbanization, aging, reduced physical activity, and increasing rates of overweight and obesity have been identified as the driving factors behind this trend[1]. DR is currently a leading cause of vision loss among adults in some developed countries[2]. As such, it has become an important public health concern[3].

Patients with DR are treated based on the type of retinal lesions they manifest. These retinal lesions may include retinal microaneurysms, macular edema, and neovascularization, for which retinal laser photocoagulation, intravitreal drug injection, and vitrectomy may be performed. Systemic control of blood glucose, blood pressure, and blood lipid levels, as well as close monitoring of at-risk conditions, such as pregnancy and obesity, are also recommended[4]. The retinal changes in DR are typically irreversible. As such, it is important to implement effective interventions before the development of DR to reduce the overall risk for blindness.

Animal and human studies have shown that inflammatory cytokines and angiogenic factors are markedly involved in the pathogenesis of DR. Animal and human studies have shown that inflammatory cytokines and angiogenic factors are markedly involved in the pathogenesis of DR[2]. During early diabetes, the retina expresses elevated concentrations of proinflammatory mediators, such as tumor necrosis factor alpha (TNF-α), vascular endothelial growth factor (VEGF), intercellular adhesion molecule (ICAM), cluster of differentiation 18 (CD18), and interleukin 6 (IL-6)[5]. These mediators can initiate complex inflammatory processes that can structurally and functionally damage diabetic retinas[6]. Signal transducer and activator of transcription 3 (STAT3) is a member of the Janus kinase/STAT signaling pathway that regulates cell cytokine signaling. In endothelial cells, the inflammatory effects of STAT3 are largely attributed to the induction of ICAM-1 and VEGF[7]. STAT3 participates in inflammation and VEGF-induced angiogenesis, leading to increased endothelial cell permeability and vascular leakage in early DR[8,9]. Retinal pericytes function as early inflammatory sensors in DR by modulating inflammation in the retinal microenvironment via the Hes family BHLH transcription factor 1 (HES1)/STAT3 pathway. STAT3 inhibition has been shown to effectively alleviate early-stage damage in DR[10]. In addition, immunohistochemistry studies of diabetic retinas (donor eyes and animal models) have demonstrated increased rates of rod cell apoptosis, Müller cell glial fibrillary acidic protein (GFAP) overexpression[11], microglial cell activation, neurotrophic factor imbalance, and GFAP upregulation[12,13], which promote DR progression[14].

Previous studies have shown that diabetic rats express significantly higher levels of vitreal VEGF at 8 days after induction of high-glucose states and peak by the fourth week. The control groups in these studies had significantly lower expression of vitreal VEGF[15,16]. VEGF overexpression increases vascular permeability, which promotes neovascularization and retinal structural and functional abnormalities[17]. Our previous study[18] showed that intravitreal ranibizumab delayed DR progression at a very early stage in streptozotocin-induced diabetic rats.

This study explored the best approach for targeted ranibizumab treatment and the mechanisms underlying its effect on the development and progression of early-stage DR.

MATERIALS AND METHODS

Cells and culture

Adult retinal pigment epithelial 19 (ARPE-19) cells were purchased from Procell (Wuhan, Hubei Province, China), cultured in Dulbecco’s Modified Eagle Medium containing 5.5 mmol/L glucose (Procell), and supplemented with 10% fetal bovine serum (HyClone, Logan, UT, United States) and 1% penicillin/streptomycin (Thermo Fisher Scientific, Wuhan, China) in a humidified atmosphere with 50 mL/L carbon dioxide at 37 °C. Human retinal microvascular endothelial cells (HRMECs) and primary human retinal endothelial cells were purchased from ScienCell (Carlsbad, CA, United States) and cultured in a low-glucose (5.5 mmol/L) primary endothelial cell culture medium (ScienCell, Wuhan, Hubei Province, China) in a humidified atmosphere containing 50 mL/L carbon dioxide at 37 °C. The high-glucose model utilized the same parameters, but the glucose concentration in the culture medium was adjusted to 25 mmol/L for 72 hours prior to the experiment.

Cell viability assay

The Cell Counting Kit-8 (CCK-8) assay was used to assess the viability of ARPE-19 cells and HRMECs. The cells were seeded at a density of 3 × 10³/cm² in 100 μL culture medium in 96-well plates. After overnight incubation at 37 °C, the cells were treated with various concentrations of ranibizumab (Novartis, Zurich, Switzerland): 0.0625 mg/mL, 0.125 mg/mL, and 0.25 mg/mL for 12 hours, 24 hours, and 48 hours. Following treatment, 10 μL CCK-8 solution was added to each well, and the plates were incubated for an additional 1 hour. Optical densities were measured at 490 nm using a microplate reader (Thermo Fisher Scientific).

Animals

A total of 75 male Sprague-Dawley (SD) rats, 8-9 weeks old and weighing 280 ± 20 g, were purchased from the Animal Center of Nanchang University (Nanchang, Jiangxi Province, China). The rats were housed under controlled conditions: Temperature 23 ± 2 °C, relative humidity 50%, and a 12-hour light/dark cycle. They had ad libitum access to sterilized standard laboratory chow and water. All procedures followed the principles of animal ethics. The rats were anesthetized with an intraperitoneal injection of pentobarbital sodium (40 mg/kg body weight; Sigma-Aldrich, Merck Millipore, Darmstadt, Germany) prior to experimentation. All experiments complied with the guidelines of the Experimental Animal Ethical Committee for Traditional Chinese Medicine (Nanchang).

Diabetes induction and experimental groups

DM was induced in SD rats via a single intraperitoneal injection of streptozotocin (Sigma-Aldrich) at 60 mg/kg body weight. Rats with blood glucose levels ≥ 13.9 mmol/L (250 mg/dL) at 24 hours post-injection, which remained hyperglycemic for four consecutive days, were considered successfully induced diabetic models. A total of 100 diabetic rats were randomly divided into five groups (Groups A-E), while 20 healthy, untreated SD rats were included as a normal control group (Group F). Insulin was administered to maintain blood glucose levels within 2.80-7.56 mmol/L for groups requiring strict glycemic control.

Group A: Intravitreal injection of 1 μL ranibizumab[19] into the right eye on day 8 post-induction, combined with strict glycemic control.

Group B: Intravitreal injection of 1 μL ranibizumab on day 8 post-induction without glycemic control.

Group C: Strict glycemic control starting on day 8 post-induction without intravitreal injection.

Group D: No intravitreal injection or glycemic control.

Group E: Intravitreal injection of 1 μL sterile saline into the right eye on day 8 post-induction without glycemic control.

Group F: Healthy SD rats with no treatment, maintained under standard feeding condition.

All rats underwent fundus fluorescein angiography (FFA) and were sacrificed at weeks 6, 8, and 10 post-induction for further analyses.

FFA

FFA is a common and primary assessment for DR. The right eyes of the rats in each group were examined using FFA (Heidelberg Spectralis HRA, Heidelberg, Germany). Each rat was weighed and anesthetized with an intraperitoneal injection of sodium pentobarbital (40 mg/kg). Compound tropicamide (Mydrin-POR; Santen, Osaka, Japan) was instilled for pupil dilation, followed by Alcaine for local anesthesia and methyl cellulose to maintain corneal moisture. During the FFA examination, the rats received an intraperitoneal injection of 10% sodium fluorescein (0.001 mL/g; International Medication Systems, Dunstable, United Kingdom) for rapid imaging.

Intraocular ranibizumab injection and tight glycemic control

The right eye of each rat was intravitreally injected with 1 μL ranibizumab under anesthesia. Ranibizumab was delivered into the center of the vitreous humor (0.5 mm posterior to the limbus) using a 5-μL microsyringe (Hamilton, Bonaduz, Switzerland). Eyes with evidence of lens or retinal injury were excluded from the analysis. Based on the blood glucose concentration of each SD rat, 2-6 units of isophane protamine biosynthetic human insulin (pre-mixed 30R; Novonordisk, Tianjin, China) were administered subcutaneously. Glycemic levels were strictly controlled between 3.0 and 10.0 mmol/L (54-180 mg/dL).

Experimental samples

To label all blood vessels, intravascular perfusion of fluorescent tomato (Lycopersicon esculentum) lectin was used. The anesthetized rats were intravenously injected with 100 μL fluorescein isothiocyanate-conjugated tomato lectin (1 mg/mL; Sigma-Aldrich). Tomato lectin binds uniformly to the luminal surface of endothelial cells and labels all blood vessels with adequate blood supply. At 15 minutes after the injection, the rats were perfused with stroke-physiological saline solution for 10 minutes through the left ventricle under anesthesia, at a pressure of 10.7-16.0 kPa (80-120 mmHg) for 5-10 minutes. The vitreous humor and retina were carefully isolated from the eyes under a 2.5 × anatomic microscope. The vitreous humor was isolated for VEGF-A enzyme-linked immunosorbent assay (ELISA) and the retina was isolated for hematoxylin and eosin (H&E) staining, periodic acid-Schiff (PAS) staining, fluorescence imaging, and proinflammatory protein expression assessments during the sixth, eighth, and tenth weeks after the induction of DM, respectively.

Estimation of VEGF-A in the vitreous humor

The isolated vitreous humor was homogenized in 185 μL sterile phosphate-buffered saline (PBS) after being frozen at -80 °C for 5 minutes. The concentration of the VEGF-A protein in the vitreous homogenates was estimated using a rat VEGF-A ELISA kit capable of detecting both VEGF-A isoforms (RayBiotech Inc., Norcross, GA, United States), according to the manufacturer’s instructions. The antibodies in the kit have > 95% cross-reactivity with the rat.

H&E-stained retinal preparations

The retinal tissues were isolated from normal and diabetic rats and fixed in 4% paraformaldehyde solution at 20 °C for 2 hours. The samples were subsequently sectioned at a thickness of 5 μm, stained with H&E, and examined under a light microscope (magnification, 400 ×; Zeiss AG, Oberkochen, Germany) to determine the number and area of retinal ganglion cells (RGCs) in the sample.

PAS-stained retinal preparations

Retinal tissues were isolated from each group and fixed in 4% paraformaldehyde solution at 20 °C for 24 hours. The samples were then placed in trypsin solution for 40 minutes at 37 °C, stained with PAS, and examined under a light microscope (magnification, 400 ×; Zeiss AG) to determine the endotheliocyte to pericyte (E/P) ratio and the number of acellular strands.

Fluorescence imaging techniques for flat retinal preparations

Retinal flat mounts were processed to visualize the vascular basement membrane by immersing them in marker solutions. Before immersion staining, the retinal flat mounts were incubated for 30 minutes at room temperature in 5% normal bovine serum in PBS containing 0.5% Triton X-100 (0.5% T-PBS) as a blocking agent. Subsequently, the flat mounts were immersed overnight at room temperature in a marker solution containing rabbit polyclonal anti-type IV collagen antibody (1:300, ab19808; Abcam, Cambridge, United Kingdom) to target the basement membrane. Fluorescent goat anti-rabbit immunoglobulin G (1:45, BA1105; Wuhan Boster Biological Technology, Ltd., Wuhan, Hubei Province, China) was used as the secondary antibody. After secondary incubation at 20 °C for 5 minutes, the retinal flat mounts were washed three times with 0.5% T-PBS, placed in 4’,6-diamidino-2-phenylindole (DAPI) for 5 minutes, and washed an additional three times with 0.5% T-PBS. The retinal flat mounts were mounted on Vectashield (Wuhan Boster Biological Technology) and analyzed using the Zeiss LSM 710 confocal laser scanning microscope to determine the number of type IV collagen strands, as well as the area and number of retinal neurocytes.

Quantitative polymerase chain reaction

To determine the mRNA expression levels of VEGF, IL-6, CD18, ICAM-1, and TNF-α in ARPE-19 cells, HRMECs, and SD rat retinal tissues, total RNA was extracted from one-fourth of the remaining retinal tissue using TRIzol reagent (Invitrogen, Carlsbad, CA, United States) and reverse transcribed with the HiFiScript cDNA Synthesis Kit (First-Strand, CoWin Biosciences, China). Polymerase chain reaction (PCR) amplification was conducted using Taq DNA polymerase (Servicebio^®, Wuhan, Hubei Province, China) on a thermal cycler (GeneAmp PCR system; Applied Biosystems, Foster City, CA, United States). The oligonucleotide sequences of the quantitative PCR (qPCR) primers are presented in Table 1.

Table 1 Oligonucleotide sequences of the quantitative polymerase chain reaction primers.

Genes	Forward primers (5’-3’)	Reverse primers (3’-5’)
VEGF	TTGCCTTGCTGCTCTACCTCCA	GATGGCAGTAGCTGCGCTGATA
IL-6	AGACAGCCACTCACCTCTTCAG	TTCTGCCAGTGCCTCTTTGCTG
CD18	AGTCACCTACGACTCCTTCTGC	CAAACGACTGCTCCTGGATGCA
ICAM-1	AGCGGCTGACGTGTGCAGTAAT	TCTGAGACCTCTGGCTTCGTCA
TNF-α	CTCTTCTGCCTGCTGCACTTTG	ATGGGCTACAGGCTTGTCACTC
STAT3	GGGGTTCTGGGAAGTCTG	ATGGCTGGGTGAGGTTG
β-actin	AGCCATGTACGTAGCCATCC	ACCCTCATAGATGGGCACAG

CD18: Cluster of differentiation 18; ICAM-1: Intercellular adhesion molecule-1; IL-6: Interleukin 6; STAT3: Signal transducer and activator of transcription 3; TNF-α: Tumor necrosis factor alpha; VEGF: Vascular endothelial growth factor.

Western blotting

After extracting protein samples, the proteins were separated on 10% sodium dodecyl sulfate-polyacrylamide gels and subsequently transferred to a 0.45 μm polyvinylidene fluoride membrane (Invitrogen). The membranes were blocked in 5% skimmed milk for 1 hour and incubated with primary antibodies overnight at 4 °C, followed by incubation with secondary antibodies for 40 minutes at room temperature.

The primary antibodies included anti-hypoxia inducible factor-1 alpha (1:100, sc-13515; Santa Cruz Biotechnology, Santa Cruz, CA, United States), anti-angiopoietin-like protein 4 (ANGPTL4, 1:200, sc-373761; Santa Cruz Biotechnology), anti-ANGPTL4 (1:500, 18374-1-AP; Proteintech Group Inc., Chicago, IL, United States), anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (1:5000, 60004-1-Ig; Proteintech Group), anti-zona occludens 1 (1:500; Affinity Biosciences, Cincinnati, OH, United States), anti-occludin (1:500; Affinity Biosciences), anti-phosphorylated STAT3 (pSTAT3, 1:1000; Cell Signaling Technology, Danvers, MA, United States), and anti-STAT3 (1:1000; Cell Signaling Technology). Bands were detected and visualized using the Immobilon Western Chemiluminescent Horseradish Peroxidase Substrate (WBKLS0100; Merck Millipore, Billerica, MA, United States). The gray band densities were normalized to GAPDH values (used as a control) using ImageJ software (National Institutes of Health, Bethesda, MD, United States).

Image processing and statistical analyses

IPP 6.0 and ImageJ 2.0 were used to process the images, and IBM SPSS 19.0 statistical software was used for statistical analyses of the obtained data. One-way analysis of variance was conducted for multiple mean values, and independent samples t-test was conducted for the data between the groups. All data are presented as the mean ± SE of the mean. P < 0.05 was considered statistically significant.

RESULTS

Ranibizumab on the viability of ARPE-19 cells and HRMECs

ARPE-19 cells and HRMECs were treated with 0 mg/mL, 0.0625 mg/mL, 0.125 mg/mL, or 0.25 mg/mL of ranibizumab for 6 hours, 12 hours, 24 hours, and 48 hours to determine its effective concentration. The cells were subsequently assessed using the CCK-8 assay. In our study, cell viability decreased with increasing doses and durations of ranibizumab (Figure 1). Ranibizumab concentrations greater than 0.125 mg/mL (24 hours) were significantly inhibitory and cytotoxic. Therefore, 0.25 mg/mL ranibizumab was used to incubate the cells for 24 hours in subsequent experiments.

Figure 1 Effect of ranibizumab on the viability of adult retinal pigment epithelial 19 cells and human retinal microvascular endothelial cells. Cell viability was assessed using the Cell Counting Kit-8 assay (n = 3, independent experiments). A: Effect of ranibizumab (0 mg/mL, 0.0625 mg/mL, 0.125 mg/mL, or 025 mg/mL) treatment on adult retinal pigment epithelial 19 (ARPE-19) cell viability; B: Effect of ranibizumab (0 mg/mL, 0.0625 mg/mL, 0.125 mg/mL, or 025 mg/mL) treatment on human retinal microvascular endothelial cell (HRMEC) viability. All results are expressed as the mean ± SD. ^dP < 0.0001. ¹P vs 24 hour group. ²P vs 48 hour group. NC: Untreated group.

Expression of cytokine mRNA in ARPE19 cells and HRMECs

ARPE-19 cells and HRMECs were cultured under normal (5.5 mmol/L) and high (25 mmol/L) glucose concentrations for 48 hours. Ranibizumab (0.125 mg/mL) was then added for 24 hours (Figure 2). qPCR technology was used to assess the mRNA expression of VEGF, IL-6, CD18, ICAM, TNF-α, and STAT3 in each group. The mRNA expression of VEGF, IL-6, CD18, ICAM, TNF-α, and STAT3 was higher in the high-glucose group than the normal glucose group (P < 0.05). The mRNA expression levels decreased relative to that of the untreated cells after treatment with ranibizumab (P < 0.05).

Figure 2 mRNA expression of vascular endothelial growth factor, interleukin 6, cluster of differentiation 18, intercellular adhesion molecule, tumor necrosis factor alpha, and signal transducer and activator of transcription 3 in adult retinal pigment epithelial 19 cells and human retinal microvascular endothelial cells. Untreated (NC) (5.5 mmol/L glucose), NC + ranibizumab (5.5 mmol/L glucose + 0.125 mg/mL ranibizumab), high glucose (Hg) (25 mmol/L glucose), and Hg + ranibizumab (25 mmol/L glucose + 0.125 mg/mL ranibizumab). A: Ratios of the mRNA expression of vascular endothelial growth factor (VEGF), interleukin 6 (IL-6), cluster of 18 differentiation (CD18), intercellular adhesion molecule (ICAM), tumor necrosis factor alpha (TNF-α), and signal transducer and activator of transcription 3 (STAT3) in different adult retinal pigment epithelial 19 (ARPE-19) groups; B: Ratios of mRNA expression of VEGF, IL-6, CD18, ICAM, TNF-α, and STAT3 in different human retinal microvascular endothelial cell (HRMEC) groups. All results are expressed as the mean ± SD. ^aP < 0.05. ^bP < 0.01. ¹P vs NC group. ²P vs Hg group.

Expression of GFAP, STAT3, and pSTAT3 proteins in ARPE19 cells and HRMECs

ARPE-19 cells and HRMECs were cultured under normal (5.5 mmol/L) and high (25 mmol/L) glucose concentrations for 48 hours. Ranibizumab (0.125 mg/mL) was then introduced for 24 hours (Figure 3). Western blotting was used to detect the expression of GFAP, STAT3, and pSTAT3 proteins in each group. The expression of GFAP, STAT3, and pSTAT3 proteins in the high-glucose group were significantly higher than in the normal glucose group (P < 0.05). Protein expression decreased after treatment with ranibizumab compared to that in untreated cells (P < 0.05).

Figure 3 Expression of glial fibrillary acidic protein, signal transducer and activator of transcription 3 (STAT3), and phosphorylated STAT3 proteins in adult retinal pigment epithelial 19 cells and human retinal microvascular endothelial cells. Untreated (NC) (5.5 mmol/L glucose), NC + ranibizumab (5.5 mmol/L glucose + 0.125 mg/mL ranibizumab), high glucose (Hg) (25 mmol/L glucose), and Hg + ranibizumab (25 mmol/L glucose + 0.125 mg/mL ranibizumab). A: Expression of glial fibrillary acidic protein (GFAP), signal transducer and activator of transcription 3 (STAT3), and phosphorylated STAT3 (pSTAT3) proteins in adult retinal pigment epithelial 19 (ARPE-19) cells were detected by Western blot analysis; B: Expression of GFAP, STAT3, and pSTAT3 proteins in human retinal microvascular endothelial cells (HRMECs) were detected by Western blot analysis; C: Ratios of GFAP, STAT3, and pSTAT3 protein expression levels in different groups of ARPE-19 cells; D: Ratios of GFAP, STAT3, and pSTAT3 protein expression levels in different groups of HRMECs. All results are expressed as the mean ± SD. ^aP < 0.05. ^bP < 0.01. ¹P vs NC group. ²P vs Hg group.

Metabolic condition of rats

The experiment was conducted using 100 diabetic rats, which were randomly divided into five groups (A, B, C, D, E), with each group consisting of 20 rats. Additionally, 20 normal rats without any treatment served as the control group (Group F). Diabetes was induced in Groups A, B, C, D, and E with an intraperitoneal injection of streptozotocin at a dosage of 60 mg/kg. On the eighth day after diabetes induction, certain groups received interventions, such as intravitreal injections or intensive blood glucose control. During the follow-up period, FFA examinations were performed on five rats from each group at 4 week intervals (weeks 4, 6, 8, and 10). At the end of each time point, five rats from each group were randomly sacrificed for further analysis. The control group (Group F) did not receive any intervention. Throughout the experiment, changes in body weight and blood glucose concentrations were recorded for all groups.

Compared with Group F, the body weight of SD rats in Groups A, B, C, D, and E significantly decreased on the fourth day after successful modeling (P < 0.05). However, after implementing strict blood glucose control on the eighth day, the body weight of rats in groups A and C gradually increased, while remaining lower than the normal group (P < 0.05). By contrast, the body weight of rats in Groups B, D, and E continued to decrease as the disease progressed. On the first day after successful modeling, the blood glucose levels in Groups A, B, C, D, and E rose sharply, showing a statistically significant difference compared with group F (P < 0.001). By the eighth day, following strict blood glucose control, the blood glucose levels in Groups A and C approached normal levels, whereas those in Groups B, D, and E remained consistently elevated (Figure 4).

Figure 4 Changes in body weights and blood glucose concentrations of Sprague-Dawley rats within 10 weeks after induction. A: Experimental flowchart: 100 diabetic rats were randomly divided into Groups A, B, C, D, and E, while 20 normal rats with no treatment were assigned to Group F; B: Changes in the body weights of Sprague-Dawley rats in each group; C: Changes in the blood glucose concentrations of Sprague-Dawley rats in each group. All results are expressed as the mean ± SD. ^aP < 0.05. ^cP < 0.001. P vs A group. FFA: Fundus fluorescein angiography; STZ: Streptozotocin.

FFA results

We performed FFAs to document any fundus changes in the rats during the 10 week observation period (Figure 5). Sodium fluorescein was injected intraperitoneally for 3-5 seconds. Retinal vein laminar flow was then observed for 5-7 seconds until it disappeared completely after 3-6 minutes. The entire skin of the SD rats was yellow at the end of examination. During the eighth and tenth weeks, there were significantly more RGCs in Groups A and B than in Groups C and D (P < 0.01) but no significant difference was noted compared with Group F (P > 0.05). Additionally, there was no difference in the RGC areas among the groups at all time points. We observed microvascular dilatation in Groups C and D during the sixth week, which worsened as the disease progressed. Neovascular buds and vascular expansion were eventually observed in Group D during the tenth week, while no such phenomena were observed in Groups A, B, and E.

Figure 5 Morphological retinal changes on fundus fluorescein angiography (scale bar: 200 μm). The yellow arrow indicates vessels with increased vascular tortuosity and dilatation. The orange arrow indicates leaky vessels with increased tortuosity and dilatation. The green arrow indicates areas of retinal vascular occlusion.

Retinal H&E staining

We observed the morphological changes in the retinal nerve fiber layer and retinal pigment epithelium during the sixth, eighth, and tenth weeks after DM induction. The number and area of RGCs in each group were counted and measured (Figure 6).

Figure 6 Hematoxylin and eosin staining and the ratio of retinal ganglion cell (RGC) number to RGC area in each group at each time point. A: Ratio of retinal ganglion cell (RGC) number to RGC area in each group based on retinal tissue hematoxylin and eosin staining (magnification: 400 ×; scale bar: 25 μm); B and C: Ratio of RGC number to RGC area in each group at each time point, respectively. The blue arrows indicate neovascularization buds, while the yellow arrows indicate abnormally dilated microvessels. All results are expressed as the mean ± SD. ^aP < 0.05. ^bP < 0.01. ^cP < 0.001. ¹P vs F group. ²P vs C group. ³P vs B group.

During the eighth and tenth weeks, there were significantly more RGCs in Groups A and B than in Groups C and D (P < 0.01), but no significant difference was noted compared with Group F (P > 0.05). Additionally, there was no difference in RGC areas among the groups at all time points. We observed microvascular dilatation in Groups C and D during the sixth week, which worsened as the disease progressed. Neovascular buds and vascular expansion were eventually observed in Group D during the tenth week, while no such phenomena were observed in Groups A, B, and E.

Retinal PAS staining

PAS staining of retinal tissue depicts elongated, light-colored retinal vascular endothelial cell nuclei and round, darker-colored pericyte nuclei. We measured the E/P ratios and the number of acellular capillaries during the fourth, sixth, eighth, and tenth weeks (Figure 7). During the eighth and tenth weeks, the E/P ratios of Groups A and B were significantly lower than in Groups C and D (P < 0.05). There was no difference in the E/P ratio in Groups C and D during the sixth and eighth weeks, but a significant difference was observed during the tenth week (P < 0.05). Acellular capillaries formed in all groups during the observation period; however, there was no significant difference between Groups A and B and Group E. There were significantly less acellular capillaries in Groups C and D. There was no difference in the E/P ratio between Groups C and D during the sixth and eighth weeks after modeling, but a significant difference was observed during the tenth week (P < 0.05).

Figure 7 Ratio between the endothelial cell to pericyte ratio and the number of acellular strands. A: Ratio between the endothelial cell to pericyte (E/P) ratio and the number of acellular strands in each group revealed by retinal periodic acid-Schiff staining (magnification: 400 ×; scale bar: 2.5 μm); B and C: Ratio between E/P and the number of acellular strands during the sixth, eighth, and tenth weeks. The yellow arrows indicate the acellular strands, while the orange arrows indicate neovascularization bud. All results are expressed as the mean ± SD. ^aP < 0.05. ^bP < 0.01. ¹P vs F group. ²P vs B group. ³P vs D group.

Fluorescence imaging techniques for retinal flat preparations

This study examined the cell and microvessel changes in the retinal tissue of SD rats using fluorescein isothiocyanate-tomato lectin, rabbit polyclonal anti-type IV collagen antibody, and DAPI-labeled retinal flat preparation (Figure 8). There were less anti-IV + collagen strands in the retinal tissues of Groups A or B than in Groups C and D at each time point over the 10 week observation period (P < 0.05), but no relative difference compared to Group E. There were also less anti-IV + collagen strands in the retinal tissue of Group C than in Groups D and E at each time point (P < 0.05).

Figure 8 Immunohistochemical imaging and the ratio of the number of type IV collagen-positive strands, retinal cells, and area of retinal cells. A: Immunohistochemical imaging of retinal tissues in Sprague-Dawley rats (magnification: 100 ×); B-D: Ratio of the number of type IV collagen-positive strands, retinal cells, and area of retinal cells during the sixth, eighth, and tenth weeks, respectively. The white arrows indicate type IV collagen-positive strands. The yellow arrows indicate areas of vascular permeability. The orange arrows indicate vascular buds. All results are expressed as the mean ± SD. ^aP < 0.05. ^bP < 0.01. ¹P vs F group. ²P vs B group. ³P vs D group. AT-IV: Anti-IV; CSs: Collagen strands; DAPI: 4’,6-diamidino-2-phenylindole; FITC: Fluorescein isothiocyanate; TL: Tomato lectin; RFC: Retinal cell.

Similarly, there were less retinal tissue cells in Groups A and B than in Group E at each time point during the 10 week observation period (P < 0.01). There was no difference in the number of retinal cells in Groups C and D during the sixth week, but a significant difference was observed with disease progression (P < 0.01).

Group A exhibited local vascular leakage during the tenth week, whereas Groups C and D showed significant vascular leakage as early as the eighth week. Group D demonstrated local neovascular bud formation during the tenth week, but the same was not observed in Group B over the duration of the study. We observed no differences in retinal tissue cell volumes among the groups at all time points.

Expression of multiple cytokine mRNAs in retinal tissue

This study used qPCR to measure the expression of various cytokine mRNAs in the retinal tissues of SD rats (Figure 9). At weeks 6, 8, and 10, VEGF mRNA expression was significantly higher in Groups A, B, and F than in Groups C, D, or E (P < 0.05). Similarly, at weeks 4, 6, and 10, IL-6 mRNA expression was significantly lower in Groups A, B, and F than in Groups C, D, or E (P < 0.05). At weeks 4, 6, 8, and 10, ICAM mRNA expression was significantly lower in Groups A, B, and F than in Groups C and D (P < 0.05). Additionally, at weeks 4, 8, and 10, TNF-α mRNA expression was significantly lower in Groups A, B, and F than in Groups C, D, and E (P < 0.05). We did not note any significant difference in the mRNA expression of CD18 and STAT3 among the groups at any of the timepoints during this study (P > 0.05).

Figure 9 mRNA expression of vascular endothelial growth factor, interleukin 6, cluster of differentiation 18, intercellular adhesion molecule, tumor necrosis factor-alpha, and signal transducer and activator of transcription 3 in the retinal tissues of Sprague-Dawley rats. A: Ratio of vascular endothelial growth factor (VEGF) mRNA expression in the retinal tissue of each group of Sprague-Dawley (SD) rats; B: Ratio of interleukin 6 (IL-6) mRNA expression in the retinal tissue of each group of SD rats; C: Ratio of cluster of differentiation 18 (CD18) mRNA expression in the retinal tissue of each group of SD rats; D: Ratio of intercellular adhesion molecule (ICAM) mRNA expression in the retinal tissue of each group of rats; E: Ratio of tumor necrosis factor alpha (TNF-a) mRNA expression in the retinal tissue of each group of rats; F: Ratio of signal transducer and activator of transcription 3 (STAT3) mRNA expression in the retinal tissue of each group of rats. All results are expressed as the mean ± SD. ^aP < 0.05. ^bP < 0.01. ¹P vs A group. ²P vs B group. ³P vs C group. ⁴P vs D group. ⁵P vs F group.

GFAP, STAT3, and pSTAT3 protein expression in retinal tissue

GFAP, STAT3, and pSTAT3 protein expression was measured by Western blotting (Figure 10). GFAP protein expression was consistently high in Group D throughout the 10 week observation period and significantly higher than in Groups E and F at all time points (P < 0.05). Comparatively, GFAP protein expression in Group C was only high during the fourth week, but it was significantly higher than in Groups D, E, and F at this time point (P < 0.05). GFAP protein expression in Groups A and B showed a significant increase by the fourth week and remained high for the duration of the study (P < 0.05).

Figure 10  Expression of glial fibrillary acidic protein, signal transducer and activator of transcription 3 (STAT3), and phosphorylated STAT3 proteins in the retinal tissues of each group of Sprague-Dawley rats. A: Detection of glial fibrillary acidic protein (GFAP), signal transducer and activator of transcription 3 (STAT3), and phosphorylated STAT3 (pSTAT3) protein expression in the retinal tissues of Sprague-Dawley (SD) rats using Western blot analysis; B: Ratio of GFAP protein expression in the retinal tissues of each group of SD rats; C: Ratio of STAT3 protein expression in the retinal tissues of each group of SD rats; D: Ratio of pSTAT3 protein expression in the retinal tissues of each group of rats. All results are expressed as the mean ± SD. ^aP < 0.05. ^bP < 0.01. ¹P vs A group. ²P vs C group. ³P vs D group. ⁴P vs F group. GAPDH: Glyceraldehyde-3-phosphate dehydrogenase.

STAT3 protein expression in Group D gradually increased with disease progression. During the fourth week, STAT3 protein expression was significantly lower in Group A than in Groups B, D, and E (P < 0.05). However, as the disease progressed, the pSTAT3 protein levels in Groups A and B were higher and showed more variations than in Groups C and F at the same time points (P < 0.05).

DISCUSSION

DR, a major ocular complication of diabetes, significantly impacts global health[20]. The mechanisms underlying its development and progression are complex and poorly understood[21]. Further elucidation of these mechanisms may aid in mitigating the progression of DR. Glucose-treated RPE cells and HRMECs are commonly used as a model for investigating DR[22]. Several studies have demonstrated that retinal microvascular damage is associated with the upregulation of various cytokines, such as IL-6, TNF-α, VEGF, CD18, and ICAM-1[23,24]. We detected the increased expression of these factors under high-glucose conditions in both ARPE-19 cells and HRMECs. The introduction of ranibizumab partially reduced their expression.

Studies have also shown that retinal neuronal dysfunction occurs prior to the onset of microvascular lesions in DR[25]. We observed a gradual decrease in RGC number in SD rats over time. Low RGC counts are associated with various degenerative eye diseases, such as DR and glaucoma[26]. While the specific molecular mechanisms underlying RGC pathologies remain unclear, inflammation, oxidative stress, and advanced glycation end products are theorized to be the primary culprits in DR[27].

The morphological changes in early DR include retinal alterations, such as pericyte loss, endothelial cell apoptosis, increased vascular permeability, and capillary dropout[28,29]. Pericyte loss is typically the earliest observed change, we theorize that pericyte loss occurs secondary to the inflammation and metabolic disturbances observed in high-glucose states. Pericyte loss then leads to increased retinal vasculature instability and permeability, as well as endothelial cell injury and apoptosis[30,31]. We theorize that pericyte loss occurs secondary to the inflammation and metabolic disturbances observed in high-glucose states. Pericyte loss then leads to increased retinal vasculature instability and permeability, as well as endothelial cell injury and apoptosis. In our study, diabetic rats showed an increase in the endothelial/pericyte ratio, acellular capillary count, and anti-IV + collagenous strand numbers after 4 weeks of successful modeling, which was stabilized by week 10. Vascular leakage and neovascular buds were observed in weeks 8 and 10.

GFAP is significantly upregulated in DR. It typically serves as a critical marker of reactive gliosis in Müller cells, which reflects the retina’s stress response to hyperglycemic environments[32]. GFAP overexpression is closely associated with Müller cells activation. It further exacerbates retinal vasculature abnormalities by inducing higher levels of inflammation and oxidative stress, which triggers retinal neurodegeneration and RGC apoptosis[14]. Studies have shown that GFAP levels are elevated during early DR. As such, changes in its serum values hold potential diagnostic value[33]. Recent work on GFAP-targeted therapies such as Müller cell suppressions or GFAP RNA inhibition, have demonstrated promising efficacy, offering novel approaches for the treatment of DR[25]. As such, GFAP seems to be an excellent molecular marker for understanding DR pathogenesis, as well as a potential therapeutic target. Future research is warranted for this.

VEGF overexpression is associated with abnormal angiogenesis and increased retinal vascular permeability, which contribute to retinal dysfunction[17]. Fluorescein (or Evans blue) studies in diabetic rat models have demonstrated a significant increase in retinal vessel permeability on day 8 (week 2) of modeling, which provides evidence that increased retinal vascular permeability occurs in early DR[34].

IL-6, a widely functional and pleiotropic cytokine, is upregulated in cases of trauma, surgery, and infection, while TNF-α is a multifunctional pro-inflammatory cytokine that plays an important role in inducing inflammatory reactions and diabetes. CD18, a type I transmembrane protein in the integrin superfamily, is involved in mediating inflammatory functions like leukocyte adhesion. CD18 mRNA expression in the retinal tissues of diabetic rats began to increase a week after successful modeling[20]. STAT3, a member of the signal transducer and activator of transcription family, is involved in the development of DR and exists as an inactive monomer in the cytoplasm of various cells, mediating cell proliferation and angiogenesis[35]. After phosphorylation, the monomeric form of STAT3 aggregates and translocates to the nucleus, upregulating the expression of multiple signaling pathway-related factors in cells[36]. VEGF and IL-6 activate STAT3, but STAT3 can likewise mediate the expression of VEGF, IL-6, ICAM-1, and TNF-α. Collectively, these molecules participate in the development and progression of DR[7,9].

IL-6, CD18, ICAM, TN-α, and STAT3 levels change as DR progresses. Our study aimed to further understand the molecular mechanisms behind early DR. After 4 weeks of successful modeling, we found no differences in the expression of VEGF, IL-6, ICAM, TNF-α, and STAT3 mRNA in the retinal tissue of SD rats in Groups D and E. The vitreal VF-A concentration in these groups were also similar. By contrast, Group D showed lower CD18 mRNA levels compared to Group E, which indicated that intravitreal ranibizumab had no significant effect on cytokine expression immediately after injection.

DR is an important sequelae of DM. In most cases, it can be prevented, treated, or controlled by insulin therapy. However, if left unchecked, severe cases can lead to blindness[37] after strict blood sugar control. In our study, Group C had higher RGCs, E/P ratios, acellular capillary count, vitreous VEGF-A concentrations, RGC counts, acellular capillary counts, anti-IV + collagenous strand counts, and retinal cell counts. Group C also consistently had higher mRNA expression levels of VEGF, CD18, and TNF-α compared to Groups D and E at the same time point after strict blood glucose control. In contrast and compared to Groups D and E, Group C showed higher levels of STAT3 mRNA only during the fourth week. Group C also only showed elevations of IL-6 and ICAM mRNA in the eighth week. This suggested that strict blood glucose control in Group C effectively delayed DR progression.

Ranibizumab is monoclonal antibody with a molecular weight of 48 kDa. It works by binding to and neutralizing VEGF-A, which downregulates VEGF levels in the eyes of patients with DR. Eyes with lower levels of VEGF are less likely to exhibit pathologic neovascularization, advanced DR, and blindness. In clinical practice, the half-lives of 0.5 mg ranibizumab in monkey and rabbit eyes are 2.6 and 2.9 days, respectively. These are shorter than that of bevacizumab (4.32 days in rabbit eyes) and aflibercept (3.92 days in rabbit eyes). Comparatively, the half-life of ranibizumab in human eyes is 7.19 days. In our study, Groups A and B received intravitreal injections of 1 μL ranibizumab on day 8 after successful modeling. Our data demonstrated that intravitreal ranibizumab during early DR effectively delayed disease progression by downregulating VEGF, IL-6, CD18, ICAM, and TNF-α mRNA expression and upregulating STAT3 mRNA expression. Nakao et al[38] reported that anti-VEGF-A drugs effectively reduce retinal inflammation. Our data supported this by showing that intravitreal ranibizumab downregulated the expression of inflammatory mediators in the eyes of SD rats with early DR. SD rats that received intravitreal ranibizumab had lower rates of retinal damage from vascular leakage, leukocyte accumulation, and general inflammation, which ultimately delayed DR. progression. STAT3 mRNA upregulation also seems to play a role in ranibizumab activity.

Our study had certain limitations. Specifically, the limitation is the reliance on an animal model, which may not fully reflect the complexities of human DR. Although SD rats are commonly used in DR research, species differences in retinal structure and disease progression could impact the applicability of the findings to human patients. Therefore, clinical studies involving human subjects or human retinal tissue models are necessary to confirm the relevance and effectiveness of ranibizumab in treating DR in humans.

In conclusion, our data demonstrated that early intravitreal ranibizumab effectively regulated the VEGF/STAT3/GFAP signaling pathway in diabetic rat models. This maintained vitreous VEGF concentration levels at normal values, as well as downregulated VEGF, IL-6, CD18, ICAM, and TNF-α mRNA expression and upregulated STAT3 mRNA expression. Inhibition of STAT3 protein upregulated GFAP and pSTAT3 protein levels, which increased the number of RGCs and maintained the retinal vascular network system. All these contributed to effectively delaying the progression of DR.

CONCLUSION

Retinal neurofunctional impairment and microvascular changes are evident in early DR. Intravitreal injection of ranibizumab has been shown to mediate the expression of various cytokines through the VEGF/STAT3/GFAP signaling pathway, which promotes retinal neuronal and vascular repair and delays DR progression. Combining intravitreal ranibizumab with glycemic control seems to provide even stronger protection against DR progression. These findings highlight the potential clinical significance of early interventions that target the VEGF pathway and offers new therapeutic strategies for the management and prevention of DR progression.